-------------------------------------------------------


NEW PAPER


Conclusion: CT severity score was associated with inflammatory levels and higher NLR and CT severity score on admission were independent risk factors for short-term progression in patients with NCP outside Wuhan. Furthermore, early admission and surveillance by CT should be recommended to improve clinical outcomes.
Objective: To determine the predictive value of CT and clinical characteristics for short-term disease progression in patients with 2019 novel coronavirus pneumonia (NCP). Materials and Methods: 224 patients with confirmed 2019 novel coronavirus (COVID-19) infection outside Wuhan who had chest CT examinations were retrospectively screened. Clinical data were obtained from electronic medical records. CT images were reviewed and scored for lesion distribution, lobe and segment involvement, ground-glass opacities, consolidation, and interstitial thickening. All included patients with moderate NCP were observed for at least 14 days from admission to determine whether they exacerbated to severe NCP (progressive group) or not (stable group). CT and clinical characteristics between the two groups were compared, and multivariate logistic regression and sensitivity analyses were performed to identify the risk factors for developing severe NCP. Results: A total of 141 patients with moderate NCP were included, of which 15 (10.6%) patients developed severe NCP during hospitalization and assigned to the progressive group. Multivariate logistic regression analysis showed that higher neutrophil-to-lymphocyte ratio (NLR) (odds ratio [OR] and 95% confidence interval [CI], 1.26 [1.04-1.53]; P = 0.018) and CT severity score (OR and 95% CI, 1.25 [1.08-1.46]; P = 0.004) on admission were independent predictors for progression to severe NCP, and sensitivity analysis confirmed the consistent results in nonimported patients but not in imported patients. However, no significant difference in lung involvement was found on CT between imported and nonimported patients (all P > 0.05). Patients who were admitted more than 4 days from symptom onset tended to have more severe lung involvement. Spearman correlation analysis showed the close association between CT severity score and inflammatory indexes (r = 0.17~0.47, all P < 0.05). The outbreak of 2019 novel coronavirus pneumonia (NCP) originated from Wuhan has shown the ability of human-to-human transmission and rapidly spread to become a world-wide emergency along with increasing imported and secondary contacted infection risk. 1 Most patients with NCP have a mild clinical course, while a proportion of patients demonstrated rapid deterioration (particularly within 7-14 days) from onset of symptoms into severe NCP with or without acute respiratory distress syndrome (ARDS). These patients have poor survival and often require intensive medical resource utilization, and the mortality of them are about 20 times higher than that of non-severe 2019 novel coronavirus (COVID-19) patients. 2,3 Thus, early identification of patients at risk of serious complications of NCP is of clinical importance. Several studies reported that the prevalence of severe NCP ranged from 15.7% to 26.1% and these cases were often associated with abnormal chest CT findings and clinical laboratory data. [3] [4] [5] Guan et al indicated that patients with severe NCP were more likely to show ground-glass opacity (GGO), local or bilateral patchy shadowing, and interstitial abnormalities on CT. 5 This likely reflects the clinical progression of disease but also offers an opportunity to investigate the clinical utility of chest CT as a predictive tool to risk-stratify the patients. Furthermore, the predictive value of chest CT in NCP prognosis is warranted as to assist the effective treatment and control of disease spread. Previous study suggested that higher CT lung score correlated with poor prognosis in patients with Middle East respiratory syndrome coronavirus (MERS). 6 Chest CT has been demonstrated to be an important approach for screening individuals with suspected NCP and monitoring treatment response according to the dynamic radiological changes of NCP. 7 Therefore, we enrolled a cohort of patients with moderate NCP on admission and observed for at least 14 days to explore the early CT and clinical risk factors for progression to severe NCP. Meanwhile, we also compared the CT and inflammatory indexes in patients with different source of infection or period from symptom onset to admission to provide deep understanding of the relationship among CT findings, epidemiological features, and inflammation. Our study was conducted in compliance with the Health Insurance Portability and Accountability Act (HIPAA). The permission of Institutional Review Board of The Third Xiangya Hospital was obtained for retrospective assessment of clinical and radiological data with waiver of informed consent. Health records were reviewed for patients who were treated at Third Xiangya Hospital, Changsha Public Health Treatment Center, and Second People's Hospital of Hunan between January 17, 2020 and February 1, 2020. Patients were included in the study if they satisfied the following criteria: 1) confirmed COVID-19 infection; 2) available chest CT scan on admission. The diagnosis of COVID-19 infection was established based on the World Health Organization (WHO) interim guidance, and a confirmed case was defined as a positive result to high-throughput sequencing or real-time reverse transcription-polymerase chain reaction (RT-PCR) assay for nasal and pharyngeal swab specimens. Patients were observed for at least 14 days from admission to determine whether they exacerbated to severe NCP (progressive group) or not (stable group). Demographic, epidemiological, and laboratory data were obtained with data collection forms from electronic medical records. According to the source of infection, patients were divided into imported group (patients who had recently been to Wuhan) and nonimported group (patients who contacted with people from Wuhan or local infected patients). The clinical classification of NCP are as follows: (1) moderate type, patients with fever, respiratory tract symptoms, and radiological evidence of confirmed pneumonia. (2) severe type, patients with one of the following: a) respiratory distress (respiratory rate ≥ 30 beats/min); b) hypoxia (oxygen saturation ≤ 93% in the resting state); c) hypoxemia (arterial blood oxygen partial pressure/oxygen concentration ≤ 300mmHg). (3) critical type, patients with one of the following: a) respiratory failure requiring mechanical ventilation; b) shock; c) intensive care unit (ICU) admission is required for combined other organs failure. In this study, severe NCP broadly included severe and critical types as above defined. The endpoint of this study was the development of severe NCP, and patients who discharged or did not developed severe NCP at enrollment were censored. The SOMATOM go.Top scanner (Siemens, Erlangen, Germany), GE Revolution CT scanner (GE Healthcare, Waukesha, USA), and TOSHIBA Aquilion 16 (Toshiba, Tokyo, Japan) were used for chest CT examinations. The images were reconstructed to 1.0-mm thickness for the transverse scans. Sagittal and coronal reconstructions with a 3.0-mm thickness were performed. All CT images were reviewed independently by two radiologists, each with over 10 years of experience in chest imaging. A third experienced radiologist was consulted if there was a disagreement in interpreting imaging results. The imaging features including ground-glass opacities (GGO), consolidation, crazy-paving, and air bronchogram defined as previous studies were recorded. 8, 9 The lesion distribution pattern, lobe and segment involvement were also assessed. The CT findings in the outer one third of the lung were defined as peripheral, and those in the inner two thirds of the lung were defined as central. Besides, the presence of discrete nodules, lymphadenopathy, and pleural effusion was recorded. Each of the five lung lobes was reviewed for opacification and consolidation. The lesions extent within each lung lobe was semi-quantitatively evaluated by scoring from 0 to 5 based on the degree of involvement: score 0, none involvement; score 1, ≤ 5% involvement; score 2, 6%~25% involvement; score 3, 26%~50% involvement; score 4, 51%~75% involvement; score 5, > 75% involvement. The total score was calculated by summing up scores of all five lobes to provide a CT severity score ranging from 0 to 25. 7 . CC-BY-NC-ND 4.0 International license It is made available under a author/funder, who has granted medRxiv a license to display the preprint in perpetuity. is the (which was not peer-reviewed) The copyright holder for this preprint . Quantitative variables are presented as median and interquartile range (IQR), and categorical variables are presented as frequency and percentage. Normality of the continuous variables was assessed with the Kolmogorov-Smirnov test. Differences between groups were examined using Student's t-test or Mann-Whitney U test for quantitative variables according to the normal distribution and Chi-square test or Fisher's exact test for categorical variables. Multivariate logistic regression with forward stepwise selection based on likelihood ratio was used to identify the risk factors for development of severe NCP, and sensitivity analysis was performed according to the source of infection. Statistical analyses were performed using IBM SPSS statistics software (version 22.0, SPSS Inc., Chicago, IL, USA) and R software (version 3.6.1). A P value of less than 0.05 considered to be statistically significant. A total of 224 patients were identified according to the inclusion criteria, of which 83 patients were excluded for having: 1) observation period from admission less than 14 days (n = 60); 2) negative CT findings or severe NCP on admission (n = 16); 3) age younger than 18 years old (n = 7). Finally, a cohort of 141 patients were included in our study (Figure 1 ). The baseline demographic, epidemiological, and laboratory characteristics of included 141 patients are presented in Table 1 . Among them, 76 (53.9%) patients who had recently been to Wuhan were imported and 72 (51.1%) were male, with a median age of 44 years and a median period symptom onset to admission of 4 days. 52 patients of them had discharged at enrollment. During the hospitalization, 15 cases progressed to severe NCP (progressive group) and the remaining patients did not (stable group). Compare with stable group, patients in progressive group were significantly older (P = 0.001), but the male gender and imported proportions were not shown statistical difference. Patients who progressed to severe NCP were more likely to have underlying hypertension (P = 0.004), but did not otherwise have significant differences in other co-morbidities including diabetes, chronic obstructive pulmonary disease (COPD), cardiovascular disease, cerebrovascular disease and chronic hepatitis B infection. The main symptoms between the two groups were not statistically different, while slightly more patients manifested digestive symptoms in progressive group, such as anorexia and diarrhea (P = 0.088 and 0.065, respectively). Patients in the progressive group had lower baseline lymphocyte count, higher neutrophil-to-lymphocyte ratio (NLR), and C-reactive protein (all P < 0.01). There were 87 (61.7%) patients with 4 to 5 lobes involved, with a median of 9 segments involved. Most lesions were bilateral (87.2%) and peripheral or mixed distributed (98.6%). The main CT characteristics of patients with NCP included GGO, GGO with crazy-paving, consolidation, and GGO with consolidation, ranging from 29.8%~95.7% (Table 1 ; Figure 2 a-f). Other infrequent features included discrete nodules, lymphadenopathy, and small pleural effusion. Compared with stable group, patients in progressive group had more lobes and segments involved, higher proportion of crazy-paving sign, and higher CT severity score (all P < 0.05; Figure 3 a). Logistic regression analysis showed that baseline NLR (odds ratio [  There was no significant difference in the CT findings between imported and nonimported patients, who had similar lesion distribution pattern, main imaging features and comparable lung involvement and CT severity score (Figure 3 b) . For patients who were admitted more than 4 days from symptom onset, more lobes or segments involved and higher CT severity score were found (Figure 3 c) , while there was no significant difference in age and inflammatory indexes (including white blood cell count, NLR, and C-reactive protein) between patients with different period (≤ 4 days vs. > 4 days) from symptom onset to admission. Further Spearman correlation analysis showed the close association between CT characteristics and inflammatory indexes, especially for CT severity score and C-reactive protein ( Figure 4 ). In this study, we retrospectively assessed the CT and clinical characteristics of 141 patients with NCP outside Wuhan and identified the baseline risk factors for clinical progression. Our results showed that higher NLR and CT severity score on admission were independent predictors for progression to severe NCP, especially in nonimported patients. In addition, patients who were admitted longer from symptom onset tended to have more severe lung involvement. With the increase of newly confirmed and severe cases, the management of severe patients brings huge challenge in efficiently controlling the COVID-19 outbreak. Some patients progressed rapidly with ARDS and subsequent multiple organ dysfunction syndrome (MODS). 2 Early identification of patients with high risk to develop ARDS or MODS and risk stratification management might be helpful to reduce mortality. In our cohort, the prevalence of severe NCP was 10.6%, which was . CC-BY-NC-ND 4.0 International license It is made available under a author/funder, who has granted medRxiv a license to display the preprint in perpetuity. is the (which was not peer-reviewed) The copyright holder for this preprint . lower than some large-scale reports. 3, 5 The main cause was because our patients were moderate on admission and enrolled from one single city. We found that progressive patients were older and had a greater number of underlying hypertension than stable patients. These data were in agreement with another recent report, which suggested that age and history of hypertension may be risk factors for short-term progression. 10 Our study further showed that progressive patients had lower baseline lymphocyte count, higher NLR and C-reactive protein. COVID-19 virus might act on lymphocytes as does severe acute respiratory syndrome coronavirus (SARS-CoV) which induces a cytokine storm and results in a series of immune responses. 11 Some studies suggest that decrease of T lymphocyte count indicate that coronavirus consumes many immune cells and inhibits the cellular immune function, and reduced but hyperactivated peripheral T cells partially accounts for the severe immune injury in COVID-19 patients. 12, 13 Thus, damage to lymphocytes might be an important factor leading to exacerbations of patients. NLR, as a simple lymphocyte related parameter to assess easily the inflammatory status, is widely used for the prediction of prognosis of patients with pneumonia. 14, 15 Higher NLR indicates damaged lymphocyte function and/or risk of bacterial infection due to low immune function. In addition, C-reactive protein is another serum maker produced by the liver in response to inflammation. Liu et al showed that C-reactive protein might be predictive of disease severity in COVID-19 infected patients. 16 Thus, our results suggested that patients with higher inflammatory levels on admission had higher risk to develop severe NCP. To explore the predictive value of CT for progression, we compared the difference of CT characteristics in stable and progressive patients, and found that progressive patients had greater lung involvement and higher CT severity score. CT severity score is used to semi-quantitatively estimate the pulmonary involvement, which is associated with both number of involved lobes and extent of lesions. 17 Pan et al reported a 4-stage dynamic pattern of NCP based on CT severity score and indicated that lung abnormalities showed greatest severity approximately 10 days after initial onset of symptoms. 7 In support of our findings, a previous report regarding MERS showed the predictive value of CT severity score for prognosis and short-term mortality. 6 Furthermore, higher proportion of progressive patients showed crazy-paving sign which reflects interstitial thickening. 18 Guan et al also reported higher proportion of severe patients with interstitial abnormalities. 5 The binding of COVID-19 spike protein to the receptor angiotensin converting enzyme II (ACE2) contributes to the downregulation of ACE2, increased pulmonary capillary permeability, and diffuse alveolar damage during NCP. 19, 20 In patients with SARS, mixed and predominant reticular patterns were noted from the second week. 8, 21 Thus, we speculated that the involvement of the interstitial vascular endothelial cells results in interlobular and intralobular septal thickening, which is associated with the disease severity. Our results showed that NLR and CT severity score on admission were significant predictors for progression to severe illness. The early predictive value of NLR has been reported in a recent study. 10 To our knowledge, this is the first study to indicate . CC-BY-NC-ND 4.0 International license It is made available under a author/funder, who has granted medRxiv a license to display the preprint in perpetuity. is the (which was not peer-reviewed) The copyright holder for this preprint . the role of CT to predict disease progression in patients with NCP. Previous study showed that the MuLBSTA score can early warn the mortality of viral pneumonia, which included lymphopenia and multilobe infiltration. 22 Our findings were consistent with theirs but more quantitatively in terms of imaging evaluation of lung involvement. Like SARS and MERS, some NCP patients progressed rapidly at about 10-14 days after onset likely dues to the cytokine storm in the body as evidenced by increased plasma proinflammatory cytokines. 11, 23 Our results further revealed the significant association between inflammatory markers and CT characteristics, especially for C-reactive protein and CT severity score, which indicated the potential value of CT to estimate pulmonary inflammation and lung damage. Furthermore, the predictive significance of NLR and CT severity score was confirmed only in nonimported patients outside Wuhan. Among the imported patients from Wuhan, the elder had higher risk to develop severe illness. However, the reason for this discrepancy between imported and nonimported patients remains to be elucidated. The patients who were admitted more than 4 days after symptom onset had more severe lung involvement, which likely attributed to the lung involvement progression as disease course extends. Song et al found that more consolidation and less GGO lesions in patients with an interval > 4 days between symptom onset and CT scan. 24 Thus, early admission and surveillance by CT is crucial for the individual management of patients with NCP. There were some limitations in our study. First, our study retrospectively included 141 patients with moderate NCP on admission from one single city outside Wuhan. Data from multiple cities are needed to further validate our findings. Second, dynamic inflammatory indexes and treatment during hospitalization were not included in the analysis. More comprehensive investigation of the relationship between CT characteristics and cytokine storm induced by COVID-2019 needs to be performed. In conclusion, our results showed that higher NLR and CT severity score on admission were independent risk factors for short-term progression in patients with NCP. The CT findings in patients with NCP were associated with inflammatory levels and both them have the potential to predict disease progression. Imported and nonimported patients outside Wuhan should be managed equally, and early admission is required for avoiding severe pulmonary damage. CT severity score 6 (4-10) 6 (3-9) 10 (7-15) 0.003 COPD, chronic obstructive pulmonary disease; CT, computed tomography; GGO, ground-glass opacities; NCP, 2019 novel coronavirus pneumonia; NLR, neutrophil-to-lymphocyte ratio. . CC-BY-NC-ND 4.0 International license It is made available under a author/funder, who has granted medRxiv a license to display the preprint in perpetuity. is the (which was not peer-reviewed) The copyright holder for this preprint . CC-BY-NC-ND 4.0 International license It is made available under a author/funder, who has granted medRxiv a license to display the preprint in perpetuity. is the (which was not peer-reviewed) The copyright holder for this preprint . https://doi.org/10.1101/2020.02. 19.20025296 doi: medRxiv preprint  The boxplots of the CT severity score between stable group and progressive group (a), between imported patients and nonimported patients (b), and between the patients who were admitted ≤ 4 days and > 4 days from symptom onset. Heatmap depicting the correlations between inflammatory markers and CT characteristics (within the red dotted box). It shows the correlation coefficients r with P < 0.05 of all pairs. It can be observed that C-reactive protein and CT severity score had the strongest correlation. CT, computed tomography. . CC-BY-NC-ND 4.0 International license It is made available under a author/funder, who has granted medRxiv a license to display the preprint in perpetuity. is the (which was not peer-reviewed) The copyright holder for this preprint . https://doi.org/10.1101/2020.02. 19.20025296 doi: medRxiv preprint 
-------------------------------------------------------


NEW PAPER


Equine arteritis virus (EAV) is the causative agent of equine viral arteritis (EVA), a respiratory, systemic, and reproductive disease of horses and other equid species. Following natural infection, 10-70% of the infected stallions can become persistently infected and continue to shed EAV in their semen for periods ranging from several months to life. Recently, we reported that some stallions possess a subpopulation(s) of CD3 + T lymphocytes that are susceptible to in vitro EAV infection and that this phenotypic trait is associated with longterm carrier status following exposure to the virus. In contrast, stallions not possessing the CD3 + T lymphocyte susceptible phenotype are at less risk of becoming long-term virus carriers. A genome wide association study (GWAS) using the Illumina Equine SNP50 chip revealed that the ability of EAV to infect CD3 +
Equine arteritis virus (EAV) is a single-stranded, positive-sense RNA virus that belongs to the family Arteriviridae in the order Nidovirales [1] [2] [3] . It is the causative agent of equine viral arteritis (EVA) a respiratory, systemic, and reproductive disease of horses [2, 4, 5] . While most naturally acquired EAV infections are clinically inapparent, relatively virulent field strains of EAV periodically emerge around the world giving rise to outbreaks of EVA [6, 7] . The disease is characterized by fever (greater than 41˚C); depression; leukopenia; rhinitis often accompanied by nasal discharge; urticaria; and edema [8] . Abortion is a frequent outcome in naïve pregnant mares and congenital infection in neonatal foals is characterized by severe, fulminating interstitial pneumonia [9] . In the stallion, EAV is shed in semen during the acute phase of the infection and in some individuals, for a short time during the convalescent period until they clear the virus entirely from all body tissues [10] . However, in contrast, EAV establishes long-term persistent infection in 10-70% of infected stallions and these constantly shed virus in their semen for extended periods (years or even life long) [8, 11, 12] . The mechanism of long-term persistence of EAV in the reproductive tract of stallions is not well understood. It has been shown that EAV persistence in sexually intact post-pubertal colts or stallions is testosterone dependent [13, 14] . Persistently infected stallions play an important role in maintenance and perpetuation of the virus in equine populations by transmitting the virus during breeding to naïve susceptible mares and can be responsible for outbreaks of EVA [8, [13] [14] [15] [16] [17] . The use of virus-infective frozen or chilled semen for artificial insemination and embryo transfer can increase the risk of spread of EAV [18] . In previous studies in our laboratory, it has been shown that the experimentally derived virulent Bucyrus strain (VBS) of EAV can infect CD3 + T lymphocytes in vitro from some but not all horses [19] . In one study, 310 horses of Thoroughbred, Standardbred, Saddlebred, and Quarter horse breeds were tested and their phenotypes identified with respect to in vitro infection of their CD3 + T lymphocytes by the VBS of EAV [20] . Those whose CD3 + T lymphocytes could be infected in vitro with this virus strain were identified as susceptible, and those whose cells were not infected were identified as resistant [20] . A genome wide association study (GWAS) identified a common haplotype associated with the in vitro CD3 + T lymphocyte susceptible phenotype in a region of equine chromosome 11 (ECA11:49572804-49643932) whose distribution was consistent with a dominant mode of inheritance [20] . Subsequently, it has also been demonstrated that stallions with the CD3 + T lymphocyte susceptibility phenotype to in vitro EAV infection are also at a significantly higher risk of becoming persistently infected carriers compared to those that lack this phenotype [21] . The primary objective of this study was to identify the specific gene(s) involved in the in vitro CD3 + T lymphocyte susceptibility to EAV infection and also the molecular mechanism(s) responsible for this phenomenon. We have combined contemporary genomics, molecular biology, and proteomics techniques to demonstrate that a specific gene in equine chromosome 11 (ECA11), equine CXCL16 (EqCXCL16), plays an essential role in the in vitro CD3 + T lymphocyte susceptibility to EAV infection. In this study, we have shown the EqCXCL16 gene has three alleles, two coding for the susceptibility phenotype (EqCXCL16Sa and EqCXCL16Sb) and one coding for the resistant phenotype (EqCXCL16R). Furthermore, compelling evidence is provided that allelic variation within EqCXCL16 is a major determining factor for establishment of long-term persistent EAV infection in the stallion reproductive tract. This study also identified key functional differences between the EqCXCL16S and EqCXCL16R proteins, highlighting the biological significance of mutations present in the EqCXCL16 gene. Non-synonymous nucleotide sequence variation within the 48M-51M region of ECA11 (SNP Discovery) Three horses were selected for whole genome sequence analysis to identify nucleotide substitutions within the 3 megabase target region of ECA11 (ECA11: 48M-51M) associated with in vitro susceptibility or resistance of CD3 + T lymphocytes to infection with EAV. Two horses (Thoroughbred-10 [TB10] and Standardbred-22 [STB22]) possessed CD3 + T lymphocytes with the EAV susceptible phenotype while the corresponding cells in Thoroughbred-3 (TB03) were resistant. Sequencing of 500 bp paired end libraries was accomplished using an Illumina HiSeq2000 platform to a depth of approximately 30× coverage. Following alignment to the reference genome (Ecab 2.0), the distribution of SNPs within ECA11: 48M-51M was compared to identify differences between susceptible and resistant horses. SNPs of interest were those which occurred in both susceptible horses but not the resistant horse. The analysis focused on identification of SNPs causing non-synonymous mutations within exons for genes annotated by Ensembl as these were considered to have a high likelihood of influencing the phenotype. In total, 12 non-synonymous SNPs were found in exons for eight annotated genes within the target region (Table 1) . No translational frame-shift insertions or deletions (indels) were found within any previously annotated exon. SNPs within the equine orthologue of CXCL16 (EqCXCL16) are associated with susceptibility of CD3 + T lymphocytes to EAV infection Potential associations between SNPs identified by genomic sequencing and the CD3 + T lymphocyte EAV susceptibility phenotype were determined by sequencing PCR amplified fragments spanning each of the 12 SNP locations from a subset of the 240 horses (resistant [n = 2] and susceptible [n = 8]) ( Table 2 ). Only the SNPs associated with the equine orthologue of the gene encoding the chemokine CXCL16 (EqCXCL16) showed a complete association; the two resistant horses were homozygous for the A, G, T, and G SNPs at (XM_00154756) c.715, c.801, c.804, and c.810, while all eight susceptible horses were either homozygous for the SNPs T, C, A, and A or were heterozygous A/T, G/C, T/A, and G/A at those positions, respectively , This observation is consistent with genetic dominance for the EAV CD3 + T lymphocyte susceptibility phenotype as reported previously [20] . To investigate the distribution of the four SNPs described above in a larger, more representative sample of the horse population, exon 1 of EqCXCL16 was PCR amplified and sequenced from 240 horses (Thoroughbred [n = 67], Standardbred [n = 60], Quarter Horse [n = 53], and Saddlebred horses [n = 60]; see S1 Table for primer sequences). The 240 horses had previously been characterized as EAV CD3 + T resistant (n = 85) or susceptible (n = 155). Based on the annotation for the equine reference CXCL16 (XM_001504756) sequence, all resistant horses (n = 85) were homozygous for the SNPs A, G, T, and G at positions 715, 801, 804, and 810. However, all susceptible horses were homozygous or heterozygous at those positions and possessed the SNPs T, C, A, A or T, C, G, A (n = 154) at the respective positions. A single (n = 1) homozygote was observed for a putative haplotype possessing the combination T, C, G, A; that individual was phenotyped as susceptible. These two susceptibility alleles (T, C, A, A and T, C, G, A) were designated EqCXCL16Sa and EqCXCL16Sb, respectively. The predicted amino acid sequences were identical for EqCXCL16Sa and EqCXCL16Sb as described below. The allele for resistance was recessive and was characterized by A, G, T, and G at positions 715, 801, 804, and 810. This allele was designated EqCXCL16R and was identical to the reference genome sequence (Ecab 2.0). No additional alleles were found in exon 1 of EqCXCL16 among 47 horses of 20 other breeds tested during the study. The gene frequencies of the three CXCL16 alleles are shown in Table 3 for Thoroughbred, Quarter Horse, Saddlebred, and Standardbred horse breeds. Predicted amino acid sequences encoded by EqCXCL16Sa, EqCXCL16Sb, and EqCXCL16R; three alleles and two proteins The originally identified SNPs were non-synonymous mutations, predicting a change of amino acids in the protein ( Table 1 ). The predicted amino acid sequences for this region of exon 1 specify phenylalanine (Phe), histidine (His), isoleucine (Ile) and lysine (Lys) at amino Association of the EqCXCL16 genotypes with the phenotypes for CD3 + T lymphocytes susceptibility The distribution of the two predicted EqCXCL16 proteins among the 240 Thoroughbred, Quarter Horses, Saddlebred, and Standardbred was compared to the distribution of their CD3 + T lymphocyte phenotypes for resistance or susceptibility to equine arteritis virus infection. As can be seen in Table 4 , the association was complete. This observation is particularly noteworthy since it occurred across a group including 4 different horse breeds. The 5 carriers with the resistance genotype came from 5 different breeds (Saddlebred, Friesian, Lusitano, Thoroughbred, and Quarter Horse) while the 11 non-carriers with susceptibility genotypes came from 8 different breeds (Belgian Warmblood, Belgian draft, Saddlebred, Standardbred (4), Paint, Friesian, Hanoverian, and Quarter Horse. Inter-species comparison of predicted amino acid sequences within exon 1 of CXCL16 The predicted amino acid sequences within exon 1 of CXCL16 was compared for seven species using the reference genome sequences from white rhinoceros (Ceratotherium simum), horse (Equus caballus), dog (Canis lupus familiaris or Canis familiaris), human (Homo sapiens), domestic rat (Rattus norvegicus), cattle (Bos taurus), and African elephant (Loxodonta africana) (Fig 1) . As noted above, the horse genome reference sequence (Ecab 2.0) is identical to EqCXCL16R (Fig 1) . Although the sequences depicted in Fig 1 have some shared elements such as a GN-GS motif (positions 2-6 Fig 1) or a C residue at position 11, this region is not highly conserved between species and different amino acid side chains are permitted at sites equivalent to those involving polymorphism within CXCL16 in the horse. However, position 40 of EqCXCL16R differs in that Tyr is hydrophilic; while in EqCXCL16S and all other species examined, this site is occupied by a non-polar residue. In addition, EqCXCL16S differs at position 49 in that His has a basic side chain; whereas in the case of EqCXCL16R, along with other Variation in CXCL16 Gene Determines the Outcome of EAV Carrier State species, this position is generally occupied by amino acids with acidic or polar side chains ( Fig  1) . In contrast, the identities of amino acids or amino acid side chain properties at positions 50/52 in EqCXCL16S and EqCXCL16R are observed in other species, with Phe in EqCXCL16R seen in humans along with another perissodactyl, the white rhinoceros, and Ile is present in cattle. Similarly, at position 52, Glu in EqCXCL16R is also present in the white rhinoceros; although Lys in EqCXCL16S is not found in any of the other species shown in Fig 1, positively charged Arg is observed in cattle and African elephants. These results suggest that exon 1 of CXCL16 has been exposed to considerable selective pressure over time and imply adaptive evolution for this region unique to each mammalian species. We previously reported an association between susceptibility of CD3 + T lymphocytes to infection and the ability of EAV virus to establish the long-term carrier state in stallions (7) . Following natural infection, some stallions continue to shed virus in the semen for long periods (more than one year to lifelong) after EAV is no longer present in blood and nasal secretions [2, 22] and as such become reservoirs for the maintenance of this virus in equid populations. Therefore, we analyzed semen from 77 stallions comprising 24 different breeds that had been infected with EAV and their status as long-term carrier animals determined (shedders; Table 5 ). Since the CXCL16 alleles showed complete association with the CD3 + T lymphocyte phenotype across several breeds, data for the carrier-status phenotype were pooled for 77 stallions from different breeds. Of these, 37 were identified as long-term carriers (stallions EAV seropositive and shedding detectable levels of virus in semen more than one year after initial infection) while 40 were identified as non-carriers (stallions seropositive for EAV, but had apparently cleared the virus from the reproductive tract [non-shedders]) based on absence of detectable virus in semen post-infection. Exon 1 of EqCXCL16 was PCR amplified and sequenced from all 77 stallions to determine each stallion's genotype. The results demonstrated a strong, although incomplete, association between long-term EAV carrier status in stallions and the presence of at least one copy of the EqCXCL16Sa or EqCXCL16Sb allele among 86% of the stallions with that phenotype (Table 6 ). In contrast, 73% of stallions that had been infected with EAV but were negative for virus in their semen following infection were homozygous for EqCXCL16R (Table 6 ). Effect of amino acid substitutions in the EqCXCL16 protein on response to EAV infection Recently, we demonstrated that the membrane-associated form of the chemokine encoded by the EqCXCL16Sa allele can function as a host-cell receptor for EAV binding and entry [23] . Since the four predicted amino acid substitutions discovered in exon 1 of EqCXCL16R compared to EqCXCL16Sa/b appear to be non-conservative, we hypothesized they might modify the functional EAV receptor properties of the resultant EqCXCL16R molecule. Consequently, as described previously for EqCXCL16S [23] , we established an HEK-293T cell line for stable expression of the transmembrane form of the EqCXCL16R protein. Stable expression of the respective proteins from HEK-EqCXCL16S and HEK-EqCXCL16R cells were compared using confocal microscopy (Fig 2A) and Western blot (WB) analysis ( Fig 2B) using guinea pig (Gp) α-EqCXCL16 antisera. The results confirmed that the variant forms of EqCXCL16 were produced in apparently similar amounts and that they were both clearly associated with the plasma membrane (Fig 2A) . Infection of HEK-EqCXCL16S with the EAV sVBSmCherry encoding the mCherry fluorescent protein confirmed our previous reported findings that by 12 hpi the fluorophore was expressed in almost every cell ( Fig 2C, panel b) . In contrast, expression of mCherry in HEK-EqCXCL16R cells infected at the same MOI with EAV sVBSmCherry was almost undetectable at the same time point and as such was equivalent to that observed in mock transfected HEK-293T cells (Fig 2C, panels a and c) . This result was confirmed in experiments where EAV gene expression following infection was detected by an indirect immunofluorescence assay (IFA) using a monoclonal antibody against viral nonstructural protein 1 (nsp-1). Although nsp-1 expression was detectable in a few EAV infected HEK-EqCXCL16R and mock transfected HEK-293T cells (Fig 2D, panels b and c) , the numbers were markedly lower than observed in similarly infected HEK-EqCXCL16S cells ( Fig 2D, panel a) . These results demonstrate that the plasma membrane-associated EqCXCL16S isoform can function as an efficient cellular receptor for EAV while EqCXCL16R does not. By employing a combination of the virus overlay protein binding assay (VOPBA) and Far-Western blot (Far-WB) techniques, we previously demonstrated that EAV VBS could directly bind with EqCXCL16S [23] . In view of the results outlined above, the experiment was repeated with the stable cell lines expressing both isoforms of EqCXCL16 (HEK-EqCXCL16S and HEK-EqCXCL16R cell lines). Equal amounts of total lysate from HEK-EqCXCL16S, HEK-EqCXCL16R, and mock transfected HEK-293T cells were subjected to polyacrylamide gel electrophoresis and transferred onto a polyvinylidene difluoride (PVDF) membrane prior to sequential denaturation and renaturation by treatment with guanidine-HCl (Gn-HCl) AC buffer. Refolded membrane-bound prey proteins were incubated with purified EAV VBS (bait protein) which was detected using a mouse monoclonal antibody directed against the GP5 envelope protein of EAV (mouse α-GP5 Ab). Two strong signals were detected with apparent molecular weights (MW) of approximately 52 and 30 kDa in HEK-EqCXCL16S, but not HEK-EqCXCL16R or mock transfected HEK-293T (naïve HEK) cell lysates ( Fig 3A, panel a) . No bands were visible in a control experiment where the EAV VBS binding step was omitted ( Fig 3A, panel b) , demonstrating the absence of non-specific interactions between mouse α-GP5 Ab and cellular lysate proteins. The monomeric form of CXCL16S has an apparent molecular weight of 30 kDa, and this was confirmed by stripping the membrane depicted in panel a of Fig 3 and re-probing with polyclonal Gp α-EqCXCL16 antisera (pAb). In this experiment, co-migrating bands with an apparent MW of 30 kDa were visible in both HEK-EqCXCL16S and HEK-CXCL16R lysates (Fig 3, panel c) . The detection of a strong signal at 52 kDa is unlikely to be the result of interactions between EAV VBS and another cellular protein because this was only visible in HEK-EqCXCL16S lysates. Therefore, while initial polyacrylamide gel separation of the cell lysates was conducted under denaturating conditions, it is conceivable that EqCXCL16S may associate strongly with other proteins. Some support for this hypothesis is provided by the fact Gp α-EqCXCL16 antisera reacts with material with apparent MWs between 52 and 30 kDa in HEK-EqCXCL16S and HEK-EqCXCL16R, but not in mock transfected HEK-293T cell lysates (Fig 3, panel c) . To further confirm these observations, biotinylated EAV VBS was incubated with HEK-EqCXCL16S, HEK-EqCXCL16R, and naïve control HEK-293T cells at 4˚C to prevent internalization. Subsequently, the bound virus particles were visualized using streptavidin Binding of EAV to EqCXCL16 is determined by amino acid substitutions within the N-terminal ectodomain of the protein encoded by exon 1. A). Effect of amino acid differences between EqCXCL16S and EqCXCL16R on EAV binding as determined by a combination of VOPBA and Far-WB analysis. Naïve HEK-293T, stable HEK-EqCXCL16S and HEK-EqCXCL16R cells were lysed in RIPA buffer, the lysate proteins separated using 12% SDS-PAGE and transferred onto a PVDF membrane. Membrane bound proteins were denatured and renatured using sequentially decreasing concentrations of Gn-HCl following which the membranes were blocked and incubated either with purified EAV VBS (15 μg/ml) in protein binding buffer (a) or with protein binding buffer without purified EAV VBS (b). After washing, membranes were incubated with α-GP5 MAb 6D10 and developed using the enhanced chemiluminescence (ECL) method. Binding of EAV VBS to EqCXCL16 protein is shown in (a [arrow]). The same membrane (a) was stripped and re-probed with Gp anti-EqCXCL16 as shown in panel c. As indicated by the arrow in panel c, EqCXCL16S and EqCXCL16R were detected at the same position on the membrane where EAV GP5 was detected in panel a. B). Amino acid substitutions between the "S" and "R" EqCXCL16 isoforms and attachment of FITC. Fluorescence microscopy revealed that significantly higher number of HEK-EqCXCL16S cells bound to biotinylated EAV VBS compared to the HEK-EqCXCL16R or control HEK-293T cells (Fig 3B) . Taken together, these results ( Fig 3A and Fig 3B) confirm that EAV VBS is capable of binding to EqCXCL16S, but that this association is almost completely abrogated by the four amino acid substitutions encoded by exon 1 of EqCXCL16R. The CXCL16 protein was originally identified as a scavenger receptor for oxidized low-density lipoprotein (OxLDL) in humans [25, 26] . Therefore, it was decided to determine if scavenger receptor activity differed between EqCXCL16 isoforms. The HEK-EqCXCL16S, HEK-EqCXCL16R, and naïve HEK-293T control cells were incubated with Dil labelled OxLDL (Dil-OxLDL) and analyzed by fluorescence microscopy. Interestingly, the highest levels of Dil-OxLDL binding and potential internalization were observed in the HEK-EqCXCL16S while amounts bound to HEK-EqCXCL16R were very low, similar to those observed in mock transfected naïve HEK-293T cells (Fig 4, panel b vs panels a and c). Furthermore, Dil-OxLDL binding to HEK-EqCXCL16S cells was significantly reduced by prior treatment with a polyclonal antibody against EqCXCL16 (Gp α-EqCXCL16 pAb; Fig 4, panel d) , suggesting a high degree of specificity in reactivity between Dil-OxLDL and the "S" form of the EqCXCL16 chemokine. These data strongly suggest that the EqCXCL16S isoform retains the function of an efficient scavenger receptor for Dil-OxLDL, whereas this functional property appears to be almost completely absent in the EqCXCL16R isoform. This confirms that four amino acid substitutions in the ectodomain of the EqCXCL16S isoform play a critical role in scavenger properties which may also favor the infection of cells with EAV as compared to the EqCXCL16R isoform. Both variants of EqCXCL16 bind to the equine CXC-chemokine receptor 6 (EqCXCR6) protein in vitro Although the experiments described above were not quantitative, they did reveal obvious differences between the two EqCXCL16 isoforms in terms of OxLDL scavenger receptor activity. Therefore, Far-WB analysis was employed to determine if similar differences existed between EqCXCL16S and EqCXCL16R in binding to the CXCR6 receptor molecule. Purified recombinant HA-tagged EqCXCR6 protein was separated in SDS-PAGE and incubated with purified recombinant EqCXCL16S or EqCXCL16R protein and binding detected using EqCXCL16 specific antibody produced in rabbits (Rb α-EqCXCL16 Ab). Very similar signal levels were observed with both EqCXCL16 isoforms (Fig 5, panels a and b) at a position approximating a MW of 25 kDa which is the predicted size for cloned EqCXCR6. This was confirmed by stripping the membrane depicted in Fig 5 (panel a) before re-probing with anti-His antibody, thus demonstrating that the location of the His-tagged EqCXCR6 protein was identical to that observed for the interaction between EqCXCL16 and EqCXCR6 (Fig 5, panel c) . There was no non-specific reactivity between the EqCXCL16 isoforms and the anti-His antibody with BSA. EAV to host cells. Equal numbers (2 x 10 6 ) of a) naïve HEK-293T cells and stable b) HEK-EqCXCL16S and c) HEK-EqCXCL16R cells were washed, resuspended in cold PBS (pH 7.4) with 2% FBS (PBS-F) and then incubated with biotinylated EAV VBS on ice for 2 h in the dark. After adsorption, excess EAV was removed by washing in cold PBS-F, and cells were then stained with Streptavidin-FITC and DAPI solution. Significantly higher number of HEK-EqCXCL16S cells were observed to bind biotinylated EAV (panel a) compared to the HEK-EqCXCL16R (panel b) or naïve control HEK-293T cells (panel c). Statistically different results are represented with different letters, a and b. Panel d shows the graphical representation of the images from panels a, b, and c. All the images depicted were representative of three independent experiments. Data were analyzed by ANOVA; P<0.001 was considered as significant. doi:10.1371/journal.pgen.1006467.g003 Variation in CXCL16 Gene Determines the Outcome of EAV Carrier State Chemoattractant properties in vitro for CD3 + T lymphocytes do not differ between the two EqCXCL16 isoforms The fact that human CXCL16 can act as a chemoattractant and recruit lymphoid cell types to sites of inflammation within the body suggests that this property may also be present in EqCXCL16 [25, 27] . Therefore, experiments to determine if EqCXCL16 possessed chemoattractant potential and if this differed between the two isoforms were conducted using purified recombinant soluble forms of EqCXCL16S and EqCXCL16R proteins produced in E.coli. Equine CD3 + T lymphocytes derived from peripheral blood mononuclear cells (PBMCs) were Effect of amino acid substitutions between "S" and "R" isoforms of EqCXCL16 on binding to the EqCXCR6 receptor protein in vitro. Interactions between purified recombinant EqCXCL16S/R-EqCXCR6 were examined using Far-WB. Equal amounts (20 μg) of His-tagged EqCXCR6 protein or BSA as a control were separated in different lanes on 10% SDS-PAGE and transferred onto a PVDF membrane. Proteins were then sequentially denatured and renatured by using different concentrations of Gn-HCl. After blocking, the membranes were incubated with soluble EqCXCL16S (panel a) or EqCXCL16R (panel b) protein (5 μg/ml) followed by Rb α-EqCXCL16 Ab. After washing, membranes were developed using the ECL method. Binding of EqCXCL16S and EqCXCL16R to EqCXCR6 is indicated by arrows (panels a and b). These interactions occurred at the same location as that occupied by EqCXCR6; this was confirmed by stripping the membrane shown in panel b and re-probing it with anti-His antibody as shown in panel c. doi:10.1371/journal.pgen.1006467.g005 enriched using anti-CD3 conjugated magnetic beads to > 95% purity based on the proportion of CD4 + and CD8 + T lymphocytes (S1A Fig top row) present in the population with the remainder comprised mainly of CD21 + B cells and CD14 + monocytes (S1A Fig bottom row) . Approximately 6% of this enriched CD3 + T lymphocyte population had detectable cell-surface expression of EqCXCR6 as determined by staining with rabbit anti-EqCXCR6 antibody (PA7511) followed by analysis using flow cytometry (S1B Fig). Incubation of enriched equine CD3 + T lymphocyte preparations at 37˚C with the recombinant chemokines separated using a Boyden chamber with a 3 μm pore size provided clearest evidence that EqCXCL16 possessed chemoattractant potential compared to the medium only control (P<0.001) (Fig 6) . Furthermore, there were no statistically significant differences between the S and R isoforms in numbers of EqCXCR6 expressing CD3 + T lymphocytes entering the chamber (P = 0.210) (Fig 6) . Evidence of the chemoattractant properties can be attributed directly to EqCXCL16 and not to potential contaminants in the recombinant protein preparations was substantiated by the fact CD3 + T lymphocyte migration into the chamber was not statistically different from the Effect of amino acid substitutions between "S" and "R" isoforms of EqCXCL16 on chemoattractant properties of EqCXCL16S and EqCXCL16R. Purified recombinant soluble forms of EqCXCL16S and EqCXCL16R proteins (2 μg/ml) in RPMI medium containing 0.5% BSA were added to the lower compartment of a Boyden chemotaxis chamber separated from the upper by a polycarbonate filter having a 3 μm pore size; CD3 + T lymphocytes (5 x10 5 ) labelled with Calcein-AM were added to the upper chamber and incubated for 6 h. The cells that passed through the filter were counted with a fluorescent microscope and represented in a bar diagram. Controls consisted of RPMI with 0.5% BSA and EqCXCL16 (S and R) pre-treated by incubation with Gp α-EqCXCL16 pAb. No statistically significant differences were found in chemoattractant potential between recombinant EqCXCL16S and EqCXCL16R proteins whereas CD3 + T lymphocyte migration in response to pretreatment of these molecules with Gp α-EqCXCL16 pAb was similar to that observed in the medium only control. Experiments were repeated independently three times. The bar diagram represents mean ± SD, P<0.00105 values were considered as significant by ANOVA. Both a and b are significantly different. medium only control where the equine chemokine isoforms were pre-incubated with Gp α-EqCXCL16 pAb (Fig 6) . Membrane-associated forms of human CXCL16 have been reported to function as cellular adhesion molecules [25] . Preliminary observations suggested differences between HEK cells expressing the two EqCXCL16 isoforms and adhesion to cell culture plates. Therefore, HEK-EqCXCL16S, HEK-EqCXCL16R, and naïve control HEK-293T cells were tested using a previously described protocol [53] to evaluate resistance to detachment from the cell culture plate following incubation for 10 min at 37˚C with 0.5M EDTA. Surprisingly, HEK cells expressing EqCXCL16S were found to be significantly (P<0.0001) more resistant to treatment with EDTA than those expressing EqCXCL16R or naïve control HEK-293T cells (Fig 7A and  Fig 7B) . Furthermore, the superior adhesion properties of HEK-EqCXCL16S were abrogated by pre-treatment prior to plating with Gp anti-EqCXCL16 polyclonal antibody providing strong evidence of resistance to EDTA-induced detachment was mediated only by the "S" isoform of the chemokine (Fig 7A and Fig 7B) . Previously, we have shown that horses can be categorized into two distinct phenotypic groups based on the susceptibility of their CD3 + T lymphocytes to in vitro infection with EAV [19, 20] . It was also discovered that stallions with EAV-susceptible CD3 + T lymphocytes were at higher risk of becoming long-term carriers following exposure to this virus than those with a resistant CD3 + T-cell phenotype [21] . Furthermore, a genome wide association study (GWAS) based on single nucleotide polymorphism (SNP) detection using the Equine SNP50 BeadChip suggested an association between these phenotypic traits and a dominant genetic marker(s) located on equine chromosome 11 (ECA11) between nucleotide positions 49,572,804 and 49,643,932 [20] . In this study, we present genetic and functional data confirming that susceptibility or resistance of CD3 + T lymphocytes to in vitro EAV infection is indeed under genetic control and can be explained based on mutations found in the chromosome region predicted by GWAS. Whole genome sequencing of one resistant and two susceptible horses revealed 12 non-synonymous nucleotide substitutions distributed among eight previously annotated genes in the targeted region on ECA11. CXCL16 was implicated as the putative cause of the phenotype based on tests of the SNPs among 10 horses (8 susceptible and 2 resistant) that excluded the seven other genes. Furthermore, genetic association studies based on typing 240 horses of different breeds demonstrated three alleles of CXCL16 encoding for two proteins, EqCXCL16R and EqCXCL16S. The gene for EqCXCL16S was dominant and correlated completely with the EAV-susceptible CD3 + T lymphocyte phenotype. Specifically, EAV susceptibility of CD3 + T lymphocyte population members was conferred by two distinct alleles (EqCXCL16Sa and EqCXCL16Sb), although the resultant proteins were predicted to possess identical amino acid sequences. These susceptibility alleles exhibited a dominant mode of inheritance, while that associated with CD3 + T lymphocyte EAV resistance (EqCXCL16R) was recessive, in that the phenotype was only found in equids homozygous for EqCXCL16R. At least three of the SNPs located within EqCXCL16Sa, EqCXCL16Sb, and EqCXCL16R are predicted to induce non-conservative or even radical amino acid substitutions within exon 1 of the resultant protein. If analogous to the human orthologue of CXCL16, exon 1 should encode the N-terminal ectodomain of the chemokine and, therefore, play a major role in its biological properties. Although there is to our knowledge no published information concerning the functional properties of EqCXCL16, the human equivalent of this protein (CXCL16) is expressed as a type-I membrane protein and is comprised of several distinct domains (chemokine, extracellular [mucin stalk], hydrophobic transmembrane, and intracellular cytoplasmic) [28] . The protein can also exist in an unbound soluble form resulting either from cleavage with cellular metalloproteinases such as ADAM10 [28] or alternative mRNAsplicing [29] . Membrane-associated forms of human CXCL16 (HuCXCL16) are expressed on dendritic cells, CD14 + monocytes/ macrophages, CD21 + B lymphocytes, endothelial cells, smooth muscle cells, and keratinocytes [30] [31] [32] [33] [34] [35] [36] [37] . Moreover, expression on these cell types is upregulated by inflammatory mediators and bacterial lipopolysaccharides [27, 29, 36, [38] [39] [40] . Both soluble and membrane-bound forms of HuCXCL16 specifically interact with its receptor CXCR6 (also known as STRL33/ BONZO/TYMSTR) expressed on the surface of CD4 + and CD8 + T lymphocytes, NKT cells, and NK cells [27, [41] [42] [43] . Binding to the CXCR6 receptor is facilitated by the mucin-stalk located within the extracellular domain of the molecule [44, 45] . Soluble HuCXCL16 has strong chemotactic potential in that it effectively recruits CXCR6 + T lymphocytes to sites of inflammation. The extracellular domain of HuCXCL16 also recognizes oxidized low-density lipoprotein (OxLDL) along with phosphatidylserine; therefore, the protein is multifunctional, acting as a scavenger receptor in addition to possessing chemokine activity [34] . Furthermore, aberrant expression of HuCXCL16 is implicated in the pathogenesis of certain viral infections, arthritis, atherosclerosis, and the metastasis of some cancers [46] [47] [48] [49] [50] . Preliminary in silico studies suggested that EqCXCL16 possesses a domain structure very similar to its human counterpart including the presence of six cysteine residues within the chemokine domain. However, a Tyr-X-Pro-Val motif in the C-terminal intracellular domain believed to act as a potential substrate for tyrosine kinase phosphorylation in human and mouse variants of the protein is replaced by Tyr-X-Pro-Val in the horse. Experiments described here demonstrate EqCXCL16 binds to the equine orthologue of CXCR6 (EqCXCR6). In addition, the soluble form of EqCXCL16 has chemotactic activity for CXCR6 expressing T lymphocytes. Although there are no detectable qualitative changes in chemotactic property between the two equine isoforms described here, the predicted amino acid substitutions between EqCXCL16R and EqCXCL16S consisting of Tyr to Phe at position (p) 40, Asp to His at p49, Phe to Ile at p50, and Glu to Lys at p52 (p.Tyr40Phe, p.Asp49His, p. Phe50Ile, and p.Glu52Lys) respectively, had considerable effects on the ability to bind OxLDL. Moreover, there were dramatic differences between HEK-293T cells expressing each of these proteins to adhere to plastic culture vessels in the presence of EDTA. Collectively, these results indicate that EqCXCL16S, in common with its human counterpart, has both chemotactic and scavenger receptor activity, whereas the latter property is likely to be substantially reduced in the EqCXCL16R isoform [51] . However, in terms of EAV, the most significant finding is that while the membrane-bound variant of EqCXCL16S can bind this virus and function as a cellular receptor [28] , these properties are completely abrogated in the EqCXCL16R isoform. This suggests amino acid residues located at positions 40 to 52 within the chemokine domain of EqCXCL16 have a direct role in viral attachment via interactions with EAV membrane surface glycoproteins. These studies confirm and extend our previous findings [28] that although some viruses such as HIV and severe acute respiratory syndrome (SARS) virus may specifically interact with CXCL16 to influence the course of viral pathogenesis [32, 33] , EAV is at present unique in its ability to utilize this chemokine as a receptor protein. However, the choice of a scavenger receptor as a portal for viral entry does have a precedent among the arteriviruses in that porcine reproductive and respiratory syndrome virus (PRRSV) uses CD163 [48] . It has been reported that CXCL16 in humans is expressed by a subpopulation(s) of T lymphocytes [49, 50] . Consequently, the simplest explanation for the two phenotypically distinct horse groups identified in our earlier studies [16] is that EAV "susceptible" animals (EqCXCL16Sa,b/EqCXCL16Sa,b or EqCXCL16Sa,b/EqCXCL16R) possess a subpopulation(s) of T lymphocytes with cell-surface expression of EqCXCL16S and are thus permissive for viral infection, while the equivalent CD3 + T lymphocyte population in those that are "resistant" express only EqCXCL16R (EqCXCL16R/EqCXCL16R). Further studies are needed to investigate whether EqCXCL16 is constitutively expressed on CD3 + T lymphocytes or if expression can be increased by mitogen, TNFα, and IFNγ stimulation and/or by EAV infection of equine PBMCs. Another caveat to this hypothesis is that EAV exhibits a broad host-cell tropism and can infect several common laboratory cell lines from different species that certainly do not express EqCXCL16. Previous studies from our laboratory and others have shown that EAV can utilize a number of different non-related, host-specified molecules as cellular entry receptors or accessory molecules [52] . Consequently, demonstrating that susceptibility of a CD3 + T lymphocyte subpopulation(s) to EAV infection results from the presence of EqCXCL16S acting directly as an entry receptor will necessitate not only showing the presence of this protein, but also that alternative receptors are either not expressed in these cells or, for some reason, not utilized. Furthermore, the exact mechanism(s) of CXCL16 in infection of equine CD3 + T lymphocytes is yet to be determined. One possibility is that CXCL16S (or CXCL16R) is constitutively expressed or can be induced to express on subpopulation of CD3 + T lymphocytes, but only a CD3 + T lymphocyte subpopulation expressing CXCL16S becomes infected with EAV (Fig 8A and 8B; Model 1) . Alternatively, viral entry is mediated by the interaction between soluble isoform of CXCL16S and its cellular receptor CXCR6 expressed on CD3 + T lymphocytes (Fig 8C and 8D; Model 2) . However, these two models are not mutually exclusive and, as such, both need to be tested to unequivocally confirm the mechanism of CD3 + T lymphocyte infection dynamics. Although differences between horses in EAV CD3 + T lymphocyte subpopulation susceptibility is an interesting observation and may correlate with the severity of acute clinical signs [53] , it is the association between the EqCXCL16 genotype and long-term carrier status in stallions that is of major importance because these animals are key to the survival of this virus in equid populations [10] . In contrast to the CD3 + T lymphocyte EAV-susceptible/resistant phenotype, the correlation between carrier status in stallions and EqCXCL16 allelic content was strong (P<0.000001 Fisher's Exact Test without regard to breed) but not absolute. This P-statistic has limited value in that the stallions in Table 6 came from diverse breeds (Table 5) , however the association of the genotype with the phenotype without regard to breed was remarkable. In an analysis of 77 EAV-infected stallions, only 14% of the shedders and 86% of non-shedders were homozygous for the resistance genotype. Conversely, 74% of the shedders and 26% of non-shedders had the allele for the EqCXCL16S protein. Although results presented here are consistent with existence of the long-term carrier status in the majority of stallions being dependent on the membrane-bound form of EqCXCL16S acting as a cellular receptor for EAV, the fact that the EqCXCL16 allelic association is not complete suggests additional genetic, immunological, and viral factors or even environmental factors also play a role in this determination. Furthermore, the EAV carrier state in stallions has a number of features that distinguish it from many other persistent viral infections of the male reproductive tract. These add additional layers of complexity and must be considered in any proposed mechanism. For example, recent virus isolation and immunohistochemistry studies have confirmed that EAV persists primarily in the ampulla along with other accessory sex glands rather than in immunologically privileged sites such as the Sertoli cells within the testis (Carossino et al. manuscript submitted) [10] . This is despite the fact that carrier stallions possess active immune responses against EAV, and the virus is not detectable in any organ or tissue except the reproductive tract in carrier animals. Indeed inflammatory infiltrates in close proximity to viral antigen-expressing cells are frequently observed in the stallion reproductive tract indicating that EAV persistence is a continuous, dynamic process that occurs in the presence of active local immune responses. Clearly these responses are not completely effective in clearance of the virus for reasons that are unknown at present. Potential explanations range from some form of localized immunosuppression based on the fact that androgens such as testosterone can down-regulate immune responses, to more specific activity such as that mediated by T regulatory lymphocytes or even to differences in functional properties between EqCXCL16S and EqCXCL16R. These are not mutually exclusive, and so the failure to eliminate EAV from the stallion's reproductive tract could result from a combination of mechanistic factors. In addition, the virus probably contributes to its own survival via antigenic drift as evidenced by the continual emergence of horses with the CD3 + T lymphocyte subpopulation susceptible to in vitro EAV infection; A) or TM-EqCXCL16R (i.e. horses that lack the CD3 + T lymphocyte subpopulation susceptible to in vitro EAV infection; B) on their cell surface and EAV could directly infect these CD3 + T lymphocytes expressing TM-CXCL16S but not those expressing TM-CXCL16R. Model 2: CD3 + lymphocytes do not express either isoform of TM-CXCL16 but EAV infects these cells via ligandreceptor (CXCL16-CXCR6) complexes occurring on the CD3 + T lymphocytes surface (indirect infection). In this model the CD14 + monocytes and macrophages from susceptible and resistant horses express TM-CXCL16S (C) or TM-CXCL16R (D). EAV binds with TM-CXCL16S but not with TM-CXCL16R and is then cleaved by metalloproteinases and the soluble CXCL16S-EAV complex (sCXCL16S-EAV) interacts with CXCR6 present on the CD3 + T lymphocytes and thus, infects the CD3 + T lymphocytes from susceptible horses (C). It is also possible that TM-CXCL16S after interaction with EAV is not cleaved but comes in contact with CXCR6 present on the CD3 + T lymphocytes cells and thus infects the CD3 + T lymphocytes. TM-CXCL16R will not bind with EAV and hence CD3 + T lymphocytes of resistant horses will not become infected (D). doi:10.1371/journal.pgen.1006467.g008 novel variants during the course of persistent infections [16, 54, 55] . However, it is also possible that variation exists between individual stallions in the efficacy of EAV-specific immune responses within the reproductive tract that can operate independently of the EqCXCL16 genotype. If so, this could explain why 25% of stallions that possess at least one EqCXCL16Sa/b allele cease shedding shortly after infection with EAV, while approximately one in six that are homozygous for EqCXCL16R become long-term (more than one year) carriers. Based on the EqCXCL16 gene polymorphisms and its association with long-term carrier status, it would be possible to develop an allelic discrimination real-time PCR assay to distinguish horses that are prone to become long-term versus short-term shedders. It is interesting that while polymorphisms of CXCL16 have been reported in, for example, the human CXCL16 gene [44, 51, [56] [57] [58] , these are located in regions other than exon 1. Furthermore, these polymorphisms have not been shown to completely abrogate the chemokine, scavenger receptor properties [43] . The horse may therefore be highly unusual in possessing allelic variants of this gene with SNPs situated in exon 1 that in the case of EqCXCL16R appear to disrupt the ability of the resultant protein to act as a scavenger receptor for OxLDL. In conclusion, these genomic studies unequivocally demonstrate that horse genomic sequences encoding the EqCXCL16 chemokine are associated with in vitro susceptibility of equine CD3 + T lymphocytes to EAV infection, as well as the establishment of long-term carrier state in stallions. Although the molecular mechanisms associated with these phenotypic traits have not been fully elucidated, there is compelling evidence the plasma membrane-associated variant of EqCXCL16S can function as a cellular entry receptor for EAV, whereas this property is absent in the EqCXCL16R isoform of the protein. It is interesting that the sequence associated with abrogation of the virus binding site costs the horse the scavenging capability of CXCL16. If we assume the EqCXCL16Sa/b form is ancestral, the virus will have made use of a functional part of EqCXCL16, potentially to deter adaptation; selective pressure may have been sufficient for evolution of a variant that does not allow virus binding at the cost of losing the scavenging capability for OxLDL. The origins and selection pressures for these variants of EqCXCL16 warrant further study. This study was performed in strict accordance with the recommendations in the Guide for the Care and Use of Laboratory Animals of the National Institutes of Health. The animal protocol involving horses was approved by the University of Kentucky Institutional Animal Care and Use Committee (IACUC; protocol number 2013-1098). The animal protocol involving rabbits and guinea pigs was approved by the Thermo Scientific, Rockford, IL IACUC (NIH OLAW assurance number: A3669-01, USDA research license registration number: 23-R-0089, and PHS assurance number: A3669-01). This study was performed according to these IACUCapproved protocols. Equine pulmonary artery endothelial cells (EECs) were maintained in Dulbecco's modified Eagle's medium (DMEM; Mediatech, Herndon, VA, USA) with sodium pyruvate, 10% fetal bovine serum (Hyclone Laboratories, Inc., Logan, UT, USA), 100 U/ml of penicillin/100 μg/ml of streptomycin (Gibco, Carlsbad, CA, USA), and 200 mM L-glutamine [59] [60] [61] . The high passaged rabbit kidney cells (HP-RK-13 [KY] P399-409, originally derived from ATCC CCL-37; American Type Culture Collection, Manassas, VA, USA) were propagated in Eagle's minimal essential medium with 10% ferritin-supplemented bovine calf serum (Hyclone Laboratories, Inc.) and 100 U/ml of penicillin/100 μg/ml of streptomycin (Gibco). Human embryonic kidney (HEK-293T) cells (ATCC CRL-3216) were propagated in DMEM with 10% ferritin-supplemented bovine calf serum (Hyclone Laboratories, Inc.),100 U/ml of penicillin, and 100μg/ ml streptomycin (Gibco). HEK-293T cells stably expressing EqCXCL16S and EqCXCL16R were maintained in DMEM with 10% ferritin-supplemented bovine calf serum (Hyclone Laboratories, Inc.) and puromycin (Clontech Laboratories Inc., Mountain View, CA, USA) at 3 μg/ml of medium. Isolation of peripheral blood mononuclear cells (PBMCs) and enrichment of equine CD3 + T lymphocytes Isolation of PBMCs from peripheral blood of horses (n = 9) was performed as described previously [19, 45] with some modifications. Briefly, blood (20 ml) was collected aseptically using Vacutainer tubes containing 0.1 ml of 15% EDTA solution (Covidien, Dublin, Ireland). PBMCs were isolated from the buffy-coat fraction by centrifugation through Ficoll-Paque Plus (Amersham Biosciences, Piscataway, NJ, USA) at 500 × g for 30 min at 25˚C. The PBMC layer was collected and washed twice with Hanks balanced salt solution (pH 7.4) (Life Technologies, Grand Island, NY, USA) by centrifugation at 300 × g for 10 min to eliminate the platelets. The cells were resuspended in RPMI-1640 medium with 2 mM GlutaMAX and 25 mM HEPES (Life Technologies) without FBS and counted using a Countess Automated Cell Counter (Life Technologies). CD3 + T lymphocytes were enriched indirectly using magnetic microbeads according to the manufacturer's protocol (Miltenyi Biotec, Bergisch Gladbach, Germany). Briefly, PBMCs were blocked with 5% normal mouse serum (Innovative Research, Novi, MI, USA) and then incubated with anti-equine CD3 (Clone UC F6G) conjugated with Alexa Fluor 647 for 30 min. After washing with MACS buffer, the cells were incubated with anti-Alexa Fluor 647 microbeads for 20 min. The cells were then washed with MACS buffer and applied to an LS column, washed three times, and the bound CD3 + T lymphocytes were eluted with MACS buffer in the absence of the magnet, washed, and counted using a Countess Automated Cell Counter. The purity of the CD3 + T lymphocytes was confirmed by flow cytometric analysis. Staining was performed following the protocol as described previously [20, 45] . These cells were used for studying the chemokine function of EqCXCL16S and EqCXCL16R. Two strains of EAV, the experimentally derived virulent Bucyrus strain (VBS) (ATCC VR-796) [4, 62] and the recombinant EAV VBS virus expressing mCherry (EAV sVBSmCherry) [24] were used. Both viruses were propagated in EECs to generate high titer working stocks as previously described [63, 64] . Briefly, EECs infected with each virus were frozen at -80˚C when 90-100% cytopathic effect (CPE) was observed. Cell lysates were clarified by centrifugation (500 × g) at 4˚C for 15 min, followed by ultracentrifugation (Beckman Coulter, Miami, FL, USA) at 121,600 × g through a 20% sucrose cushion in NET buffer (150 mM NaCl, 5 mM EDTA, and 50 mM Tris-HCl, pH 7.5) at 4˚C for 4 h to pellet the virus. Purified preparations of each strain of EAV were resuspended in phosphate buffered saline (PBS, pH 7.4) and frozen at -80˚C. Virus stocks were titrated by standard plaque assay in RK-13 cells and titers were expressed as PFU/ml [65] . A total of 240 horses from four different breeds, Thoroughbred (TB; n = 67), American Saddlebred (ASB; n = 60), Standardbred (STB; n = 60), and Quarter Horses (QH; n = 53) were randomly selected for sequencing. These horses were a random subset of horses randomly selected from farms in Central Kentucky and previously tested for CD3 + T lymphocyte phenotyping [15] . A panel of 77 archived semen samples from EAV carrier stallions (n = 37) and non-carrier (n = 40) that had been stored at -80˚C was tested. These semen samples were previously submitted to the EVA OIE Reference Laboratory at the Maxwell H. Gluck Equine Research Center, Lexington, KY for testing. To confirm carrier status, isolation of EAV from equine semen samples was attempted in a high passage (P399-409) rabbit kidney-13 (RK-13) (HP-RK-13 [KY]) cell line according to the OIE described protocol [66] . Serum neutralizing antibodies to EAV were demonstrated by microneutralization assay in both carrier and non-carrier stallions to confirm that they were seropositive for EAV [67] . Based on clinical histories, none of the stallions had been vaccinated against EAV, and seroconversion was the result of natural infection. EAV long-term shedders were defined as those horses that had detectable EAV in their semen for more than one year following infection where date of exposure was known. Nonshedders of EAV were those which likely shed EAV in their semen during the acute phase of infection but that had ceased shedding virus at the time of initial testing for presence of the carrier state. These horses came from a wide range of breeds including Warmbloods, Standardbreds, Thoroughbreds, Quarter Horses, Belgian draft, Andalusian, Friesian, Rocky Mountain, Selle Francais, Tennessee Walking Horse, Arabian, Lusitano and American Saddlebred horses ( Table 5 ). Comparisons of shedders to non-shedders for EqCXCL16 genotypes were made using Fisher's Exact Test. The susceptible or resistant phenotype of each animal was defined by dual-color flow cytometric analysis of in vitro EAV-infected CD3 + T lymphocytes as described previously [28] . Horses were classified as susceptible or resistant to in vitro EAV infection based on the in vitro susceptibility or resistance of their CD3 + T lymphocytes. Genomic DNA (gDNA) was obtained from PBMCs or semen of each animal by using the Puregene whole-blood extraction kit (Qiagen, Valencia, CA, USA) in accordance with the manufacturer's instructions as previously described [20] . DNA quality and concentration were assessed using Nanodrop (Thermo Scientific, Wilmington, DE, USA) at an absorbance ratio of optical density at 260 nm/280 nm (OD260/280). Three horses were selected from those SNP genotyped and phenotyped for the CD3 + T lymphocyte susceptibility (S) or resistance (R) phenotype based on dual-color flow cytometric analysis in the previous study [19] ; specifically, they were TB03 (R), TB10 (S), and ST22 (S). Genomic DNA was submitted to BGI Americas (Davis, CA, USA) for sequencing from each of these three horses. Approximately 20 μg of DNA was submitted for construction of short insert (500 bp) libraries for sequencing using the Illumina HiSeq2000. To achieve 20-35 GB of raw data per lane,100 bp paired end sequencing across 7 lanes was conducted. Approximately 30× coverage was obtained per sample; there were 874,258,138 clean reads of 95.86 Q20 with ST22; 838,009,546 clean reads of 95.61 Q20 with TB03; and 867,086,552 clean reads of 95.96 Q20 with TB10. Reads were mapped to the horse genome reference sequence (Ecab 2.0) [68] using CLC workbench 8.0.1(CLC Bio, Boston, MA, USA). The whole genome sequence data from the three horses have been submitted to the Sequence Read Archive and can be found under Bio-Sample/experiment accession numbers SAMN03838869/SRX1097022, SAMN03838867/ SRX1097495 and SAMN03838868/SRX1097492 for CXCL16 of TB10, TB3, and ST22, horses respectively. The variant discovery and genotyping were done with the Genome Analysis Toolkit UnifiedGenotyper using arguments -nt 4 -gt_mode DISCOVERY-validation_strictness LENIENT. [69] . Genome annotation was from Ensembl version. Equus caballus. EquCab 2.75. The primer sequences for the eight candidate genes found using the target region (ECA11: 48M-51M) are shown in S2 Table. A PCR was performed in a final volume of 18 μL and the reaction for amplification consisted of: 100 ng of genomic DNA, 0.15 mM of each primer, and 10 μL of AmpliTaq GoldFast PCR Master Mix (Applied Biosystems, Foster City, CA, USA). Amplifications consisted of the following steps: initial denaturation at 95˚C for 10 min; 30 cycles of denaturation at 95˚C for 30 sec, annealing of primers at 56˚C for 30 sec, extension at 72˚C for 30 sec; and a final extension at 72˚C for 10 min. Amplicons were shipped overnight for sequencing to Eurofins Genomics (Louisville, KY, USA). DNA sequences were analyzed using Chromas Lite (Technelysium Pty Ltd, South Brisbane, Australia). For annotation of bases for equine CXCL16, the sequence XM_001504756 was used. The NCBI database was used to obtain and compare reference genome predicted-protein sequences of CXCL16 for white rhinoceros (Ceratotherium simum; XM_004433427), horse (Equus caballus; XM_001504756), dog (Canis lupus familiaris; XM_844211), human (Homo sapiens; AY358909), domestic rat (Rattus norvegicus; DQ025528), cattle (Bos taurus; NM_001046095) and African elephant (Loxodonta Africana; XM_003416737). Cloning and expression of EqCXCL16S, EqCXCL16R, and EqCXCR6 in E. coli Recombinant plasmids for cloning EqCXCL16 and EqCXCR6 were designed by CLC Main Workbench 7 programs using horse genome sequences obtained as part of this study. Synthetic sequences encoding the predicted exposed part of EqCXCL16 (aa [45] and the entire EqCXCR6 were produced by IDT (Coralville, IA, USA). These sequences were designed to have flanking BamHI and XhoI sites. In order to produce recombinant plasmids encoding the full size of S and R versions of CXCL16, synthetic sequences flanked by PstI and XhoI were also produced. These sequences encoded the N terminus half (aa 1-143) of both versions of EqCXCL16 (S and R). Following digestion with BamHI and XhoI or PstI and XhoI (Thermo Scientific, Rockford, IL, USA), fragments of amplicon and synthetic DNA were separated in E-Gel EX 1% agarose (Life Technologies) and extracted from gel using Zymoclean Gel Recovery Kit (Zymo Research, Irvine, CA, USA). Purified fragments encoding aa 17-247 and aa 25-199 of EqCXCL16 and entire EqCXCR6 were ligated into pET15b (Novagen, Temecula, CA, USA) followed by transformation into E. coli NovaBlue (Novagen). Ligation and transformation were performed using Rapid DNA Ligation and TransformAid kits (Thermo Scientific), respectively. Recombinant plasmids p15-16A (aa 17-247), p15-16B (aa , and p15-R6 were isolated from ampicillin-resistant clones using ZR Plasmid Miniprep™ Kit (Zymo Research). Plasmids p15-16R and p15-16S encoding R and S versions of full EqCXCL16 were produced by substitution of the smaller PstI/XhoI fragment of p15-16A to synthetic sequences encoding an N terminal half of R and S versions of this protein, respectively. In order to express recombinant polypeptides, plasmids were transformed into E. coli BL21(DE3) expression vector (Novagen). Several ampicillin-resistant clones were grown overnight in 1 ml of MagicMedia (Life Technologies) supplemented with 50 μg/ml of ampicillin (Sigma-Aldrich). Production of polypeptides was confirmed by SDS-PAGE electrophoresis following mini-scale isolation of recombinant proteins using Talon Metal Affinity Resin (Clontech Laboratories Inc.). For large scale production of recombinant EqCXCL16 (S and R) and EqCXCR6, 500 ml cultures of BL21(DE3) strain of E. coli with respective plasmids were grown overnight at 37˚C in MagicMedia supplemented with 50 μg/ml of ampicillin. Following centrifugation at 6000 × g for 15 min, the cell pellet was resuspended in Buffer A (50 mM sodium phosphate, 6 M guanidine-HCl, and 300 mM NaCl; pH 7.0) and subjected to several short cycles of sonication to reduce viscosity. The lysate was centrifuged at 16,000 × g for 30 min at 4˚C to remove debris. The His-tagged recombinant EqCXCL16 and EqCXCR6 proteins were purified from the supernatant by affinity chromatography using Talon Superflow Metal Affinity Resin (Clontech Laboratories Inc.) in combination with an FPLC apparatus (Amersham Pharmacia Biotech Inc.). Columns were equilibrated and washed with Buffer A, while proteins were eluted using Buffer B (45 mM sodium phosphate, 5.4 mM Gn-HCl, 270 mM NaCl, and 150 mM imidazole; pH 7.0). Eluted recombinant protein was dialyzed against PBS using Slide-A-Lyzer Dialysis Cassettes (Thermo Scientific). The purity and integrity of recombinant EqCXCL16S, EqCXCL16R, and EqCXCR6 proteins were evaluated by subjecting them to electrophoresis on a 4-20% gradient gel SDS followed by staining with PageBlue Protein Staining Solution (Thermo Scientific) and WB analysis. Protein concentration was determined with BCA Protein Assay (Thermo Scientific) using BSA as the standard. Protein-specific rabbit antipeptide sera (Rb α-EqCXCL16 [rabbit Ab, PA7509]), and guinea pig polyclonal antibody (Gp α-EqCXCL16 pAb) to detect EqCXCL16 proteins (EqCXCL16S and EqCXCL16R) were generated by immunizing rabbits with two synthetic peptides and guinea pigs with recombinant EqCXCL16 expressed in E.coli as previously described [45] . For this study anti-EqCXCR6 peptide antibody was generated by immunizing two rabbits with the 14 amino acid peptide (amino acid residues 17-30: DSSQEHERFLQFKK). These antibodies were extensively characterized by ELISA, confocal microscopy, and WB analysis. The monoclonal antibody (MAb) to equine CD3 surface molecule, UC F6G, was kindly provided by Dr. Jeff Stott, University of California, Davis. The R-PE conjugated F(ab 0 ) 2 fragment of goat antimouse IgG1 (Southern Biotech, Birmingham, AL, USA) was used as the secondary antibody. Mouse EAV α-GP5 and mouse EAV α-nsp-1 monoclonal Abs (MAb 6D10 and MAb 12A4, respectively) have been described previously [70, 71] . Detection of EAV antigen in infected cells was conducted using Alexa Fluor 488-labeled MAb against nonstructural protein 1 (nsp1; MAb 12A4) [19, 70] . Goat α-rabbit IgG (H+L)-HRP and goat α-mouse IgG (H+L)-HRP were purchased from Cell Signaling Technology, Inc. (Danvers, MA, USA). Goat α-guinea pig IgG (H+L)-HRP, goat α-rabbit IgG (H+L) conjugated to Alexa Fluor 488, and goat α-guinea pig conjugated to Alexa Fluor 488 were purchased from Life Technologies. Streptavidin conjugated to FITC was purchased from Southern Biotech. Stable HEK-293T transfectants expressing EqCXCL16S (HEK-EqCXCL16S cells) and EqCXCL16R (HEK-EqCXCL16R cells) were generated as described earlier [45] . Briefly, for expression of the EqCXCL16R protein in eukaryotic cells, a codon-optimized full-length EqCXCL16R sequence (obtained from whole genome sequencing in this study) was commercially synthesized and cloned into the pJ609 plasmid, into which the puromycin resistance gene was incorporated, by DNA2.0 (Menlo Park, CA, USA). This molecular construct was identified as pJ609-EqCXCL16R and used to transform E. coli DH10B cells (Life Technologies). For the establishment of stable cells, the HEK-293T cells were seeded in 6-well plates (2 x 10 6 cells/well) and transfected with 3 μg of codon-optimized pJ609-EqCXCL16R plasmid DNA (DNA2.0) using lipofectamine 3000 (Life Technologies) following the manufacturer's instructions. At 24 h post transfection, the medium was replaced with fresh medium containing 4 μg/ml of puromycin (Clontech Laboratories Inc.) and cells were incubated at 37˚C in a 5% CO 2 incubator. This process was repeated every other day until only puromycin-resistant colonies remained. These puromycin resistant cells were cloned by limiting dilution in 96-well plates and screened by IFA, after which clones showing the highest level of EqCXCL16S and EqCXCL16R protein expression were frozen in commercial cell-freezing medium (Recovery Cell Culture Freezing Medium; Life Technologies) and stored in liquid nitrogen until needed. At every 5 th passage and up to the 50 th serial passage, cells were analyzed by IFA using Gp α-EqCXCL16 pAb to confirm the expression of EqCXCL16R. All the experiments in HEK-EqCXCL16S and HEK-EqCXCL16R cells were performed within passage levels 5 to 10. Naïve HEK-293T and stable HEK-EqCXCL16S and HEK-EqCXCL16R cells in 8-well Thermo Scientific Lab-TeK chamber slides were washed in cold phosphate buffered saline (PBS, pH 7.4) and fixed in 4% paraformaldehyde (PFA; Sigma-Aldrich, St Louis, MO, USA) for 30 min at room temperature (RT). Cells were then stained as described previously [72] . Following fixation, cells were washed 5 times in ice-cold 10 mM glycine (Sigma-Aldrich) in PBS, pH 7.4 (PBS-Glycine) and were then permeabilized with 0.2% saponin (Sigma-Aldrich) in PBS or left untreated with detergent where examination of surface staining was required. All cells were washed again in 10 mM PBS-glycine and blocked with 5% normal goat serum (MP Biomedicals, Santa Ana, CA, USA) for 30 min at RT prior to incubation with specific primary antibodies (1:100 dilution) for 1 h at 37˚C in a humidified chamber. After washing in 10mM PBSglycine, the cells were incubated with anti-mouse or anti-guinea pig IgG(H+L) secondary antibodies conjugated with Alexa Fluor 488 (AF488, 1:200 dilution) for 1 h at 37˚C in a humidified chamber maintained in total darkness. After washing, slides were mounted in Vectashield mounting medium containing 4 0 , 6-diamidino-2-phenylindole (DAPI; Vector Laboratories, Burlingame, CA, USA). The slides were observed either under a Leica TSP SP5 confocal microscope in an environmental chamber at the University of Kentucky imaging core facility or with an inverted fluorescence microscope (ECLIPSE Ti; Nikon, Melville, NY, USA). Cells were lysed in RIPA lysis buffer (Santa Cruz Biotechnology, Dallas, TX, USA) in Halt protease and phosphatase inhibitor cocktails (Thermo Scientific). The solubilized proteins were mixed with Pierce lane marker reducing 5X sample buffer containing 100 mM dithiothreitol (DTT; Thermo Scientific) and heated for 5 min at 95˚C. Samples were resolved in SDS-polyacrylamide gel (5% stacking and 12% resolving; Bio-Rad) at 200 V for 45 min and then transferred onto a PVDF membrane (Bio-Rad, Hercules, CA, USA) at 100 V for 1 h using the transblot transfer system (Bio-Rad) [73, 74] . The membrane was blocked with 5% non-fat milk powder (Bio-Rad) in TBS-T (10 mM Tris-HCl [pH 7.6], 150 mM NaCl, and 0.1% Tween 20) for 1 h at RT and incubated with primary antibodies (Abs): rabbit α-EqCXCL16 PA7509 (1:500), guinea pig α-EqCXCL16 (1:1000), mouse monoclonal α-EAV GP5 MAb 6D10 (1:2000) , and mouse monoclonal α-EAV nsp-1Mab (12A4). The Abs were diluted in TBS-T with 5% bovine serum albumin (Sigma-Aldrich) overnight at 4˚C. The following day, the membranes were washed with TBS-T and then incubated with anti-rabbit, anti-mouse, or anti-guinea pig IgG, as appropriate, and conjugated with horseradish peroxidase (HRP, 1:3000; Cell Signaling Technology, Inc.) for 1 h at RT. The membranes were washed again and antibody binding was visualized with an ECL-detection system using SuperSignal West Pico chemiluminescent substrate (Thermo Scientific). The cell adhesion assay was performed in accordance with a published protocol [75] . Approximately 1 x 10 5 HEK-EqCXCL16S, HEK-EqCXCL16R, and naïve HEK-293T cells were plated in a 96-well plate. After 24 h incubation, the cells were washed in 37˚C PBS. Cells were then incubated with 0.5M EDTA for 10 min at 37˚C following which they were washed again with PBS and fixed with 4% PFA for 15 min at RT. Cells were then washed in distilled water and stained with 0.1% crystal violet solution for 20 min at RT, visualized with an inverted light microscope and photographed. After washing with distilled water, the cells were air-dried, then incubated with 10% acetic acid for 20 min with shaking prior to the transfer of 50 μl of the lysate to a new 96-well ELISA plate for determination of OD 595nm values using a Synergy H1MD microplate reader (BioTek Instruments Inc., Winooski, VT, USA). Cell migration in response to EqCXCL16S and EqCXCL16R proteins was determined using a Chemotaxis assay kit (Cell Biolabs, Inc., San Diego, CA, USA). Briefly, FPLC purified soluble recombinant EqCXCL16S and EqCXCL16R proteins were diluted in serum-free RPMI medium with 0.5% BSA (cell culture grade) at a concentration of 2 μg/ml in 500 μl of RPMI and were added to the lower well of a 24-well chemotaxis chamber; the lower and the upper wells were separated by a polyvinylpyrrolidone-free polycarbonate filter insert with the pore size of 3 μm. To the upper wells of the chamber, 100 μl of purified equine CD3 + T lymphocytes (5 x 10 5 cells/ml) labelled with Calcein-AM (Life Technologies) was added. After 6 h of incubation at 37˚C in 5% CO 2 , media from the inside of the insert was removed and insert was placed in a clean well with 400 μl of Cell Detachment solution and incubated at 37˚C for 30 min. After complete dislodging of the cells from the underside of the insert, 400 μl of the cells containing Cell Detachment solution was mixed with 400 μl of the migratory cells from the original well. Then 50 μl of the mixed cell solution was added onto a glass slide, and cells that passed through the filter were counted under a fluorescent microscope (40× objective) and represented in a bar diagram (averages of six different fields were included). Stable HEK-293T cells expressing EqCXCL16S and EqCXCL16R and naïve HEK-293T cells were seeded in 12-well plates. At about 85% confluency, cells were washed with warm (37˚C) PBS (pH 7.4) and replenished with warm (37˚C) growth medium containing Dil-OxLDL at 10 μg/ml (Kalen Biomedical LLC, Montgomery Village, MD, USA) and incubated at 37˚C for 3 h. Cells were washed with ice-cold PBS and examined under an inverted epifluorescence microscope for the uptake of Dil-OxLDL by cells. Approximately 100 μg of total protein lysate from naïve HEK-293T cells or stable HEK-EqCXCL16 cells were separated in 12% SDS-PAGE and transferred onto a PVDF membrane for Far-WB analysis following a modification of the published protocol [76] . The bound proteins were then denatured and gradually renatured on the membrane by sequential incubation with 6 M, 3 M, 1 M, and 0.1 M Gn-HCl in freshly prepared AC buffer (100 mM NaCl, 20 mM TRIS [pH 7.5], 10% glycerol, 0.5 mM EDTA, 0.1% Tween-20, 2% non-fat dry milk, and 5 mM DTT) for 30 min at RT or with only AC buffer in the absence of Gn-HCl overnight at 4˚C. The membrane was blocked with 5% non-fat dry milk in TBS-T (0.1% Tween-20) and overlaid with purified EAV VBS (15 μg/ml) and incubated overnight at 4˚C. The next day, the membrane was washed vigorously (3 washes each of 10 min) and incubated with mouse monoclonal Ab (α-GP5; MAb 6D10) directed against EAV GP5 envelope glycoprotein. Monoclonal antibody binding was detected by the ECL-detection system using SuperSignal West Pico chemiluminescent substrate (Thermo Scientific). To study EqCXCL16-EqCXCR6 interaction in a separate Far-Western blot experiment, 20 μg of purified recombinant hemagglutinin (HA)-tagged EqCXCR6 was separated on 12% SDS-PAGE and transferred onto a PVDF membrane which was incubated with recombinant purified EqCXCL16S or EqCXCL16R proteins (5 μg/ml). The membrane was then developed using Rb-anti EqCXCL16 antibody. Labeling of EAV with biotin and EAV binding assay EAV VBS was purified by ultracentrifugation (121,600 × g for 4 h) through a 20% sucrose cushion and protein concentration determined using the BCA protein assay kit. About 2 mg of purified EAV was biotinylated using EZ-Link Sulfo-NHS-Biotin (Thermo Scientific) following the manufacturer's protocol. Excess unbound biotin was removed by filtering through a Zeba desalt spin column (MWCO 7000; Thermo Scientific) equilibrated in PBS (pH 7.4). The naïve HEK-293T and HEK-EqCXCL16 cells were washed in cold PBS (pH 7.4) and removed from the culture dish using a non-enzymatic cell dissociation solution (Cellstripper; Mediatech Inc.). Cells were resuspended in cold PBS (pH 7.4) containing 2% FBS (PBS-F), centrifuged at 1000 × g for 5 min at 4˚C and incubated with biotinylated EAV at an MOI of 100 on ice for 2 h in total darkness. Excess EAV was removed by washing three times in cold PBS-F. Subsequently, the cells were stained with Streptavidin-FITC (1:100) and incubated at 4˚C for 30 min in total darkness. Cells were washed in PBS-F at 1000 × g for 5 min at 4˚C, transferred onto glass microscope slides using a Shandon CytoSpin III Cytocentrifuge with Shandon single cytofunnel with white filter cards (Thermo Scientific) and incubated with DAPI solution to permit visualization of cell nuclei. Cells were then analyzed using a Nikon inverted fluorescence microscope and the percentage of cells bound to EAV was calculated. Statistical tests for association of CXCL16 genotypes with phenotypes, including susceptibility and resistance as well as carrier status, were conducted using the Fisher's Exact Test. Differences among multiple treatment groups were analyzed by statistical analysis software Sigmaplot 12.3 (SystatSoftware Inc., San Jose, CA, USA), by ANOVA with pairwise multiple comparison procedures by the Holm-Sidak method. P-values less than 0.05 were considered to be statistically significant. S1 Fig. A) Enrichment of CD3 + T lymphocytes from PBMCs. T lymphocytes were enriched from PBMCs using anti-equine CD3 (Clone UC F6G) and micro magnetic beads conjugated to anti-mouse IgG1. The enriched cells were found to be > 95% T lymphocytes based on the proportion of CD4 + and CD8 + cells (top row) present in the enriched fraction. Furthermore, both CD21 + B lymphocytes and CD14 + monocytes (bottom row) were found to be a very minor component of the enriched T lymphocyte fraction. B) Expression of CXCR6 on equine CD3 + T lymphocytes. Magnetic bead purified CD3 + T lymphocytes were stained with α-EqCXCR6 antibody. Percentage of CD3 + T lymphocytes cells stained by pre-bleed rabbit sera as a negative control (shown as a). Percentage of CD3 + T lymphocytes expressing CXCR6 (shown as b). (TIF) S1 
-------------------------------------------------------


NEW PAPER


Background: The outbreak of novel coronavirus pneumonia in China began in December 2019. Studies on novel coronavirus disease (COVID-19) were less based on pediatric patients. This study aimed to reveal the clinical characteristics of COVID-19 in children. Method: This study retrospectively analyzed the clinical symptoms, laboratory results, chest CT, and treatment of children with laboratory-confirmed COVID-19(ie, with samples that were positive for 2019 novel coronavirus[2019-nCoV]) who were admitted to Result: Nine patients had no obvious clinical symptom. 11 patients developed fever. Other symptoms, including cough(in eleven of seventeen patients), rhinorrhea(in two), diarrhea(in two), vomiting(in two), were also observed. A small minority of patients had lymphocytopenia. Alanine transaminase or transaminase increased in three cases. According to chest CT scan, 11 patients showed unilateral pneumonia, 8 patients had no pulmonary infiltration. No serious complications such as acute respiratory syndrome and acute lung injury occurred in all patients. Conclusion: The clinical characteristics of 2019-nCoV infection in children were different from adult. The overall condition of children were mild and have a good prognosis.
Pneumonia caused by the novel coronavirus was discovered in Wuhan, Hubei, China and then the epidemic broke out. It was defined by the World Health Organization as "an international public health emergency of public concern" [1, 2] . Recently, many researches in the lancet reported the epidemiological, clinical symptoms, laboratory data, imaging features, treatment methods and clinical outcomes of non-pregnant adults and pregnant women infected with 2019-nCOV [3, 4] . However, the clinical characteristics of children have not been well disclosed, and whether there are differences in treatment methods has not been determined. We retrospectively analyzed the clinical data of 28 children with COVID-19 tested positive for 2019-nCOV by use of RT-PCR on samples. We reviewed 26 cases of children(>1 year old and <14 years old) in the Third People's Hospital of Shenzhen from January 16 to February 8, 2020 . Diagnosis of Program(6th edition) issued by the National Health and Health Committee of China [5] . The study was approved by the instututional of research ehtics committee of the Third People's Hospital of Shenzhen ([2020-063]). We collected the clinical manifestations, laboratory results, chest CT, treatment methods, and outcomes of 26 children, and the data was updated until February 19. To ensure the accuracy of the data, the data were independently checked by two clinicians. Each . CC-BY-NC-ND 4.0 International license It is made available under a is the author/funder, who has granted medRxiv a license to display the preprint in perpetuity.  Investigations into the sources of exposure indicate a history of contact with person from Hubei or family clusters in each case. There were 9 males (35%) and 17 females (65%), with an average age of 6.9 (0.7) years, ranging from 1 to 13 years (table 1). None of the above children had underlying diseases such as congenital diseases, dysplasia, diabetes, etc. Because some children cannot expresss independently, the clinical symptoms were mainly objective symptoms. Most patients had fever and cough at admission. About one-third of the patients had no clear symptoms, and other symptoms included vomiting, diarrhea, and rhinorrhoea (table 1). According to the standard,8 cases(31%) were mild and 18(69%) were ordinary. No severe complications such as acute respiratory distress syndrome (ARDS) and acute lung injury (ALI) occurred in the cases. Laboratory results revealed that most of the blood routine showed normal or decreased white blood cell counts, lymphocytes generally increased, and hemoglobin . CC-BY-NC-ND 4.0 International license It is made available under a is the author/funder, who has granted medRxiv a license to display the preprint in perpetuity. (which was not peer-reviewed) The copyright holder for this preprint . https://doi.org/10.1101/2020.03.08.20029710 doi: medRxiv preprint 6 and platelets were mostly unaffected (table 2) . There were no significant abnormalities in coagulation function. In terms of biochemistry, Alanine transaminase (ALT) combine with glutamic oxaloacetic transaminase (AST) increased in two cases(8%),ALT or AST increased in one case(4%).In addition, all the above cases had not exceed 50% of the normal high value. Most of the myocardial enzymes(myoglobin, troponin, kinase, creatine kinase) were not abnormal, with only a significant increase in LDH(46%). In the field of inflammatory mediators (table 2) , procalcitonin was normal, and 5(19%) patients had elevated C-reative protein. 19 patients were tested for interleukin-6 and only one had an increase. T-lymphocyte subset analysis showed no abnormalities in CD3+CD4+/CD3+CD8+.Humoral immunity showed that IgA, IgG, and C3c were partially reduced (table 2) , and IgM and C4 were normal. According to chest X-ray and chest CT examination, 8 patients(31%) had no pulmonary infiltration. There were 11 cases(42%) of lateral pulmonary infiltration and 7 cases(27%) of bilateral pulmonary infiltration. All patients were treated in isolation. Medicines for treating COVID-19 include oseltamivir, ribavirin, interferon, kaletra and traditional Chinese medicine.4 patients (15%) were treated with interferon alone and 1 patient (4%) was treated with kaletra alone.26 patients discharge (13.6±1.03 days on average) or improvement after treatment. There was no significant increase in inflammatory indexes and mild clinical symptoms in 26 patients, all of them were mild and common type cases, no severe pneumonia cases and death cases, and the outcome was similar to that of SARS. Among them, 11 patients (about 42%) had fever, which was inconsistent with 98% of the . CC-BY-NC-ND 4.0 International license It is made available under a is the author/funder, who has granted medRxiv a license to display the preprint in perpetuity. The copyright holder for this preprint . https://doi.org/10.1101/2020.03.08.20029710 doi: medRxiv preprint clinical symptoms of adult patients reported in [2, 3] . Other symptoms such as nasal obstruction, runny nose, vomiting and diarrhea [1, 3, 6] were less common than adults. All of the above suggest that the clinical symptoms of children infected with 2019-CoV were not typical. It is considered that most of the children might be infected after exposure to patients with COVID-19, which belongs to the second or third generation of infected cases, and the pathogenicity of the virus is decreased. In addition, the changes of leukocytes in children were the same as in adults, and most of them appear normal or decreased. Lymphocytes in adult patients had decreased significantly, and this study found that most of the children's lymphocytes increased beyond the normal range, which might be related to the higher percentage of The literature reported that the prognosis of elderly patients and those with chronic . CC-BY-NC-ND 4.0 International license It is made available under a is the author/funder, who has granted medRxiv a license to display the preprint in perpetuity. The copyright holder for this preprint . https://doi.org/10.1101/2020.03.08.20029710 doi: medRxiv preprint underlying diseases is poor, which was presumed to be associated with low immunity [3] . We know that most T cells develop in the thymus. In this study, the prognosis of children with COVID-19 was good, and the prognosis was different from that of middle-aged and elderly patients in other studies. Consider that the difference was related to thymic immune function. In the treatment of SARS patients, drugs such as glucocorticoids and thymosin were used [7] . Although there was no clinical evidence to support the use of glucocorticoids [8] , whether early use of thymosin and gamma globulin for immunomodulation in patients with severe and critical COVID-19 can reduce cytokine storms, reduce clinical symptoms and improve prognosis requires further exploration. Except that LDH increased in the myocardial zymogram of children with COVID-19 in the study, other enzymes such as creatine kinase and troponin I were normal. In addition to being derived from the heart muscle, LDH was also found in the liver and kidneys. We speculated that the increase in LDH may not originate from the myocardium, suggesting that children's infections were less likely to affect the myocardium. Similar to other studies [1, 3, 4] , this study found that ALT and AST increased, so we cannot rule out the effect of 2019-nCoV on the liver and induce the increase of liver enzymes. As patients often use antiviral drugs in the early stage of hospitalization, it was also necessary to consider the increase of liver enzymes caused by drug factors. Angiotensin converting enzyme 2 (ACE2) had been shown to be one of the main receptors that mediate the entry of 2019-nCoV into human cells [9, 10] , while ACE2 was at high expression levels in the kidney, vas deferens and testicular mesenchymal cells [11] .SARS had been shown to cause orchitis and affect spermatogenesis function [11, 12] .Although acute renal impairment had not yet occurred in pediatric cases, . CC-BY-NC-ND 4.0 International license It is made available under a is the author/funder, who has granted medRxiv a license to display the preprint in perpetuity. The copyright holder for this preprint . https://doi.org/10.1101/2020.03.08.20029710 doi: medRxiv preprint children with COVID-19 must be alert to potential risks and follow-up on reproductive function after discharge from hospital. Patients in this study were tested with the 2019-nCOV nucleic acid test. The specimens included nasopharyngeal swabs, blood, sputum, and anal swabs. We found that nasal swabs and anal swabs were positive even without respiratory symptoms such as cough and gastrointestinal symptoms such as vomiting and diarrhea Therefore, the source of samples cannot be selected based on clinical manifestations, and should be submitted multiple times and at multiple sites to avoid misdiagnosis of patients with mild imaging without imaging manifestations. Although there is no clinical evidence that the virus is transmitted through the conjunctival pathway [13] , mother-infant, and breast-milk [4] , researchers have detected the virus in urine [14] , saliva [15] , suggesting that novel coronavirus has a wide range of transmission routes. The clinical classification of children was mostly mild or ordinary type, so the treatment plan is relatively simple. In some cases, the nucleic acid test turned negative after treatment with interferon atomization or lopinavir ritonavir. For mild children, symptomatic treatment or single use of traditional Chinese medicine needs to be further verified. The sample size of this study was only 26, and it was a retrospective study, so it has certain limitations. On the one hand, the article only included children aged 1-14 years. The characteristics of newborns infected with COVID-19 are not clear, and whether there are differences in the disease characteristics of infants, preschool and school-age children need further research; On the other hand, the changes of T lymphocyte subsets in children need to be tracked and observed during the course of the disease. . CC-BY-NC-ND 4.0 International license It is made available under a is the author/funder, who has granted medRxiv a license to display the preprint in perpetuity. The copyright holder for this preprint . https://doi.org /10.1101 /10. /2020 In summary, COVID-19 children showed asymptomatic or mainly respiratory symptoms such as fever and dry cough, some of them have digestive tract symptoms, and their clinical symptoms are mild. In our study, the clinical types were mild and common, and the probability of severe complications such as myocarditis and ARDS was low, which provided important information for understanding the clinical characteristics of children with COVID-19. . CC-BY-NC-ND 4.0 International license It is made available under a is the author/funder, who has granted medRxiv a license to display the preprint in perpetuity. The copyright holder for this preprint . https://doi.org/10.1101/2020.03.08.20029710 doi: medRxiv preprint All data used during the study appear in the submitted article. The authors thank the department of Shenzhen Center for Disease Control and . CC-BY-NC-ND 4.0 International license It is made available under a is the author/funder, who has granted medRxiv a license to display the preprint in perpetuity. The copyright holder for this preprint . https://doi.org/10.1101/2020.03.08.20029710 doi: medRxiv preprint 937-47. . CC-BY-NC-ND 4.0 International license It is made available under a is the author/funder, who has granted medRxiv a license to display the preprint in perpetuity. (which was not peer-reviewed) The copyright holder for this preprint . https://doi.org/10.1101/2020.03.08.20029710 doi: medRxiv preprint . CC-BY-NC-ND 4.0 International license It is made available under a is the author/funder, who has granted medRxiv a license to display the preprint in perpetuity. (which was not peer-reviewed) The copyright holder for this preprint . https://doi.org/10.1101/2020.03.08.20029710 doi: medRxiv preprint Remain in hospital 9 (35%) 
-------------------------------------------------------


NEW PAPER


Background: Since mid-December 2019, a cluster of pneumonia-like diseases caused by a novel coronavirus, now designated COVID-19 by the WHO, emerged in Wuhan city and rapidly spread throughout China. Here we identify the clinical characteristics of COVID-19 in a cohort of patients in Shanghai.
In December 2019, a cluster of patients with pneumonia of unknown cause was linked to a seafood wholesale market in Wuhan, China 1 At the time of this analysis, only a little over a month has passed since COVID-19 was first reported 1 . There are reasonable concerns about whether COVID-19 has undergone rapid or marked genomic mutation during transmission. Yet large cohort reports from outside Wuhan (Hubei Province) are not yet available, and the clinical characteristics of COVID-19 remain largely unclear. In this article, we describe the initial clinical, laboratory, and radiological characteristics of patients confirmed to have COVID-19 in Shanghai, and we compare the clinical features between patients with less severe illness and those requiring critical care. All rights reserved. No reuse allowed without permission. author/funder, who has granted medRxiv a license to display the preprint in perpetuity. The copyright holder for this preprint (which was not peer-reviewed) is the . https://doi.org/10.1101/2020.03.04.20030395 doi: medRxiv preprint consent was waived in light of the urgent need to collect clinical data. The severity of COVID was defined based on the criteria established by China's National Health Commission 4 . 1. Mild. Minor symptoms only, without evidence for pneumonia by chest X-ray. 2, Moderate. Fever and respiratory symptoms are present, and there is evidence for pneumonia by chest X-ray. 3. Severe. Defined by any of the following conditions. 1) Dyspnoea, respiratory rate ≥ 30 /min, 2) resting hypoxia SaO2 ≤ 93%, 3) PaO2/FiO2 ≤ 300 mmHg. 4. Critical. The presence of any of the following conditions. 1) Respiratory failure, require mechanical ventilation, 2) shock, 3) other acute organ failure. Median and interquartile range (IQR) or mean and standard deviation were calculated for continuous variables. Count and percentages were presented for categorical variables. Statistical Analysis was performed by using SPSS21.0. All rights reserved. No reuse allowed without permission. author/funder, who has granted medRxiv a license to display the preprint in perpetuity. The copyright holder for this preprint (which was not peer-reviewed) is the . https://doi.org/10.1101/2020.03.04.20030395 doi: medRxiv preprint Demographic and clinical characteristics. The demographic and clinical characteristics are shown in with diabetes, 12 (6.0%) with cardiovascular diseases, 4 (2.0%) with malignancy, and 6 (3.0%) with thyroid diseases. The most common symptom identified was fever (86.9%). Patients' body temperatures were most commonly in a range from 38-38.9 °C (IQR). 10 (8.8%) patients presented with high fever greater than 39 . Less than half of the patients presented with respiratory systems including cough (46.4%), sputum production (23.2%), itchy or sore throat (9.8%), shortness of breath (4.5%) and chest congestion (8.0%). Myalgia or fatigue occurred in almost one-third (31.3%) of the patients. In addition, 5 (4.4%) patients had diarrhoea, which is suggestive of digestive system involvement (Table 1) . No infections of health care workers were detected during the time interval studied. A history of recent travel to Wuhan, and contact with people from Wuhan was documented in 126 (63.6%) of the patients. There was no difference in severity of disease between cases with Wuhan contact history and no-Wuhan contact history. Of these 198 patients, 179 (90.4%) were admitted to isolation wards, including 3 mild and 176 moderate cases. 19 (9.5%) were admitted and transferred to the ICU because of the development of respiratory failure or other organ dysfunction, including 9 severe cases and 10 critical cases. 1 patient died All rights reserved. No reuse allowed without permission. author/funder, who has granted medRxiv a license to display the preprint in perpetuity. The copyright holder for this preprint (which was not peer-reviewed) is the . https://doi.org/10.1101/2020.03.04.20030395 doi: medRxiv preprint without collecting sufficient data, and therefore was excluded from this study. Patient age differed significantly between the two groups (63.7 ± 16.8 for ICU vs 48.6 ± 15.6 for non-ICU patients). Male patients were significantly more common in ICU as compared with non-ICU cases (89.5% vs. 46.9%). Moreover, patients with underlying cardiovascular diseases were significantly more common in ICU cases as compared with non-ICU cases (26.3% vs. 3.9%, P<0.01). Compared with the non-ICU, patients admitted to the ICU were more likely to have high fever with temperature over 38.5 o C (78.6% vs 37.6%), shortness of breath (36.8% vs 2.2%), and longer waiting period from onset of symptom to hospital admission (7 vs 4 days).  All rights reserved. No reuse allowed without permission. author/funder, who has granted medRxiv a license to display the preprint in perpetuity. author/funder, who has granted medRxiv a license to display the preprint in perpetuity. The copyright holder for this preprint (which was not peer-reviewed) is the . https://doi.org/10.1101/2020.03.04.20030395 doi: medRxiv preprint time (PT) (median 13.3, 12.9-13.8) and activated partial thromboplastin time (APTT, median 39.5, 36. 7-42.8) were normal in most patients. However, a marked rise in fibrinogen and CRP was observed in 108 (55.4%) and 152 (78.4%) cases, respectively. The majority of patients (65.4%) had decreased blood calcium concentration, while 19.5% cases had decreased blood sodium concentration. Compared with the non-ICU patients, patients admitted to the ICU were more likely to have increased neutrophil count (15.8% vs 5.2%), decreased lymphocyte count (84.2% vs 0.6%), increased prothrombin time (26.3% vs 7.4%), increased APTT (42.1% vs 15.3), increased fibrinogen (78.9% vs 52.8%) and increased D-dimer (63.2% vs 29.5%), as well as decreased levels of blood sodium (42.1% vs 17.0%) and calcium (100% vs 61.4%). By contrast, patients admitted to the ICU were less likely to have increased C-reactive protein (57.9% vs 80.6%). Table 2 : Admission blood cell count and coagulation function of patients with COVID-19  All rights reserved. No reuse allowed without permission. author/funder, who has granted medRxiv a license to display the preprint in perpetuity. The copyright holder for this preprint (which was not peer-reviewed) is the  All rights reserved. No reuse allowed without permission. author/funder, who has granted medRxiv a license to display the preprint in perpetuity. author/funder, who has granted medRxiv a license to display the preprint in perpetuity. author/funder, who has granted medRxiv a license to display the preprint in perpetuity. All rights reserved. No reuse allowed without permission. author/funder, who has granted medRxiv a license to display the preprint in perpetuity. All rights reserved. No reuse allowed without permission. author/funder, who has granted medRxiv a license to display the preprint in perpetuity.  All rights reserved. No reuse allowed without permission. author/funder, who has granted medRxiv a license to display the preprint in perpetuity. The copyright holder for this preprint (which was not peer-reviewed) is the Radiological findings. Of 198 patients who underwent chest computed tomography on admission, 98.5 % cases manifested abnormalities suggestive of pneumonia. The most common patterns on chest computed tomography were ground-glass opacity and bilateral patchy consolidation. Figure 1 illustrates representative findings with different degrees of radiological abnormalities from seven patients. To determine the major clinical features that appeared during COVID-19 progression, the dynamic changes in 11 clinical laboratory parameters, including haematological and biochemical parameters, were tracked from day 1 to day 15 after the onset of the disease at 2-day intervals (Fig. 2 &3) . ICU patients were subdivided into two subgroups as described in the Methods section: severe and critical. The levels of CRP and procalcitonin in critically ill patients showed a trend for sustained elevation 9 days after admission (Fig. 2A&B) . White blood cells and neutrophils counts were higher in critical patients than in severe patients (Fig. 2C&D ). Most significantly, critical cases had more severe lymphopenia than severe cases (Fig 2E, < 0 .6 x 10 9 /L, P<0.05). Critical cases had significantly more organ dysfunction and coagulation abnormalities than severe cases. Cardiac troponin I and myoglobin was higher in critical patients than in severe cases (Fig. 3A&B) , indicating heart injury. The levels of blood urea nitrogen (BUN) and creatinine level were higher in critical cases than in severe cases (Fig. 3C&D) , suggesting kidney injury. The level of D-dimer was higher in All rights reserved. No reuse allowed without permission. author/funder, who has granted medRxiv a license to display the preprint in perpetuity. The copyright holder for this preprint (which was not peer-reviewed) is the . https://doi.org/10.1101/2020.03.04.20030395 doi: medRxiv preprint critically ill than in severely ill (Fig. 3E) , however, there was no marked difference in fibrinogen level between these two groups (Fig. 3F ). All rights reserved. No reuse allowed without permission. author/funder, who has granted medRxiv a license to display the preprint in perpetuity. The copyright holder for this preprint (which was not peer-reviewed) is the . https://doi.org/10.1101/2020.03.04.20030395 doi: medRxiv preprint In this single centre cohort of confirmed COVID-19 cases in Shanghai, the median duration of symptoms from disease onset to hospital admission was 4 days. Compared to the previously reported duration of 7-10 days between symptom onset and hospitalization in Wuhan 3 5 6 , patients in Shanghai received earlier in-ward observation and medical service, most likely due to improved monitoring and increased awareness of the outbreak. As a result, our data likely were recorded closer to the onset date of the disease, which may explain some of the differences between the results reported from Wuhan and Shanghai. In terms of patient's characteristics, we observed a comparatively equal male to female ratio in this study, which differs from the data reported in Wuhan 3 5 6 . This may be because most of these initial cases in Shanghai had a history of exposure in Wuhan (63.6%), and the disease thereafter appeared in a family cluster pattern. Almost half of the patients in Shanghai were under 50 years old, in contrast with the predominantly older patient population reported in Wuhan. In addition, 69 (34.8%) patients in our cohort had pre-existing comorbidities. Whether the disease is more likely to manifest in patients with comorbidities and/or older patients remains unclear and requires further study. Regarding laboratory evaluation, low or normal levels of white blood cell (WBC) and neutrophil counts were present in COVID-19 patients upon admission. ESR and CRP levels were elevated in most patients. Although procalcitonin level was increased in over one-third (28.4%) of the patients, the median 0.03 (0.02-0.06) was close to normal range. These results are similar with that of two Wuhan cohort studies already published 3 5 . We further observed that serum lactate levels were above the normal upper range in the majority of patients (83.7%), indicating the likely presence of metabolic acidosis. The sensitivity of LACT in COVID-19 infection is similar to ESR and CRP. All rights reserved. No reuse allowed without permission. author/funder, who has granted medRxiv a license to display the preprint in perpetuity. The copyright holder for this preprint (which was not peer-reviewed) is the . https://doi.org /10.1101 /10. /2020 In addition, we observed a marked reduction in T and T H subtype cells in infected patients, suggesting that SARS-Cov-2 infection may impair cellular immunity. It is known that MERS-CoV is able to infect both CD4 + and CD8 + primary human T cells and, upon infection induces T cell apoptosis in vitro 7 . While the cell surface receptor for SARS-Cov-2 has now been identified as ACE2 8 , other receptors may also exist on immune cells to bind to coronavirus [9] [10] [11] . As there is no proven antiviral treatment yet available, strategies to enhance the immune system may be considered. Since the most common haematological changes in COVID-19 patients were lymphopenia and immunodeficiency, we postulate that hematopoietic growth factors such as G-CSF, by mobilizing endogenous blood stem cells and endogenous cytokines, may represent a potential haematological treatment for COVID-19 patients 12 . Interestingly, a noticeable increase of glutathione reductase (GR) level occurred in 40.2% of the COVID-19 patients in our cohort, while ALT and AST are normal in most patients. GR is an essential enzyme that recycles oxidized glutathione back to the reduced form 13 . GR is known to participate in an oxidative defence system required for effective immune responses against bacteria 14 15 . Whether GR is involved in host defence systems against viruses such as SARS-Cov-2 remains to be determined. We observed that 65.4% of COVID-19 patients had decreased serum calcium levels. Calcium influx regulates both clathrin-mediated and clathrin-independent endocytosis during viral infection, thus constituting a key mechanism for regulation of influenza A virus internalization and infection 16 . Based on this theory, calcium channel blockers, including amlodipine, verapamil, and diltiazem, as well as BAPTA-AM, have been proven effective in inhibiting IAV replication in a dose-dependent manner in a canine kidney cell model 17 . Moreover, diltiazem may have both prophylactic and therapeutic effects in IAV treatment according to both ex vivo and in vivo testing 17 . We therefore propose the hypothesis that Ca 2+ levels and/or Ca 2+ channels may play a role in endocytosis and infection of SARS-Cov-2. Further studies are needed to characterize the functional importance of this potential pathway. All rights reserved. No reuse allowed without permission. author/funder, who has granted medRxiv a license to display the preprint in perpetuity. The copyright holder for this preprint (which was not peer-reviewed) is the . https://doi.org/10.1101/2020.03.04.20030395 doi: medRxiv preprint The patients admitted to the ICU were more likely to be older, male, with temperature over 38.5 o C, symptom of dyspnoea, underlying cardiovascular disease, and longer waiting period from onset of symptom to hospital admission, compared to those not admitted to the ICU. This suggests that age, sex, high fever, admission time and co-morbidity are risk factors for disease severity. Compared with non-ICU patients, patients who received ICU care had numerous laboratory abnormalities. These abnormalities suggest that SARS-Cov-2 infection can be associated with cellular immune deficiency, coagulation activation, myocardial injury, hepatic injury and kidney injury. These laboratory abnormalities are similar to those previously observed in patients with MERS-Cov and SARS-Cov infection 18 19 . The longitudinal profile of laboratory findings was dynamically tracked for those patients requiring ICU care. When these patients were subdivided into severe and critical ill subgroups, we observed that white blood cell count, neutrophil count, D-dimer, BUN, creatinine, myoglobulin and troponin I levels were higher in critical cases than in severe cases. Neutrophilia may be associated with secondary infection, while coagulation activation could be related to sustained inflammatory responses. Acute heart and kidney injury could be related to direct effects of the virus or possibly hypoxia 9 20 . Our study has several limitations. First, although we obtained data from the initial 198 patients with laboratory-confirmed COVID-19 in Shanghai, the cohort is still relatively small. More patients need to be analysed to provide a comprehensive and precise description of the spectrum of disease associated with this infection. Secondly, our study reports primarily baseline results from patients upon hospital admission, and more longitudinal data regarding disease progression and clinical outcomes will require further collection and study. Despite these limitations, our study provides data from the first large cohort outside Wuhan and adds important laboratory information to the rapidly accumulating body of information about COVID-19. These results will assist in multi-centre monitoring of the disease. All rights reserved. No reuse allowed without permission. author/funder, who has granted medRxiv a license to display the preprint in perpetuity. The copyright holder for this preprint (which was not peer-reviewed) is the . https://doi.org/10.1101/2020.03.04.20030395 doi: medRxiv preprint In conclusion, COVID-19 affects a wide-range of patients, from youth to the elderly. Fever is consistently the most common symptom of onset, but multiple other clinical manifestations occur, including a spectrum form T cell deficiency to symptoms of digestive system involvement. Older age, male, fever over 38.5 o C, symptom of dyspnoea, the presence of underlying cardiovascular disease, and longer waiting period from onset of symptom to hospital admission are risk factors associated with the severity of disease. All rights reserved. No reuse allowed without permission. author/funder, who has granted medRxiv a license to display the preprint in perpetuity. The copyright holder for this preprint (which was not peer-reviewed) is the . https://doi.org/10.1101/2020.03.04.20030395 doi: medRxiv preprint author/funder, who has granted medRxiv a license to display the preprint in perpetuity. The copyright holder for this preprint (which was not peer-reviewed) is the . https://doi.org/10.1101/2020.03.04.20030395 doi: medRxiv preprint Fig. 1 All rights reserved. No reuse allowed without permission. author/funder, who has granted medRxiv a license to display the preprint in perpetuity. The copyright holder for this preprint (which was not peer-reviewed) is the . https://doi.org/10.1101/2020.03.04.20030395 doi: medRxiv preprint Fig. 2 All rights reserved. No reuse allowed without permission. author/funder, who has granted medRxiv a license to display the preprint in perpetuity. The copyright holder for this preprint (which was not peer-reviewed) is the . https://doi.org/10.1101/2020.03.04.20030395 doi: medRxiv preprint author/funder, who has granted medRxiv a license to display the preprint in perpetuity. The copyright holder for this preprint (which was not peer-reviewed) is the . https://doi.org/10.1101/2020.03.04.20030395 doi: medRxiv preprint
-------------------------------------------------------


NEW PAPER


A recently developing pneumonia caused by SARS-CoV-2 was originated in Wuhan, China, and has quickly spread across the world. We reported the clinical characteristics of 82 death cases with COVID-19 in a single center.
The bulk of death cases had comorbidity (76.8%). Respiratory failure remained the leading cause of death, following by sepsis syndrome/MOF, cardiac failure, hemorrhage, and renal failure. Most patients had a high neutrophil-to-lymphocyte ratio, high systemic immune-inflammation index, and increased levels of proinflammatory cytokines. SARS-CoV-2 causes a cluster of severe respiratory illness which is similar to another two fatal coronavirus infection caused by SARS-CoV and MERS-CoV. Death is more likely to occur in older male patients with comorbidity. Infected patients might develop acute respiratory distress and respiratory failure which was the leading cause of death, but damages of other organs and systems, including cardiac, hemorrhage, hepatic, and renal also contribute to the death. These damages might be attributable to indirect cytokines storm initiated by immune system and direct attack from SARS-CoV-2 itself. In December 2019, the first acute respiratory disease caused by severe acute respiratory syndrome coronavirus (SARS-CoV)-2, and recently officially named as Corona Virus Disease 2019 (COVID- 19) by World Health Organization (WHO) occurred in Wuhan, China. 1, 2 Person-to-person transmission has been identified through respiratory droplets or likely feces. [3] [4] [5] By February 14, 2020, more than 60,000 confirmed cases and close to 2,000 dead cases have been documented in China, with hundreds of imported patients found in other countries. [4] [5] [6] [7] Generally, the incubation period of COVID-19 was 3 to 7 days. Fever, cough, and fatigue were the most common symptoms. 1 Approximately 20-30% of cases would develop severe illness, and some need further intervention in intensive care unit (ICU) . 8, 9 Organ dysfunction including acute respiratory distress syndrome (ARDS), shock, acute cardiac injury, and acute renal injury, can happen in severe cases with COVID-19. 1, 8, 9 It has been reported that critical ill patients were more likely to be older, had underlying diseases, and were more likely to have a symptom of dyspnea. 9 Oxygen therapy, mechanical ventilation, intravenous antibiotics and antiviral therapy were usually applied in clinical management, but presently there were no effective drugs for improving the clinical outcome of COVID-19, especially for severe cases. 1, 8, 9 ARDS, a rapidly progressive disease, is the main cause of death for the patients infected with previously recognized corona virus infection such as SARS-CoV and Middle Eastern respiratory syndrome coronavirus (MERS-CoV). 10, 11 In this context, it was initially considered that lung is the most commonly damaged organ by All rights reserved. No reuse allowed without permission. author/funder, who has granted medRxiv a license to display the preprint in perpetuity. The copyright holder for this preprint (which was not peer-reviewed) is the . https://doi.org/10.1101/2020.02.26.20028191 doi: medRxiv preprint SARS-CoV-2 infection since human airway epithelia express angiotensin converting enzyme 2 (ACE2), a host cell receptor for SARS-CoV-2 infection. 12, 13 However, increasing clinical cases indicated cardiac, renal and even digestive organ damage in the patients with COVID-19, 9 which is consistent with the findings that kidney, colon and the other tissues also express ACE2 besides airway epithelia. 14, 15 The above clinical phenomenon and basic research suggest more complicated pathogenesis of COVID-19. Hence, analyzing clinical characteristics of death cases with COVID-19 is urgently needed to improve the outcome of infected patients. We retrospectively collected epidemiological and clinical features of laboratory-confirmed COVID-19 dead patients from January 11, 2020 to February 10, 2020 in the Renmin Hospital, Wuhan University. The confirmed diagnosis of COVID-19 was defined as a positive result by using real-time reverse-transcriptase polymerase-chain-reaction (RT-PCR) detection for routine nasal and pharyngeal swab specimens. This study received approval from the Research Ethics Committee of the Renmin Hospital of Wuhan University, Wuhan, China (approval number: WDRY2020-K038). The Research Ethics Committee waived the requirement informed consent before the study started because of the urgent need to collect epidemiological and clinical data. We analyzed all the data anonymously. The clinical features, including clinical symptoms, signs, laboratory analyses, radiological findings, treatment, and outcome, were obtained from the hospital's electronic medical records according to previously designed standardized data All rights reserved. No reuse allowed without permission. author/funder, who has granted medRxiv a license to display the preprint in perpetuity. The copyright holder for this preprint (which was not peer-reviewed) is the . https://doi.org/10.1101/2020.02.26.20028191 doi: medRxiv preprint collection forms. Laboratory analyses included complete blood count, liver function, renal function, electrolytes test, coagulation function, C-reactive protein, lactate dehydrogenase, myocardial enzymes, procalcitonin, and status of other virus infection. Radiological analyses comprised of X-ray and computed tomography. The date of onset of symptoms, initial diagnosis of COVID-19, and death were recorded accurately. The incubation period was defined as the time from the contact of transmission origin to the onset of different symptoms and signs. Onset survival time was defined as the period between the onset of different symptoms and signs and the time of death. To increase the accuracy of collected data, two researchers independently reviewed the data collection forms. We also directly communicated with patients or their family members to ascertain the epidemiological and symptom data. Inflammation markers were calculated using specific parameters of blood tests. Neutrophil-to-lymphocyte ratio (NLR) was calculated by dividing the absolute neutrophil count by the lymphocyte count. Systematic inflammatory index (SII) was defined as platelet count × neutrophil count/ lymphocyte count (/μL). Interleukin  All rights reserved. No reuse allowed without permission. author/funder, who has granted medRxiv a license to display the preprint in perpetuity. The copyright holder for this preprint (which was not peer-reviewed) is the . https://doi.org/10.1101/2020.02.26.20028191 doi: medRxiv preprint Descriptive analyses were used to determine the patients' epidemiological and clinical features. Continuous variables were presented as median and interquartile range (IQR), and categorical variables were expressed as the percentages in different categories. The Chi-squared test or Fisher's exact test was adopted for category variables. The association between the different clinical variables and the time from initial symptom to death was evaluated using Spearman's rank correlation coefficient. Statistical analyses in this study were performed with use of STATA 15.0 software (Stata Corporation, College Station, TX, USA). A two-sided p value less than 0.05 was considered statistically significant. From January 11, 2020 to February 10, 2020, a total of 1,334 patients with a diagnosis of laboratory-confirmed COVID-19 were recorded in the Renmin Hospital, Wuhan University, while 6.2% (82/1334) of patients with this disease were dead. In the same period, the rate of mortality for all causes and non-COVID-19 in this hospital, were 2.3% (162/7119) and 1.4% (80/5785), respectively. The mortality rate of COVID-19 was higher than that of non-COVID-19 (p<0.001). author/funder, who has granted medRxiv a license to display the preprint in perpetuity. The copyright holder for this preprint (which was not peer-reviewed) is the . https://doi.org/10.1101/2020.02.26.20028191 doi: medRxiv preprint of death cases had comorbidity (75.6%), including hypertension (56.1%), heart disease (20.7%), diabetes (18.3%), cerebrovascular disease (12.2%), and cancer (7.3%). 30 out of 82 dead patients (30.5%) had 2 or more underlying diseases. We analyzed the causes of mortality of patients with laboratory-confirmed SARS-CoV-2 infection (table 2) . Respiratory failure remained the leading cause of death (69.5%), following by sepsis syndrome/multiple organ dysfunction syndrome (MOF) (28.0%), cardiac failure (14.6%), hemorrhage (6.1%), and renal failure (3.7%). Furthermore, respiratory, cardiac, hemorrhage, hepatic, and renal damage were found in 100%, 89%, 80.5%, 78.0%, and 31.7% of patients, respectively. A majority of patients (75.6%) had 3 or more damaged organs or systems following the infection with SARS-CoV-2. were observed in patients. All rights reserved. No reuse allowed without permission. author/funder, who has granted medRxiv a license to display the preprint in perpetuity. The copyright holder for this preprint (which was not peer-reviewed) is the . https://doi.org/10.1101/2020.02.26.20028191 doi: medRxiv preprint Furthermore, in the last 24 hours of the death, lymphopenia (73.7%), neutrophilia (100%), and thrombocytopenia (63.2%) were continuously present. Increased C-reactive protein level, high NLR, increased lactate dehydrogenase, and increased D-dimer were found in all patients. The incidence of neutrophilia was increased from 74.3% to 100%, and the incidence of lymphopenia was reduced from 89.2% to 73.7%. The incidence of high creatinine and blood urea nitrogen were increased from 15.3% The median time from initial symptom to death was 15 days (IQR 15-20) and a significant association between aspartate aminotransferase (p=0.002), alanine aminotransferase (p=0.037) and time from initial symptom to death were interestingly observed (figure 1A-C). To our knowledge, this is the first study to describe the clinical characteristics of dead patients with COVID-19. The mortality of 6.2% from current study was lower than that of SARS infection in 2003. However, the mortality rate from this center is a little All rights reserved. No reuse allowed without permission. author/funder, who has granted medRxiv a license to display the preprint in perpetuity. The copyright holder for this preprint (which was not peer-reviewed) is the . https://doi.org/10.1101/2020.02.26.20028191 doi: medRxiv preprint bit higher than previously reported. 9 We speculated the reason might be that fewer patients in our study were transferred to the ICU in time when their condition rapidly worsened. On the other hand, limited death cases, less than 15 patients were included in their cohort, 8, 9 while much more death cases were included in the present study. Our study firstly focused on the epidemiological characteristics of dead patients with COVID-19. Several factors were responsible for the death of these patients included in this study. A majority of patients were older than 60 years in our study, and a borderline significant association between age and time from initial symptom to death were interestingly observed. These results are consistent with that older age was more likely occurred in critically ill patients. 16 Moreover, we found underlying diseases such as hypertension, heart disease and diabetes were very common in our death cases, and 30.5% of patients had 2 or more comorbidities. These features are consistent with previous report that patients with underlying diseases more likely developed to severe illness. 9 Cancer patients is comprised of 7.3% in our cohort, much higher than cancer morbidity, suggest that cancer patients more likely develop to severe disease, even death. These results are consistent with the findings from a national wide analysis in China. 17 Immune deficiency to virus infection seems to be the common features in older males with comorbidities. We further analyzed the cause of death case with COVID-19 and found that respiratory failure remained the leading cause of death. It has been reported that the binding receptor for SARS-CoV-2, ACE2 is mainly expressed in blood vessels and lung alveolar type II (AT2) epithelial cells, 13 Similar to the SARS-CoV and MERS-CoV, SARS-CoV-2 can directly attack ACE2-expressing cells. 13 Indeed, All rights reserved. No reuse allowed without permission. author/funder, who has granted medRxiv a license to display the preprint in perpetuity. The copyright holder for this preprint (which was not peer-reviewed) is the . https://doi.org/10.1101/2020.02.26.20028191 doi: medRxiv preprint pathological findings indicated that infected lungs with SARS-CoV-2 present as ARDS, pulmonary edema with hyaline membrane formation, evident desquamation of pneumocytes. 18 Therefore, our finding that respiratory failure is the leading cause of death, is consistent with the underlying pathological mechanism of COVID-19. Besides respiratory failure, cardiac failure, hemorrhage, renal failure and even MOF were also recognized as the cause of death by COVID-19 in our study. Laboratory findings also revealed cardiac, hepatic, and renal damage in some of patients. We also observed a significant association between aspartate aminotransferase, alanine aminotransferase and time from initial symptom to death. These clinical phenomena could be explained by virus itself attack and cytokine release storm (CRS) mediated tissue damage. First, ACE2 expression is also found in the kidney, heart, and liver etc, therefore SARS-CoV-2 could invade the cells of above tissues, reproduce and damage these organ. 14, 15 Second, virus infection and subsequent tissue damage either in the lung or other target organ could elicit immune cells to produce pro-inflammatory cytokines, namely CRS, ultimately injury the tissue and cause target organ failure. Obviously, increased amounts of cytokines, including IL-1β, IL-6, and monocyte chemotactic protein-1 (MCP-1), are associated with severe lung injury in patients infected with SARS-CoV and MERS-CoV. 19, 20 A recent study showed high levels of IL-1β, interferon γ -induced protein 10 , and MCP-1 occurred in serum of patients infected with SARS-CoV-2, which probably leaded to the activation of T-helper-1 cell response 1 . In the present report, high level of IL-6 of more than10U/L and C-reactive protein were detected in all patients, even in the last 24 hours prior to death. All rights reserved. No reuse allowed without permission. author/funder, who has granted medRxiv a license to display the preprint in perpetuity. The copyright holder for this preprint (which was not peer-reviewed) is the . https://doi.org /10.1101 /10. /2020 We also depicted the immune status of COVID-19 patients with severe illness. Most of patients in our study presented as neutrophilia and lymphopenia on admission, specifically reduced CD3+, CD4+, and CD8+ T-cell counts were observed in some patients. A high NLR was also observed on the admission and 24 hours before the death. These results, consistent with the previous findings found in the patients with severe illness, suggest perturbation of immune system contribute the pathogenesis of SARS-CoV-2. These observations could be also explained why older males with comorbidities likely succumb to COVID-19. Our study has some limitations. Firstly, some patients did not receive timely supportive interventions such as admission to ICU, because increasing number of severe patients occurred in a short period. However, present data could partially be scenario where COVID-19 patients progress in a natural pathophysiology rather than outcome from intervention by treatment. Secondly, consecutive detection of cytokines was lacking, which fail to truly monitor the severity of CRS. Thirdly, organ damage could originate from a history of medication including nonsteroidal anti-inflammatory drugs, antibiotics, and traditional Chinese medicine which are associated with renal or liver injury. 21, 22 In our study, all patients received intravenous of antibiotics and anti-virus drugs. Overall, from the point of view of the causes of death, we presented the clinical characteristics of patients with COVID-19. Lung injury begins with an insult to the lung epithelium mainly attacked by SARS-CoV-2 itself because of ACE2 expressed in the lungs, which leads to most common respiratory failure. Other organs or tissues, more or less, are potentially damaged through direct attack from SARS-CoV-2. In All rights reserved. No reuse allowed without permission. author/funder, who has granted medRxiv a license to display the preprint in perpetuity. The copyright holder for this preprint (which was not peer-reviewed) is the . https://doi.org/10.1101/2020.02.26.20028191 doi: medRxiv preprint addition, damages of multiple systems including the lungs, might originate with systemic damage due to CRS following SARS-CoV-2 infection. Considering the pandemic potential and moderate threaten of COVID-19 for population with multiple underlying diseases, further studies are required to focus on pathology and pathophysiology of tissue injury caused by SARS-CoV-2 infection, especially on the activation process of immune response and cytokines storm. JW, BZ and QS had the idea for and designed the study and had full access to all data in the study and take responsibility for the integrity of the data and the accuracy of the data analysis. BiZh, YZ and YQ contributed to writing of the report. BZ contributed to critical revision of the report. JW and BiZh contributed to the statistical analysis. All authors contributed to data acquisition, data analysis, or data interpretation, and reviewed and approved the final version. All authors declare no competing interests. The data that support the findings of this study are available from the corresponding author on reasonable request. Participant data without names and identifiers will be made available after approval from the corresponding author and National Health Commission. After publication of study findings, the data will be available for others to request. The research team will provide an email address for communication once the data are approved to be shared with others. The proposal with detailed description All rights reserved. No reuse allowed without permission. author/funder, who has granted medRxiv a license to display the preprint in perpetuity. The copyright holder for this preprint (which was not peer-reviewed) is the . https://doi.org/10.1101/2020.02.26.20028191 doi: medRxiv preprint of study objectives and statistical analysis plan will be needed for evaluation of the reasonability to request for our data. The corresponding author will make a decision based on these materials. Additional materials may also be required during the process. author/funder, who has granted medRxiv a license to display the preprint in perpetuity. The copyright holder for this preprint (which was not peer-reviewed) is the . https://doi.org/10.1101/2020.02.26.20028191 doi: medRxiv preprint author/funder, who has granted medRxiv a license to display the preprint in perpetuity. The copyright holder for this preprint (which was not peer-reviewed) is the author/funder, who has granted medRxiv a license to display the preprint in perpetuity. The copyright holder for this preprint (which was not peer-reviewed) is the . https://doi.org/10.1101/2020.02.26.20028191 doi: medRxiv preprint
-------------------------------------------------------


NEW PAPER


The aim of the study was to analyze the incidence of COVID-19 with early renal injury, and to explore the value of multi-index combined detection in diagnosis of early renal injury in COVID-19.
This study intends to use a number of laboratory test indexes, including serum creatinine, blood urea nitrogen, urine creatinine, urine microalbumin and urine microglobulin et al to comprehensively assess renal function and determine the incidence of COVID-19 with early renal injury. The related risk factors of COVID-19 with early renal injury were also analyzed.  A standardized case collection form was designed to collect laboratory data of the included patients. It mainly included the following items: (1) blood routine; (2) biochemical indexes of kidney, liver, and heart function; (3) coagulation function; (5) infection indexes. Serum creatinine (Scr), blood urea nitrogen (BUN), urine creatinine (Ucr), urine protein (PRO), urine microalbumin (UMA), α 1-microglobulin (A1M), urine immunoglobulin G (IGU), and urine transferrin (TRU) were detected. The estimated glomerular filtration rate (eGFR), endogenous creatinine clearance (Ccr), and urine microalbumin/creatinine ratio (UACR) were calculated. All rights reserved. No reuse allowed without permission. author/funder, who has granted medRxiv a license to display the preprint in perpetuity. In this study, two or more abnormalities of eGFR, Ccr and UACR were defined as early renal injury. The area under curves (AUC) of receiver operating characteristic (ROC) were calculated for predictive analysis. Spearman rank correlation coefficient was used to analyze the linear correlation between two sets of continuous variables. It was considered statistically significant when the P value was less than 0.05. All statistical analyses were processed using SPSS 25.0 statistical software. Patients were not directly involved in the design, planning and conducting of this study. Of the 12 patients with COVID-19, 2 were severe patients, 8 were general patients, and 2 were light patients. Laboratory testing items included blood routine, electrolyte, metabolism, heart, liver, kidney function indicators, coagulation function and infection indicators. As shown in Table 1 , the common abnormal indicators (abnormal rate ≥ 50%) included: increased neutrophil ratio (50%), increased monocyte ratio All rights reserved. No reuse allowed without permission. author/funder, who has granted medRxiv a license to display the preprint in perpetuity. The copyright holder for this preprint (which was not peer-reviewed) is the . https://doi.org/10.1101/2020.03.07.20032599 doi: medRxiv preprint (75%), hypokalemia (50%), hyponatremia (50%), hypoproteinemia (75%), and increased C-reactive protein (83.3%). As shown in Table 2 , Scr and BUN were not significantly abnormal in 12 patients. However, the calculation of eGFR and Ccr showed that early renal injury was common in COVID-19 patients. Among them, 66.7% patients had reduced eGFR and 41.7% patients had reduced Ccr. Urinary microprotein test also confirmed that COVID-19 with early renal injury was common, with UACR> 30mg / g accounting for 41.7%. Among the 12 patients, the abnormally elevated rates of UMA, A1M, IGU, and TRU were 33.3%, 33.3%, 41.7%, and 16.7%, respectively. The positive rate of the four-index combined was 58.3%. The results suggested that case 3, 5, 6, and 7 were significant early kidney injury cases. The AUC of UMA, A1M, IGU, and TRU for the diagnosis of early renal injury in COVID-19 were 0.813, 0.813, 0.750, and 0.750 respectively, while the AUC of the combined index of UMA + A1M + IGU + TRU was 0.875, suggesting that the multi-index combined was helpful for the diagnosis of early renal injury in COVID-19, as shown in Fig 1.  In this study, Spearman correlation coefficients were used to evaluate the correlation between renal function indicators (eGFR, Ccr, and UACR) and the obviously abnormal biochemical indicators (C-reactive protein (CRP), neutrophil ratio (NER), monocyte ratio (MOR), serum potassium (K), serum sodium (Na), and albumin (ALB)). As shown in Table 3 , it was found that eGFR was negatively correlated with CRP and NER, and positively correlated with serum K, Ccr was negatively correlated All rights reserved. No reuse allowed without permission. author/funder, who has granted medRxiv a license to display the preprint in perpetuity. The copyright holder for this preprint (which was not peer-reviewed) is the . https://doi.org/10.1101/2020.03.07.20032599 doi: medRxiv preprint with CRP and NER, and positively correlated with MOR and serum Na, and UACR was positively correlated with CRP, and negatively correlated with serum K (Fig 2) . The results suggested that the more severe the infection, the more obvious the early renal injury, and the early renal injury in COVID-19 can often cause hypokalemia and hyponatremia. Coronaviruses are single-stranded positive-strand RNA viruses. In 2014, the Peng Zhou et al [11] reported the identification and characterization of 2019-nCoV. Through full-length genome sequences analysis, they found that the whole genome of 2019-nCoV shared 79.5% sequence identify to SARS-CoV. The pairwise protein sequence analysis of seven conserved non-structural proteins showed that this virus belongs to the species of severe acute respirator syndrome-related All rights reserved. No reuse allowed without permission. author/funder, who has granted medRxiv a license to display the preprint in perpetuity. The copyright holder for this preprint (which was not peer-reviewed) is the . https://doi.org/10.1101/2020.03.07.20032599 doi: medRxiv preprint coronaviruses (SARSr-CoV). In addition, they also confirmed that 2019-nCov used the same cell entry receptor, Angiotensin converting enzyme II (ACE2), as SARS-CoV. The high degree of similarity in gene sequence and cellular mechanism of 2019-nCoV and SARS-CoV suggests that the risk factors of mortality could also be similar. SARS is an acute respiratory infectious disease with pulmonary parenchyma and/or interstitium as the main site of invasion and multiple organ injury. Previous clinical studies have found that, similar to SARS, the causes of death caused by 2019-nCov are not only lung tissue damage, but also heart, liver, kidney and other organ dysfunction or even failure, which is one of the important causes of aggravation and death. [8, 9] In the previous SARS case study [12] , acute renal injury was found to be the top risk factor of mortality, even higher than acute respiratory distress. In that case study, all patients who eventually died had a progressive rise of Scr, and the rise of Scr was rapid in those who succumbed early in their illness. Previously, an ongoing case study reported 59 patients infected by 2019-nCoV, including 28 severe cases and 3 death. In that study, 63% of the patients exhibited proteinuria, and 19% and 27% of the patients had an elevated level of Scr and BUN respectively. Computed tomography (CT) scans revealed abnormal renal imaging in all patients. The results suggested that renal injury was common in COVID-19, which may contribute to multiorgan failure and death eventually. [13] In this study, 12 COVID-19 cases were analyzed. Although all the patients still had normal level of Scr and BUN, the high incidence of early renal injury in COVID-19 was found by calculating eGFR, CCR and UACR. The abnormal rates of eGFR, Ccr, and UACR were 66.7%, 41.7%, and 41.7%, respectively. Combined detection of UMA, A1M, IGU, and TRU could be helpful for the diagnosis of early renal injury in COVID-19. Furthermore, the study also found that the degree of early renal injury was significantly related to C-reactive protein (CRP) and neutrophil ratio (NER), suggesting that the more severe the infection, the more obvious the early renal injury. Hypokalemia and hyponatremia were common in patients with COVID-19, and there was a correlation with the degree of renal injury. Xin Zou et al [14] analyzed All rights reserved. No reuse allowed without permission. author/funder, who has granted medRxiv a license to display the preprint in perpetuity. The copyright holder for this preprint (which was not peer-reviewed) is the . https://doi.org/10.1101/2020.03.07.20032599 doi: medRxiv preprint the single-cell RNA sequencing datasets to explore the expression of ACE2 in the main physiological systems of the human body, including the respiratory, cardiovascular, digestive and urinary system. The study showed that heart, esophagus, kidney, bladder, and ileum have similar or higher ACE2 expression than in alveoli. In the analysis of specific cell types, the expression of ACE2 in renal proximal tubule cells was about four times higher than that of type II alveolar cells (AT2). The results suggest that the kidney may be one of the primary targets of attack for the 2019-nCov. α 1-microglobulin can pass through the glomerular filtration membrane, and 95% to 99% is reabsorbed in the proximal tubule, and it is not affected by pH. In the early stage of renal injury, the renal tubule reabsorption function reduced, resulting in urine α 1-microglobulin excretion increased, and the elevation is consistent with the degree of renal tubular injury. [16] Due to the large molecular weight, immunoglobulin G is All rights reserved. No reuse allowed without permission. author/funder, who has granted medRxiv a license to display the preprint in perpetuity. The copyright holder for this preprint (which was not peer-reviewed) is the . https://doi.org/10.1101/2020.03.07.20032599 doi: medRxiv preprint difficult to pass through the glomerular filtration membrane. When renal function is impaired, it can lead to increased permeability of the glomerular filtration membrane. The excretion of albumin in urine increases, and as the lesions worsen, the excretion of immunoglobulin G in urine also increases. Therefore, the detection of urine immunoglobulin G can help to judge the degree of damage on the glomerular filtration membrane. [17] The main physiological function of transferrin is to transport iron ions. It is a negatively charged protein, and its isoelectric point is one unit higher than that of albumin. Normally, it cannot pass through the positively charged glomerular filtration membrane. Urine transferrin is one of the indicators of early glomerular injury, which mainly reflects the damage of glomerular filtration membrane charge selection barrier. [18]  Strengths of this study are as follows: Firstly, this study first investigated the incidence of early renal injury in COVID-19 and revealed the prevalence of early renal injury in COVID-19 patients. Secondly, this study clarified the value of multi-index combined detection for the diagnosis of early renal injury in COVID-19, and provided a basis for early detection and early intervention. The present study has some limitations that must be taken into account when considering its contribution. First and most significantly is the small sample size of this study. COVID-19 is a newly emerging epidemic. The cases found in this city are all imported cases. After strict prevention and control, there are fewer confirmed cases in this city. Therefore, the number of cases included in this research is also small. In addition, considering COVID-19 is a completely new disease, and we do not yet fully know the characteristics of its cases, the type of research used in this study is designed as a descriptive study, which initially explores the basic characteristics of the disease and does not set the control group. This study found that early renal injury was common in patients with COVID-19. Combined detection of UMA, A1M, IGU, and TRU could be helpful for the diagnosis All rights reserved. No reuse allowed without permission. author/funder, who has granted medRxiv a license to display the preprint in perpetuity. The copyright holder for this preprint (which was not peer-reviewed) is the . https://doi.org/10.1101/2020.03.07.20032599 doi: medRxiv preprint of early renal injury in COVID-19. The 12 cases of COVID-19 reported in this study were all healed and discharged after being treated with antivirals, maintaining water and electrolyte balance, enhancing immunity, oxygen therapy, and traditional chinses medicine treatment. In the process of diagnosis and treatment, the assessment of early author/funder, who has granted medRxiv a license to display the preprint in perpetuity. The copyright holder for this preprint (which was not peer-reviewed) is the . https://doi.org/10.1101/2020.03.07.20032599 doi: medRxiv preprint Legend Table 1. Laboratory test results of the COVID-19 patients   Table 2 . Renal function indexes of the COVID-19 patients Table 3 . Laboratory indexes related to early renal injury in COVID-19 author/funder, who has granted medRxiv a license to display the preprint in perpetuity. The copyright holder for this preprint (which was not peer-reviewed) is the . author/funder, who has granted medRxiv a license to display the preprint in perpetuity. The copyright holder for this preprint (which was not peer-reviewed) is the All rights reserved. No reuse allowed without permission. author/funder, who has granted medRxiv a license to display the preprint in perpetuity. The copyright holder for this preprint (which was not peer-reviewed) is the All rights reserved. No reuse allowed without permission. author/funder, who has granted medRxiv a license to display the preprint in perpetuity. The copyright holder for this preprint (which was not peer-reviewed) is the All rights reserved. No reuse allowed without permission. author/funder, who has granted medRxiv a license to display the preprint in perpetuity. The copyright holder for this preprint (which was not peer-reviewed) is the 
-------------------------------------------------------


NEW PAPER


2019 novel Coronavirus (2019-nCov) had typical clinical manifestations (fever and cough) ,and presented with characteristic chest CT imaging features (multiple lesions in both lungs, often accompanied by GGO, vascular enlargement and cobblestone/reticular pattern), which are helpful to the radiologist in the early detection and diagnosis of this emerging global health emergency. In addition, changes in these main CT features can indicate development of the disease.
Background and Objective: WHO Director-General declared that the 2019-nCoV outbreak constitutes a Public Health Emergency of International Concern,and the outbreak is still on-going.Chest CT had been a key component of the diagnostic workup for patients with suspected infection. In this retrospective study, we attempt to summarize and analyze the chest CT features of 2019-nCov infections, and to identify the typical features to improved the diagnostic accuracy of new coronavirus pneumonia (NCP). Methods:Chest CT scans and Clinical data of 21 patients confirmed NCP in our hospital were enrolled.These patients were divided into mild and sever group according to clinical manifestations described by the 6th clinical practice guideline of NCP in China. Main clinical and chest CT features were analyzed and identify. Results: Fever (85.7%) and cough (80.9%) were the two main symptoms of NCP patients.More significantly higher incidence (85.7%) of shortness of breath in the severe cases. Multiple lesions in both lungs and with incidence of GGO(100%),vascular enlargement (76.5%) and cobblestone/reticular pattern(70.6%) were the major feature.The incidence of consolidation, mixed pattern and vascular enlargement features were up to 100% in the severe group, significantly higher than that of patients in mild group. In addition, the incidence of air-bronchogram, dilated bronchi with thickened wall and fibrosis in the severe group was significantly higher than that in the mild group. Conclusions: Fever and cough are the typical clinical features of NCP patients, and chest CT mainly manifested as multiple lesions in both lungs, often accompanied by GGO, vascular enlargement and cobblestone/reticular pattern.Changes in these main CT features can indicate development of the disease Keywords: 2019-nCov; New coronavirus pneumonia; Chest CT; Ground glass opacification The atypical pneumonia case, caused by a novel coronavirus (2019-nCov), was first reported and confirmed in Wuhan, China in December 31, 2019 (Pneumonia of unknown cause-China,2020) [1] . Initially confined to Wuhan, the infection has spread elsewhere, sporadic cases exported from Wuhan were reported in many foreign countries.On January 30, the WHO Director-General declared that the 2019-nCov outbreak constitutes a Public Health Emergency of International Concern [2] . So far, the outbreak is still on-going, the number of reported cases of 2019-nCov infection continue to increase.Several relevant literatures estimated the transmissibility of 2019-nCoV, the mean estimate of R0 for the new coronavirus pneumonia (NCP) ranges from 1.4 to 3.58 [3, 4] . As of 18 February 2020, NCP has caused 73400 confirmed cases including 1870 deaths, 6242 suspected cases, and 12734 cured cases in mainland released by the National Health Commission of China. Now Japan has become the second most affected country with 519 confirmed cases, and followed by Singapore with 77 confirmed cases. Person-to-person transmission of NCP has been confirmed and with several ways of transmission, moreover the asymptomatic individual is identified as a potential source of infection. Fever and cough were reported as main clinical manifestations of NCP patients [5] . Based on the clinical observations and summary, the main diagnostic criteria [6] of 2019-nCov infections are : 1) Contact history of the epidemic area and related patients; 2) Nucleic acid is positive; 3) Fever, cough, shortness of breath and other clinical symptoms; 4) Laboratory examination showed that white blood cells were normal or decreased and lymphocytes decreased; 5) Lung lesions. Especially,the 1th and 2th standards were the most important for 2019-nCov infected confirm. However, in the diagnosis and treatment process, it was found that the above diagnostic criteria could not fully meet the needs of clinical prevention and control, and some atypical clinical manifestations appeared [5, 7, 8] , which brought great difficulties and challenges to the prevention and control of the epidemic. Due to the large and complex population movements, the diversity of transmission pathways [5, 8, 9] (droplets,contact, fecal-oral and aerosol pathways),and natural and intermediate hosts are not fully understood, this leads to ambiguous contact history information. In addition, some patients do not have typical symptoms of fever and cough, and there may be basic lesions or complex infections that lead to the lack of changes in the number of typical white blood cells and lymphocytes, which all increase the difficulty of clinical diagnosis. At present, nucleic acid detection was still as gold standard to confirmed NCP, but the sensitivity and specificity of NCP was not evaluated [10] .According to current clinical observation, the positive rate is only 30%-50%. It has been found that patients with high clinical suspicion still have negative viral nucleic acid test for 2 times, while chest CT has displayed obvious lesions [11] . Unlucky, at the beginning, these patients failed to timely and effectively prevent and control due to negative nucleic acid test, leading to the expansion of the infection. CT diagnosis specificity is not as good as nucleic acid, but its sensitivity is significantly higher than nucleic acid. Therefore, to fight the epidemic front in Wuhan, doctor Zhang urgently called for " NPC diagnosis should not blindly believe in nucleic acid testing, chest CT findings should take as the main diagnostic basis". Therefore, the fifth edition of NCP diagnostic All rights reserved. No reuse allowed without permission. the author/funder, who has granted medRxiv a license to display the preprint in perpetuity. The copyright holder for this preprint (which was not peer-reviewed) is . https://doi.org/10.1101/2020.03.03.20030775 doi: medRxiv preprint criteria took chest CT as the main diagnostic basis [6] . Radiologists will encounter NCP patients with the number of patients continues to increase. Therefore, radiologists must be familiar with the CT features of NCP. In this study, we attempt to summarize and analyze the chest CT imaging data of patients confirmed with 2019-nCov infections, and to identify the typical imaging features to improved the diagnostic accuracy of NCP, and to reduced missed diagnosis and misdiagnosis. From January 20, 2020 to February 6, 2020, 21 patients in the First people's hospital of Foshan in Guangdong Province, China, with confirmed 2019-nCov were enrolled. these patients were divided into mild and sever group according to clinical manifestations described by the 6th clinical practice guideline of NCP in China [6] . All patients were underwent chest CT examination on an Aquilion ONE scanner (Toshiba Medical Systems;Tokyo,Japan), and imaged with 1mm slice thickness, Max 512×512. All patients were positive for 2019-nCov via laboratory testing of respiratory secretions. Our institutional review board approved this retrospective study analyzing existing patient data with patient information de-identified and all subjects gave written informed consent. And this study involved no potential risk to patients. The HRCT findings of NCP, according to previously described [9] and clinical observation, each of the 21 patients, the initial chest CT findings were categorized the predominant pattern as followed: ground glass opacification (GGO),consolidation, cobblestone/reticular pattern; mixed pattern (combination of consolidation, GGO and reticular opacities); air-bronchogram; vascular enlargement in lesion; nodule (D<30mm) and thorn pear signs; slightly dilated bronchi with thickened wall; fibrosis,etc. The the main HRCT characteristics and related exposition were showed in Fig.1 . In addition, the distribution characteristics of lesions, the extent of pulmonary involvement, mediastinal lymph node status and pleural effusion were also analyzed. All CT images were reviewed on PACS by two cardiothoracic radiologists with 10 years ( Dr. Xu) and 8 years (Dr.Wu) of clinical experience, respectively. Each patient's chest CT findings was reviewed independently and final decisions reached by consensus. If the interpretations of features extracted by two primary radiologist are inconsistent, a third fellowship trained senior radiologist with 25years of experience (Dr.Pan) adjudicated a final decision. The calinical manifestations of NCP patients, such as fever,cough, shortness of breath, weak,diarrhoea and muscle pain were analyzed. All rights reserved. No reuse allowed without permission. the author/funder, who has granted medRxiv a license to display the preprint in perpetuity. The copyright holder for this preprint (which was not peer-reviewed) is . https://doi.org/10.1101/2020.03.03.20030775 doi: medRxiv preprint We used statistical software SPSS 2.0. The differences among the constituent ratios of various signs in NCP patients were analyzed with the X 2 test. The differences of the constituent ratios of various signs between mild group and severe group were also analyzed with X 2 test test. Differences of p<0.05 were considered statistically significant. Table 1 shows the clinical characteristics of 21 patients with NCP. 21 patients all had clear contact history, including 10 men and 11 women with average age was 43.1 years. Fever (85.7%) and cough (80.9%) were the two main symptoms, followed by shortness of breath (85.7%) and weak (28.6%).Besides,diarrhoea and muscle pain were rare. Compared with patients in mild group, male proportion and onset age of patients in serve group were both higher (p<0.05). Fever and cough were found in 100% of the severe cases, but there was no significant difference between these two groups. The incidence of shortness of breath was 85.7% in the severe cases, significantly higher than 21.4% in the mild cases. Lung lesions occurred in 17 (80.9%) patients with NCP, while normal lungs occurred in 4 (19.1%) patients. Among the 17 patients with pulmonary lesions, 14 (82.4%) cases showed multiple lesions in both lungs ( Fig.2A) . Besides,10 patients had lesions distributed in the peripheral zone (subpleural)( Fig.2A) , and 7 cases showed both peripheral zone and medium-to internal zone were all involved.15 (88.2%) cases showed simultaneous involvement of multiple pulmonary lobes (≥2) (Fig.2B) , among them 12 cases (70.6%) with more than 4 lobes were involved.6 patients (85.7%) in serve group displayed 4-5 lobes were involved simultaneously, which was significantly higher than that in the mild group (60% ). However,there was no difference in lesion location and whether bilateral lung involvement between these two group. Table 2 shows the manifestations of various major chest CT features in 17 patients with NCP. GGO, Vascular enlargement and Cobblestone/Reticular pattern were the three major feature of the disease (Fig.3) , with an incidence of 100%, 76.5% and 70.6%, respectively and followed by Consolidation and Mixed pattern, accounting for 52.9%.In addition, mediastinal lymphadenopathy and hydrothorax had the lowest incidence of 17.6% and 5.9%, respectively. All rights reserved. No reuse allowed without permission. the author/funder, who has granted medRxiv a license to display the preprint in perpetuity. The copyright holder for this preprint (which was not peer-reviewed) is . https://doi.org/10.1101/2020.03.03.20030775 doi: medRxiv preprint Table 3 shows the manifestations of various CT features between the two groups. GGO were present in all patients(100%) in both groups, and Cobblestone/Reticular pattern occurred in mild and sever group were 60% and 85.7%, respectively, without significant different. The consolidation, Mixed pattern and Vascular enlargement features were up to 100% in the severe group, significantly higher than 20%, 20% and 60% in the mild group. Consolidation and Mixed pattern had significant differences between the two groups (p =0.002). In addition, the incidence of Air-bronchogram, Dilated bronchi with thickened wall and Fibrosis in the severe group was significantly higher than that in the mild group (10%). Nodule/Thorn pear signs were found 50% in mild cases, significantly higher than those in severe cases (p<0.01). Mediastinal lymphadenopathy and hydrothorax were rare,but both occurred in severe cases. Fig.4 and Fig.5 showed the CT features of mild case and serve case underwent NCP. All 21 patients in our study all had a history of recent travel to Wuhan or directly contact with NCP patients.This results was convincing proof that the disease is characterized by human-to-human transmission. A new study revealed that the reproductive number of COVID-19 is up to 3.28 [3] . However, in this study, the patients did not be acquired a clear transmission route, because the virus may have multiple transmission modes [12, 13] . Previous study demonstrated that common symptoms at onset of the illness were fever, cough and shortness of breath and with less common symptoms were sputum production, headache and diarrhoea etc [4, 5, 7, 12, 13] . Our results revealed that main symptom of NCP patients manifested as fever (85.7%), cough (80.9%), followed by shortness of breath(42.9%)and fatigue(28.6%),diarrhea and muscle pain,which was consistent with previous studies.However,a large clinical cases(1099 cases) study by professor zhong et al [7] found that only 43.8% showed fever in the early stage, but up to 87.9% showed fever after hospitalization.It dmonstrated that fever is closely related to the development of the disease, therefore fever as a screening indicator may miss a large number of early patients. In addition,our study and all previous showed that diarrhea was rare, and we found that diarrhea only occurred in serve patients. Therefore, it is necessary to pay attention to the damage of digestive tract function in these patients and be alert to the risk of fecal-oral infection. 85.7% of severe cases manifested shortness of breath, were significant higher than that of mild cases. Shortness of breath would be a better indicators to reflect the disease's progression. A new study [12] base on 72314 patients revealed that middle-aged(30-79 years old) and elderly men (51.4%) were prone to the NCP,and looking at the sex ratio, men are more likely to die (2.8%) than women (1.7%). our results showed NCP patients with average 43.1 years old, and men are more likely to be worse(57.1%), showed similar characteristics. Lung lesions in NCP patients mainly manifested as subpleural multiple lesions in bilateral lung [14, 15] ,our results showed 82.4% cases displayed multiple lesions in All rights reserved. No reuse allowed without permission. the author/funder, who has granted medRxiv a license to display the preprint in perpetuity. The copyright holder for this preprint (which was not peer-reviewed) is . https://doi.org/10.1101/2020.03.03.20030775 doi: medRxiv preprint bilateral lungs, and all cases showed significant peripheral distribution characteristics, wich was consistent with previous studies [14] [15] [16] .We speculated that the main reasons for the distribution characteristics of NCP as followed: 1)The 2019-nCov infect alveolar epithelial cells through angiotensin-converting enzyme 2 (ACE2) and cause acute lung injury [17] .The peripheral lobe of the lung is relatively well developed, abundant alveolar epithelial cells may express more ACE2. 2) The ventilation function of alveolar tissue in the peripheral part of the lung is relatively weak, which may more easily lead to the aggregation of high concentration of virus particles and the onset of disease. In addition, some scholars believe that virus particles are extremely small (2019-nCov is about 120nm), which can directly enter the distal alveoli with respiration. In this study, CT features of NCP patients mainly displayed GGO (100%), vascular enlargement sign (76.5%) and cobblestone/reticular pattern (70.6%), as followed by consolidation and mixed pattern (GGO and consolidation) with about 52.9%. While mediastinal lymph nodes and pleural effusion were rare, which was basically consistent with literature reports [14, 15] .In general, the lung characteristics of NCP patients were mainly associated with pulmonary interstitial inflammation, which was similar to Middle East respiratory syndrome(MERS) [18] and Severe Acute Respiratory Syndrome(SARS) [19] caused by coronavirus. In fact, the pathological characteristics of NCP were confirmed to be similar to SARS and MERS [20] Previous reports revealed that GGO as the key CT feature.It can exists for a long time throughout the course of the disease, and may be unique CT feature in some early patients [15, 16] . In the recent study, GGO feature was observed in all NCP patients. Pathological basis of GGO is mainly diffuse alveolar injury, accompanied by cell fibromyxoid exudate,and lung transparent membrane formation. GGO at the early stage could be completely absorbed and improved after effective treatment [21] . With the development of the disease, the interlobular septa thickened on the basis of GGO, and showed typical cobblestone/reticular pattern,and when increased exudation in the alveoli and alveolar collapse resulted in consolidation. Compared with the mild group patients, the incidence of consolidation, mixed pattern, air-bronchogram, dilated bronchi with thickened wall and fibrosis significantly increased in severe patients. In addition, cobblestone/reticular pattern and vascular enlargement sign were both common findings in severe patients without significant different. It is suggested that the changes of pulmonary imaging characteristics can effectively reflect the changes of lung tissue and functional injury.Vascular enlargement in or around the lesions were considered to be an important features for NCP diagnosis,76.5% patients in our study presented this features. This may be due to inflammatory vascular congestion, or to thickening or fluid buildup in the perivascular interstitium [20] . Additionally,all these findings suggest the presence of fibrosis, which also presented commonly in patients underwent SARS and MERS.In this study, consolidation and mixed manifestations were observed in 52.9% of NCP patients, and the incidence rate was up to 100% in severe patients. Combined with previous literature results, lung tissue appeared multiple consolidation with the aggravation of the disease. That's the main difference from MERS [18] . 6 patients (85.7%) in serve group displayed 4-5 lobes were involved simultaneously, which was significantly higher than that in the mild group (60% ).Our result demonstrated that almost all lobes involvement exists in severe patients, which well explains the pathologic basis for the high incidence of dyspepsia. It was found that CT can not only detect early and diagnose NCP accurately, All rights reserved. No reuse allowed without permission. the author/funder, who has granted medRxiv a license to display the preprint in perpetuity. The copyright holder for this preprint (which was not peer-reviewed) is . https://doi.org/10.1101/2020.03.03.20030775 doi: medRxiv preprint but also follow up the change trend of CT features can effectively predict the outcome of NCP patients. The main limitations of this study, firstly, the small number of NCP cases enrolled in this study and additional study cases are needed in the following study. Besides, This study lacks patient follow-up data, so further follow-up is needed to better observe the relationship between CT signs and clinical course. Fever and cough are the main clinical features of NCP. CT showed multiple peripheral lesions in bilateral lungs, mainly characterized by GGO, paving stones and vascular thickening. There were some differences in CT manifestations between mild cases and severe cases. In severe cases, pulmonary consolidation, mixed consolidation, slight bronchial dilatation and thickening, bronchial gas image and fibrosis were significant, which were consistent with NCP as pathological manifestations of pulmonary interstitial lesions. According to the characteristics of chest CT, early accurate diagnosis and prognosis of NCP can be achieved. Figure1. The key chest CT findings of NCP. . 57-year old man with history of recent travel to Wuhan, presenting with fever, cough and shortness of breath. Axial thin-section non-contrast CT scan shows diffuse bilateral confluent and patchy ground-glass, air-bronchogram and consolidation, characterized by peripheral distribution. Multiple lobes were involved. All rights reserved. No reuse allowed without permission. the author/funder, who has granted medRxiv a license to display the preprint in perpetuity. The copyright holder for this preprint (which was not peer-reviewed) is . https://doi.org/10.1101/2020.03.03.20030775 doi: medRxiv preprint Figure 3 . 76-year old man with history of recent travel to Wuhan, presenting with fever and productive cough. Axial thin-section non-contrast CT image shows diffusion lesions in bilateral lung, mainly manifested as ground-glass opacification (red arrow and red line around region),and cobblestone/Reticular pattern( blue arrow and blue line around region). Besides,Vascular enlargement sign was observed in the lesions(yellow arrow). the author/funder, who has granted medRxiv a license to display the preprint in perpetuity. The copyright holder for this preprint (which was not peer-reviewed) is . https://doi.org/10.1101/2020.03.03.20030775 doi: medRxiv preprint Figure 5 . 68-year old female as serve NCP patients with history of recent travel to Wuhan, presenting with fever and productive cough. Axial thin-section non-contrast CT image shows diffusion lesions in bilateral lung. A showed many CT features such as ground-glass opacification,cobblestone/reticular pattern(blue arrow and bule line around region), frbrosis and dilated bronchi with thickened wall (the enlarged image in the upper right corner, red arrow). B showed consolidation (pink arrow) and mixed pattern (yellow arrow and yellow line around region) in the bilateral lower lobes. the author/funder, who has granted medRxiv a license to display the preprint in perpetuity. The copyright holder for this preprint (which was not peer-reviewed) is . https://doi.org/10.1101/2020.03.03.20030775 doi: medRxiv preprint GGO: Ground glass opacification; Con: Consolidation; Mix-P: Mixed pattern; Cob/Ret: Cobblestone/Reticular pattern; Air-Bro:Air-bronchogram; VEs:Vascular enlargement sign; Nod/Tho-: Nodule/Thorn pear signs; Dil-Bro: Dilated bronchi with thickened wall; Fib: Fibrosis; Med-Hym: Mediastinal lymphadenopathy; Hyd: Hydrothorax Table 3 . The Differences of major CT Characteristics between two groups. the author/funder, who has granted medRxiv a license to display the preprint in perpetuity. The copyright holder for this preprint (which was not peer-reviewed) is . https://doi.org/10.1101/2020.03.03.20030775 doi: medRxiv preprint
-------------------------------------------------------


NEW PAPER


Retrospective case series.
Coronavirus disease 2019 (COVID-19) is highly contagious and spreads rapidly through human-to-human transmissions. 5, 6 As of 20 February, 75 571 confirmed cases had been reported in Mainland China and 1083 confirmed cases in 24 other countries and regions. Amongst the cases in Mainland China, 2239 cases died, 11 cases of death was reported from out of Mainland China. 7 There were 41 initial cases of COVID-19 that were directly or indirectly linked to the Wuhan Huanan Seafood Wholesale Market, as reported by Huang and colleagues. 8 The clinical features include fever, dry cough, dyspnea, myalgia, fatigue, decreased leukocyte counts and computed tomography (CT) evidence of pneumonia. 8 Subsequently, Chen and colleagues reported 99 cases from a single centre of Wuhan, but the severe and non-severe cases were not compared. 9 And then Wang and colleagues published a study based on 138 hospitalized patients from Wuhan. 10 However, all these studies are based on cases identified in Wuhan. Recently, Guan and colleagues reported 1099 cases with laboratory-confirmed SARS-CoV-2 from 552 hospitals across 31 provinces/provincial municipalities. 11 They reported that the median age was 47.0 years, 41.90% were females, 31.30% had been to Wuhan and 71.80% had contacted people from Wuhan, and the average incubation period was 3.0 days. The most common symptoms were fever and cough. 11 Together, several articles about cases from Wuhan have reported epidemiologic and clinical manifestations and provided important initial background upon which we seek to furnish further in this paper. [8] [9] [10] 12 The clinical features of COVID-19 of cases outside Wuhan is still largely unknown. Zhejiang Province has consistently been one of the top three provinces with most cases in China and therefore provide a good basis to learn how COVID-19 spreads outside Wuhan. The cluster events, public transport transmissions, and clinical diagnosis of this new infectious disease are of great importance to be reported as it shred lights on of cases that occur outside Wuhan. Here, we report the epidemiologic and clinical characteristics of COVID-19 patients from five hospitals in Zhejiang province, China. We performed a retrospective, multicentre study on the epidemiologic history, clinical records, laboratory results, and chest radiological features of 88 laboratory-confirmed and 3 clinical-diagnosed patients with COVID-19 that were diagnosed from 20 January 2020 to 11 February 2020. Final follow-up for this report lasted until 16 February 2020. The primary method of diagnosis is to perform real-time reversetranscriptase polymerase-chain-reaction (RT-PCR) assay test using throat swab specimens that were collected from upper respiratory tracts. This test is done twice at 24-hour interval. Of the 91 cases reported in here, 88 cases were tested positive for SARS-CoV-2 at least once. These laboratory confirmation assays for SARS-CoV-2 were performed at CDCs of various cities and at Ningbo First Hospital following the standard protocol. 8 Three further cases were reported in Ningbo cohort as clinical-diagnosed COVID-19 pneumonia because of their epidemiological history, signs, symptoms and chest CT evidence according to National Health Commission of the People's Republic of China guidance, though they tested negative for the SARS-CoV-2. 13,14 The incubation period was defined as the time from the exposure to the confirmed or suspected transmission source to the onset of illness. A team of doctors who had been treating these patients extracted the medical records of these patients and sent the data to working group in Ningbo to further examine. When the data were not clear or missing, the working group in Ningbo would clarify the details with the doctors in charge of treating these patients. The study has been reviewed and approved by the Medical Ethical Committees (2020-R018). The requirement for written informed consent was waived because of the urgent need to collect clinical data and no harm could potentially be done to patients. Doctors who treated the patients collected and recorded the epidemiological characteristics by interviewing each patient on their activity history during the two weeks before symptoms onset or admission into hospital. All patients underwent chest CT scans. The clinical symptoms, chest CT and laboratory findings on admission were extracted from electronic medical records. Laboratory results included blood routine, blood chemistry, arterial blood gas, fibrinogen, liver and renal function, electrolytes, C-reactive protein, and procalcitonin. Patients were divided into the diagnosed as severe group and mild group according to national treatment guideline. 13,14 Questionnaires of MuLBSTA scores 15 were recorded by attending physicians according to six indexes, which are multilobular infiltration, lymphopenia, bacterial co-infection, smoking history, hypertension, and age. All data were checked by two experienced physicians (GQ and NY). We present the summary statistics of continuous variables using the means and standard deviations (SD) or median (IQR), comparison across groups were performed using the Mann-Whitney U test. Categorical variables were expressed as the counts and percentages in each category. Chi-square tests was used for categorical variables as appropriate. All analyses were analyzed by IBM SPSS statistics version 26.0. This was a retrospective case series study, no patients and public were involved in the design, or conduct, or reporting, or dissemination plans of our research. Of all 91 patients recruited as of 11 February, we detected 88 (96.70%) laboratory-confirmed COVID-19 pneumonia with throat swab samples that positive for SARS-CoV-2 and 3(3.30%) clinical-confirmed COVID-19 pneumonia since they had definite demographic history, typical symptoms and chest CT images. As shown in Table 1 , The median age of the 91 patients was 50 years (IQR, 36.5 to 57.0), ranged from 5 to 96 years. There was 1 child (5 years . CC-BY-NC-ND 4.0 International license It is made available under a is the author/funder, who has granted medRxiv a license to display the preprint in perpetuity. The copyright holder for this preprint . old), 1 student who were 17-years of age, and six cases of elderly patients aged 70 or above. Adults account for most of the cases, with breakdown as follows: aged 18-39 (26 cases, 28.57%), aged 40-49 (16 cases, 17.58%), aged 50-59 (28 cases, 30.77%) and aged 60-69 (13 cases, 14.29%). There were 54 female patients (59.34%), with none of them pregnant. As for epidemiologic characteristics, 31 patients had been to Wuhan/Hubei within the past two weeks (G0). Local cases are defined as follows: 8 patients had within the past two weeks had contact with personnel who had been in Wuhan/Hubei, these are the first generation of local cases (G1). Following that 40 patients contracted the disease after having had contact with the local cases, these are defined as G2 local cases. In particular 23 of the G2 cases were all derived from an event where people collectively prayed for good luck in a temple for the new lunar year. There were 11 patients who were infected while air travelling. It is impossible to establish how the remaining 1 patient contracted the disease. None of the patients had a history of exposure to the Huanan seafood wholesale market. No healthcare workers were found in our study. Moreover, none of the health workforce in  On admission, the most common symptoms were fever (65, 71.43%), cough (55, 60.44%) and fatigue (40, 43.96%) ( Table 2 ). Among them, 34 (37.36%) cases have temperature between 38.1-39℃, but none had a very high fever (body temperature > 41℃). Other symptoms included expectoration (30, 32.97%), anorexia (23, 25.27%), diarrhea (21, 23.08%), neausea (11, 12 .09%) and vomiting (6, 6.59%). In a case, her abdominal discomfort was the only symptom. Nine patients were diagnosed as severe pneumonia because of the development of pneumonia. As of 16 February, 60 (65.93%) patients were still isolated in our hospitals, and 31 (34.07%) patients had been discharged and no patients had died so far. The median of incubation period is 6 (IQR, 3-8) days, and number of days from first visit to a doctor till the case is confirmed is 1 (1-2). . CC-BY-NC-ND 4.0 International license It is made available under a is the author/funder, who has granted medRxiv a license to display the preprint in perpetuity. The copyright holder for this preprint According to data of laboratory tests, half of patients (49, 53.85%) demonstrated elevated levels of C-reactive protein (CRP), but elevated levels of procalcitonin were detected in only a minority (14, 15.39%). Data from blood routine showed that 14 (15.39%) patients had below the normal range of leucocyte, and 3 (3.97%) cases with elevated levels ( Table 3) The serum level of albumin was suppressed in 43 (47.25%) patients. ALT (7, 7.69%) and AST (9, 9.89%) were slightly elevated in minority of cases. Suppressed calcium concentration and serum sodium concentration were less common. Some patients had abnormality in arterial blood gas analysis, with decreased PaO2 in 18 cases, increased levels of lactic acid in 25 cases, and increased PH in 14 cases. Our study has stratified patients with COVID-19 pneumonia based on the severity of symptoms on admission according to national guidelines. 13, 14 Nine patients were diagnosed as severe pneumonia because of the development of pneumonia. Compared with those mild patients, patients diagnosed as severe pneumonia had numerous differences in laboratory results (Table 4) , including higher neutrophil, lower lymphocytes count, hyponatremia, hypocalcemia, as well as higher levels of CRP. MuLBSTA score is used to predict mortality in viral pneumonia, amongst the 9 severe pneumonia patients their MuLBSTA scores were significantly higher, indicating they had higher mortality risks. This report, to our knowledge, is the largest case study to date of hospitalized patients with COVID-19 in Zhejiang province, which is outwith of Wuhan and Hubei. As of 16 February, 2020, of the 91 cases in the present study, 9 were diagnosed as severe pneumonia, 31 were discharged, 60 . CC-BY-NC-ND 4.0 International license It is made available under a is the author/funder, who has granted medRxiv a license to display the preprint in perpetuity. The copyright holder for this preprint The most common symptoms at the onset of COIVD-19 were fever, cough, and fatigue. The median of incubation period was 6 (IQR, [3] [4] [5] [6] [7] [8] days and from first visit to a doctor to confirmed diagnosis was only 1 (1-2) days. Our study provided further evidence that rapid human-to-human transmission of SARS-CoV-2 outside Wuhan had occurred. None of the patient in this study had direct contact with wildlife or from Huanan Seafood Wholesale Market. While 8.79% had contact with people who had travelled from Wuhan, more than 52.75% were local Zhejiang residents with no travel history to Wuhan and Hubei. In particular, 23 cases of 38 cases of Ningbo Cohort were related to the outbreak of a temple cluster, including 11 patients had directly participated in a temple activity. Our findings have provided strong evidence that large social events should be cancelled in order to prevent infection. Notably, there is 54 (59.34%) female cases in our study compared to 27% of the first reported study 8 or 41.8% of a recently study 11 , and 44% of a recently study from Zhejiang Province. 16 However, it has been found that more males were infected by Middle East respiratory syndrome (MERS)-CoV and SARS-CoV. 17, 18 Whether it is related to endocrinology 9 , social activities, or religious activities still needs further research. Fever (71.43%), cough (60.44%) and fatigue (43.96%) are the most comment clinical characteristics in our study and is similarly to the cohorts reported in the published literature. 8, 10, 16 Furthermore, it is reported only 43.8% of COVID-19 had fever onset and 87.9% reported having had fever during hospitalization. 11 Fever is less widespread amongst those infected with COVID-19 than those with SARS-CoV (99%) and MERS-CoV (98%). 19 As screening heavily on fever detection, patients that do not have fever might get suspected as being wrongly diagnosed. Some COVID-19 cases had atypical symptoms or were asymptomatic. Furthermore, asymptomatic persons are potential sources of SARS-CoV-2 transmission. 20 It appears that transmission is possible during the incubation period, and the carrier cannot be spotted. 20 Recently, a study reported that they detected SARS-CoV-2 in stool samples from patients with abdominal symptoms. 21 Interestingly, we had detected SARS-CoV-2 in rectal swab of . CC-BY-NC-ND 4.0 International license It is made available under a is the author/funder, who has granted medRxiv a license to display the preprint in perpetuity. The copyright holder for this preprint . https://doi.org/10.1101/2020.02.23.20026856 doi: medRxiv preprint patients who were twice tested negative by throat swab specimens using RT-PCR. Therefore, it is helpful to detect atypical symptoms COVID-19 using both throat swab and rectal swab, particularly for patients who suffer symptoms with their digestive tract, like diarrhea. Only 28 (30.77%) cases had lymphopenia in our study, while half the patients (49, 53.85%) demonstrated elevated levels of C-reactive protein, decreased platelet in 10 cases. Most cases had normal serum levels of procalcitonin on admission (procalcitonin <0.04 ng/ml), except 14 cases had procalcitonin level higher than 0.04 ng/ml. All patients were scanned by chest CT scan. Of the 91 patients, 61 (67.03%) had multilobe infiltration. The typical manifestation of lung CT were ground-glass opacification (GGO) of bilateral multiple lobular and subsegmental areas of consolidation (Figures 1 and 2 ). Our study provided three cases as clinical-confirmed COVID-19 pneumonia because of their epidemiological history, signs, symptoms and chest CT evidence according to guidance, though they tested negative for the SARS-CoV-2. Thus, CT could be used as an important differential method in cities where they are lacking nucleic acid kits, like Wuhan, and help to bring forward the isolation period for the infected. Currently, no anti-viral agents have been proven to be effective treatment for COVID-19. It is reported that combination of Lopinavir and Ritonavir had been applied to SARS-CoV-2 patients with substantial clinical benefit. 22 As an emerging virus, all the patients received anti-viral agents in Ningbo cohort, including Kaletra (Lopinavir/Ritonavir) and Umifenovir, 9 cases received methylprednisolone to treat high fever, SpO2 ≤93% and hypoxiemia. The dose of methylprednisolone depends on disease severity between 1mg/kg to 2mg/kg. However, the outcome is still unclear. As of 16 February 2020, no death has been reported in Zhejiang province as the government authority has taken unprecedented and effective effort to reduce the risk of transmission. Early diagnosis, early isolation and early management all contributed to reducing transmission and mortality in Zhejiang. However, 60 patients are still hospitalization and their cases should be followed up in the future. Our study has some limitations. Firstly, two thirds of patients are still hospitalized. For those who had been released their cases are being followed up for 2 weeks, and further information can be learnt from these individuals. Secondly, we just have recruited 91 cases. Increasing the number of cases is good for observing COVID-19 outside of Wuhan/Hubei. Thirdly, we did not . CC-BY-NC-ND 4.0 International license It is made available under a is the author/funder, who has granted medRxiv a license to display the preprint in perpetuity. The copyright holder for this preprint . https://doi.org/10.1101/2020.02.23.20026856 doi: medRxiv preprint have many severe cases to compare the differences of epidemiological and clinical features. . CC-BY-NC-ND 4.0 International license It is made available under a is the author/funder, who has granted medRxiv a license to display the preprint in perpetuity. The copyright holder for this preprint . . CC-BY-NC-ND 4.0 International license It is made available under a is the author/funder, who has granted medRxiv a license to display the preprint in perpetuity. . CC-BY-NC-ND 4.0 International license It is made available under a is the author/funder, who has granted medRxiv a license to display the preprint in perpetuity. The copyright holder for this preprint . . CC-BY-NC-ND 4.0 International license It is made available under a is the author/funder, who has granted medRxiv a license to display the preprint in perpetuity. (which was not peer-reviewed) The copyright holder for this preprint . https://doi.org/10.1101/2020.02.23.20026856 doi: medRxiv preprint 
-------------------------------------------------------


NEW PAPER


Corona Virus Disease 2019 has spread rapidly to more than 70 countries and regions overseas and over 80000 cases have been infected, resulting in more than three thousand deaths. Rapid diagnosis of patients remains a bottleneck in containing the progress of the epidemic. We used automated chemiluminescent immunoassay to detect serum IgM and IgG antibodies to 2019-nCoV of 736 subjects. COVID-19 patients were becoming reactive(positive) for specific antibodies from 7-12 days after the onset of morbidity. Specific IgM and IgG increased with the progression of the disease. The areas under the ROC curves of IgM and IgG were 0.988 and 1.000, respectively. Specific antibody detection has good sensitivity and specificity.
Coronaviruses (CoVs) are enveloped single-stranded positive-sense RNA viruses, which are widely distributed in humans and other mammals. Coronaviruses usually cause respiratory, digestive and nervous system diseases in humans and animals (1) . In the past 20 years, coronavirus has caused two global epidemics of severe respiratory infectious diseases, one of which was severe acute respiratory syndrome (SARS) (2, 3) the author/funder, who has granted medRxiv a license to display the preprint in perpetuity. The copyright holder for this preprint (which was not peer-reviewed) is . https://doi.org/10.1101/2020.03.04.20030916 doi: medRxiv preprint swab (5-9). However, in the actual diagnosis and treatment, the sensitivity of nucleic acid detection was not ideal enough. Only 30-50% of the confirmed COVID-19 cases had positive results, moreover, in some confirmed case, nucleic acid testing often took four or more tests to get a positive result. It is necessary to use a fast and convenient method to realize the rapid diagnosis of 2019-nCoV infection. After the virus infects the organism, the immune system carries on the immune defense to the virus and produces the specific antibody. In the laboratory diagnosis of infectious diseases, the detection of specific antibodies to pathogens is a sensitive method for fast diagnosis. However, how the 2019-nCoV antibody produced and changed during COVID-19 progression is still unclear. In this study, we used automated chemiluminescent immunoassay to detect serum IgM and IgG antibodies to 2019-nCoV, to understand the process of antibody production in disease progression, and to evaluate the value of antibody detection in the laboratory diagnosis of COVID-19. The study was conducted in accordance with the International Coordinating Council for Clinical Trials and the Helsinki Declaration, and was approved by the Hospital Ethics Review Committee (Ethics No 2020PS038K), and the patient's informed consent was exempted. All rights reserved. No reuse allowed without permission. the author/funder, who has granted medRxiv a license to display the preprint in perpetuity. The copyright holder for this preprint (which was not peer-reviewed) is Another 222 outpatients with other diseases in the same period, 63 medical staffs worked for fever clinic and 223 healthy physical examinees in 2018 were collected and were named other disease group, medical staff group and health control group, respectively. According to the unified form, two residents collected clinical data from medical records separately. Blood sampling Fasting venous blood (5ml) was collected from all the subjects and put into the yellow head vacuum tube containing separation gel. After centrifugation, the serum samples were stored at-20 ℃. The CT value of 2019-nCoV nucleic acid test results should be interpreted according All rights reserved. No reuse allowed without permission. the author/funder, who has granted medRxiv a license to display the preprint in perpetuity. The copyright holder for this preprint (which was not peer-reviewed) is . https://doi.org/10.1101/2020.03.04.20030916 doi: medRxiv preprint to the recommendations of the manufacturer's instructions, and the suspicious results should be notified for clinical re-sampling and re-examination. In order to be diagnosed as positive in laboratory test results, it is necessary to meet the standard that 2019-nCoV ORF1ab and N gene of same sample shows at least one target specific RT-PCR test result is positive. The diagnosis was made according to the "Diagnosis and Treatment plan of Quantitative variables were expressed as median (P99). The normality of variables All rights reserved. No reuse allowed without permission. the author/funder, who has granted medRxiv a license to display the preprint in perpetuity. The copyright holder for this preprint (which was not peer-reviewed) is . https://doi.org/10.1101/2020.03.04.20030916 doi: medRxiv preprint was tested using Kolmogorov-Smirnov test. and LDS-t-test for comparison among groups. A receiver operating characteristic curve (ROC) was plotted to evaluate the diagnostic performance and correlations determined by Spearman's rank correlation. Z test was used to compare AUC between two groups. All tests were two sided and P values < 0.05 were considered statistically significant. Of the 3 cases of confirmed case, 2 were male and 1 was female. The age ranged from 39 to 57 years. 2 patients had diabetes and hypertension respectively. 1 was a common case, 2 were severe cases. 1 case had history of Wuhan contact and the other 2 cases had no clear epidemiological history (Table 1 ). The main laboratory findings of COVID-19 patients were normal or slightly low white blood cells and lymphocytes, elevated inflammatory indicators such as interleukin-6, procalcitonin, C-reactive protein, serum amyloid A, erythrocyte sedimentation rate; and normal myocardial markers ( Table 2 ). the author/funder, who has granted medRxiv a license to display the preprint in perpetuity. The copyright holder for this preprint (which was not peer-reviewed) is . https://doi.org/10.1101/2020.03.04.20030916 doi: medRxiv preprint showed that the level of anti-2019-nCoV IgG was continue higher than that of IgM, and the other two cases showed that the level of anti-2019-nCoV IgM increased more than that of IgG from 2 weeks of morbidity (Table 3 ). In non-COVID-19, other disease, medical staff and health control groups, there were a few cases reactive for 2019-nCoV IgM and IgG, all the cases were single reactive for IgM or IgG. The sensitivities of IgM and IgG were 100%, as for specificities of IgM and IgG were all over 97% (Table 4 ). Of 225 non-COVID-19 cases, 2 cases were detectable for influenza A RNA and 2 cases were detectable for influenza B RNA, respectively, 4 cases were detectable for adenovirus DNA, 17 cases were detectable for mycoplasma pneumonia DNA (Table   4 ). We also compared the anti-2019-nCoV antibodies values distributions in different groups. The anti-2019-nCoV IgM levels in non-COVID-19 was higher than that of healthy control group, the difference was statistically significant (Table 4 ; The area under the curve was 0.988 and 1.000, and the best cut-off value was 10.14 and 15.99, respectively ( Figure 3 ). With China's growing surveillance network and laboratory capacity, the All rights reserved. No reuse allowed without permission. the author/funder, who has granted medRxiv a license to display the preprint in perpetuity. The copyright holder for this preprint (which was not peer-reviewed) is . https://doi.org/10.1101/2020.03.04.20030916 doi: medRxiv preprint outbreak was identified within a few weeks and the viral genome sequence was announced (10), effectively promoting in vitro diagnostic tests. At present, the main diagnostic method is to detect 2019-nCoV nucleic acid by real-time quantitative fluorescent PCR. In the early stage of this epidemic, due to the insufficient production of nucleic acid detection kits and the high requirement of technical norms for nucleic acid detection, the application of nucleic acid testing as a diagnostic standard was limited, When the body's immune system produces antibodies, the clinical signs and symptoms is obvious, the virus is likely to decline without being detected. All rights reserved. No reuse allowed without permission. the author/funder, who has granted medRxiv a license to display the preprint in perpetuity. The copyright holder for this preprint (which was not peer-reviewed) is Recently, a study of 138 patients showed a high proportion (41%) of suspected nosocomial infection (11) , and it is also noteworthy that the presence of asymptomatic patients with latent mild pneumonia may be an important source of infection for outbreak transmission (12) . Therefore, it remains critical to apply a fast and convenient detection method to distinguish and trace suspicious case or contacts as early as possible in order to prevent super-transmission events. Antibodies are the products of humoral immune response after infection with viruses. As a new infectious disease, while the detection of nucleic acid cannot be used widely, specific antibodies to 2019-nCoV can be used to determine whether the patient has been recently infected with 2019-nCoV or not. It has been reported that serum samples from 5 patients were detected by self-made 2019-nCoV IgG and IgM ELISA kits, and the antigen could cover 92% of the 2019-nCoV NP amino acids (13) . Generally speaking, the immune response of pathogenic microorganisms is usually stimulated by the rise of IgM after infection, IgG usually appears 1-2 weeks after IgM, and has been rising and maintaining high levels in the body for a long time. Because COVID-19 is a new infectious disease and the immunological test reagent has just been developed, there is still no report on how IgM and IgG antibodies were produced and developed after 2019-nCoV infection. In our study, we found that specific antibodies reactive to 2019-nCoV appeared from 7-12days after the onset of morbidity in all 3 patients. Unlike previous experience that IgG usually appears 1-2 weeks after IgM, the presence of anti-2019-nCoV IgM antibodies in COVID-19 cases was followed by the presence of All rights reserved. No reuse allowed without permission. the author/funder, who has granted medRxiv a license to display the preprint in perpetuity. The copyright holder for this preprint (which was not peer-reviewed) is 2019-nCoV is highly infectious and the population is generally susceptible to 2019-nCoV.The most common symptoms after infection include fever, fatigue, dry cough, and muscle pain, with expiratory dyspnoea occurring in more than half of patients (14) . Severe cases are prone to rapidly progress to acute respiratory distress syndrome, septic shock, high risk of admission to intensive care units, and even death. Therefore, how to closely observe the condition after morbidity and find severe cases as soon as possible is the key to reduce the mortality of critically ill patients. According to our findings, it seems that the time and speed of production of specific anti-2019-nCoV IgM antibodies correlate with disease severity. But because the number of cases is so small, more research is needed to confirm it. In addition to COVID-19, the fever patient of non-COVID-19, other disease, All rights reserved. No reuse allowed without permission. the author/funder, who has granted medRxiv a license to display the preprint in perpetuity. The copyright holder for this preprint (which was not peer-reviewed) is . https://doi.org/10.1101/2020.03.04.20030916 doi: medRxiv preprint medical staffs and healthy controls were also studied. Non-COVID-19 group included several other respiratory viruses such as influenza A, B and adenovirus infection cases, these cases were negative for anti-2019-nCoV specific antibody detection, indicating that the antibody detection has a good ability to resist interference and differential Considering that COVID-19 has broken out in many countries around the world, more than 80000 people have been diagnosed and the number is growing rapidly, the main problem at present is the need for highly sensitive tests to screen the suspected cases and to prevent missed diagnosis by nucleic acid tests, lower false positive rates for antibody testing are acceptable. In the meantime, for patients with morbidity for a week or more, simultaneous positive of anti-2019-nCoV IgM and IgG will be helpful to improve the specificity. Compared with nucleic acid test, which requires respiratory tract samples and complex testing procedures, the operation requirement of serum antibody detection in clinical laboratory is lower than that of nucleic acid detection, which can be detected quickly (30min) and in large quantities, and can be completed in common P2 Biosafety Laboratory. When the morbidity is more than a week, nucleic acid detection is not convenient, serological dynamic monitoring can be carried out, once positive, it All rights reserved. No reuse allowed without permission. the author/funder, who has granted medRxiv a license to display the preprint in perpetuity. The copyright holder for this preprint (which was not peer-reviewed) is . https://doi.org/10.1101/2020.03.04.20030916 doi: medRxiv preprint is strongly recommended to use nucleic acid diagnosis immediately. The disadvantage of nucleic acid detection is the existence of relative high false negative rate, and serological antibody detection has the advantage of high sensitivity, so the combination of the two will be a good diagnostic means. It can be inferred that after the future epidemic situation has been controlled to a certain extent, as a convenient method, antibody detection is still necessary to make differential diagnosis of other respiratory pathogens infection. It must be emphasized that independent results of specific antibodies testing should not be used as a diagnostic criteria, especially when the epidemiological history is unclear, and must be combined with the patient's morbidity time and clinical signs.It must be emphasized that independent results of specific antibodies testing should not be used as a diagnostic basis, especially when the epidemiological history is unclear, and must be combined with the patient's morbidity time and clinical signs. To our knowledge, little has been reported about the specific antibody production process in the course of COVID-19 disease, and little has been reported about the different situation of antibodies in fever non-COVID-19 population, other diseases, special contact population such as medical staff and healthy population. This study provides data on the regularity of antibody production in the course of COVID-19, and provides some understanding of the basic data of specific antibodies in different populations. The results of this study help to provide evidence for rapid screening of suspected cases through the serological testing to curb the rapid progress of the epidemic globally. Just on the day of this manuscript was submitted, the China All rights reserved. No reuse allowed without permission. the author/funder, who has granted medRxiv a license to display the preprint in perpetuity. The copyright holder for this preprint (which was not peer-reviewed) is Considering that Liaoning Province, where this study was conducted, is a low epidemic area, the possibility of asymptomatic infection would be very small. In this paper, we studied the producing process of specific antibody in patients with COVID-19, compared and evaluated the diagnostic value of antibody in different populations, which is beneficial for doctors to use in the process of diagnosis and treatment. As a useful complement to nucleic acid detection，the detection of specific anti-2019-nCoV antibodies will be able to draw a more comprehensive, rapid and accurate diagnosis to COVID-19, so as to effectively distinct between COVID and non-COVID-19 patients and curb the rapid spread of 2019-nCoV in the global epidemic period. All rights reserved. No reuse allowed without permission. the author/funder, who has granted medRxiv a license to display the preprint in perpetuity. The copyright holder for this preprint (which was not peer-reviewed) is . https://doi.org/10.1101/2020.03.04.20030916 doi: medRxiv preprint We declare no competing interests. the author/funder, who has granted medRxiv a license to display the preprint in perpetuity. The copyright holder for this preprint (which was not peer-reviewed) is the author/funder, who has granted medRxiv a license to display the preprint in perpetuity. the author/funder, who has granted medRxiv a license to display the preprint in perpetuity. the author/funder, who has granted medRxiv a license to display the preprint in perpetuity. The copyright holder for this preprint (which was not peer-reviewed) is . https://doi.org/10.1101/2020.03.04.20030916 doi: medRxiv preprint Table2. Laboratory findings of 3 patients with COVID-19   Normal   range   case1  case2  case3 white blood cell count,x10 9 /L 3.5-9.5 5.0 6.3 4.0 neutrophil count,x10 9 /L 1.9-7.2 3.7 4.1 3.4 Lymphocyte count,x10 9 /L 1. the author/funder, who has granted medRxiv a license to display the preprint in perpetuity. The copyright holder for this preprint (which was not peer-reviewed) is . https://doi.org/10.1101/2020.03.04.20030916 doi: medRxiv preprint Table 3 . Anti-2019-nCoV antibody production 
-------------------------------------------------------


NEW PAPER


The ongoing outbreak of coronavirus disease 2019 (COVID-19) caused by severe acute respiratory syndrome coronavirus 2 (SARS-CoV-2) started in the end of 2019 in China has triggered a global public health crisis. Previous studies have shown that SARS-CoV-2 infects cells by binding angiotensin-converting enzyme 2 (ACE2), which is the same as SARS-CoV. The expression and distribution of ACE2 in the pancreas are unknown. At the same time, the injury of pancreas after SARS-CoV-2 infection has not been concerned. Here, we collected public datasets (bulk RNA-seq and single-cell RNA-seq) to indicate the expression and the distribution of ACE2 in pancreas (in both exocrine glands and islets). And further, clinical data including mild and severe patients with COVID-19 demonstrated there existed mild pancreatitis. In the 67 severe cases, 11 patients (16.41%) showed elevated levels of both amylase and lipase, and 5 patients (7.46%) showed imaging alterations. Only one patient (1.85%) showed elevated levels of both amylase and lipase in 54 mild cases, without imaging changes. Our study revealed the phenomenon and possible cause of mild pancreatic injury in patients with COVID-19. This suggests that pancreatitis after SARS-CoV-2 infection should also be paid attention in clinical work.
Severe acute respiratory syndrome coronavirus 2 (SARS-CoV-2) , a novel coronavirus that causes coronavirus disease 2019 in humans, began to occur in Wuhan (China) in December 2019 and has the trend of spreading around the world [1, 2, 3] . By February 28, 2020, a total of 78962 people had been infected in China with 2791 (3.53%) deaths, and 4838 people had been infected with 76 (1.57%) deaths in other countries. It was confirmed by etiology that SARS-CoV-2 belonged to the novel coronavirus of the subgenus Sarbecovirus (Beta-CoV lineage B), and had 79.5% similarity with SARS-CoV [4] , which caused the global outbreak in 2003. Spike (S) protein, one of the main structural proteins of SARS-CoV-2, binds angiotensin-converting enzyme 2 (ACE2) protein of the host cell membrane to fuse into the cell for nucleic acid replication just similar to SARS-CoV [5] . Through a variety of bioinformatics and experimental verification, ACE2 is not only expressed in alveolar epithelial cells, but also in the heart, gastrointestinal tract, kidney, testis and other organs, which means that SARS-CoV-2 is likely to enter other tissues and organs through ACE2 binding, causing multiple organ damage [6, 7] . In recent studies, SARS-CoV-2 has been found to cause heart, kidney, liver injury and gastrointestinal symptoms in addition to lung lesions such as ARDS [6] . However, there is lacking attention in pancreatic injury, which may accelerate patients' progression. In our clinical observation, we also found that some patients showed signs of pancreatic injury, such as high levels of amylase and lipase in serum and urine. . CC-BY-NC-ND 4.0 International license It is made available under a author/funder, who has granted medRxiv a license to display the preprint in perpetuity. is the (which was not peer-reviewed) The copyright holder for this preprint . https://doi.org /10.1101 /10. /2020 In this study, we used public datasets to explore the expression of ACE2 in the pancreas and various types of pancreatic cells. Combined with clinical data, we demonstrated the pancreatic injury in some patients with COVID-19, and explained for the first time that the pancreas was also a vulnerable organ after SARS-CoV-2 infection. We collected bulk RNA-seq data from GTEx (https://gtexportal.org), which contains normal tissue and organ sequencing data from multiple individuals. The data analysis was performed online. In addition, we collected data from single-cell RNA-seq (scRNA-seq) of the pancreas NCBI-GEO (GSE85241 (4 donors with 2,126 pancreatic cells), GSE84133 (4 donors with 8569 pancreatic cells)). The details of the donors are available in the previous report [8, 9] . The scRNA-seq data processing process is as follows: Unique molecular identified (UMI) expression count matrix was obtained from the database, and Seurat object was created. Further quality control was performed, cells with high mitochondrial gene expression > 5% were filtered. The data was normalized and log-transformed with the method "LogNormalize" in NormalizeData function. Before we performed linear dimensional reduction, the data was scaling. And then principal component analysis (PCA) was performed to the data and determined the dimensionality. Finally, after clustering cells based on graph-based clustering approach, the non-linear dimensional reduction based on uniform manifold . CC-BY-NC-ND 4.0 International license It is made available under a author/funder, who has granted medRxiv a license to display the preprint in perpetuity. is the (which was not peer-reviewed) The copyright holder for this preprint . https://doi.org/10.1101/2020.02.28.20029181 doi: medRxiv preprint approximation and projection (UMAP) was performed the data to analyze and visualize the data. All of scRNA-seq data analysis was based on Seurat R package (version: 3.1.4) with the default parameters [10] . The annotation of cell types was completed based on the featured genes of each cluster and the cell markers of each type pancreas cell from CellMarker database [11] and previously report [8, 9] . This is a retrospective and observational study from Wuhan Tongji hospital and Wuhan Jin Yin-tan hospital. We retrospectively analyzed patients diagnosed with COVID-19 from January 1, 2020 to February 15, 2020. The criteria for the diagnosis and severity of the patients were followed by the diagnosis and treatment scheme for COVID-19 (trial version 6) issued by National Health Commission of the People's Republic of China. We collect hospital admissions, laboratory tests, and imaging tests from clinical electronic medical records. In this study, we collected clinical information including age, gender, amylase and lipase in serum, and the imaging results including bedside ultrasound and abdominal CT. Mild COVID-19 patients with serum amylase and lipase in the normal range did not undergo the imaging evaluation of the pancreas, and their pancreas was assumed to be normal. Bedside ultrasound of the abdomen was all performed in critically ill patients, and CT was added if there were any abnormalities. . CC-BY-NC-ND 4.0 International license It is made available under a author/funder, who has granted medRxiv a license to display the preprint in perpetuity. is the (which was not peer-reviewed) The copyright holder for this preprint . https://doi.org/10.1101/2020.02.28.20029181 doi: medRxiv preprint To determine the expression of ACE2 in normal pancreas of humans, we used the GTEx database to explore ACE2 expression. We found ACE2 was expressed in many organs or tissues except most brain tissues (Fig. 1A) , which suggests that these organs or tissues such as unreported ovary and thyroid might also be targets for SARS-CoV-2. For comparison, we compared ACE2 expression in lung and pancreatic tissues, since the lung is known to be the first target organ to be attacked by SARS-CoV-2. The mRNA level of ACE2 in pancreas was higher than that in lung ( Fig. 1A , P < 0.001, Wilcoxon signed rank test). To investigate the distribution of ACE2 in the pancreas, we analyzed two scRNA-seq datasets. We identified different types of pancreatic cells, such as alpha cell, beta cell, acinar cell, ductal cell and several other types of cells (Fig. 1B, D) . In GSE85241, ACE2 was expressed in 55 cells (2.59%) among the total 2126 cells, most of which (81.82%) were expressed in exocrine gland (duct cells and acinar cells), and a few (16.36%) were expressed in pancreatic islets (alpha, beta, delta and PP cells) (Fig. 1C , (Fig. 1E , Table 1 ). Individual differences between donors from different sources may lead to differences in the analysis results of the two scRNA-seq datasets. These data still indicate that ACE2 is expressed in both the exocrine glands and islets, which may be the main potential part targeted by SARS-CoV-2 in the pancreas, resulting in . CC-BY-NC-ND 4.0 International license It is made available under a author/funder, who has granted medRxiv a license to display the preprint in perpetuity. is the (which was not peer-reviewed) The copyright holder for this preprint . https://doi.org/10.1101/2020.02.28.20029181 doi: medRxiv preprint pancreatic injury. In our study cohort, among the indicators of pancreatitis we focused on, the main ones were the elevation of serum amylase and lipase. In mild cases, 1.85% (1/54) of patients showed elevated levels of both amylase and lipase, while in severe cases, the proportion of increased amylase was 17.91% (12/64) and the proportion of increased lipase was 16.41% (11/64) ( Table 2 ). However, in the imaging evaluation, only 5 severe patients (7.46%) showed changes in the pancreas, mainly focal enlargement of the pancreas or dilatation of the pancreatic duct, without acute necrosis (Fig. 2) . These clinical data show that there exist mild pancreatic lesions in some patients with COVID-19, mainly in severe cases. spread in China, and now has the tendency to spread all over the world, causing a surge in infections and deaths [6] . So far, there is no specific drug for SARS-CoV-2, and challenges remain. Although there are asymptomatic infected people, who can also be the source of transmission [1, 12] , in clinical practice, patients infected with SARS-CoV-2 mainly suffer from systemic symptoms like fever and fatigue, respiratory symptoms like cough and expectoration, digestive symptoms such as diarrhea [2, 13] . In some patients, especially critically ill cases, damage to other organs, such as the heart, kidneys, and liver, has also been found, which brings great inconvenience to manage patients [2, 14] . As the receptor of SRAS-CoV-2, the . CC-BY-NC-ND 4.0 International license It is made available under a author/funder, who has granted medRxiv a license to display the preprint in perpetuity. is the (which was not peer-reviewed) The copyright holder for this preprint . https://doi.org/10.1101/2020.02.28.20029181 doi: medRxiv preprint expression of protein ACE2 in these organs may provide an approach for SARS-CoV-2 infection [7] , leading to tissue injuries. However, the injury of pancreas has been ignored in clinic, since the symptoms of mild pancreatic injury are not specific to other systemic lesions, such as fever, nausea, vomiting and abdominal pain. In this study, we focused on the expression of ACE2 in pancreas and the damage to the pancreas in a proportion of patients with SARS-CoV-2 infection. Although many previous studies have been reported on the expression of ACE2 tissues and organs [15] , including bulk RNA-seq and scRNA-seq data, this analysis in the pancreas is still lacking. Firstly, we found ACE2 was mostly expressed in the pancreas of normal people, and it was slightly higher in the pancreas than in the lungs, indicating that when SARS-CoV-2 appeared in the circulation, it might also combine with ACE2 in the pancreas to enter cells and cause pancreatic injury. Further, we performed scRNA-seq datasets of the pancreas to explore the localized expression of ACE2 in the pancreas. ACE2 is expressed in both exocrine glands and islets. At the same time, analysis results from scRNA-seq data were consistent with elevated levels of amylase and lipase in the plasma of our patients. In our cases, about 1-2% of the non-severe patients with SARS-CoV-2 had pancreatic lesions, and about 17% of the severe patients suffered from pancreatic injuries, which was not previously noticed. Although their imaging alterations suggested that pancreatitis was not severe, the problem should not be ignored, especially in severe patients. In the clinic, the consequences of pancreatic injury can still be very serious, . CC-BY-NC-ND 4.0 International license It is made available under a author/funder, who has granted medRxiv a license to display the preprint in perpetuity. is the (which was not peer-reviewed) The copyright holder for this preprint . https://doi.org/10.1101/2020.02.28.20029181 doi: medRxiv preprint such as aggravating systemic inflammation, especially in the lungs and accelerating the occurrence of acute respiratory distress syndrome (ARDS) [16] , and even developing into chronic pancreatitis without early intervention, which will have a serious impact on the health and quality of life of patients. On the other hand, Yang et al. reported that patients infected with SARS-CoV who had never used glucocorticoids suffered from hyperglycemia, which might be caused by SARS-CoV damaging the pancreatic islets through ACE2 [16, 17] . Since SARS-CoV-2 has the same receptor as SARS-COV, whether this result will occur requires more attention. Therefore, increased attention should be paid to the pancreas in patients with SARS-CoV-2 infection, especially in severe cases. In our study, there are still some limitations. There is no direct evidence that the slight damage to the pancreas was caused by SARS-COV-2 binding to ACE2 in the pancreas, and no studies have shown that nucleic acids of SARS-COV-2 have been detected in the pancreas. Due to the particularity of the location of the pancreas, this study could not be carried out. At present, the body anatomy has been conducted in China, and we hope to see this result in the future. At the same time, the systemic inflammatory response in severe COVID-19 patients might also cause mild damage to the pancreas. However, in the causes of pancreatitis, virus infection factors such as mumps, herpes simplex virus should not be ignored [18] . Our study provides a new idea and a possible mechanism. . CC-BY-NC-ND 4.0 International license It is made available under a author/funder, who has granted medRxiv a license to display the preprint in perpetuity. is the (which was not peer-reviewed) The copyright holder for this preprint . https://doi.org/10.1101/2020.02.28.20029181 doi: medRxiv preprint . CC-BY-NC-ND 4.0 International license It is made available under a author/funder, who has granted medRxiv a license to display the preprint in perpetuity. is the (which was not peer-reviewed) The copyright holder for this preprint . https://doi.org/10.1101/2020.02.28.20029181 doi: medRxiv preprint . CC-BY-NC-ND 4.0 International license It is made available under a author/funder, who has granted medRxiv a license to display the preprint in perpetuity. is the (which was not peer-reviewed) The copyright holder for this preprint . https://doi.org/10.1101/2020.02.28.20029181 doi: medRxiv preprint . CC-BY-NC-ND 4.0 International license It is made available under a author/funder, who has granted medRxiv a license to display the preprint in perpetuity. is the (which was not peer-reviewed) The copyright holder for this preprint . https://doi.org/10.1101/2020.02.28.20029181 doi: medRxiv preprint
-------------------------------------------------------


NEW PAPER


The novel coronavirus disease first identified in 2019 in Wuhan, has become a serious global public health concern. One current issue is the ability to adequately screen for the virus causing COVID-2 (SARS-CoV-2). Here we demonstrate the feasibility of shotgun proteomics as a SARS-CoV-2 screening method, through the detection of viral peptides in proteolytically digested body fluids. Using in silico methods, we generated trypsin-based shotgun proteomics methods optimized for LCMS systems from 5 commercial instrument vendors (Thermo, SCIEX, Waters, Shimadzu, and Agilent). First, we generated protein FASTA files and their protein digest maps. Second, the FASTA files were used to generate spectral libraries based on experimental data. Third, transition lists were derived from the
The detection of viral proteins in body fluids can be a rapid and specific diagnostic for infection in severe acute respiratory syndrome (SARS). 1-3 During the 2003 (SARS) outbreak, non-MS based methods of protein detection proved to be more successful 4,5 than LCMS methods. [6] [7] [8] Non-MS based methods, such as western blots, enzyme-linked immunosorbent assays (ELISAs), and protein arrays, rely on antibodies for the detection of proteins. Given recent studies concerning high variability in antibody production, LCMS-based methods are an attractive alternative approach for the rapid identification of small molecules, proteins, and peptides in clinical settings where consistency is paramount. 9, 10 In the 15 years since the 2003 SARS outbreak, LCMS technology has experienced a revolution led primarily by increases in the speed, sensitivity, and resolution of MS instruments. Today, protein array and antibody-based methods are falling out of favor in both research and clinical diagnostics, due in large part to the improvements in LCMS technology. 11, 12 A review of this growth by Grebe and Singh described a clinical lab with no LCMS systems in 1998 that completed over 2 million individual LCMS clinical assays in 2010. 13 Incremental improvements in rapid sample preparation techniques, chromatography, and data processing have also contributed to the increasing use of LCMS-based clinical testing. A 2013 study demonstrated the level of advance by identifying 4,000 yeast proteins in one hour of LCMS run time, identifying approximately 75 proteins/min at a rate 100 times faster than studies a decade prior. 14 LCMS methods can measure protein quantity by either the intact protein (protein-centric) or the analysis of proteolytic peptides (peptide-centric). While the mass to charge (m/z) ratio of a peptide or protein (MS1) may be a specific diagnostic in some materials, the majority of LCMS methods employ tandem MS in which the peptide or protein parent ion is subjected to gas phase collision to produce fragment ions. The measurement of the fragment ions (MS2) has a higher specificity and lower level of false positives and is the method of choice in clinical diagnostics. 12, 13 Protein-centric assays, such as intact parallel reaction monitoring or proteoform monitoring, performs well on small GTPases like KRAS, but larger proteins cannot currently be monitored with the same relative ease and accuracy. 15 For this reason, peptide-centric assays are currently the dominant proteomics approach. In peptide-centric assays, commonly called "bottom-up" or "shotgun" proteomics, proteins are subjected to proteolytic cleavage to produce smaller peptide sequences prior to LCMS, often using chemically modified trypsin. [16] [17] [18] Proteolytic digestion of a protein mixture increases the total number of molecules present and thereby increases the relative background noise of the sample. However, the production of multiple peptides from the protein of interest typically results in the production of highly selective peptide targets. Through a priori selection of peptide targets that are biologically unique or in some way chemically distinct, extremely specific assays can be rapidly designed. 19, 20 When multiple selective peptides are available for a given protein, the independent peptide measurements can be combined to provide replicate measurements to increase certainty in protein presence and relative abundance. A universal in all LCMS assays is the rule of "breadth vs. depth," simply stated, increasing the number of targets in an experiment results in a decrease in the overall sensitivity of each measurement. The most common example is that untargeted assays which may observe thousands of peptide ions per experiment invariably have a lower limit of detection (LOD) and quantification (LOQ) compared to assays where a smaller number of ions are targeted. [21] [22] [23] Improvements in each subsequent generation of hardware can mitigate this compromise, but the improvement is limited. Today, the only way to truly offset this rule is to increase the total LCMS run time. A 2014 study by Majchrzykiewicz-Koehorst et al. described an untargeted LCMS assay that could discern three viruses in samples, both in vitro and ex vivo. However, the run time for this experiment was six hours per sample using nanoESI-quadrupole TOF technology. 24 Targeted peptide-centric assays are advantageous when sensitivity is paramount over the quantity of identified targets. Targeted assays often rely on tandem MS with high speed, but relatively low accuracy, quadrupoles. 21 Quadrupoles can be used to select ions for fragmentation with quantification of fragment ions by other quadrupoles in single reaction monitoring (SRM). They can also be used in conjunction with high resolution systems in single ion monitoring (SIM) and parallel reaction monitoring (PRM). SRM relies on fragmentation which requires a priori the mass to charge ratio (m/z) of both the peptide ion of interest and its dominant ions produced during fragmentation. Commonly, the collision energy is optimized for each peptide fragment in order to maximize efficiency and signal. SIM uses higher resolution scans in lieu of fragmentation, which requires a priori only an approximate m/z ratio for the peptide. In SIM targeted assays, the peptide ion's exact mass is extracted post-run during data processing. Two studies have shown that SIM scans with ≥60,000 resolution produce selectivity comparable to SRM. 25, 26 PRM targeted assays combine fragmentation and high resolution molecule monitoring. In contrast to SIM, isolated ions are subjected to fragmentation. Post-run data processing then calculates quantification from the highresolution accurate mass of the fragment ions. For proteins with only one available peptide target, PRM is the most selective technique in modern proteomics. 27, 28 Untargeted assays can be divided into two broad categories, data dependent (DDA) and data independent analyses (DIA). In DDA, the masses of all ions are observed in a relatively wide m/z range (MS1); the MS1 peptide ions meeting user-defined thresholds are subjected to fragmentation. 29 The resulting fragment ions are scanned, typically in a wide range for obtaining the most complete coverage of the peptide fragment ions. Common user-defined thresholds in DDA include selecting for peptides with: (1) sufficient signal for sequencing and (2) isotopic distributions matching models of typical peptides for that m/z. 30 The ion selection is performed in real time and automatically by the instrument, and has required less user input in method design with each generation of hardware. 31,32 DDA data analysis methods requires the least a priori information of all current LCMS methods. Although full de novo identification of peptide sequences is possible, it remains computationally demanding, thus typical post-instrument analysis relies on peptide search engines. Peptide search engines require the user to provide a database containing the protein sequences the user expects. As the MS1 and MS2 data is directly compared to this expected database, results are impacted by the accuracy of the database as well as on the availability of comprehensive proteome sequence information for the organism(s). Much of proteomics research relies on peptide search engines that predate the LCMS instrumentation by decades. Such software can consider multiple scenarios, such as alternative charge states, but cannot account for post translational modifications (PTMs) unless defined by the user a priori. 33 Recent developments in next-generation search engines (i.e., MetaMorpheus 34 , Bolt 35 , Byonic 36 , and MSFragger 37 ) can identify PTMs without the user providing predicted PTMs, but these tools have yet to be widely adopted by the field. Unlike DDA, DIA does not perform automatic real-time decision making processes. Instead, MS1 and MS2 scans are acquired with set mass widths (MS2 windows) that cover the entire peptide mass range of interest. The size of the MS2 windows vary based on the instrument speed and sensitivity, but the same instrument run method may be applied to any peptide experiment. Thus, DIA involves minimal optimization of the instrument run method, but this generalized data collection approach puts the onus on the post-run data analysis of the resulting MS2 fragments. Large MS2 windows may display peptide ion fragments originating from hundreds of unique peptide ions. Peptides are detected in DIA MS2 windows by matching the experimental results against peptide fragment spectral libraries. Spectral libraries are annotated peptide fragments, typically produced by previous DDA experiments. MS2 spectra selected for libraries typically contain complete fragment sequence coverage, with an MS2 fragment representing the product of breaking each peptide bond within the peptide of interest. In a common DIA workflow, a portion of all peptides from the study are pooled and chemically fractionated before being subjected to DDA LCMS analysis. The DDA experiments are used to create the spectral library for the experiment and each individual sample is separately analyzed by DIA. Quantification of the peptides and proteins occurs in the individual DIA experiments and the spectral libraries from the pools serve as a reference for identification of the peptides being quantified. [38] [39] [40] While spectral libraries are essential for DIA, they can also be utilized in DDA and can results in the most specific SRM and PRM targeted assays. Peptide fragmentation follows specific energetic patterns, resulting primarily in fragments caused by separation at the peptide bond. It is therefore possible to create theoretical spectral libraries in silico from peptide sequence alone. Theoretical spectral libraries are especially useful when the biological samples are unavailable. New tools that employ deep learning algorithms have been demonstrated to produce theoretical MS2 spectra superior to previous prediction models and, in the absence of true experimental data, are the best resources currently available. 41, 42 These deep learning algorithms can learn from vast libraries of experimental data to predict the fragmentation patterns of new peptide sequences that they are given. One such algorithm, PROSIT, uses the vast synthetic human peptide libraries, from the ProteomeTools project 43 for its training dataset. Due to the high quality of the 450,000 synthetic peptides experimentally fragmented in ProteomeTools to date, PROSIT has been demonstrated to create spectral libraries that are, in some cases, superior to experimentally derived in-house spectral libraries. 44 In this study, we describe a in silico approach to develop LCMS screening assays of virus peptides in human samples. Currently, there is no publicly available proteomics data from COVID-19 infected human tissues and we are restricted from access to these materials. Our in silico developed materials facilitate both global and targeted analysis by providing all necessary materials for both DDA and DIA investigation of these materials through the production of FASTA databases, spectral libraries and a list of predicted PTMs investigators should consider when searching with historic peptide search engines. The peptide spectral libraries are further utilized to create transition lists optimized for hardware from 5 instrument vendors and complete PRM methods for all 3 quadrupole Orbitrap architectures. All materials and methods described herein are available as supplemental material to this publication as described in Table 1 . The supplemental material includes protein sequences (FASTA files), predicted PTMs, theoretical MS2 spectral libraries, instrument methods and targeted method data processing templates. Our work demonstrates not only the feasibility of this approach, but also its ability to rapidly develop methods even in the face of limitation of access to sample experimental data. We use the example of SARS-CoV-2 viral protein detection to underscore the utility of this approach in responding to an urgent public health crisis. Coronavirus FASTA databases . CC-BY 4.0 International license author/funder. It is made available under a The copyright holder for this preprint (which was not peer-reviewed) is the . https://doi.org/10.1101/2020.03.08.980383 doi: bioRxiv preprint At the date of this writing, only theoretical protein sequences for SARS-CoV-2, are available. These sequences are being acquired and annotated and the result of translation of genomic sequence information. All sequences in this study were obtained from NCBI accession: txd2697049, https://www.ncbi.nlm.nih.gov/protein/?term=txid2697049). Using Proteome Discoverer 2.4 (Thermo), the protein sequences were combined into a single protein FASTA database (2019-nCOVpFASTA1; Supplemental Information), and added to human proteome sequences (UniProt SwissProt Human database; downloaded 2/15/2020) to produce a database including both human and COVID-19 protein sequences (Human_plus_2019-nCOVpFASTA2; Supplemental Information). Publicly available experiments on other coronavirus experiments were found by searching the ProteomeXchange Consortium web interface (http://www.proteomexchange.org/). 45, 46 Clarification of the identity of unpublished data from Pacific Northwest National Laboratory (PNNL) was provided by Dr. Michael Monroe. The 2019-nCOVpFASTA1 (Supplemental Information) was converted to PROSIT peptide format with the EncyclopeDIA 47 software, resulting in an in silico peptide digestion (parameters: Charge Range = 2-3; Max missed cleavage = 1, m/z range = 400-1,500; default NCE = 30eV; default charge = +2). PROSIT peptide fragmentation prediction libraries were generated using the PROSIT online web portal (https://www.proteomicsdb.org/prosit/) with the spectral library modeling interface (options: prediction model = Prosit_2019_intensity; iRT prediction model = Prosit_2019_irt). The libraries were exported in NIST MSP text format for use in Skyline. 48, 49 While preparing this manuscript, Proteome Sciences reported the creation of PROSIT libraries for use in the ScaffoldDIA software (cite ScaffoldDIA?) (https://support.proteomesoftware.com/hc/en-us/articles/360035151172-Prosit-Derived-Spectral-Libraries-for-Scaffold-DIA-Searches). For SRM transitions, the 2019-nCOVpFASTA1 (Supplemental Information) was imported into Skyline v20.1.0.28 (University of Washington) along with the PROSIT tryptic peptide spectral library. Peptide settings and transitions were optimized within Skyline to reflect the vendor optimization requirements. For Agilent systems, the 20ms default dwell time was selected for the transition settings. For SCIEX instruments, the same dwell time was utilized as well as automatic optimization of the declustering potential and compensation voltage from the transition settings menu. For Waters, Thermo, and Shimadzu systems, no further settings were required for transition list generation. All transition lists were exported as unscheduled 15min methods. For PRM methods, three peptides were selected for each protein due to the increased time per scan relative to SRM methods. Instrument-specific Skyline files are included in the Supplemental information as described in Table 1. interface can accept a maximum of 5,000 amino acids. In order to analyze YP_009724389.1 the predicted sequence had to be divided into five sequences, using a 100 amino acid overlap to avoid disrupting potential large motifs. The results of ModPred were compiled into a single spreadsheet with all modifications of all confidence levels. A second sheet was created that contained only the high confidence PTMs predicted, as well as a final summary for the counting of predicted high confidence PTM occurrence, provided as Supplemental Material as described in Table 1 . In shotgun proteomics, proteins are first digested into smaller peptide fragments that are more easily detected by the instrument. Given its widespread use, high efficiency and speed of digestion, we chose to develop methods that exclusively use the proteolytic enzyme trypsin, which produces "tryptic peptides." Sequencing grade trypsin exhibits high efficiency cleavage at unmodified (1) arginine and (2) lysine residues unless followed by a proline. Trypsin also has the advantage of leaving a terminal basic residue at the cleavage site, which increases the likelihood of complete fragment ion coverage from the charged terminal. 18 Due to these reasons, trypsin is utilized unless the protein sequence has an abnormally high or low number of lysine or arginine residues. A very high frequency of the residues (such as Lysine-rich proteins) will create very short peptides that could be uninformative for protein identification. A very low frequency of the residues will create very large peptides, or undigested ("intact") proteins in some cases, that are difficult to detect and fragment. Our theoretical trypsin digest of the 2019-nCoVpFASTA1 database produced tryptic peptides with average lengths of 8 to 18 amino acids. These results indicate that trypsin digestion is an appropriate choice for detection of these viral proteins. To widen use of our developed methods, we used the Skyline software to create transition lists optimized specifically for each of the Skyline-compatible triple quadrupole instruments. While minor modifications are required for SCIEX, Agilent, and Shimadzu instruments, Waters Xevo and Thermo instruments use identical parameters for transition list design. Most modern triple quadrupole instruments are capable of 500 SRMS/sec and fully permit the use of 2,000 transition lists, as provided here. For older instruments that lack this scan speed or that require higher dwell times, the transition lists included in the supplemental methods may be reduced by the end user accordingly. PRM methods monitor multiple transitions simultaneously but at a time cost. The highest scan speed currently available in Orbitrap instruments is 48 scans per second and is only available on the Exploris 480 system (data not shown). In order to achieve maximum sensitivity, higher fill times are often required for these instruments. We chose to utilize three peptides/protein for these methods. Alternative peptides can be selected from the Skyline files provided (Supplemental Information) or by selecting peptide mass targets from the other transition lists.  The Prosit spectral libraries (Supplemental Material) enable the interrogation of DIA data and may be used for DDA experiments that employ tools such as the MSPepSearch (NIST). 51 DDA data requires only the protein FASTA file and a list of PTMs that may be present in the sample. Our analysis using ModPred predicted 17 possible PTMs (Figure 1) . Amidation was the most frequent predicted PTM, but there is no known biological mechanism that we could derive from a survey of the literature. Palmitoylation, the second most frequent predicted PTM, is a well characterized viral PTM with critical functions in human immunodeficiency virus (HIV), human herpes virus (HHV), and influenza virus infectivity. [52] [53] [54]  The ProteomeXchange Consortium is an open-access platform for the rapid sharing of proteomics data. The prominent proteomic technical journals, the Journal of Proteome Research and Molecular and Cellular Proteomics, strictly require that all unprocessed instrument files and processed results are made publicly available through these services. While attempting to obtain coronavirus data, we identified unpublished metabolomics, lipidomics, and proteomics data generated from human samples infected with Middle Eastern respiratory syndrome (MERS-CoV). We provide this data, as well as a list of other proteomic studies of note, for use in comparative studies with SARS-CoV-2. Using in silico methods, we have developed methods for the detection of SARS-CoV-2 in human samples. In vitro validation of this method is required and outside the scope of this paper given our lack of access to such samples. We have provided the minimum materials for data processing for both DDA and DIA untargeted proteomics methods with FASTA databases, spectral libraries and by predicting relevant PTMs for consideration. To broaden the number of labs that can apply our methods, we optimized run parameters for widely used LCMS systems compatible with Skyline, representing instruments from five companies. We will continue to refine these resources and post updates to these methods to LCMSmethods.org and invite researchers anywhere in the world to contact us for assistance in further optimization to address this emerging threat.
-------------------------------------------------------


NEW PAPER


Since COVID-19 emerged in early December, 2019 in Wuhan and swept across China Mainland, a series of large-scale public health interventions, especially Wuhan lock-down combined with nationwide traffic restrictions and Stay At Home Movement, have been taken by the government to control the epidemic. Based on
December 1, 2019 in Wuhan City, Hubei Province, China (Huang et al, 2020) , the overall confirmed cases in China had reached 78,959 by the end of February 27, 2020, and a total of 2,791 people had died of the disease. COVID-19 had also spread to other 50 country with the confirmed cases and deaths were 4,696 and 67 respectively, by the end of February 27, 2020. The World Health Organization declared the COVID-19 epidemic as an international public health emergency on January 30, 2020. To prevent further dissemination of SARS-CoV-2, 31 Provinces in China Mainland had raised their public health response level to the highest state of emergency (level-1) by January 29, 2020. The Chinese government has implemented a series of large-scale public health interventions to control the epidemic, many of which have far exceeded what International Health Regulations required, especially Wuhan lock-down, nationwide traffic restrictions and Stay At Home Movement. Wuhan had prohibited all transport in and out of the city as of 10:00 on January 23, 2020, this is maybe the largest quarantine/movement restriction in human history to prevent infectious disease spread (Tian et al, 2020) . Hundreds of millions Chinese residents, including 9 million Wuhan residents, have to reduce even stop their inter-city travel and intra-city activities due to these strict measures. Due to Wuhan lock-down, Kucharski et al (2020) estimated that the median daily reproduction number had declined from 2.35 of January 16 to 1.05 of January 30, Tian et al. (2020) estimated that the dispersal of infection to other cities was deferred 2.91 days (CI: 2.54-3.29). However, Read et al (2020) suggested that travel restrictions from and to Wuhan city are unlikely to be effective in halting transmission across China; with a 99% effective reduction in travel, the size of the epidemic outside of Wuhan may only be reduced by 24.9% on February 4. Do these large-scale public health interventions really work well in China? Besides, hundreds of officials were fired or appointed rapidly according to their incompetent or outstanding performances during the epidemic. How to assess the efforts of different regions in China Mainland against COVID-19? Here we present a simple yet effective model based on Baidu Migration data and the confirmed cases data to quantify the consequences and importance of Wuhan lock-down combined with nationwide traffic restrictions and Stay At Home Movement on the ongoing spread of COVID-19 across China Mainland, and preliminarily assess the efforts of 29 Provinces and 44 prefecture-level cities during the epidemic. . CC-BY-NC-ND 4.0 International license It is made available under a is the author/funder, who has granted medRxiv a license to display the preprint in perpetuity. (which was not peer-reviewed) The copyright holder for this preprint . https: //doi.org/10.1101 //doi.org/10. /2020 Due to Wuhan lock-down, more than 9 million residents were isolated in Wuhan City since January 23, 2020, only 1.2 million travelers from or to Wuhan during January 24 to February 15, 2020 according to Baidu Migration. The travelers were down 91.61% and 91.62%, compared to the same period last year (14 million) and the first 23d (14 million from January 1 to January 23, 2020), respectively (Fig.1) . CC-BY-NC-ND 4.0 International license It is made available under a is the author/funder, who has granted medRxiv a license to display the preprint in perpetuity. The copyright holder for this preprint . https://doi.org/10. 1101 Due to Stay At Home Movement, the mean intensity of intra-city activities for 316 cities was 2.61/d during January 24 to February 15, 2020 according to Baidu Migration. It was down 42.42% and 50.27% when compared to the same period last year (4.53/d) and the first 23d (5.25/d), respectively (Fig.3) . Obviously, COVID-19 greatly reduced the human mobility of China Mainland. We consider 44 regions in China Mainland which accept travelers from Wuhan, including 29 Provinces (Tibet was excluded since only one confirmed case was reported) and 15 prefecture-level cities in Hubei province. We noticed that the number of confirmed cases between non-Hubei and Hubei were closer in the early period. For example, the number of cumulative confirmed cases by the end of January 26 in Chongqing (non-Hubei) and Xiaogan (Hubei) were 110 and 100, respectively. Their cumulative confirmed cases by the end of February 27, however, were 576 and 3,517, respectively. We surmise that this is partly because Xiaogan has received more infected cases from Wuhan than Chongqing since the human-to-human transmission risk of COVID-19 was confirmed and announced on January 20. This surmise was confirmed by Fig.4 . The proportion of travelers from Wuhan accepted by Hubei regions to the total travelers from Wuhan increased rapidly from 70% (before January 19) to 74% (January 20), even over 77% after January 26. So we concluded that the first key factor affecting the later (e.g. February 27) cumulative confirmed cases of each non-Wuhan region is the sum immigrants from Wuhan during January 20 to January 26 (few immigrants after January 27). These immigrants had higher infected probability but lower transmission . CC-BY-NC-ND 4.0 International license It is made available under a is the author/funder, who has granted medRxiv a license to display the preprint in perpetuity. The copyright holder for this preprint . https://doi.org/10.1101/2020.02.29.20029561 doi: medRxiv preprint ability because the susceptible strengthened protection awareness and measures after the declaration of human-to-human transmission. The second key factor is the sum number of the infected immigrants from Wuhan before January 19. According to the recent report, there is a mean 10-day delay between infection and detection, comprising a mean about 5 day incubation period and a mean 5 day delay from symptom onset to detection of a case (Imai et al, 2020; Yang et al, 2020; Li et al, 2020) . So the second key factor can be represented with the number of cumulative confirmed cases by the end of January 29. These "seed cases" had higher transmission ability because the susceptible had no any protection measure. A simple regression model was established as follow. y= 70.0916+0.0054×x1+2.3455×x2 (n = 44, R 2 = 0.9330, P<10 -7 ) Here, y is the number of cumulative confirmed cases by February 27 of each non-Wuhan region, x1 is the sum number of immigrants from Wuhan during January 20 to January 26 of each non-Wuhan region, and x2 is the number of cumulative confirmed cases by January 29 of each non-Wuhan region. The standard regression coefficients of x1 and x2 are 0.6620 and 0.3796 respectively, indicating that x1 is more important than x2 for determining y. The observed and expected values of the cumulative confirmed cases by is the author/funder, who has granted medRxiv a license to display the preprint in perpetuity. The copyright holder for this preprint . https://doi.org/10.1101/2020.02.29.20029561 doi: medRxiv preprint To evaluate the effect of Wuhan lock-down, x2, the number of cumulative confirmed cases by January 29, should be fixed. Then we assumed three different lock-down plans besides the real plan (Table 1) . I20  I21  I22  I22  I22  I22  I23  I24  I25  I26 Strict * : none left Wuhan since the date of lock-down. I20 ~ I29 are the real emigrants from January 20 to January 29, respectively. It's not fair to assess the effort of different regions only based on the final number of cumulative confirmed cases due to the difference of the number of immigrants from Wuhan. The aforementioned interpretative model has taken this difference into account and the results are listed with 5 grade (Excellent, Good, Normal, Poor, Very poor) according to standard residuals (Table 2 and Table 3 ). We emphasize that this is only a preliminary evaluation and the results are for reference only. The human mobility data on inter-city travel and intra-city activity from January 1, 2020 to February 21, 2020 (including the same period data in 2019) in China Mainland is from Baidu Migration (http://qianxi.baidu.com). The inter-city travel population of each city is represented with immigration and emigration index, the travelers proportions of different destination and departure, on the level of Province and city (but only the top 100 cities), are also listed. The intra-city activity intensity of each city is represented with the index (not proportion) of . CC-BY-NC-ND 4.0 International license It is made available under a is the author/funder, who has granted medRxiv a license to display the preprint in perpetuity. The copyright holder for this preprint . https://doi.org/10.1101/2020.02.29.20029561 doi: medRxiv preprint . CC-BY-NC-ND 4.0 International license It is made available under a is the author/funder, who has granted medRxiv a license to display the preprint in perpetuity. The copyright holder for this preprint . https: //doi.org/10.1101 //doi.org/10. /2020 activity population to total population. Nanjing, Qingdao, Shenzhen and Foshan during the 2019 Spring Festival travel rush (40d) are from their Official website of the Transportation Commission. We estimate one Baidu Migration index is about equal to 56,137 travelers (Fig.6) . The confirmed case data on each Province and prefecture-level city are from the National Health Commission of China (http://www.nhc.gov.cn/xcs/yqtb/list_gzbd.shtml) and its affiliates. Baidu Migration recorded more types of spatial displacement, including airplane, high-speed rail, ship, coach and private car, so it theoretically has higher accuracy. The real inter-city travel population of each city, however, is to the last day before lock-down (January 22) is 95.98, then the two model estimated travelers are 2.9 million and 5.4 million, respectively. News Release Conference of Wuhan on 26 January confirmed that more than 5 million people had left Wuhan after 10 January due to the Spring Festival travel rush and epidemic. Our estimated value is closer to official reports. It is worth noting that the Baidu Migration index still has the following disadvantages for estimating real traveler population. 1) The mobility behavior of a large number of groups un-connected to Baidu Map and third-party . CC-BY-NC-ND 4.0 International license It is made available under a is the author/funder, who has granted medRxiv a license to display the preprint in perpetuity. The copyright holder for this preprint . https: //doi.org/10.1101 //doi.org/10. /2020 users has not been recorded. 2) The spatial displacement of users is recorded only within 8 hours. 3) Most of the trips are disassembled and not fully identified. For example, one user travel from A, pass through B and arrive final destination C, and the user also has location information in C by coincidence, then this trip will be disassembled into A-B and B-C. Regardless, big data has and will continue to play an important role in public health. There are 9.48 million residents in Wuhan around January 26, the cumulative confirmed cases are 2,261 by January 29. We estimate at least 56,916 people were infected in Wuhan according to our model (up to February 27 the confirmed cases are 48,137). In other words, more than 8,000 undetected but infected individuals still wait for checking out in the center of epidemic storm. Wuhan still has much to do. SARS-CoV-2 has diversity transmission approaches, including respiratory (mouth foam) and contact routes which have been confirmed, as well as aerosol and digestive (fecal-oral) routes which cannot be ruled out (National Health Commission of China, 2020). It is also highly concealed according to the recent report of transmission of the virus from asymptomatic and mild individuals (Zhang et al, 2020; Sanche et al, 2020 Yang et al (2020) , respectively. At the beginning of the outbreak, the infected individuals may be greatly underestimated due to the asymptomatic transmissions, insufficient sensitivity of diagnostic reagents and delayed . CC-BY-NC-ND 4.0 International license It is made available under a is the author/funder, who has granted medRxiv a license to display the preprint in perpetuity. The copyright holder for this preprint . https://doi.org/10. 1101 diagnosis. The latest estimated value of R0 and the control reproductive number is 4.7-6.6 (Steven et al, 2020) and 6.47 (CI: 5.71-7.23) (Tang et al, 2020), respectively. SARS-CoV-2 is highly contagious, Wang et al (2020) projected that without any control measure the infected population would exceed 200,000 in Wuhan by the end of February. As for Steven et al (2020) and Read et al (2020) , the estimated number was 233,400 (CI: 38,757-778,278) by the end of January and 191, 529 (CI: 132, 649) by February 4, respectively. Up to February 27, the confirmed cases are 48,137. The only lesson that humans have learned from history is that humans have not learned anything from history. Clearly, Wuhan has not learned anything from the SARS epidemic in 2003, now she is suffering from her early delays. Fortunately, China Government has implemented a series of large-scale public health interventions to control the epidemic. In fact, many prevention and control measures taken by China, especially Wuhan lock-down, nationwide traffic restrictions and Stay At Home Movement, go far beyond the requirements for responding to emergencies, setting new benchmarks for epidemic prevention in other countries. The Chinese method has proven to be successful. The strategy adopted by China has changed the fast-rising curve of newly diagnosed cases, and the simplest and most direct thing that can explain this is the data (Fig.7) . The SARS-CoV-2 epidemic is still rapidly growing and spread to more than 42 countries as of February 27, 2020. At present, the most serious countries outside China are South Korea, Italy, Iran, and Japan. It is worrying that although some measures have been taken, the current prevention and control measures of these countries may still be insufficient. None of them has reached the level of prevention and control in China's moderately affecting . CC-BY-NC-ND 4.0 International license It is made available under a is the author/funder, who has granted medRxiv a license to display the preprint in perpetuity. (which was not peer-reviewed) The copyright holder for this preprint . https://doi.org/10. 1101 
-------------------------------------------------------


NEW PAPER



Since December 2019, an increasing number of patients with pneumonia occurred in Wuhan, Hubei province, China, which attracted much attention not only within China but across the world 1,2 . The novel pneumonia was named as Corona Virus by World Health Organization (WHO) (https://www.who.int/docs/default-source/coronaviruse/situationreports/20200211-sitrep-22-ncov.pdf?sfvrsn=fb6d49b1_2), the common symptoms of COVID-19 at illness onset were fever, fatigue, dry cough, myalgia, and dyspnea 3 . In addition, some patients might suffer from headache, dizziness, abdominal pain, diarrhea, nausea, and vomiting 3 . Onset of disease may lead to progressive respiratory failure due to alveolar damage and even death 4 . Scientists then isolated a novel coronavirus from human airway epithelial cells, which was named 2019-nCoV 5 . Lu et al. 6 found that 2019-nCoV was closer to bat-SL-CoVZC45 and bat-SL-CoVZXC21 at the whole-genome level, and the external subdomain of the 2019-nCoV receptor-binding domain (RBD) was more similar to that of severe acute respiratory syndrome (SARS) coronavirus (SARS-CoV). Study of Zhou et al. 4 indicated that the angiotensin-converting enzyme II (ACE2) is likely the cell receptor of 2019-nCoV, which were also the receptor for SARS-CoV and HCoV-NL63 7,8 . Zhou et al. 4 also proved that 2019-nCoV does not use other coronavirus receptors, aminopeptidase N, and dipeptidyl peptidase 4. The study of Xu et al. 9 found that the RBD domain of the 2019-nCoV S-protein supports strong interaction with human ACE2 molecules. These findings suggest that the ACE2 plays an important role in cellular entry, thus ACE2-expressing cells may act as target cells and are susceptible to 2019-nCoV infection 10 . The expression and distribution of the ACE2 in human body may indicate the potential infection routes of 2019-nCoV. Through the developed single-cell RNA sequencing (scRNA-Seq) technique and single-cell transcriptomes based on the public database, researchers analyzed the ACE2 RNA expression profile at single-cell resolution. High ACE2 expression was identified in type II alveolar cells (AT2) of lung [10] [11] [12] , esophagus upper and stratified epithelial cells, absorptive enterocytes from ileum and colon 12 , cholangiocytes 13 , myocardial cells, kidney proximal tubule cells, and bladder urothelial cells 10 . These findings indicated that those organs with high ACE2-expressing cells should be considered as potential high risk for 2019-nCoV infection 10 . In order to investigated the potential routes of 2019-nCov infection on the mucosa of oral cavity, we explored whether the ACE2 is expressed and the ACE2-expressing cell composition and proportion in oral cavity based on the public bulk RNA-seq profiles from two public databases and single-cell transcriptomes from an independent data generated in-house. The result showed that the ACE2 could be expressed in the oral cavity, and was highly enriched in epithelial cells. Moreover, among different oral sites, ACE2 expression was higher in tongue than buccal and gingival tissues. These findings indicate that the mucosa of oral cavity may be a potentially high risk route of 2019-nCov infection. Public bulk RNA-seq dataset analysis NA-seq profile data of 13 organs including 695 para-carcinoma normal tissues as control from public TCGA were obtained for our analysis, and Fig. 1a showed that ACE2 could be expressed in various organs, the mean expression of different organs could be found in Table 1 . According to the mean expression of ACE2, the mucosa of oral cavity could express ACE2, and the results were validated by the data of normal tissues from the FANTOM5 CAGE dataset (Fig. 1b) . To investigate the ACE2 expression on mucosa of oral cavity, we looked into the ACE2 expression in different oral sites. According to the site information provided by the TCGA, among the 32 adjacent normal tissues, 13 tissues located in the oral tongue, 2 tissues located in the base of tongue, 3 tissues located in the floor of mouse, and 14 tissues did not definite the site and were just put into the category of oral cavity. The mean expression distribution of different sites was shown in Fig. 1c . When we combined the base of tongue, floor of mouth and oral cavity as other sites, and compared them with oral tongue, we found the obvious tendency that the mean expression of ACE2 was higher in oral tongue (13 tissues) than others (19 tissues) (Fig. 1d ), while may due to the limitation of the sample size, the p value was not significant (P = 0.062). Single cell RNA-seq analysis of oral tissues Single cell RNA-seq was utilized for four oral tissues, and the data was analyzed to confirm the above results and assess the cell type-specific expression of ACE2. After the data preprocessing (shown in section "Materials and methods"), 22 969 cells were acquired and 7 cell types were identified (Fig. 2a) , including epithelial cells (marker genes including SFN, KRT6A, and KRT10), fibroblasts (marker genes including FAP, PDPN, COL1A2, DCN, COL3A1, COL6A1), T cells (marker genes including CD2, CD3D, CD3E, and CD3G), macrophages (marker genes including CD163, CSF1R, CD68, and FCGR2A), mast cells (marker genes including CMA1, MS4A2, TPSAB1, TPSB2), B cells (marker genes including SLAMF7, FCRL5, and CD79A) and endothelial cells (marker genes including PECAM1, VWF, and ENG). The heatmap of main cell markers across the cell types can be found in Fig. 2b . According to Fig. 2c , d, we confirmed the ACE2 was expressed in oral tissues (0.52% ACE2-positive cells), and higher in oral tongue than buccal and gingival tissues (95.86% ACE2-positive cells located in oral tongue). Figure 2e shows that the ACE2-positive cells could be found in oral tissues including epithelial cells (1.19% ACE2-positive cells), T cells (<0.5%), B cells (<0.5%), and fibroblast (<0.5%), and the ACE2 was highly enriched in epithelial cells, of which 93.38% ACE2positive cells belong to epithelial cells (Fig. 2f) . The above results indicated that the ACE2 could be expressed on the epithelial cells of the oral mucosa and highly enriched in tongue epithelial cells. In the last two decades, coronavirus has caused two large-scale pandemics, SARS in 2002 and the Middle East respiratory syndrome In this study, the analysis of public bulk-seq RNA datasets showed that the mucosa of oral cavity could express the ACE2 and was higher in tongue than other oral sites. The results of this study were consistent with the study of Zou et al. 10 in general, many organs with higher expression of ACE2 than lung, such as intestine, heart, and kidney. According to the study of Zhao et al. 11 , the ACE2 expression in lung is concentrated in a small population of type II alveolar cells (AT2), that may cause the relatively low ACE2 expression of lung in bulk-seq RNA datasets analysis. Even though, the result of Zou et al. indicated that the respiratory tract should also be considered as a vulnerable target to 2019-nCoV infection 10 . The results of our single cell RNA-seq profiles validated the ACE2 expression in oral cavity, and the level of ACE2 expression in oral tissues was higher in tongue than buccal or gingival tissues. Furthermore, we have also demonstrated that the ACE2-positive cells were enriched in epithelial cells, which was also reported by previous study [10] [11] [12] 16 . These findings indicated that oral cavity could be regarded as potentially high risk for 2019-nCov infectious susceptibility. Interestingly, we found that the ACE2 also expressed in lymphocytes within oral mucosa, and similar results were found in various organs of the digestive system and in lungs 11, 12 . Whether those facts have reminded the 2019-nCoV attacks the lymphocytes and leads to the severe illness of patients needs more in vitro and in vivo evidence and validations, though the proportion of ACE2-positive lymphocytes is quite small. Previous studies have investigated the ACE2 mRNA and protein expression in various tissues by bulk samples 17, 18 , however, the distribution of ACE2 through bulk data could not indicate the cell type-specific expression of ACE2. Recently developed single-cell RNAsequencing technology enabled the generation of vast amounts of the transcriptomic data at cellular resolution 19 . The ACE2 expression profile in various organs, tissues, and cell types, provides the bioinformatics evidence for the potential infection routes of 2019-nCov, which might also be associated with presented symptoms. Although studies have reported multiple symptoms of hospitalized patients with 2019-nCoV infection 3, 20 , some cases at home might be asymptomatic. It is worth noting that, a previous study showed that 99% of the patients had no clinical manifestation of oral human papillomavirus (HPV), but HPV DNA was detected in 81% of oral mucosa samples, and anti-HPV IgA was detected in the saliva of 44% of the patients 21 . Likewise, although 2019-ncov infection hardly presented oral symptoms, the ACE2 expression in the oral cavity indicated that the oral infection route of 2019-nCoV cannot be excluded. Moreover, a latest pilot experiment showed that 4 out of 62 stool specimens tested positive to 2019-nCoV, and another four patients in a separate cohort who tested positive to rectal swabs had the 2019-nCoV being detected in the gastrointestinal tract, saliva, or urine 20 . Thus, our results support that in addition to the respiratory droplets and direct contact, fecal-oral transmission might also be the route of transmission of 2019-nCoV. Our results are mainly based on public datasets and single cell RNA-sequencing data of in-house oral tissues with minimal diseased lesion which from our previous project found no significant expression difference among the common epithelial markers in our past study and other previous study [22] [23] [24] . It is warrant that further histological methods are used to confirm our results and enhance the persuasion of the conclusion. The ACE2-expressing cells in oral tissues, especially in epithelial cells of tongue, might provide possible routes of entry for the 2019-nCov, which indicate oral cavity might be a potential risk route of 2019-nCov infection. Those preliminary findings have explained the basic mechanism that the oral cavity is a potentially high risk for 2019-nCoV infectious susceptibility and provide a piece of evidence for the future prevention strategy in clinical practice as well as daily life. Bulk RNA-seq data of para-carcinoma normal tissues which were taken as control tissues in the studies were downloaded from The Cancer Genome Atlas (TCGA; https://www.cancer.gov/tcga), and 695 para-carcinoma normal tissues distributed in different organs were obtained for this study, which included intestine (51 tissues), kidney (129 tissues), stomach (35 tissues), bile duct (9 tissues), liver (50 tissues), oral cavity (32 tissues), lung (110 tissues), thyroid (59 tissues), esophagus (11 tissues), bladder (19 tissues), breast (113 tissues), uterus (25 tissues), and prostate (52 tissues). The RNA-seq data were batch effects normalized and log2-transformed for the subsequent analysis. Violin plot was used to show the distribution of ACE2 expression among different organs, t test was performed to compare the ACE2 expression between two different groups shown in boxplot. Besides, Bulk RNA-seq data of normal tissues were downloaded from Functional Annotation of The Mammalian Genome Cap Analysis of Gene Expression (FANTOM5 CAGE) dataset 25 , as only two samples of this dataset which owns 60 samples in total located in the tongue, we just downloaded 14 organ types to validate the ACE2 expression in oral cavity with bar plot, including colon, ovary, breast, cerebellum, epididymis, esophagus, gallbladder, heart muscle, kidney, liver, lung, pancreas, prostate, and tongue. All analyses were performed in R (R version 3.6.0) and significant level was set as 0.05. Oral tissues of in-house cohort Due to our previous project about oral potential malignant disorders, four tissues of oral mucosa were obtained from patients after informed consent and ethical approval from West China Hospital of Stomatology, Sichuan University. These oral tissues had been sent for single cell RNA sequence. The four were taken from three patients with an average age of 50, which were all diagnosed as hyperkeratosis without dysplasia by pathologists, just showing an increase in cell number in the spinous layer and/or in the basal/ parabasal cell layers without cellular atypia, and its genetic profiles would be much more closed to normal tissue than malignant Single cell RNA-sequencing library preparation To generate single-cell Gel Beads-in-Emulsion (GEMs), the 10× chromium platform was used to capture and barcode cells. Cells were partitioned into GEMs along with gel beads coated with oligonucleotides and cDNAs with both barcodes were amplified, and a library was constructed using the 10× Genomics Chromium Single Cell Kit (v3 chemistry) for each sample. The resulting libraries were sequenced on an Illumina NovaSeq 6000 System. Single cell RNA-seq data preprocessing and analysis The FASTQ files were analyzed with the Cell Ranger Software Suite (version 3.1; 10× Genomics). The Seurat (version 3.0) was applied to read the gene-barcode matrix of four tissues. To control quality, we removed cells with <200 genes and 500 UMI counts, and as well as the cells with mitochondrial content higher than 5%. Besides, the genes detected in <3 cells were filtered out. The "sctransform" wrapper in Seurat was applied to normalize the data and remove confounding sources of variation, the "IntegrateData" was used for integrated the Seurat objects from four tissues. The uniform manifold approximation and projection (UMAP) was used for dimensionality reduction and clustering the cells, cell types were assigned based on their canonical markers. UMAP plots, heatmap, and violin plots were generated with Seurat in R.
-------------------------------------------------------


NEW PAPER



Since December 2019, an increasing number of patients with pneumonia occurred in Wuhan, Hubei province, China, which attracted much attention not only within China but across the world 1,2 . The novel pneumonia was named as Corona Virus by World Health Organization (WHO) (https://www.who.int/docs/default-source/coronaviruse/situationreports/20200211-sitrep-22-ncov.pdf?sfvrsn=fb6d49b1_2), the common symptoms of COVID-19 at illness onset were fever, fatigue, dry cough, myalgia, and dyspnea 3 . In addition, some patients might suffer from headache, dizziness, abdominal pain, diarrhea, nausea, and vomiting 3 . Onset of disease may lead to progressive respiratory failure due to alveolar damage and even death 4 . Scientists then isolated a novel coronavirus from human airway epithelial cells, which was named 2019-nCoV 5 . Lu et al. 6 found that 2019-nCoV was closer to bat-SL-CoVZC45 and bat-SL-CoVZXC21 at the whole-genome level, and the external subdomain of the 2019-nCoV receptor-binding domain (RBD) was more similar to that of severe acute respiratory syndrome (SARS) coronavirus (SARS-CoV). Study of Zhou et al. 4 indicated that the angiotensin-converting enzyme II (ACE2) is likely the cell receptor of 2019-nCoV, which were also the receptor for SARS-CoV and HCoV-NL63 7,8 . Zhou et al. 4 also proved that 2019-nCoV does not use other coronavirus receptors, aminopeptidase N, and dipeptidyl peptidase 4. The study of Xu et al. 9 found that the RBD domain of the 2019-nCoV S-protein supports strong interaction with human ACE2 molecules. These findings suggest that the ACE2 plays an important role in cellular entry, thus ACE2-expressing cells may act as target cells and are susceptible to 2019-nCoV infection 10 . The expression and distribution of the ACE2 in human body may indicate the potential infection routes of 2019-nCoV. Through the developed single-cell RNA sequencing (scRNA-Seq) technique and single-cell transcriptomes based on the public database, researchers analyzed the ACE2 RNA expression profile at single-cell resolution. High ACE2 expression was identified in type II alveolar cells (AT2) of lung [10] [11] [12] , esophagus upper and stratified epithelial cells, absorptive enterocytes from ileum and colon 12 , cholangiocytes 13 , myocardial cells, kidney proximal tubule cells, and bladder urothelial cells 10 . These findings indicated that those organs with high ACE2-expressing cells should be considered as potential high risk for 2019-nCoV infection 10 . In order to investigated the potential routes of 2019-nCov infection on the mucosa of oral cavity, we explored whether the ACE2 is expressed and the ACE2-expressing cell composition and proportion in oral cavity based on the public bulk RNA-seq profiles from two public databases and single-cell transcriptomes from an independent data generated in-house. The result showed that the ACE2 could be expressed in the oral cavity, and was highly enriched in epithelial cells. Moreover, among different oral sites, ACE2 expression was higher in tongue than buccal and gingival tissues. These findings indicate that the mucosa of oral cavity may be a potentially high risk route of 2019-nCov infection. Public bulk RNA-seq dataset analysis NA-seq profile data of 13 organs including 695 para-carcinoma normal tissues as control from public TCGA were obtained for our analysis, and Fig. 1a showed that ACE2 could be expressed in various organs, the mean expression of different organs could be found in Table 1 . According to the mean expression of ACE2, the mucosa of oral cavity could express ACE2, and the results were validated by the data of normal tissues from the FANTOM5 CAGE dataset (Fig. 1b) . To investigate the ACE2 expression on mucosa of oral cavity, we looked into the ACE2 expression in different oral sites. According to the site information provided by the TCGA, among the 32 adjacent normal tissues, 13 tissues located in the oral tongue, 2 tissues located in the base of tongue, 3 tissues located in the floor of mouse, and 14 tissues did not definite the site and were just put into the category of oral cavity. The mean expression distribution of different sites was shown in Fig. 1c . When we combined the base of tongue, floor of mouth and oral cavity as other sites, and compared them with oral tongue, we found the obvious tendency that the mean expression of ACE2 was higher in oral tongue (13 tissues) than others (19 tissues) (Fig. 1d ), while may due to the limitation of the sample size, the p value was not significant (P = 0.062). Single cell RNA-seq analysis of oral tissues Single cell RNA-seq was utilized for four oral tissues, and the data was analyzed to confirm the above results and assess the cell type-specific expression of ACE2. After the data preprocessing (shown in section "Materials and methods"), 22 969 cells were acquired and 7 cell types were identified (Fig. 2a) , including epithelial cells (marker genes including SFN, KRT6A, and KRT10), fibroblasts (marker genes including FAP, PDPN, COL1A2, DCN, COL3A1, COL6A1), T cells (marker genes including CD2, CD3D, CD3E, and CD3G), macrophages (marker genes including CD163, CSF1R, CD68, and FCGR2A), mast cells (marker genes including CMA1, MS4A2, TPSAB1, TPSB2), B cells (marker genes including SLAMF7, FCRL5, and CD79A) and endothelial cells (marker genes including PECAM1, VWF, and ENG). The heatmap of main cell markers across the cell types can be found in Fig. 2b . According to Fig. 2c , d, we confirmed the ACE2 was expressed in oral tissues (0.52% ACE2-positive cells), and higher in oral tongue than buccal and gingival tissues (95.86% ACE2-positive cells located in oral tongue). Figure 2e shows that the ACE2-positive cells could be found in oral tissues including epithelial cells (1.19% ACE2-positive cells), T cells (<0.5%), B cells (<0.5%), and fibroblast (<0.5%), and the ACE2 was highly enriched in epithelial cells, of which 93.38% ACE2positive cells belong to epithelial cells (Fig. 2f) . The above results indicated that the ACE2 could be expressed on the epithelial cells of the oral mucosa and highly enriched in tongue epithelial cells. In the last two decades, coronavirus has caused two large-scale pandemics, SARS in 2002 and the Middle East respiratory syndrome In this study, the analysis of public bulk-seq RNA datasets showed that the mucosa of oral cavity could express the ACE2 and was higher in tongue than other oral sites. The results of this study were consistent with the study of Zou et al. 10 in general, many organs with higher expression of ACE2 than lung, such as intestine, heart, and kidney. According to the study of Zhao et al. 11 , the ACE2 expression in lung is concentrated in a small population of type II alveolar cells (AT2), that may cause the relatively low ACE2 expression of lung in bulk-seq RNA datasets analysis. Even though, the result of Zou et al. indicated that the respiratory tract should also be considered as a vulnerable target to 2019-nCoV infection 10 . The results of our single cell RNA-seq profiles validated the ACE2 expression in oral cavity, and the level of ACE2 expression in oral tissues was higher in tongue than buccal or gingival tissues. Furthermore, we have also demonstrated that the ACE2-positive cells were enriched in epithelial cells, which was also reported by previous study [10] [11] [12] 16 . These findings indicated that oral cavity could be regarded as potentially high risk for 2019-nCov infectious susceptibility. Interestingly, we found that the ACE2 also expressed in lymphocytes within oral mucosa, and similar results were found in various organs of the digestive system and in lungs 11, 12 . Whether those facts have reminded the 2019-nCoV attacks the lymphocytes and leads to the severe illness of patients needs more in vitro and in vivo evidence and validations, though the proportion of ACE2-positive lymphocytes is quite small. Previous studies have investigated the ACE2 mRNA and protein expression in various tissues by bulk samples 17, 18 , however, the distribution of ACE2 through bulk data could not indicate the cell type-specific expression of ACE2. Recently developed single-cell RNAsequencing technology enabled the generation of vast amounts of the transcriptomic data at cellular resolution 19 . The ACE2 expression profile in various organs, tissues, and cell types, provides the bioinformatics evidence for the potential infection routes of 2019-nCov, which might also be associated with presented symptoms. Although studies have reported multiple symptoms of hospitalized patients with 2019-nCoV infection 3, 20 , some cases at home might be asymptomatic. It is worth noting that, a previous study showed that 99% of the patients had no clinical manifestation of oral human papillomavirus (HPV), but HPV DNA was detected in 81% of oral mucosa samples, and anti-HPV IgA was detected in the saliva of 44% of the patients 21 . Likewise, although 2019-ncov infection hardly presented oral symptoms, the ACE2 expression in the oral cavity indicated that the oral infection route of 2019-nCoV cannot be excluded. Moreover, a latest pilot experiment showed that 4 out of 62 stool specimens tested positive to 2019-nCoV, and another four patients in a separate cohort who tested positive to rectal swabs had the 2019-nCoV being detected in the gastrointestinal tract, saliva, or urine 20 . Thus, our results support that in addition to the respiratory droplets and direct contact, fecal-oral transmission might also be the route of transmission of 2019-nCoV. Our results are mainly based on public datasets and single cell RNA-sequencing data of in-house oral tissues with minimal diseased lesion which from our previous project found no significant expression difference among the common epithelial markers in our past study and other previous study [22] [23] [24] . It is warrant that further histological methods are used to confirm our results and enhance the persuasion of the conclusion. The ACE2-expressing cells in oral tissues, especially in epithelial cells of tongue, might provide possible routes of entry for the 2019-nCov, which indicate oral cavity might be a potential risk route of 2019-nCov infection. Those preliminary findings have explained the basic mechanism that the oral cavity is a potentially high risk for 2019-nCoV infectious susceptibility and provide a piece of evidence for the future prevention strategy in clinical practice as well as daily life. Bulk RNA-seq data of para-carcinoma normal tissues which were taken as control tissues in the studies were downloaded from The Cancer Genome Atlas (TCGA; https://www.cancer.gov/tcga), and 695 para-carcinoma normal tissues distributed in different organs were obtained for this study, which included intestine (51 tissues), kidney (129 tissues), stomach (35 tissues), bile duct (9 tissues), liver (50 tissues), oral cavity (32 tissues), lung (110 tissues), thyroid (59 tissues), esophagus (11 tissues), bladder (19 tissues), breast (113 tissues), uterus (25 tissues), and prostate (52 tissues). The RNA-seq data were batch effects normalized and log2-transformed for the subsequent analysis. Violin plot was used to show the distribution of ACE2 expression among different organs, t test was performed to compare the ACE2 expression between two different groups shown in boxplot. Besides, Bulk RNA-seq data of normal tissues were downloaded from Functional Annotation of The Mammalian Genome Cap Analysis of Gene Expression (FANTOM5 CAGE) dataset 25 , as only two samples of this dataset which owns 60 samples in total located in the tongue, we just downloaded 14 organ types to validate the ACE2 expression in oral cavity with bar plot, including colon, ovary, breast, cerebellum, epididymis, esophagus, gallbladder, heart muscle, kidney, liver, lung, pancreas, prostate, and tongue. All analyses were performed in R (R version 3.6.0) and significant level was set as 0.05. Oral tissues of in-house cohort Due to our previous project about oral potential malignant disorders, four tissues of oral mucosa were obtained from patients after informed consent and ethical approval from West China Hospital of Stomatology, Sichuan University. These oral tissues had been sent for single cell RNA sequence. The four were taken from three patients with an average age of 50, which were all diagnosed as hyperkeratosis without dysplasia by pathologists, just showing an increase in cell number in the spinous layer and/or in the basal/ parabasal cell layers without cellular atypia, and its genetic profiles would be much more closed to normal tissue than malignant Single cell RNA-sequencing library preparation To generate single-cell Gel Beads-in-Emulsion (GEMs), the 10× chromium platform was used to capture and barcode cells. Cells were partitioned into GEMs along with gel beads coated with oligonucleotides and cDNAs with both barcodes were amplified, and a library was constructed using the 10× Genomics Chromium Single Cell Kit (v3 chemistry) for each sample. The resulting libraries were sequenced on an Illumina NovaSeq 6000 System. Single cell RNA-seq data preprocessing and analysis The FASTQ files were analyzed with the Cell Ranger Software Suite (version 3.1; 10× Genomics). The Seurat (version 3.0) was applied to read the gene-barcode matrix of four tissues. To control quality, we removed cells with <200 genes and 500 UMI counts, and as well as the cells with mitochondrial content higher than 5%. Besides, the genes detected in <3 cells were filtered out. The "sctransform" wrapper in Seurat was applied to normalize the data and remove confounding sources of variation, the "IntegrateData" was used for integrated the Seurat objects from four tissues. The uniform manifold approximation and projection (UMAP) was used for dimensionality reduction and clustering the cells, cell types were assigned based on their canonical markers. UMAP plots, heatmap, and violin plots were generated with Seurat in R.
-------------------------------------------------------


NEW PAPER


South Korea, and attempts have been made to isolate the pathogen from these patients. Methods: Upper and lower respiratory tract secretion samples from putative patients with COVID-19 were inoculated onto cells to isolate the virus. Full genome sequencing and electron microscopy were used to identify the virus. Results: The virus replicated in Vero cells and cytopathic effects were observed. Full genome sequencing showed that the virus genome exhibited sequence homology of more than 99.9% with SARS-CoV-2 which was isolated from patients from other countries, for instance China. Sequence homology of SARS-CoV-2 with SARS-CoV, and MERS-CoV was 77.5% and 50%, respectively. Coronavirus-specific morphology was observed by electron microscopy in virus-infected Vero cells. Conclusion: SARS-CoV-2 was isolated from putative patients with unexplained pneumonia and intermittent coughing and fever. The isolated virus was named BetaCoV/Korea/KCDC03/2020.
Coronavirus is an RNA virus consisting of positive-sense single-stranded RNA of approximately 27-32 kb. Coronavirus belong to the family Coronaviridae, which comprises of alpha, beta, delta, and gamma coronaviruses [1, 2] . As the name indicates, the spherical external spike protein displays a characteristic crown shape when observed under an electron microscope [3, 4] . The virus is known to infect a wide range of hosts including humans, other mammals, and birds. Infected hosts exhibit different clinical courses, ranging from asymptomatic to severe symptoms in their respiratory, digestive, and genital organs [1, 2] . There are 6 known coronaviruses that typically cause infection in humans. Among these, coronavirus 229E, OC43, NL63, and HKU1 generally cause mild cold-like symptoms, whereas severe acute respiratory syndrome-coronavirus (SARS-CoV) in 2003, and Middle East respiratory syndrome-coronavirus (MERS-CoV) in 2012, caused severe respiratory diseases such as pneumonia and death [5, 6] . Following the first outbreaks of unexplained pneumonia in *Corresponding author: Myung-Guk Han Wuhan, China, in late 2019, a new coronavirus was identified as the causative agent in January 2020 [7] . As of February 11 th , 2020, a total of 45,000 cases of pneumonia, have been reported from 26 countries, including China. Approximately 96.8% of all cases have been reported in China, and patients in the Hubei province account for 75% of all cases [8] [9] [10] . In Korea, the first case was reported on January 20, 2020, when SARS-CoV-2 was detected in a traveler entering Korea from Wuhan, China [11] . This study reports the full genome sequencing of SARS-CoV-2 isolated from putative the 2019 novel coronavirus disease  The optimal concentration of primers and probes, which were synthesized using a published sequence [12] , was determined with the RNA transcripts of SARS-CoV. The primer and probe sequences used for RNA-dependent RNA polymerase gene detection were: 5′-GTGARATGGTCATGTGTGGCGG-3' (Forward),  T h e v i r u s wa s i s o l a te d f ro m n a s o p h a r y n ge a l a n d oropharyngeal samples from putative COVID-19 patients. Oropharyngeal samples were diluted with viral transfer medium containing nasopharyngeal swabs and antibiotics (Nystadin, penicillin-streptomycin 1:1 dilution) at 1:4 ratio and incubated for 1 hour at 4°C, before being inoculated onto Vero cells. Inoculated Vero cells were cultured at 37°C, 5% CO 2 in 1× Dulbecco's modified Eagle's medium (DMEM) supplemented with 2% fetal bovine serum and penicillin-streptomycin. Virus replication and isolation were confirmed through cytopathic effects, gene detection, and electron microscopy. Viral culture of SARS-CoV-2 was conducted in a biosafety Level-3 facility according to laboratory biosafety guidelines of Korea Centers for Disease Control and Prevention. Using reverse transcriptase, cDNA was synthesized from RNA extracted from the cultured cell medium in which the virus was replicated. A next generation sequencing (NGS) library was constructed after amplifying the full-length genes of the isolates using the synthesized cDNA and primers designed based on published SARS-CoV-2 DNA sequence. The prepared library was purified and analyzed with Miseq 150 PE. De novo assembly was performed on the sequenced product using Megahit to secure a full-length genome. Gene sequencing was performed using CLC Main Workbench 7.9.1. Alignment was conducted using human and animal coronavirus sequences registered in Global Initiative on Sharing All Influenza Data (GISAID) and NCBI GenBank. The phylogenetic tree was analyzed using MEGA6 with the neighbor-joining method, maximum composite likelihoodparameter distance matrix, and bootstrap values of 1,000 replicates. For transmission electron microscopy, the inoculated cells were prefixed by incubating in 2% paraformaldehyde and 2.5% glutaraldehyde in 0.1 M phosphate buffer (pH 7.4) to prevent the autolysis of the cells infected with virus. To minimize the chemical reaction between pre-and post-fixation, the slides were washed 3 times using the same buffer as in the fixative solution and post-fixed with 1% osmium tetroxide. After washing 3 times with deionized water, en bloc staining was performed using 0.5% uranyl acetate. Thereafter, 30%, 50%, 70%, 80%, 90%, and 100% ethanol were used sequentially in ascending concentration for dehydration, which was substituted with propylene oxide. The slides were then embedded in Epon812 plastic resin, and polymerized at 70°C for 48 hours. The prepared plastic block was cut to 70-nm thick sections using an ultramicrotome and mounted on a 100mesh nickel grid, and electrostained with 5% uranyl acetate. The sections were observed with a transmission electron microscope (Libra120, Carl Zeiss, Germany) at an acceleration voltage of 120 kV [13] [14] [15] . copies in the samples before inoculation was 7.6 × 10 8 and 3.9 × 10 8 copy/mL, respectively, and increased by 10-70-fold to 5.4 × 10 10 and 4.2 × 10 9 , respectively, in the cell culture supernatants. The structure of the virus in the cytoplasm of 3-day postinoculation cells was examined by electron microscopy ( Figure 2 ). Coronavirus-specific morphology was observed. Virus particle size ranged from 70-90 nm and the virus was observed in a wide range of intracellular organelles, especially in vesicles. After inoculating cells with the nasopharyngeal and oropharyngeal samples, RNA was extracted from the virusreplicated cell culture medium. The RNA was amplified with primers for full-length gene analysis, and NGS was performed 
-------------------------------------------------------


NEW PAPER


A new coronavirus (CoV) identified as COVID-19 virus is the etiological agent responsible 31 for the 2019-2020 viral pneumonia outbreak that commenced in Wuhan 1-4 . Currently there 32 is no targeted therapeutics and effective treatment options remain very limited. In order to 33 rapidly discover lead compounds for clinical use, we initiated a program of combined 34 structure-assisted drug design, virtual drug screening and high-throughput screening to 35 identify new drug leads that target the COVID-19 virus main protease (M pro ). M pro is a key 36 CoV enzyme, which plays a pivotal role in mediating viral replication and transcription, 37 making it an attractive drug target for this virus 5,6 . Here, we identified a mechanism-based 38 inhibitor, N3, by computer-aided drug design and subsequently determined the crystal 39 structure of COVID-19 virus M pro in complex with this compound. Next, through a 40 combination of structure-based virtual and high-throughput screening, we assayed over 41 10,000 compounds including approved drugs, drug candidates in clinical trials, and other 42 pharmacologically active compounds as inhibitors of M pro . Six of these inhibit M pro with 43 IC50 values ranging from 0.67 to 21.4 μM. Ebselen also exhibited strong antiviral activity 44 in cell-based assays. Our results demonstrate the efficacy of this screening strategy, which 45 can lead to the rapid discovery of drug leads with clinical potential in response to new 46 infectious diseases where no specific drugs or vaccines are available. 47 48 author/funder. All rights reserved. No reuse allowed without permission. : bioRxiv preprint CoVs infect humans and other animal species, causing a variety of highly prevalent and 49 severe diseases, including Severe Acute Respiratory Syndrome (SARS) and Middle East 50 Respiratory Syndrome (MERS) 7 . The COVID-19 virus genome is comprised of ~30,000 51 nucleotides; its replicase gene encodes two overlapping polyproteins, pp1a and pp1ab, 52 required for viral replication and transcription 3,4 . The functional polypeptides are released 53 from the polyproteins by extensive proteolytic processing, predominantly by a 33.8-kDa 54 main protease (M pro ), also referred to as the 3C-like protease. M pro digests the polyprotein 55 at no less than 11 conserved sites, starting with the autolytic cleavage of this enzyme itself 56 from pp1a and pp1ab 8 . The functional importance of M pro in the viral life cycle, together 57 with the absence of closely related homologues in humans, identify the M pro as an attractive 58 target for antiviral drug design 9 . 59 To facilitate the rapid discovery of antiviral compounds with clinical potential, we 60 developed a strategy combining structure-assisted drug design, virtual drug screening and 61 high-throughput screening to repurpose existing drugs to target COVID-19 virus M pro . 62 Establishing a high-throughput activity assay 63 Recombinant COVID-19 virus M pro with native N and C termini was expressed in 64 Escherichia coli and subsequently purified (Extended Data Fig. 1a, b). The molecular 65 weight of COVID-19 virus M pro as determined by mass spectroscopy is 33797.0 Da, 66 consistent with its theoretical molecular weight 33796.8 Da. In order to characterize its 67 enzymatic activity and to carry out high-throughput screening of inhibitors, we developed 68 a fluorescence resonance energy transfer (FRET) assay. To do this, a fluorescently labeled 69 substrate, MCA-AVLQ↓SGFR-Lys(Dnp)-Lys-NH2, derived from the N-terminal auto-70 cleavage sequence of the viral protease was designed and synthesized for time-dependent
kinetic analysis (Extended Data Fig. 1e ). The catalytic efficiency (kcat/Km) for COVID-19 72 virus M pro was measured to be 28,500 M −1 s −1 which is slightly higher than that for SARS- 73 CoV M pro (kcat/Km=26,500 M −1 s −1 ) 10 , but over 30-fold higher than that of human rhinovirus 74 3C protease (kcat/Km=920 M −1 s −1 ) 11 . In a previous study, we designed a Michael acceptor inhibitor N3 using computer-aided 77 drug design (CADD) (Extended Data Fig. 1c) , which can specifically inhibit multiple CoV utilized to evaluate the inhibition as an approximation of the pseudo second-order rate be 11,300±880 M -1 s -1 , suggesting this Michael acceptor has potent inhibition. 95 The crystal structure of COVID-19 virus M pro in complex with N3 96 In order to elucidate the inhibitory mechanism of this compound, we determined the crystal 97 structure of COVID-19 virus M pro in complex with N3 to 2.1-Å resolution. The asymmetric 98 unit contains only one polypeptide (Extended Data Table 1 ). However, two of these 99 associate to form a dimer by a crystallographic 2-fold symmetry axis (the two molecules 100 are designated protomer A and B) (Fig. 1b) . All residues (residues 1-306) are visible in 101 electron density maps. Each protomer is composed of three domains (Fig. 1a) . Domains Ⅰ 102 (residues 8-101) and Ⅱ (residues 102-184) have an antiparallel β-barrel structure. Domain 103 Ⅲ (residues 201-303) contains five α-helices arranged into a largely antiparallel globular 104 cluster, and is connected to domain Ⅱ by means of a long loop region (residues 185-200). strand155-168 on one side, and with residues 189-191 of the loop linking domains Ⅱ and Ⅲ . 111 Here we detail the specific interactions of N3 with M pro (Fig. 1c, d) . The electron density 112 shows that the Sγ atom of C145-A forms a covalent bond (1.8-Å) with the Cβ of the vinyl 113 group, confirming that the Michael addition has occurred. The S1 subsite has an absolute 114 requirement for Gln at the P1 position. The side chains of F140-A, N142-A, E166-A, H163-115 A, H172-A, S1-B (from protomer B), and main chains of F140-A and L141-A are involved 116 in S1 subsite formation, which also includes two ordered water molecules (named W1 and W2). The lactam at P1 inserts into the S1 subsite and forms a hydrogen bond with H163-118 A. The side chain of Leu at P2 site inserts deeply into the hydrophobic S2 subsite, which 119 consists of the side chains of H41-A, M49-A, Y54-A, M165-A, and the alkyl portion of the 120 side chain of D187-A. The side chain of Val at P3 is solvent-exposed, indicating that this the S1′ site, forming van der Waals interactions with T24-A and T25-A. In addition, N3 126 forms multiple hydrogen bonds with the main chain of the residues in the substrate-binding 127 pocket, which also helps lock the inhibitor inside the substrate-binding pocket. conformation. This may be attributed to an ordered water (W1) in SARS-CoV M pro -N1 133 structure, which makes a long distance hydrogen bond to the carboxylate oxygen of the 134 ester and also forms two hydrogen bonds from the backbone NH of G143 and the side 135 chain of N142. In our previous study, we proposed that all the CoV M pro s share a highly 136 conserved substrate-recognition pocket, which could serve as a drug target for the design 137 of broad-spectrum inhibitors 12 . The recent discovery of new CoVs and accumulation of 138 structural data for CoV M pro s from various species provided the opportunity to further 139 examine this hypothesis. Superposition of the 12 crystal structures of M pro s 12-21 have shown that the most variable regions were the helical domain Ⅲ and surface loops, but the 141 substrate-binding pockets located in a cleft between domains Ⅰ and Ⅱ are still highly 142 conserved among all CoV M pro s, suggesting the antiviral inhibitors targeting this site should 143 have wide-spectrum anti-CoV activity (Fig. 2b, c) . μM. However, in a detergent-based assay 24 , TDZD-8 was found to be an aggregate-based 163 author/funder. All rights reserved. No reuse allowed without permission. The copyright holder for this preprint (which was not peer-reviewed) is the . https://doi.org/10.1101/2020.02.26.964882 doi: bioRxiv preprint inhibitor, which might non-specifically inhibit M pro (Extended Data Fig. 5 ) and was 164 therefore not considered for further investigation. Since our structural data is based on N3, 165 we investigated if molecular docking could predict how other six inhibitors bind to this 166 protein. In all cases, reasonable docking poses were found, demonstrating that they could 167 fit inside the substrate-binding pocket (Extended Data Fig. 6 ). Next, we set out to identify 168 the potential covalent inhibitors among these compounds through tandem MS/MS analysis. The MS/MS data shows that ebselen, PX-12 and carmofur are all able to covalently bind Fig. 7) . Since ebselen has even stronger inhibition than the others, 173 there is a possibility that ebselen could also inhibit M pro through non-covalent binding. compound has been investigated for the treatment of multiple diseases, such as bipolar 186 author/funder. All rights reserved. No reuse allowed without permission. The copyright holder for this preprint (which was not peer-reviewed) is the . https://doi.org/10.1101/2020.02.26.964882 doi: bioRxiv preprint disorders 25 and hearing loss 26,27 . Ebselen has extremely low cytotoxicity (LD50 in rats > 187 4,600 mg/kg, per os) 28 and its safety in humans has been evaluated in a number of clinical 188 trials 26,27,29 . These data strongly suggest the clinical potential of ebselen for CoV treatment. It is also interesting that cinanserin displayed moderate inhibition against COVID- 19 virus 190 with an EC50 value of 20.61 μM from qRT-PCR analysis (Extended Data Fig. 4) , which is 191 superior to that in the enzymatic inhibition assay, suggesting that cinanserin might have phenotypic screening has proven to be valuable 30 , but the complexity of this approach is 207 not readily compatible with high-throughput pipelines, and it cannot identify the molecular 208 target or mechanism of action 31 . In this study, the convergence of structure-based ab initio 209 author/funder. All rights reserved. No reuse allowed without permission. The copyright holder for this preprint (which was not peer-reviewed) is the . https://doi.org/10.1101/2020.02.26.964882 doi: bioRxiv preprint drug design, virtual screening and high-throughput screening proved to be an efficient 210 strategy to find antiviral leads against COVID-19 virus. The methods presented here can 211 greatly assist in the rapid discovery of drug leads with clinical potential in response to new 212 emerging infectious diseases that currently lack specific drugs and vaccines. The copyright holder for this preprint (which was not peer-reviewed) is the . https://doi.org/10.1101/2020.02.26.964882 doi: bioRxiv preprint The copyright holder for this preprint (which was not peer-reviewed) is the . https://doi.org/10.1101/2020.02.26.964882 doi: bioRxiv preprint The copyright holder for this preprint (which was not peer-reviewed) is the . https://doi.org/10.1101/2020.02.26.964882 doi: bioRxiv preprint for ebselen and N3 in the plaque-reduction assay, respectively. All data are shown as mean 324 ± s.e.m., n = 4 technical replicates. 325 326 author/funder. All rights reserved. No reuse allowed without permission. The copyright holder for this preprint (which was not peer-reviewed) is the . https://doi.org/10.1101/2020.02.26.964882 doi: bioRxiv preprint The full-length gene encoding COVID-19 virus M pro was optimized and synthesized for The copyright holder for this preprint (which was not peer-reviewed) is the . https://doi.org/10.1101/2020.02.26.964882 doi: bioRxiv preprint initial complexes. Finally, these complexes were further optimized and re-scored by using 419 MM-GBSA module 42 of Schrödinger, and the residues within 5 Å around the ligand were 420 refined. discussion. We also thank the staff from beamlines BL17U1, BL18U1 and BL19U1 at 554 author/funder. All rights reserved. No reuse allowed without permission. The copyright holder for this preprint (which was not peer-reviewed) is the . https://doi.org/10.1101/2020.02.26.964882 doi: bioRxiv preprint Shanghai Synchrotron Radiation Facility (China). This work was supported by grants from The copyright holder for this preprint (which was not peer-reviewed) is the . https://doi.org/10.1101/2020.02.26.964882 doi: bioRxiv preprint
-------------------------------------------------------


NEW PAPER



A novel coronavirus, severe acute respiratory syndrome coronavirus 2 (SARS-COV-2) associated with severe human infected disease (COVID-19) outbroke starting from Wuhan city, in China. 1 Most recently, Wang D et al reported 138 hospitalized SARS-COV-2 pneumonia cases with a hospital-associated transmission rate of 41%, among whom 70% were medical staffs. 2 To date, nearly 80, 000 people have been infected by COVID-19, including over 3000 medical staffs. Therefore, in order to curb the spread of SARS-COV-2, and reduce the nosocomial infection, the exact routes of transmission need to be further studied and confirmed. Although primarily affecting the respiratory tract, it also involves extra-pulmonary sites, including the digestive tract and other organs. 3 SARS-CoV-2 infected patients develop respiratory illness, with the first symptoms of fever, cough and fatigue that quickly progress to pneumonia. COVID- 19 were also contributed to extra-pulmonary manifestations in a number of patients at the onset of the illness, such as headache, diarrhoea, nausea and vomiting, 4 or even presented with asymptomatic infection. 8 Despite the intensive work that has been focused on an understanding of the transmission of SARS-CoV-2, the spread mode of the SARS-CoV-2 is unclear. To our knowledge, limited investigations have been conducted to date into the clinical features of SARS-COV-2 in ocular surface. A few of COVID-19 patients with conjunctivitis, or even as the first symptom, make the early diagnosis and effective protection a great difficulty. A recent research has demonstrated that SARS-COV-2 can be detected in the conjunctival sac of patients with COVID-19 patients or suspected. 5 However, none of these patients had ocular symptoms. We therefore conducted a study to describe the clinical spectrum of ocular symptoms and laboratory test in conjunctival swab samples, we found a rare case of nosocomial SARS-COV-2 infection with conjunctivitis in a nurse, which suggests that ocular transmission may be a potential route of nosocomial transmission of the SARS-COV-2. All rights reserved. No reuse allowed without permission. the author/funder, who has granted medRxiv a license to display the preprint in perpetuity. The copyright holder for this preprint (which was not peer-reviewed) is Serum was collected in a serum separator tube and then centrifuged in accordance with CDC guidelines. Laboratory-confirmed cases were identified with the criteria of at least one positive result from a respiratory specimen using RT-PCR assays. All patients invited to participate in the study provided consent for the nasopharyngeal and conjunctival swab samples with synthetic fiber swabs. Then, these swab samples were detected via real-time RT-PCR assays as previously described. 7 forward primer 5′-ACTTCTTTTTCTTGCTTTCGTGGT-3′; reverse primer 5′-GCAGCAGTACGCACACAATC-3′; and the probe 5′CY5-CTAGTTACACTAGCCATCCTTACTGC-3′BHQ1. The statistical analysis was performed using SPSS software version 22.0 (SPSS Inc, All rights reserved. No reuse allowed without permission. the author/funder, who has granted medRxiv a license to display the preprint in perpetuity. The copyright holder for this preprint (which was not peer-reviewed) is Only two patients (2.78%) with conjunctivitis were identified from 72 patients with a laboratory confirmed COVID-19. However, SARS-CoV-2 was found in ocular discharges by RT-PCR only in one patient, the symptoms and laboratory test were shown in Fig. 2 . Briefly, a 29-year-old nurse working in the Emergency Department at Tongji hospital, Wuhan City, China was referred to the Department of Ophthalmology at Tongji Hospital on February 1st, 2020 due to excessive tearing and redness in both eyes (Fig. 3A) . No other systemic symptoms except for a moderate fever of 38.2℃ was reported on January 31st, 2020. The ocular examination revealed conjunctival congestion and watery discharges in both eyes with normal best corrected visual acuity, normal corneal epithelium, quiescent anterior chamber and no tenderness or enlargement of the preauricular lymph node. We therefore excluded the possibility of conventional conjunctivitis, such as bacterial conjunctivitis, hemorrhagic conjunctivitis, allergic conjunctivitis, according to her clinical symptom and sign. She clarified medical N95 respirators continuously wore during operation, while occasionally worked with a dislocated eye mask touching her eyelids. Considering the occupational exposure by SARS-COV-2, SARS-COV-2 related assay were arranged. Surprising, the chest CT showed multiple peripheral ground-glass opacities in both lungs (Fig. 3D-F) . The conjunctival and oropharyngeal swabs tested All rights reserved. No reuse allowed without permission. the author/funder, who has granted medRxiv a license to display the preprint in perpetuity. The copyright holder for this preprint (which was not peer-reviewed) is . https://doi.org/10.1101/2020.02.26.20027938 doi: medRxiv preprint for SARS-COV-2 were both positive. The blood Rt showed a normal total white blood cells with elevated monocyte counts. On the basis of her epidemiologic characteristics, clinical manifestations, chest images, and laboratory findings, this patient was diagnosed with SARS-COV-2 infected acute viral conjunctivitis and pneumonia. Before admission, the patient reported persistent a 4-day history of conjunctivitis and a 3-day history of fever. Ganciclovir eye drops was used to control her conjunctivitis, the patient's conjunctivitis vital signs remained stable during her home quarantine, apart from the development of intermittent fevers, accompanied by periods of cough from day 4 to 9 post of illness (Fig. 2) . Although, SARS-COV-2 RT-PCR assay for the nasopharyngeal and conjunctival swab were negative 5 days after the conjunctivitis taken a turn for the better, a fever reappear, and the Chest Radiographs showed that her pneumonia aggravated at day 10 post illness. Then, she was arranged to the hospital treated according to the guidance of CDC, and then discharged 11 days post hospitalization, accompanied with a slight cough. the author/funder, who has granted medRxiv a license to display the preprint in perpetuity. The copyright holder for this preprint (which was not peer-reviewed) is In the current study, we found that two patients (2.78%) with conjunctivitis was identified from 72 patients with a laboratory confirmed COVID-19. However, SARS-CoV-2 was found in ocular discharges by RT-PCR only in one COVID-19 patient. Although the incidence of conjunctivitis is extremely low, these results demonstrated that SARS-CoV-2 have shown a capacity to use the eye as a portal of entry and cause ocular disease. To our knowledge, several anatomical and mucosal immune properties permitted the eye as both a potential site of virus infected site as well as a gateway for respiratory infection. [12] [13] In coincidence with our result, SARS-CoV was detected by RT-PCR in tear samples from three probable cases. 14 Resemble as SARS-CoV, entry of SARS-CoV-2 via the host functional receptor is mediated by ACE2. 15 Moreover, Sun and colleagues 16 found that ACE2 expressed in human cornea and conjunctival tissues, which providing strong evidence for our diagnosis. Furthermore, Dr. Guangfa Wang, a member of the national expert panel on pneumonia, reported that he was infected by SARS-COV-2 during the inspection in Wuhan, through unprotected eye exposure. 17 Similarly, an anesthesiologist with insufficient eye protection confirmed with COVID-19 and also presented conjunctivitis as the initial symptom. The laboratory test revealed that the nasopharyngeal swab was positive while the conjunctival swab was negative. 5 On the contrary, our case All rights reserved. No reuse allowed without permission. the author/funder, who has granted medRxiv a license to display the preprint in perpetuity. The copyright holder for this preprint (which was not peer-reviewed) is There are several limitations need to be considered. First, the conjunctival scraping test should be done as early as possible if we find ocular symptoms in a suspected one. The mean time for conjunctival sampling in our study are 18.15 days, in which exceeded the optimal detection time. Second, how to increase the positive rate. The most commonly used chest CT examination has certain limitations for the special suspected groups, such as pregnant women. In addition, we found that 2~3 repeated tests did in suspected patients of nasopharyngeal swabs before the SARS-CoV-2 was All rights reserved. No reuse allowed without permission. the author/funder, who has granted medRxiv a license to display the preprint in perpetuity. The copyright holder for this preprint (which was not peer-reviewed) is . https://doi.org/10.1101/2020.02.26.20027938 doi: medRxiv preprint obtained, which may be related to the false-negative result of RT-PCR. Therefore, the low viral load in the ocular surface and the lower positive rate of RT-PCR makes early diagnosis of SARS-CoV-2 a challenge. Recently, Doan 21 reported that the influenza virus and rubella virus found in patients' conjunctival sac or tears by the next generation sequencing rapidly, which provides us with a feasible direction for future study. In conclusion, we suspect the incidence of SARS-COV-2 infection through the ocular surface is extremely low in the general population, however, considering the common feature of the positive cases which are occupational exposure accompanied with conjunctivitis in the early stage, the higher viral aerosol load in the hospital and more opportunities to contact with COVID-19 patients, we highlight that ocular transmission is a potential important way of occupational exposure for medical staff. Therefore, To lower the risk of SARS-COV-2 nosocomial infection, the protective goggles should be wore in all the health care workers, especially who work in the Fever Outpatient and Infection Wards. Patients with conjunctivitis in the epidemic area should also be treated seriously to rule out the COVID-19. Moreover, to date, the lower positive rate of conjunctival sac may be ascribed to the lower viral concentration, the sampling time lag, and the inefficient detection methods. Our study highlights the need for further research into the efficient detection methods of SARS-COV-2. All authors report no relevant conflicts of interest, financial or otherwise.
-------------------------------------------------------


NEW PAPER


We tested samples collected from nine patients diagnosed with coronavirus disease 2019 (COVID-19). The virus was found in urine, blood, anal swabs and oropharyngeal swabs. It is the first time for SARS-CoV-2 found in urine, though no urinary irritation was found.
2019 Novel Coronavirus (SARS-CoV-2) infection outbroke and spread to the world from 2019. By January 28, 2020, a total of 4593 global confirmed cases, of which 4537 were from China, had been reported (1) . On January 30, 2020, SARS-CoV-2 infection was declared as a global health emergency by World Health Organization (WHO). The mechanism of SARS-CoV-2 infection and organ invasion are unclear, which directly leads to difficulties and blindness in clinical diagnosis and treatment. We aim to detect SARS-CoV-2 nucleic acid from urine, blood, anal swab and oropharyngeal swab samples. We hope to provide evidences of the multiple organ invasion of the virus and figure out its relation with clinical manifestations. Nine patients confirmed diagnosed with SARS-CoV-2 infection(2) were included in this prospective study. Urine, blood, anal swabs and oropharyngeal swabs from All rights reserved. No reuse allowed without permission. author/funder, who has granted medRxiv a license to display the preprint in perpetuity. The copyright holder for this preprint (which was not peer-reviewed) is the . https://doi.org/10.1101/2020.02.21.20026179 doi: medRxiv preprint enrolled patients were obtained and detected SARS-CoV-2 RNA level by quantitative real-time polymerase chain reaction (qRT-PCR). Patients' demographic data and clinical characteristics were recorded. This study was approved by the institutional review board (IRB) of the Third Affiliated Hospital of Sun Yat-sen University. All patients voluntarily signed an informed consent form approved by the IRB before participation. Patient 7, a 31 years old female without any urinary irritation, had positive results of SARS-CoV-2 in both urine and oropharyngeal swab on the 7 th day after symptom onset. Patient 1 and 5 had negative results in oropharyngeal swab, on the 10 th and 15 th day after onset. Patient 8 had three positive results in blood, anal swab and oropharyngeal swab on the 3 rd day after onset. Viral load was roughly higher in anal swab than oropharyngeal swab in patient 9 (Table) . Eight of nine patients had an obvious history of epidemiological exposure to SARS-CoV-2. All patients presented with a fever. Other symptoms were cough, sore throat, fatigue and diarrhea. Four patients had a lymphopenia. Elevated C-reactive protein and erythrocyte sedimentation rate were found in 6 and 5 cases respectively. All patients had normal serum levels of procalcitonin, alanine transaminase, creatinine, myocardial enzyme and arterial partial oxygen pressure. All rights reserved. No reuse allowed without permission. author/funder, who has granted medRxiv a license to display the preprint in perpetuity. The copyright holder for this preprint (which was not peer-reviewed) is the . https://doi.org/10.1101/2020.02.21.20026179 doi: medRxiv preprint The pathogenic mechanism of SARS-CoV-2 infection is still unclear. Current evidences indicate that it can invade multiple organ systems, including the respiratory system, digestive system and hematological system (3) (4) (5) . But whether it can invade the urinary system has not been reported. In our study, urine, blood, anal swab and oropharyngeal swab from 9 patients were retested by qRT-PCR. It is the first time for the SARS-CoV-2 found in urine, though no urinary irritation was found. In addition, the quantity of the virus in the anal swab was closer to that of the oropharyngeal test, and those in the blood and urine were less than that of the oropharyngeal test, indicating that the "clearing effect" of the digestive tract on the virus is not obvious. Nevertheless, the relative symptoms, including diarrhea and urinary irritation did not happen to every patient with virus in anal swab and urine specimens. All samples were negative at the earliest time of the 10 th day after onset in that mild patient. Therefore, we believed that SARS-CoV-2 can invade the urinary system, hematological system and digestive system other than the respiratory system, not always with relative symptoms. Mild patients can be self-limiting. This will prompt clinicians to pay attention to the clinical manifestations of multiple systems, even if the corresponding clinical symptoms do not appear. In addition, it is necessary to carry out multiple examinations of various specimens to assess changes in disease and prognosis. All rights reserved. No reuse allowed without permission. author/funder, who has granted medRxiv a license to display the preprint in perpetuity. The copyright holder for this preprint (which was not peer-reviewed) is the  All authors declare no competing interests. All rights reserved. No reuse allowed without permission. author/funder, who has granted medRxiv a license to display the preprint in perpetuity. All rights reserved. No reuse allowed without permission. author/funder, who has granted medRxiv a license to display the preprint in perpetuity. The copyright holder for this preprint (which was not peer-reviewed) is the . https://doi.org/10.1101/2020.02.21.20026179 doi: medRxiv preprint
