In the latest study, a team from Zhejiang University reported the functional characteristics of 11 virus isolates from new crown patients, all of which had at least one mutation. Importantly, when Vero-E6 cells were infected, these virus isolates showed significant differences in cytopathic effect and viral load, with a difference of up to 270 times.
On April 19, local time, the research team of the National Key Laboratory of infectious disease diagnosis and treatment, the First Affiliated Hospital of Zhejiang University Medical College, the Institute of life sciences of Zhejiang University, and the Key Laboratory of pancreatic disease research of Zhejiang Province published a research paper patient-derived nutrition simpactpathogenicity of sars-cov-2 online on medrxiv, the medical preprint platform. The corresponding authors of this paper are Professor Li Lanjuan, academician of the Chinese Academy of engineering and member of the high level expert group of the National Health Commission, Jiang Chao, researcher of the Institute of life sciences of Zhejiang University, Professor Wu Nanping and Professor Zheng Min, State Key Laboratory of infectious diseases diagnosis and treatment of the First Affiliated Hospital of Zhejiang University Medical College.
At present, it is believed that the transmembrane spike glycoprotein (S-protein) mediates the virus into the host cell, and S-protein is divided into two functional units according to the structure and function of the protein, namely S1 and S2 protein subunits. S1 is responsible for binding with host cell receptor, S2 is responsible for fusion of virus and cell membrane. Both new coronavirus and SARS CoV use ACE2 to enter target cells. ACE2 was expressed in human nasal epithelial cells, lung, spermatogonia, stromal cells, supporting cells, stomach, duodenum and rectum epithelial cells.
It is mentioned that the receptor binding domain (RBD) of S protein is the most variable part of the genome of u03b2 - coronavirus, and some sites of S protein may be positively selected. However, despite the large number of mutations, the key question remains: do these mutations have any real functional impact on the pathogenicity of the new coronavirus?
The researchers believe that the above issues are essential to understand the mechanism of virus infection and guide the development strategy of drugs and vaccines to prepare for the next phase of the pandemic. To solve this problem, the researchers isolated 11 new coronavirus strains from patients admitted to the First Affiliated Hospital of Zhejiang University.
The researchers sequenced 11 virus isolates on novaseq6000 platform, and found that there were 1-5 mutations in the coding sequence of the virus isolates, and also observed a mixed population (representing quasispecies). They identified a total of 33 mutations (including 10 observed in the mixed population), 19 of which were new based on a comparison with 1111 genomic sequences provided by gisaid on March 24, 2020.
The researchers also infected Vero-E6 cells with 11 virus isolates and quantitatively evaluated their viral load at 1, 2, 4, 8, 24 and 48 hours after infection (P.I.), as well as their viral cytopathological effects (CPE) at 48 and 72 hours after infection.
Their results showed that when Vero-E6 cells were infected, the observed mutations had a direct impact on viral load and cell pathological effects, with a maximum difference of 270 times. This finding shows that the virus mutation observed in the study can significantly affect the pathogenicity of the new coronavirus, which may exist in virus strains collected around the world.
Overall, this study provides direct evidence that mutations currently occurring in the genome of the new coronavirus have the functional potential to affect the pathogenicity of the virus. In addition to the accumulated genome sequencing data, virus monitoring should be carried out at the cellular level as much as possible, the researchers cautioned. Finally, similar to influenza, although the development of drugs and vaccines is urgent, the impact of these cumulative mutations, especially ancestral mutations, needs to be taken into account to avoid potential defects.
Patients epidemiological history
The samples of 11 patients involved in this study were collected in the early stage of the outbreak in China from January 22, 2020 to February 4, 2020.
Ten of the 11 patients had a clear relationship with Wuhan. Five of them worked or traveled in Wuhan before the diagnosis. The other five had close contact with people living in Wuhan. The remaining one had contact with the new crown patients. It is worth noting that patients zju-4, zju-5 and zju-9 all attended the same business meeting attended by Wuhan colleagues.
Therefore, according to the epidemiological history of these 11 patients, they constitute the first and second generation of infection cases.
Among the 11 patients, 8 were male and 3 were female, ranging in age from 4 months to 71 years. Except for one patient, all the other patients had moderate or more severe symptoms, 3 had complications, and 1 needed ICU treatment.
As of the writing of this paper, all patients have recovered.
There are many mutations in the virus isolates collected from 11 patients
In order to evaluate the mutation spectrum of these 11 virus isolates, the researchers carried out super deep sequencing of the virus genome RNA on the platform of illuminanovaseq6000, with an average of 245 million readings / 67.16 GB per sample.
Specifically, g11083t and g26144t were found in zju-1, both of which are considered to be the original mutations of a large population of viruses. C8782t and t28144c mutations were found in zju-2 and zju-8. These two mutations are considered to be the original mutations of another large population virus isolate.
T22303g mutations were found in five virus isolates, zju-2, zju-5, zju-9, zju-10 and zju-11. Zju-5 and zju-9 were exposed to the same potential infection source in a commercial meeting. It is worth mentioning that t22303g mutation was previously identified only in the virus strains isolated in Australia.
It is worth noting that there is a new mutation a22301c in the virus isolate of zju-4 from the patients who attended the same meeting with zju-5 and zju-9, which results in the same missense mutation as t22303g at the protein level (s247r in S protein).
Finally, zju-11 had four mutations in orf7b gene, three of which were continuous, and two mutations were introduced at protein level.
It is worth noting that the researchers believe that although the sequence data in gisaid is very helpful to track the inter individual variation of virus, we still know little about the intra individual evolution dynamics of virus. For example, in zju-4 and zju-10, the allele frequency distribution of the two independent loci is very similar, which indicates that the two loci may be related, suggesting that there are at least two haplotypes in the virus population.
Overall, although only 11 virus strains from patients were analyzed in this study, researchers still observed a variety of mutations, including several basic mutations in different major virus clusters currently circulating around the world. The mutation spectrum of this diversity is consistent with their relatively early sampling time and relatively close to Wuhan city. However, due to the limited sample size, the complete variation diversity of early viruses in Wuhan is still unclear.
The researchers discussed that a variety of mutations were found in 11 virus isolates, including two sets of ancestral mutations of two major categories of viruses currently infecting the global population. In addition, 19 of the 31 identified mutations were new, although the sampling dates were relatively early, indicating that the true diversity of the virus strains is still largely unknown.
It is worth noting that the trinucleotide mutation of zju-11 is unexpected. The researchers noted that this particular virus isolate appeared to be very powerful in the viral load and CPE determination in the study, and the patients from it were also very rare to remain positive for 45 days, recently discharged from the hospital.
They think it would be very interesting to study the functional effects of this trinucleotide mutation. It is worth noting that another trinucleotide mutation (g2881a, g2882a and g2883c) has been identified in the current database, which also leads to missense mutations at the two protein levels.
The researchers also cautioned that, in contrast to the recently reported inability to obtain live viruses from fecal samples, three virus isolates in this study were isolated from fecal samples, indicating that the new coronavirus can be replicated in fecal samples.
Phylogenetic analysis reveals different evolutionary histories
The researchers constructed a maximum likelihood phylogenetic tree containing 736 virus sequences, and the phylogenetic tree obtained is basically consistent with the updated phylogenetic analysis on gisaid.
A considerable number of ancestral mutations have been observed. Specifically, in this phylogenetic analysis, the researchers noted the following three largest clusters: first, three nucleotide mutations c241t, c14408t and a23403g (s-d614g cluster) were found in 231 virus sequences, most of which were isolated in Europe; second, two nucleotide mutations c8782t and t28144c (orf8-l84s cluster) were found in 208 virus sequences, in this analysis They are not single source (figures 2a and S3). However, in the 92 genome sequences of orf8-l84s cluster, we can find obvious subclade, which is mainly composed of virus sequences from Seattle, USA (orf8-l84s-usa-wa-clade); thirdly, we found two nucleotide mutations g11083t (l3606f in orf1a) and g26144t (g251v in orf3a), most of which are from the Netherlands and the UK.
The researchers integrated 11 virus isolates into phylogenetic analysis, and they were scattered throughout the phylogenetic space. Zju-1 was clustered with orf1a-l3606f and orf3a-g251v, which had two typical mutations at the same time. On the other hand, zju-2 clusters with zju-8 and orf8-l84s, both of which have two ancestral mutations. Due to the t22303g mutation, zju-9 and zju-11 clustered with an Australian isolate. Other strains either had fewer mutations or no new mutations with any known large population aggregation, which reflected the wide diversity of these 11 samples.
To sum up, the researchers believe that some single source virus groups do show obvious geographical patterns (especially in Europe and the United States), but this may be due to the ancestor effect of their mutations in the early stage of the pandemic.
The researchers first tested whether the virus isolate could successfully bind to Vero-E6 cells as expected, and visually identified the virus particles with the crown formed by S protein. They then infected Vero-E6 cells with all 11 patient derived virus isolates and collected the cells 1, 2, 4, 8, 24 and 48 hours after infection. At 48 hours and 72 hours, the cells DIC micrographs were taken to evaluate CPE.
They used specific real-time transcription polymerase chain reaction (RT-PCR) for orf1a, e, and N genes to detect the presence of the new coronavirus. Cycle threshold CT is used to quantify the viral load, and a lower value indicates a higher viral load. Because the results of these three genes are highly consistent, they only discuss the results of orf1a gene.
In brief, the CT values of all virus isolates remained stable at 1, 2 and 4 hours after infection, with only small fluctuations. At these earlier time points, virus particles combine to invade cells, and virus replication rarely occurs at this stage.
Eight hours after infection, the researchers observed that the CT values of zju-6, zju-7, zju-9, zju-10 and zju-11 decreased significantly, which means that the viral load increased. 24 hours after infection, it was observed that the CT values of all virus isolates except zju-2 and zju-7 decreased significantly, and some of them, zju-10 and zju-11, decreased much faster than other viruses. After 48 hours of infection, except for zju-10 and zju-11, all virus isolates decreased slightly. The researchers speculated that zju-10 and zju-11 were stable at 24 hours.
It is worth noting that the viral load of the members of orf-8-l84s cluster, zju-2 and zju-8 (mostly from Seattle, Washington, USA), decreased significantly 24 hours after infection. On the other hand, the load of zju-1 virus in the same cluster with s-d614g branch (mainly found in Europe) is 19 times that of zju-2 and zju-8. In addition, 24 hours after infection, the viral load of zju-10 and zju-2 was nearly 270 times different.
These differences became statistically significant at 48 hours and were repeatable in the analysis of gene E and N data, the paper noted. Therefore, the researchers believe that different virus isolates show significant changes in viral load when they infect Vero-E6 cells, which is determined by different mutations in their genome.
Next, the researchers tested whether an increase in viral load would cause more cell death. Under the microscope at 48 hours and 72 hours after infection, CPE or cell death rate was highly consistent with the viral load data, indicating that the higher the viral load, the higher the cell death rate.
They mentioned in the discussion that it will be very difficult to establish the relationship between genotype and phenotype in patients due to the extremely diverse clinical symptoms. In vitro cell lines provide an ideal system for examining the effects of mutations in different isolated virus strains.
They believe that although Vero-E6 cell line is not from human, the ACE2 protein of Vero-E6 cell line is highly similar to that of human, which directly proves that the new coronavirus can infect the cell line.
Source: Qiao JunJing, editor in charge of surging news