key: cord-0858952-nn0ahc6y
authors: Prince, Tessa; Dong, Xiaofeng; Penrice-Randal, Rebekah; Randle, Nadine; Hartley, Catherine; Goldswain, Hannah; Jones, Benjamin; Semple, Malcolm G.; Baillie, J. Kenneth; Openshaw, Peter J. M.; Turtle, Lance; Hughes, Grant L.; Anderson, Enyia R.; Patterson, Edward I.; Druce, Julian; Screaton, Gavin; Carroll, Miles W.; Stewart, James P.; Hiscox, Julian A.
title: Sequence analysis of SARS-CoV-2 in nasopharyngeal samples from patients with COVID-19 illustrates population variation and diverse phenotypes, placing the in vitro growth properties of B.1.1.7 and B.1.351 lineage viruses in context
date: 2021-03-30
journal: bioRxiv
DOI: 10.1101/2021.03.30.437704
sha: 9b921cef23a37e6aeebff07e1a3f0e224e553871
doc_id: 858952
cord_uid: nn0ahc6y

New variants of SARS-CoV-2 are continuing to emerge and dominate the regional and global sequence landscapes. Several variants have been labelled as Variants of Concern (VOCs) because of perceptions or evidence that these may have a transmission advantage, increased risk of morbidly and/or mortality or immune evasion in the context of prior infection or vaccination. Placing the VOCs in context and also the underlying variability of SARS-CoV-2 is essential in understanding virus evolution and selection pressures. Sequences of SARS-CoV-2 in nasopharyngeal swabs from hospitalised patients in the UK were determined and virus isolated. The data indicated the virus existed as a population with a consensus level and non-synonymous changes at a minor variant. For example, viruses containing the nsp12 P323L variation from the Wuhan reference sequence, contained minor variants at the position including P and F and other amino acids. These populations were generally preserved when isolates were amplified in cell culture. In order to place VOCs B.1.1.7 (the UK ‘Kent’ variant) and B.1.351 (the ‘South African’ variant) in context their growth was compared to a spread of other clinical isolates. The data indicated that the growth in cell culture of the B.1.1.7 VOC was no different from other variants, suggesting that its apparent transmission advantage was not down to replicating more quickly. Growth of B.1.351 was towards the higher end of the variants. Overall, the study suggested that studying the biology of SARS-CoV-2 is complicated by population dynamics and that these need to be considered with new variants. Importance SARS-CoV-2 is the causative agent of COVID-19. The virus has spread across the planet causing a global pandemic. In common with other coronaviruses, SARS-CoV-2 genetic material (genomes) can become quite diverse as a consequence of replicating inside cells. This has given rise to multiple variants from the original virus that infected humans. These variants may have different properties and in the context of a widespread vaccination program may render vaccines less ineffective. Our research confirms the degree of genetic diversity of SARS-CoV-2 in patients. By isolating viruses from these patients, we show that there is a 100-fold range in growth of even normal variants. Interestingly, by comparing this to the pattern seen with two Variants of Concern (UK and South African variants), we show that at least in cells the ability of the B.1.1.7 variant to grow is not substantially different to many of the previous variants.

SARS-CoV-2 emerged late 2019 in Wuhan, China and causes . This can be a 59 fatal infection with severe immunopathology in the respiratory system (2). The virus has since 60 isolated from a swab sampled from a patient in the UK in March 2020 ( Figure 2 ) now possessed 133 the D614G and P323L substitutions in the spike glycoprotein and NSP12, respectively. These are 134 in contrast to the B.1.1.7 variant, which emerged later in 2020 and is characterised by the 135 presence of 23 amino acid differences from the reference genome. Analysis of the virus 136 population present in the nasopharyngeal isolate of SCV2-009 illustrated the diversity associated 137 with the virus. For example, taking the P323L substitution in NSP12, out of an amino acid 138 coverage of 202, 170 amino acids mapped to L, 12 to P and 9 to F. For the D614G substitution in 139 the spike glycoprotein, out of an amino acid coverage of 3452, 3360 amino acids mapped to G, 140 24 to S and 21 to V. This general pattern is reflected in other clinical isolates. For example, in 141 isolate SCV2-010, in NSP12, out of an amino acid coverage of 285, 273 mapped to L, 50 to I and 142 amino acid is still present, but other amino acids such as F may be common (Supplementary Table  145 1), and subject to selection pressure. In some clinical swabs, for example in N at position 204, the 146 second most common feature is a stop codon. SCV2-011 and SCV2-018 were variants isolated 147 from clinical swabs taken from the same patient but three days apart, these did not vary at the 148 consensus between each other, but did at the minor variant level. SCV2-007 and SCV2-017 were 149 also variants isolated from clinical swabs taken from the same patient but three days apart and 150 did not vary at the consensus between each other in swabs, but did at the minor variant level. 151

In order to assess the biology of the viruses isolated from the clinical swabs and compare 154 their growth to B.1.1.7 and B.1.351, sufficient stocks had to be grown. To isolate SARS-CoV-2 155 from the clinical swabs, the nasopharyngeal sample was filtered and placed on VeroE6 cells with 156 antibiotics and antifungals until CPE was observed. The supernatant was collected from these 157 cells to generate sufficient stocks for infectivity assays and comparisons. 158

Growing virus for stocks may have introduced or selected for specific variants. One of 159 these, that has been characterised for SARS-CoV-2, is a deletion of the furin cleavage site in the 160 spike glycoprotein when grown in Vero E6 cell (19) . Therefore, viral stocks were sequenced to 161 ensure they did not possess the deletion and to determine if variation occurred compared to 162 when the virus was sequenced directly from clinical swabs. Comparator viruses of known 163 provenance were obtained from collaborators. The comparator viruses were the B.1.1.7 ('Kent' 164 UK VOC) virus (termed SCV2-019 in this study) obtained at P4, the SARS-CoV-2/Victoria/01/2020 165 (an isolate from Australia) obtained at P3 (termed SCV2-021), and the B.1.351 virus ('South 166

African' VOC) (termed SCV2-022 in this study). These were grown in Vero/hSLAMS as a precaution 167 to prevent selection for the furin deletion. These were also sequenced to ensure they had the 168 variant defining mutations present. For these three comparator viruses, the sequencing showed 169 at the consensus level the furin cleavage site was intact and the other defining variations 170 separating these variants from the Wuhan reference sequence were present ( Figure 3) . 171

Analysis of the genome diversity between viruses sequenced in swabs from patients and 172 the virus stock used to infect cells indicated that most consensus variations from the Wuhan 499, P with a depth of 6 and F with a depth of 5, indicating that the consensus level amino acid 177 was still present with P and F at a minor level. For some stock viruses, variation from the 178 reference sequence was lost during preparation of the stock virus. The growth of these viruses 179 from the stocks was compared to a B.1.1.7. and a B.1.351 lineage virus and SARS-CoV-180 2/Victoria/01/2020, obtained from near the start of the COVID-19 pandemic. Vero E6 (commonly used to grow viral stocks and initial isolates from clinical samples), 188

Vero/hSLAM (reported to prevent deletion of the furin cleavage site in the spike glycoprotein) 189 and hACE2-A549 cells. This latter cell line is based on A549 cells, which are respiratory epithelium 190 in origin, commonly used to study respiratory viruses in cell culture but overexpress the ACE2 191

In Vero E6 cells, eleven SARS-CoV-2 variants followed a similar pattern of growth, with 193 the exception of SCV2-021 (SARS-CoV-2/Victoria/01/2020) which grew at significantly reduced 194 levels compared to other variants by 72 hours post infection (mean 5.7 x 10 4 PFU/ml, p=0.006) 195

( Figure 4A ). A similar pattern of growth was observed in all twelve variants in Vero/hSLAM cells, 196 but there was no significant difference (p>0.05) in the titres of any of the viruses produced by 72 hours post infection ( Figure 4B ). In contrast, in hACE2-A549 cells, there was more heterogeneity 198 observed between variants, with the range of viral titres being much lower (7.9 x 10 1 -3.01 suggest that the study of specific genotypes requires either plaque purification or reverse 293 genetics. However, the study suggests that the viral population (consensus and minor variants) 294 should be taken into account when studying the transmission of SARS-CoV-2.

Cells. African green monkey kidney C1008 (Vero E6) cells (Public Health England, PHE) were 298 cultured in Dulbecco's minimal essential medium (DMEM) (Sigma) with 10% foetal bovine serum 299 (FBS) (Sigma) and 0.05mg/ml gentamicin at 37°C/5% CO2. Vero/hSLAM cells (PHE) were grown in 300 DMEM with 10% FBS and 0.05mg/ml gentamicin (Merck) with the addition of 0.4mg/ml Geneticin 301 (G418; Thermofisher) at 37°C/5% CO2. Human ACE2-A549 (hACE2-A549), a lung epithelial cell line 302 which overexpresses the ACE-2 receptor, were the kind gift of Oliver Schwartz (23) and were 303 cultured in DMEM with 10% FBS and 0.05mg/ml gentamicin with the addition of 10µg/ml 304 Blasticidin (Invitrogen). Only passage 3-10 cultures were used for experiments. 305 gentamicin at 37°C/5% CO2 and harvested 48 hours post inoculation. Virus stocks were aliquoted 310 and stored at -80°C. 311

Viruses named SCV2-007 to SCV2-018 were grown from nasopharyngeal swabs of 312 patients using the following method. One hundred microlitres of viral transport media from the 313 swab was mixed with 100µl DMEM with 4% FBS, 0.05mg/ml gentamicin, 25µg/ml plasmocin 314 (Invivogen) and 2.5µg/ml amphotericin B (Merck). These were then filtered using ultrapure MC 315 0.22µm filters (Merck) and the filtrate placed onto cells in a 24 well plate of Vero E6 cells for 1 316 hour. After one hour, the media was topped up with DMEM (2% FBS, 0.05 mg/ml gentamicin, respective selective antibiotics for each cell line (6 wells per timepoint). Plates were incubated at 341 37°C/5% CO2 for one hour. The inoculum was removed, and cells were washed once with PBS 342 (Sigma). The respective media with 2% FBS (100µl) was added to each well. The cell supernatant 343 was removed from wells and combined (0hrs post infection) and plates incubated further. the ISARIC-4C study were sequenced using an amplicon based approach on the Oxford Nanopore 

Coronaviridae Study Group of the International Committee 488 on Taxonomy of V. 2020. The species Severe acute respiratory syndrome-related 489 coronavirus: classifying 2019-nCoV and naming it SARS-CoV-2

493 on the spike protein structure and function

MERS-CoV with a variant spike protein cleavage site and distinct fusion activation 528 properties

Tracking Changes in SARS-CoV-2 Spike: Evidence that 535 D614G Increases Infectivity of the COVID-19 Virus

The basis of a 538 more contagious 501Y.V1 variant of SARS-COV-2. bioRxiv : the preprint server for biology 539 around the dominant viral genome sequence contributes to viral load and outcome in 570 patients with Ebola virus disease

Genomic surveillance of Nevada 573 patients revealed prevalence of unique SARS-CoV-2 variants bearing mutations in the 574

Comparison viruses included the Australian Victoria variant (SCV2-021). (A) Growth of viruses in 611 plaque-forming units (PFU) per ml over times in Vero E6 African green monkey kidney cells

Growth of viruses in Vero cells expressing the human signalling lymphocytic activation module 613 (SLAM) gene (Vero/hSLAM). (C) Growth of viruses in human ACE-2 expressing A549 cells

All experiments were repeated in triplicate using supernatant from 6 wells (n=3)