Main

Since its emergence in 20194, more than 227 million SARS-CoV-2 infections have occurred worldwide, with more than 4.6 million fatalities. Despite exhibiting proofreading activity5, coronaviruses mutate frequently6, leading to thousands of mutations among the different strains. The most crucial mutations have involved the spike protein, which mediates binding to the angiotensin-converting enzyme 2 (ACE2) receptor via its S1 domain and membrane fusion via the S2 domain7. Spike substitutions influence host range, tissue tropism, transmission and pathogenesis8.

The first dominant spike substitution, D614G, enhances replication in human airway epithelial (HAE) cells and transmission in animal models9,10,11,12. Additional spike mutations, including the N501Y mutation that occurred convergently in the UK, South Africa, Brazil and elsewhere, have been identified. The Alpha variant, which contains seven additional spike substitutions, emerged in September 2020 in the UK. It was estimated to be 70–80% more transmissible than its ancestor1 and quickly spread worldwide2,3. In vitro studies show enhanced replication of Alpha in human airway cells13. Although it is effectively neutralized by vaccine-induced antibodies14,15,16, and is not consistently associated with more severe COVID-1917 disease, increased Alpha transmissibility could counteract increasing herd immunity, exacerbating the pandemic.

To investigate mechanisms of increased Alpha transmission, we used reverse genetics, a hamster model and HAE cultures to probe the effects of spike mutations. We determined the phenotypes of individual mutations as well as of the entire spike gene using competition fitness assays. Our results suggest that N501Y is a critical determinant of enhanced infection of the upper airway and transmission.

Results

To investigate phenotypes of individual Alpha spike substitutions, we infected hamsters intranasally with each mutant virus, as well as with chimeric viruses containing the entire spike gene (Extended Data Fig. 1a–c). These mutations were engineered into the USA_WA1/2020 cDNA clone containing the dominant D614G spike substitution10 using site-directed mutagenesis18 (resulting in USA_WA1/2020-G614, hereafter referred to as the wild-type (WT) virus). These mutations did not affect Vero cell plaque phenotypes (Extended Data Fig. 1d) and the RNA:plaque-forming units (PFU) ratios were similar to those for the virus expressing wild-type spike (Extended Data Fig. 1e). We tested fitness of viruses expressing the mutant and Alpha spike proteins in an in vivo competition assay by mixing WT viruses expressing mutant and wild-type spike at a PFU ratio of 1:1 for intranasal infection of hamsters (Fig. 1a). When assessed by quantification of viral RNA, these ratios were in a range of 0.72–1.54, presumably owing to minor variations in RNA:PFU ratios (Supplementary Table 1). Although the enrichment of N501Y mutants after competition overlapped with this range in some cases, the increases in viral RNA and typically larger changes in PFU ratio indicated robust fitness advantages. This competition approach has been used extensively to study microbial fitness19,20,21,22 including for SARS-CoV-210,13, owing to its precision and reproducibility23. Major advantages of this approach include the internally controlled replication of each virus, which eliminates host-to-host variation that can reduce experimental power, and the ability to assay strain ratios with more precision than when using individual virus titres. To verify the accuracy of estimates across a range of strain ratios, we performed experiments with each mixture using Sanger sequencing, and found a high degree of consistency between PFU- and genome-based estimates (Extended Data Fig. 2). We also evaluated all virus stocks used for experiments by next-generation sequencing to confirm that there were no single-nucleotide polymorphisms (SNPs) with a frequency above 3%; sequencing after two serial passages in hamsters did not detect additional SNPs. These results indicate that the mutant mixtures were genetically stable and that the PFU ratios were robust.

Fig. 1: The screening of the SARS-CoV-2 Alpha variant spike substitutions in hamsters by competition assay.
figure 1

a, Design of the hamster competition fitness studies. The mutant viruses were mixed 1:1 (PFU ratio) with WT virus and inoculated into donor hamsters intranasally at a total titre of 105 PFU per hamster. The donor hamsters were co-housed with recipient hamsters 1 day after infection. After 8 h of contact, the donors were removed. All hamsters were subjected to nasal washes daily until 4 days after infection and organs were collected 4 days after inoculation or contact. b–g, Competition of different Alpha mutants with WT virus. b, d, f, Final:inoculum PFU ratio of WT virus expressing 8 individual spike mutations or the Alpha spike protein that includes all 8 mutations: nasal washes (b), tracheae (d) and lungs (f) in donor hamsters 4 days after inoculation. c, e, g, Final:donor inoculum ratios of WT virus expressing 8 individual spike mutations or the Alpha spike protein: nasal washes (c), tracheae (e) and lungs (g) in recipient hamsters 4 days after contact. In b–g, red dots represent individual hamsters (n = 5). In catseye plots, the horizontal line shows the mean and the shaded region represents s.e.m. Numbers along the top indicate the relative fitness estimates. P values were calculated for the group (strain) coefficient for each linear regression model.

Source data

Full size image

Hamsters were sampled up to four days after infection, following peak viral loads10, using nasal washes and necropsied tracheae and lungs (Fig. 1a). One day after infection, hamsters were co-housed with recipients for 8 h to assess transmission; recipients were maintained for an additional 4 days for equivalent sampling. Inocula and hamster samples were analysed by PCR with reverse transcription (RT–PCR) spanning the mutations, and the amplicons were analysed by Sanger sequencing and electropherogram peak analyses to determine mutant:wild-type ratios. Ratio changes during infection reflected the relative fitness of the mutants.

Competition results from donor nasal washes indicated that of the 8 mutations examined, only the deletion of codons 69 and 70 (Δ69–70), N501Y and the combined Alpha spike mutations conferred significant fitness advantages, with other substitutions showing inconsistently reduced fitness or no difference from the wild-type spike (Fig. 1b). The same fitness trends were observed in recipient hamsters (Fig. 1c). Examination of donor tracheae showed similar results, except that Δ69–70 had no effect on fitness (Fig. 1d), whereas A570D increased fitness and S982A and D1118H decreased fitness. Following transmission, the same overall effects on fitness were maintained in recipient tracheae (Fig. 1e), consistent with results from nasal washes (Fig. 1c). Results from day-4 lungs were distinct from those in the upper airway (nasal washes and tracheae), with only slight fitness effects of the spike mutations; even the combined Alpha-spike mutations did not increase fitness significantly with respect to replication in donor lungs (Fig. 1f). However, there were significant fitness gains for N501Y, A570D and Alpha-spike mutations in recipients, suggesting more efficient transmission, whereas S982A decreased fitness in recipients (Fig. 1g). We also found that N501Y conferred an advantage across days 1–4 of nasal shedding, beginning on day 3 in donors, that was maintained throughout days 1–4 in recipients (Fig. 2a, b). Among the other mutations, only Δ69–70 and the Alpha-spike mutations showed consistent fitness gains (Extended Data Fig. 3).

Fig. 2: The SARS-CoV-2 spike N501Y mutant has a consistent advantage over wild-type spike in upper airway replication in hamsters.
figure 2

a, b, Competition assays between N501Y and wild-type (WT) spike assessed by sampling nasal washes of donor (a) and recipient (b) hamsters on days 1 to 4 after inoculation (donors) or contact (recipients). c–f, Competition assays between N501Y and wild-type spike in the tracheae and lungs of donor (c, d) and recipient hamsters (e, f) on days 2 and 4 after inoculation (donors) or contact (recipients). g–j, Competition assays between Alpha-FL virus and WT virus inoculated intranasally into hamsters. Nasal washes were taken on days 1 to 4 after inoculation (donors) (g) or contact (recipients) (h) and tracheae and lungs of donor (i) and recipient hamsters (j) 4 days after inoculation (donors) or contact (recipients). k–n, Competition between Alpha-FL and Alpha (WT spike) assessed by sampling nasal washes on days 1 to 4 after inoculation (donors) (k) or contact (recipients) (l) and tracheae and lungs of donor (m) and recipient hamsters (n) 4 days after inoculation (donors) or contact (recipients). The fitness advantages of the N501Y substitution and Alpha-FL compared with the wild-type both during infection (in donors) and after transmission of the virus to recipients are shown by the changes in ratios between the collected samples and inocula. Red dots represent individual hamsters (all n = 5, except n = 10 for day-1 nasal wash samples of N501Y group). In catseye plots, the horizontal line represents the mean, the shaded region represents s.e.m. Numbers along the top indicate the relative fitness estimates. P values were calculated for the group (strain) coefficient for each linear regression model.

Source data

Full size image

Because it conferred the clearest phenotype among spike mutations, we examined N501Y in trachea and lungs sampled on days 2 and 4. Similar to nasal washes, recipient samples showed a consistent fitness advantage for N501Y at both timepoints, but no difference was detected in donor lungs (Fig. 2c–f). Together, these results demonstrated a clear fitness advantage for N501Y in terms of upper airway shedding during days 1–4, including after intra-cage transmission.

To further ensure that the Sanger sequencing method did not introduce errors in our estimates of mutant ratios, all experiments with N501Y were repeated using next-generation sequencing. As shown in Extended Data Fig. 4, results using next-generation sequencing were very similar and statistically indistinguishable from those obtained with Sanger sequencing.

To examine whether the spike protein was the main determinant of Alpha replication and transmission, we generated a full-length Alpha infectious clone (Alpha-FL) (EPI_ISL_999340) incorporating all Alpha mutations, as well as chimeric viruses in which the spike was swapped (Extended Data Fig. 1c). In competition assays with WT virus, Alpha-FL had a marked fitness advantage in nasal washes from donor hamsters on days 1 and 2, when titres were maximal (Extended Data Fig. 5); similar results were observed in recipient hamsters (Fig. 2g, h). Alpha-FL showed increased fitness in the trachea and lungs of both donors and recipients on day 4, after titres had peaked (Fig. 2i, j, Extended Data Fig. 5). This difference in fitness of Alpha at different sites over time could reflect the complex effects of innate host defences and/or lower stability in the nasal cavity. In competition against Alpha expressing wild-type spike (Alpha (WT spike)), Alpha-FL showed a consistent advantage on days 1–4 in nasal washes of both donor and recipient hamsters (Fig. 2k, l) and on day 4 in tracheae and lungs (Fig. 2m, n). This again indicated that the Alpha spike was a major fitness determinant during infection. Alpha-FL showed consistently reduced fitness in competition assays with Alpha (WT spike) in nasal washes of donors and recipients, and in trachea and lungs (Extended Data Fig. 6), again suggesting that some non-spike Alpha genome regions confer a fitness disadvantage in vivo.

We also assayed relative fitness of WT virus expressing N501Y or Alpha spike in recipient hamsters using the PFU ratio in donor hamster day-1 nasal wash (representing the transmitted population) as the starting PFU ratio rather than that of the original donor inoculum, with little effect on fitness estimates (Extended Data Fig. 7).

To examine the fitness effect of the Alpha spike substitutions in human cells, we used HAE and other human cell cultures (Fig. 3a). Replication of WT virus expressing N501Y and complete Alpha-spike mutants was significantly faster than that of WT virus at early stages of infection, as measured by PFU (Fig. 3b), but there was little difference in the number of RNA copies (Fig. 3c). The RNA:PFU ratios were significantly lower for virus expressing N501Y and Alpha spike compared with wild-type spike on days 1–3 (Fig. 3d), suggesting that viruses with these mutant spike proteins show increased specific infectivity in HAE cultures. In competition infection assays, N501Y showed significantlyhigher fitness at nearly all time points in Vero E6 and Calu-3 cells (Fig. 3e, f). In HAE cells, N501Y had a fitness advantage over wild-type spike and Alpha-FL spike had a fitness advantages Alpha (WT spike) (Fig. 3g, h, Extended Data Fig. 8). Alpha-FL also showed a consistent fitness advantage over Alpha with wild-type spike, suggesting that, in vitro, the Alpha spike is the major fitness determinant of the Alpha variant (Fig. 3i). Overall, these in vivo and in vitro results suggest that N501Y is the major, but not the only, determinant of fitness of the Alpha variant.

Fig. 3: The spike N501Y mutation enhances viral replication in primary human airway cells and confers an advantage in competition with wild-type spike in vitro.
figure 3

a, Experimental scheme. Viruses expressing wild-type spike, and N501Y and Alpha spike mutants were inoculated onto HAE cells at a multiplicity of infection (MOI) of 5 PFU per cell. After a 2-h incubation, the culture was washed with Dulbecco’s PBS 3 times and maintained for 5 days. The secreted viruses were collected in Dulbecco’s phosphate-buffered saline (DPBS) by 30 min incubation at 37 °C every day. b–d, Viral replication kinetics and genomic RNA:PFU ratios. The amounts of infectious virus (b) and genomic RNA (c) were quantified by plaque assay and quantitative PCR with reverse transcription (RT–qPCR), respectively. The genomic RNA:PFU ratio (d) was calculated as an indication of virion infectivity. Dots represent individual biological replicates (n = 6) pooled from 2 independent experiments. Data are mean ± s.e.m. Non-parametric two-tailed Mann–Whitney test with Bonferroni correction to account for multiple comparisons. Differences were considered significant if P < 0.025. e–g, Competition assay between WT virus expressing spike N501Y and WT virus on Vero E6 (e), Calu-3 (f) and HAE (g) cells. h, i, Competition assay between Alpha-FL and WT virus (h) and Alpha-FL and Alpha virus expressing wild-type spike protein (i) on HAE cells. e–i, The Vero E6, Calu-3 and HAE cells were infected with mixed viruses at MOI of 0.01, 0.1 and 5, respectively. Red dots represent individual cell cultures (n = 6), pooled from two independent experiments. In catseye plots, the horizontal line represents the mean, the shaded region represents s.e.m. Numbers along the top indicate the relative fitness estimates. P values were calculated for the group (strain) coefficient for each linear regression model.

Source data

Full size image

To assess virulence, we examined hamsters infected with each mutant virus individually for weight loss; we found no significant differences between wild-type, N501Y and Alpha spike proteins on weight loss (Extended Data Fig. 9a, b). However, the trend of less weight loss in hamsters infected with WT virus suggested that differences may emerge in experiments with a larger cohort. Nasal washes showed consistently higher titres on days 1–3 and 5 for both N501Y and Alpha spike, but only when measured in PFU. These differences were significant only on day 1 (Extended Data Fig. 9c, d), reflecting the lower sensitivity of individual infections compared with competition assays. Genomic RNA:PFU ratios were higher for virus with wild-type spike than for mutant spike (Extended Data Fig. 9e), again suggesting that the viruses with mutant spike have greater specific infectivity. Additional studies are needed to determine whether these specific infectivity differences are biologically relevant. Trachea and lung samples showed no significant differences in titre when measured as PFU or number of copies of viral RNA (Extended Data Fig. 9f–h). Together, these results again suggest that N501Y results in increased viral fitness in the hamster upper airway.

As the SARS-CoV-2 pandemic has progressed, several mutations have occurred convergently in multiple variants; of the 8 Alpha spike substitutions, only Δ69–70 and N501Y have evolved convergently, consistent with our fitness findings (Extended Data Table 1). The frequency of N501Y worldwide increased markedly after October 2020 (Fig. 4a). These data corroborate our results indicating that the N501Y substitution confers a major fitness advantage.

Fig. 4: The spike N501Y substitution spread quickly and increases spike protein binding affinity for the human ACE2 receptor.
figure 4

a, The frequency of the N501Y substitution over time in all genomic SARS-CoV-2 sequences available from the GISAID database worldwide up to May 2021. The blue bars represent the total numbers of SARS-CoV-2 genomes sequenced worldwide. The red line indicates the percentage of N501Y variant in total SARS-CoV-2 genomes. b, The predicted binding site of spike N501 and Y501 residues on the human ACE2 receptor. c, d, Binding affinities of wild-type spike (c) and spike(N501Y) (d) to the human ACE2 receptor. KD, dissociation constant; koff, dissociation rate constant; kon, association rate constant. The affinity of ACE2 to the N501Y mutant RBD is below the detection limit and is plotted as <10−12. Data are derived from a single experiment.

Source data

Full size image

We next examined the mechanism behind the fitness advantage of N501Y, focusing on its effect on receptor binding. N501Y is located in the receptor-binding domain (RBD) of spike (Fig. 4b) and is proposed to result in increased interactions with the human ACE2 receptor at K353, increasing the affinity for the receptor. To test this, we performed binding assays using recombinant spike RBD and human ACE2 proteins on a bio-layer interferometry system (Fig. 4c, d). The N501Y substitution increased binding, as indicated by the more than 530-fold improvement in KD and more than 819-fold improvement in koff. These results are consistent with earlier work demonstrating that spike residue 501 is a critical residue for ACE2 receptor binding24 and suggest that N501Y increases viral fitness for replication in the upper airway, resulting in enhanced transmission, via enhanced spike–receptor interactions.

Discussion

To corroborate epidemiological findings of more efficient transmission of the Alpha variant1 of SARS-CoV-2, and to identify genetic determinants of this phenotype, we performed experimental infections of hamsters and HAE cells with each of the eight individual spike gene mutations found in Alpha. Of these mutations, only N501Y and the deletion of codons 69 and 70 resulted in consistent fitness advantages for replication in the upper airway of hamsters (with increased shedding in nasal secretions) as well as in human cells. In most experiments, N501Y alone had a phenotype similar to that of the chimeric WT virus containing the complete Alpha spike or the complete Alpha-FL, suggesting that N501Y is the major determinant driving increased transmission. Finally, our studies of chimeric SARS-CoV-2 with swapped spike genes suggest—as in another recent study25—that non-spike genomic regions may interact in vivo with innate immune responses to both positively and negatively affect replication in a site- and time-dependent manner.

Although the data are not directly comparable, the typical relative fitness values of 2–3 that we observed with the N501Y substitution in the hamster or HAE models are consistent with reproduction numbers between 1.4 and 1.8 estimated by epidemiological modelling26. This substitution, which is also present in other variants as a result of convergent evolution, is therefore a major adaptive spike mutation of concern. The N501Y mutation also enables SARS-CoV-2 infection of mice27. Its continued evolution in human viral strains only in combination with other spike mutations suggests epistatic interactions. The mechanism of enhanced N501Y fitness is likely to be related to increased replicationand shedding in the nasal cavity, leading to more efficient airborne and/or fomite-mediated transmission, both of which are important for the spread of SARS-CoV-2. More detailed transmission experiments are needed to discriminate between these two modes. In addition to upper airway shedding, increased transmission could involve differences in virus stability and infectivity (for example, the 50% infectious dose).

Tests to date with N501Y indicate that this mutation results in minimal changes in susceptibility to immunity, reflected in neutralizing antibody titre (PRNT50) values14,15,16, suggesting that it does not increase resistance to vaccine protection. These data along with ours suggest that selection of Alpha in nature has occurred owing to more efficient receptor binding rather than immune escape.

Although Δ69–70 has also been found in a mink variant as well as in wild-type SARS-CoV-2 strains (Extended Data Table 1), it has not evolved convergently in major variants. Furthermore, in the USA, despite its presence since October, 2020, it has not increased substantially in frequency28. These findings corroborate our results indicating that Δ69–70 has less effect on transmission than N501Y. However, the frequent co-occurrence of this deletion29 with N501Y suggests possible epistasis. Furthermore, initial epidemiologic estimates from the UK have indicated that N501Y alone increased transmissibility by about 10%, whereas strains with Δ69–70 showed 70–80% higher transmissibility1, suggesting that the deletion is also important for Alpha transmission.

Considering that only two of the eight spike substitutions conferred consistent fitness effects on infection of the upper airway and transmission, the origins and roles of the other six mutations deserve consideration. The first of these to appear were deletions of codons 69–70 and 145, followed by all the other except for S982A by March 2020 (Supplementary Table 2). However, from September 2020, all eight mutations—and especially N501Y—began increasing in prevalence (Fig. 4a, Supplementary Table 2). Our data indicate that of these only Δ69–70 and N501Y have a consistent phenotype, and some of the other mutations reduced fitness in certain experiments. This suggests that these other mutations occurred through drift mechanisms such as founder and hitchhiking effects, and were maintained by linkage to Δ69–70 and N501Y, or by recombination, which occurs frequently during SARS-CoV-2 replication6. Other possibilities include selection by convalescent plasma or antiviral treatments during chronic infections. It is also likely that mutations in other parts of the genome contribute to the fitness of the Alpha variant, as indicated by our results from chimeric virus.

The hamster is a useful model for SARS-CoV-2 infection in many respects10, but its ACE2 receptors differ slightly the human receptors30. The spike binding studies should be repeated with the hamster receptors to confirm our affinity findings. Although we used infection of primary HAE cells to address this limitation, the use of a primate model would further substantiate our in vivo findings.

In conclusion, we used authentic SARS-CoV-2 to demonstrate that only two out of eight spike substitutions in the Alpha variant are major contributors its enhanced transmission, and the N501Y substitution appears to be the major determinant. This mutation is present in several global regions and should be closely monitored to anticipate public health measures needed to control the spread of SARS-CoV-2.

Methods

Ethics statement

Hamster studies were performed in accordance with the guidance for the Care and Use of Laboratory Animals of the University of Texas Medical Branch (UTMB). The protocol was approved by the Institutional Animal Care and Use Committee at UTMB. All the hamster operations were performed under anaesthesia by isoflurane to minimize animal suffering.

Animals and Cells

The Syrian golden hamsters (HsdHan:AURA strain) were purchased from Envigo. African green monkey kidney epithelial Vero E6 cells (ATCC) were grown in Dulbecco’s modified Eagle’s medium (DMEM; Gibco/Thermo Fisher) with 10% fetal bovine serum (FBS; HyClone Laboratories) and 1% antibiotic/streptomycin (Gibco). Human lung adenocarcinoma epithelial Calu-3 cells (ATCC) were maintained in a high-glucose DMEM containing sodium pyruvate and GlutaMAX (Gibco) with 10% FBS and 1% penicillin/streptomycin at 37 °C with 5% CO2. The EpiAirway system is a primary human airway 3D tissue model purchased from MatTek Life Science (Ashland). This EpiAirway system was maintained with the provided culture medium at 37 °C with 5% CO2 following the manufacturer’s instruction. All other culture medium and supplements were purchased from Thermo Fisher Scientific. All cell lines were verified and tested negative for mycoplasma.

Generation of SARS-CoV-2 mutant viruses

Generation of individual point mutant viruses and WT virus expressing Alpha spike

Individual point mutations in the spike gene (Δ69–70, Δ145, N501Y, A570D, P681H, T716I, S982A and D1118H) or 8 spike mutations were introduced into a subclone pcc1-CoV-2-F5–7 containing the spike gene of SARS CoV-2-D614G17 by overlap fusion PCR.

Generation of Alpha-FL and Alpha-FL (WT spike)

The sequence for constructing Alpha-FL and Alpha-non-spike was downloaded from GISAID under accession ID EPI_ISL_999340. In brief, individual point mutations (NSP3: P153L, T183I, A890D, I1412T; NSP6: SGF106-108del; NSP12: P323L; Spike: HV69-70del, Y145del, N501Y, A570D, D614G, P681H, T716I, S982A, D1118H; ORF8: Q27stop, R52I, Y73C, S84L; N: D3L, R203K, G204R, S235F) were introduced into the three subclones pcc1-Alpha-F1–3, puc57-Alpha-F4 and pcc1-Alpha-F5–7 by overlap fusion PCR. To generate the Alpha-FL (WT spike), the spike gene of Alpha-FL was replaced with the spike gene of SARS CoV-2-D614G. All the primers used for the construction are listed in Supplementary Table 3.

The full-length infectious clones of the variant SARS-CoV-2 viruses were assembled by in vitro ligation of contiguous DNA fragments the protocol as previously described31. In vitro transcription was then preformed to synthesize full-length genomic RNA. For recovery of the mutant viruses, the RNA transcripts were electroporated into Vero E6 cells. The viruses from electroporated cells were collected at 40 h after electroporation and served as P0 stocks. All viruses were passaged once on Vero E6 cells for subsequent experiments and subjected to next-generation sequencing after RNA extraction to confirm the introduction and stability of substitutions. Viral titres were determined by plaque assay on Vero E6 cells. All virus preparation and experiments were performed in a biosafety level 3 facility. Viruses and plasmids are available from the World Reference Center for Emerging Viruses and Arboviruses (WRCEVA) at the UTMB.

RNA extraction, RT–PCR, RT–qPCR and Sanger sequencing

Cell culture supernatants or clarified tissue homogenates were mixed with a fivefold excess of TRIzol LS Reagent (Thermo Fisher Scientific). Viral RNAs were extracted according to the manufacturer’s instructions. The extracted RNAs were dissolved in 20 μl nuclease-free water. For the validation of mutant viruses, two microlitres of RNA samples were used for reverse transcription by using the SuperScript IV First-Strand Synthesis System (Thermo Fisher Scientific) with random hexamer primers. Nine DNA fragments flanking the entire viral genome were amplified by PCR. The resulting DNAs were cleaned up by the QIAquick PCR Purification Kit, and the genome sequences were determined by Sanger sequencing at GENEWIZ (South Plainfield).

To quantify mutant ratios for competition assays, RT–PCR products were amplified from extracted RNA using a SuperScript III One-Step RT–PCR kit (Invitrogen). A 20-μl reaction was assembled in PCR 8-tube strips through the addition of 10 μl 2× reaction mix, 0.4 μl SuperScript III RT/Platinum Taq Mix, 0.8 μl Forward Primer (10 μM), 0.8 μl reverse primer (10 μM), 2 μl RNA, and 6 μl Rnase-free water. Primers are listed in Supplementary Table 3. Reverse transcription and amplification were completed using the following protocol: (1) 55 °C, 30 min; 94 °C, 2 min; (2) 94 °C, 15 s; 60 °C, 30 s; 68 °C, 1 min; 40 cycles; (3) 68 °C, 5 min; (4) indefinite hold at 4 °C. The presence and size of the desired amplicon was verified with 2 µl of PCR product on an agarose gel. The remaining 18 µl were purified by a QIAquick PCR Purification kit (Qiagen) according to the manufacturer’s protocol.

Sequences of the purified RT–PCR products were generated using a BigDye Terminator v3.1 cycle sequencing kit (Applied Biosystems). The sequencing reactions were purified using a 96-well plate format (EdgeBio) and analysed on a 3500 Genetic Analyzer (Applied Biosystems). The peak electropherogram height representing each mutation site and the proportion of each competitor was analysed using the QSVanalyser program32.

Next generation sequencing

The competition results between N501Y and WT in hamsters and HAE cells derived via Sanger sequencing were confirmed using next generation sequencing methods. In brief, the viral RNAs from the N501Y–WT competition groups were used for a specific one-step RT–PCR that containing the A23063T mutation site, primers for the RT–PCR are listed in Supplementary Table 3. The PCR products were then purified by a QIAquick PCR Purification kit (Qiagen) according to the manufacturer’s protocol. Dual-indexed adapter sequences (New England BioLabs) were added with 5 cycles of PCR. Samples were pooled and sequenced by the UTMB Next Generation Sequencing Core on an Illumina MiniSeq Mid-Output flow cell with the paired-end 150 base protocol. The reads were filtered for Q-scores of 37 at the A23063T mutation site and adjacent bases and counted. The input ratios and output ratios of the two viruses were obtained and used for the relative replicative fitness analysis.

Plaque assay

Approximately 1.2 × 106 Vero E6 cells were seeded to each well of 6-well plates and cultured at 37 °C, 5% CO2 for 16 h. Virus was serially diluted in DMEM with 2% FBS and 200 µl diluted viruses were transferred onto the monolayers. The viruses were incubated with the cells at 37 °C with 5% CO2 for 1 h. After the incubation, overlay medium was added to the infected cells per well. The overlay medium contained DMEM with 2% FBS, 1% penicillin/streptomycin and 1% sea-plaque agarose (Lonza). After a 2-day incubation, plates were stained with neutral red (Sigma-Aldrich) and plaques were counted on a light box. The detection limitation of the plaque assay was 10 PFU ml−1.

Quantitative real-time RT–PCR assays

RNA copies of SARS-CoV-2 samples were detected by quantitative real-time RT–PCR (RT–qPCR) assays were performed using the iTaq SYBR Green One-Step Kit (Bio-Rad) on the QuantStudio 7 Flex system (Applied Biosystems) following the manufacturers’ protocols. Primers are listed in Supplementary Table 3. The absolute quantification of viral RNA was determined by a standard curve method using an RNA standard (in vitro transcribed 3,480 bp containing genomic nucleotide positions 26,044 to 29,883 of SARS-CoV-2 genome).

Viral infection of cell lines

Approximately 3 × 105 Vero E6 or Calu-3 cells were seeded onto each well of 12-well plates and cultured at 37 °C, 5% CO2 for 16 h. N501Y or Alpha-spike mutants were mixed with wild-type viruses and inoculated onto Vero E6 and Calu-3 cells at a MOI of 0.01 and 0.1, respectively. The mixed virus was incubated with the cells at 37 °C for 2 h. After infection, the cells were washed with DPBS 3 times to remove any un-attached virus. One millilitre of culture medium was added into each well for the maintenance of the cells. At each time point, 100 µl of culture supernatants were lysed in TRIzol LS reagent for the detection of competition assay, and 100 µl of fresh medium was added into each well to replenish the culture volume. The cells were infected in triplicate for each group of viruses. All samples were stored at −80 °C until analysis.

Viral infection in a primary human airway cell culture model

The EpiAirway system is a primary human airway 3D mucociliary tissue model consisting of normal, human-derived tracheal/bronchial epithelial cells. For viral replication kinetics, wild-type, N501Y or Alpha-spike mutant viruses were inoculated onto the culture at a MOI of 5, respectively. After 2 h infection at 37 °C with 5% CO2, the inoculum was removed, and the culture was washed 3 times with DPBS. The infected epithelial cells were maintained without any medium in the apical well, and medium was provided to the culture through the basal well. The infected cells were incubated at 37 °C, 5% CO2. From 1–5 days, 300 μl of DPBS were added onto the apical side of the airway culture and incubated at 37 °C for 30 min to elute the released viruses. Titres of samples taken immediately after washing the inoculum were not plotted because initial experiments showed the presence of very little or no detectable virus. All virus samples in DPBS were stored at −80 °C.

Spike RBD and ACE2 binding

The human ACE2 protein was purchased from Sino Biological (catalogue (cat.) no. 10108-H08H) and the human IgG1 Fc-tagged RBD proteins were made in-house as previously described33. The affinity measurement was performed on the ForteBio Octet RED 96 system (Sartorius). In brief, wild-type or N501Y mutant RBD proteins (20 µg ml−1) were captured onto protein A biosensors for 300 s. The loaded biosensors were then dipped into the kinetics buffer for 10 s for adjustment of baselines. Subsequently, the biosensors were dipped into serially diluted (1.23~300 nM) human ACE2 protein for 200 s to record association kinetics and then dipped into kinetics buffer for 400 s to record dissociation kinetics. Kinetic buffer without ACE2 was used to correct the background. The Octet Data Acquisition 9.0 software was used to collect affinity data. For fitting of KD values, Octet Data Analysis software V11.1 was used to fit the curve by a 1:1 binding model and use of the global fitting method.

Hamster infections

Four- to six-week-old male golden Syrian hamsters, strain HsdHan:AURA (Envigo), were randomly selected (although the investigators were not blinded to the assignments) and inoculated intranasally with 100 µl SARS-CoV-2. For transmission competition assays, 5 donor hamsters received a mixture containing 5 × 104 PFU of WT virus and an equal amount of variant virus. One day later, 1 infected donor hamster was co-housed with one naive hamster for 8 h (10 pairs for N501Y group, 5 pairs for the other groups) and the donors were returned to their cages. For virus replication assays, the hamsters received DMEM with 2% FBS and 1% penicillin/streptomycin (Mock, n = 4), WT virus (n = 9), N501Y mutant virus (n = 9), Alpha-spike mutant virus (n = 9) at a dose of 104 PFU per hamster. The infected hamsters were weighed and monitored for signs of illness daily. Nasal washes were collected in 400 µl sterile DPBS at indicated time points. For the transmission competition study, hamsters were humanely euthanized for organ collections at 2 or 4 days after inoculation (donors) or after contact (recipients). For the virus replication study, 4 hamsters in each group were euthanized for organ collection 2 days after infection and others were euthanized 7 days after infection. The collected tracheae and lungs were placed in a 2-ml homogenizer tube containing 1 ml of maintenance medium (DMEM supplemented with 2% FBS and 1% penicillin/streptomycin) and stored at −80 °C. Samples were subsequently thawed, lung or tracheae were homogenized for 1 min at 26  s−1, and debris was pelleted by centrifugation for 5 min at 16,100g. Infectious titres were determined by plaque assay. Genomic RNAs were quantified by RT–qPCR. Ratios of mutant:wild-type RNA were determined by RT–PCR with quantification of Sanger peak heights.

Competition assays

For the competition assay on Vero E6, Calu-3 and primary human epithelial airway (HAE) cells, WT and mutant viruses were mixed and inoculated onto the cells at a final MOI of 0.01, 0.01 and 5, respectively. For the competition in hamsters, 100 μl mixtures of WT and variant viruses (total 1 × 105 PFU per hamster) were inoculated intranasally into 4–6-week-old Syrian hamsters. On 1–4 days after inoculation (donors) or after contact (recipients), the infected hamsters were sampled for competition detection. An aliquot of the inoculum used for both hamster and cell infections was back-titred to confirm the initial ratio of viruses. All samples were stored in −80 °C freezer prior to analysis.

Validation of competition assay by Sanger sequencing

To validate the consistency and accuracy of competition assay by Sanger sequencing, the WT and mutant viruses were mixed at ratios of 10:1, 5:1, 3:1, 1:1, 1:3, 1:5 and 1:10 based on their PFU titres (total 106 PFU viruses) or mixed with 106, 105, 104, 103 and 102 PFU of the two viruses at a ratio of 1:1. The total RNA of these mixed viruses was isolated and amplified by RT–PCR. The ratios of WT:mutant were calculated by the peak heights of Sanger sequencing. Data were analysed by linear regression with correlation coefficients (r) and significance (P).

Statistics

Male hamsters were randomly allocated into different groups. The investigators were not blinded to allocation during the experiments or to the outcome assessment. No statistical methods were used to predetermine sample size. Dead hamster were excluded from sample collections and data analysis. Descriptive statistics have been provided in the figure legends. For in vitro replication kinetics, Kruskal–Wallis analysis of variance was conducted to detect any significant variation among replicates. If no significant variation was detected, the results were pooled for further comparison. Differences between continuous variables were assessed with a two-tailed non-parametric Mann–Whitney test. The PFU, genomic copies and RNA:PFU ratios were analysed using non-transformed values. The weight loss data are shown as mean ± s.d. and statistically analysed using two-way ANOVA with Turkey’s multiple comparison. Analyses were performed in Prism version 9.0 (GraphPad).

For virus competition experiments, relative replicative fitness values for mutant virus compared to G614 WT virus were analysed according to w = (f0/i0), where w is relative fitness, i0 is the initial mutant:WT ratio and f0 is the final mutant:WT ratio after competition. Sanger sequencing (initial timepoint T0) counts for each virus strain being compared were based on average counts over three replicate samples of inocula per experiment, and post-infection (timepoint T1) counts were taken from samples of individual subjects. For cell culture samples, multiple experiments were performed, so that f0/i0 was clustered by experiment. To model f0/i0, the ratio T0/T1 was found separately for each subject in each strain group, log (base 10) transformed to an improved approximation of normality and modelled by analysis of variance with relation to group, adjusting by experiment when appropriate to control for clustering within experiment. Specifically, the model was of the form log10(Ratio at T1/Ratio at T0) ~ Experiment + Group. Fitness ratios between the two groups (the model’s estimate of w = (f0/i0)) were assessed per the coefficient of the model’s Group term, which was transformed to the original scale as 10coefficient. This modelling approach compensates for any correlation due to clustering within experiment similarly to that of corresponding mixed effect models and is effective since the number of experiments was small. Statistical analyses were performed using R statistical software (R Core Team, version 3.6.1). In all statistical tests, two-sided α = 0.05. Catseye plots34, which illustrate the normal distribution of the model-adjusted means, were produced using the catseyes package35.

Reporting summary

Further information on research design is available in the Nature Research Reporting Summary linked to this paper.