## Main

The Omicron variant of severe acute respiratory syndrome coronavirus 2 (SARS-CoV-2), first identified in November 2021 in South Africa and Botswana10, has been shown by us1 and others2,3,4,5,6,7,8,9 to have extensive but incomplete escape from neutralizing immunity elicited by vaccines and previous infection, with boosted individuals showing better neutralization. In South Africa, Omicron infections led to a lower incidence of severe disease relative to other variants11,12, although this can be at least partly explained by pre-existing immunity13. The first Omicron sublineage to appear was BA.1, which was supplanted by the BA.2 sublineage in many countries14.

How Omicron BA.1 infection will interact with vaccination to protect against the previously dominant Delta variant, emerging variants such as BA.2 and other variants is still unclear. We isolated live Omicron BA.1, Omicron BA.2, ancestral, Beta and Delta viruses and neutralized viruses with plasma from participants enrolled and longitudinally sampled during the Omicron BA.1 infection wave in South Africa, with all participants having a confirmed diagnosis of SARS-CoV-2 by quantitative PCR. To quantify neutralization, we used a live virus neutralization assay and calculated the focus reduction neutralization test (FRNT50) titre, the inverse of the plasma dilution required for 50% neutralization, as measured by the reduction in the number of infection foci. We enrolled 41 participants who reported symptoms from late November 2021 to January 2022. We successfully sequenced the infecting virus in 26 participants, and all sequences corresponded to Omicron BA.1 (Extended Data Table 1). Two participants had advanced human immunodeficiency virus (HIV) disease on the basis of a low CD4 count (<200 cells μl−1 throughout the study) and unsuppressed HIV infection, and we excluded these participants because of our previous data showing an atypical response to SARS-CoV-2 in advanced HIV disease15. Extended Data Table 2 summarizes the characteristics of the remaining 39 participants.

Of the 39 participants, 27 were admitted to hospital because of coronavirus disease 2019 symptoms. Seven required supplemental oxygen and one died. Fifteen participants were vaccinated and had a breakthrough Omicron BA.1 infection. The median time post vaccination was 139 days (interquartile range (IQR) 120–178), a time interval that would predict considerable waning of the vaccine-elicited immune response16, which may have contributed to the breakthrough infections. Eight participants were vaccinated with two doses of Pfizer's BNT162b2 and seven were vaccinated with Johnson & Johnson's Ad26.CoV2.S (six with a single dose and one with two doses; Extended Data Table 1). The length of hospital stay was shorter in the vaccinated (3.5 days) relative to unvaccinated (8 days; Extended Data Table 2) participants. Three participants self-reported having a previous SARS-CoV-2 infection (Extended Data Table 1).

Participants were sampled at enrolment at a median of 6 days (IQR 3–9 days) after symptom onset, and again at weekly follow-up visits that were attended as practicable because of the Christmas holidays in South Africa. The last follow-up visit was a median of 23 days (IQR 19–27 days) post-symptom onset (Extended Data Table 1). Examining neutralization at all available time points per study participant showed that neutralization of the Omicron BA.1 variant increased substantially in most participants from enrolment to the time of the last follow-up (Extended Data Fig. 1), consistent with developing a neutralizing antibody response to Omicron BA.1 infection. We therefore analysed neutralization at enrolment (baseline for the study) and the last follow-up visit to quantify the increase in neutralization capacity after Omicron infection.

We observed that Omicron BA.1 neutralization increased in vaccinated individuals from a low geometric mean titre (GMT) FRNT50 of 42 at the enrolment visit to 575 at the last follow-up visit about 2 to 3 weeks later, a 13.6-fold change (95% CI confidence interval (CI) 3.7–50.2; Fig. 1a). The samples from unvaccinated participants at the study baseline neutralized Omicron BA.1 at a similar starting level of 46 and reached a final level of 272 at the last follow-up, a 6.0-fold increase (95% CI 2.2–16.1; Fig. 1b). Neutralization of the Delta virus also increased during this period. At enrolment, neutralization capacity against the Delta virus was 192 and reached a final level of 1,091 at the last follow-up visit in vaccinated participants, a 5.7-fold increase (95% CI 1.7–18.4; Fig. 1c). Unvaccinated participants had lower Delta neutralization at baseline with Delta virus FRNT50 = 28, and reached FRNT50 = 91, a 3.2-fold increase (95% CI 1.3–8.1; Fig. 1d).

We next compared Omicron BA.1 to Omicron BA.2, Delta, Beta (ref. 17) and ancestral virus neutralization at the last available follow-up visit in three sets of paired experiments, each comparing Omicron BA.2, Delta or ancestral and Beta virus neutralization to Omicron BA.1 neutralization. The range of Omicron BA.1 neutralization shown in Fig. 2a for different experiments (FRNT50 = 516 to 646 for vaccinated samples and 266 to 271 for unvaccinated samples) is the result of experimental variation. BA.2 neutralization was moderately and not significantly lower relative to BA.1 neutralization in both vaccinated and unvaccinated participants. Testing only participants with sequence-confirmed Omicron BA.1 infection gave a similar result (Extended Data Fig. 2). The trend for the other variants and the ancestral virus was that neutralization was higher relative to Omicron BA.1 in vaccinated participants but lower relative to Omicron BA.1 in unvaccinated participants, although the differences were mostly not significant (Fig. 2a). As a result of the relatively moderate fold change, higher participant numbers would probably be required to make the trends statistically significant.

The comparison of the other variants to Omicron BA.1 within the vaccinated or unvaccinated group does not indicate the differences in neutralization capacity elicited by Omicron BA.1 between the vaccinated and unvaccinated participants. We therefore compared neutralization of each variant between the vaccinated and unvaccinated groups at the last time point directly (Fig. 2b). The smallest difference between vaccinated and unvaccinated participants was in neutralization of Omicron BA.1, the infecting variant, with the vaccinated participants showing 2.2-fold higher neutralization. For the other variants, neutralization was higher in vaccinated participants by a factor of 4.8-fold for Omicron BA.2, 9.6-fold for Beta, 12.0-fold for Delta and 17.9-fold for ancestral (Fig. 2b). All differences were significant, and the 95% CIs for the GMT FRNT50 of vaccinated and unvaccinated participants did not overlap for BA.2, Beta, Delta or ancestral virus neutralization (Fig. 2b). For the unvaccinated participants, the absolute neutralization capacity against the BA.2, Beta, Delta and ancestral viruses was low18, with GMT FRNT50 being about or below FRNT50 = 100 (Fig. 2b).

We also tested neutralization of Omicron BA.1 by Delta-variant-elicited immunity. We collected 18 plasma samples from 14 participants (including pre-vaccination and post-vaccination samples from 4 participants) previously infected in the Delta variant wave in South Africa, 8 of whom were vaccinated either before or after infection (Extended Data Table 3). Confirming previously reported results19, we observed similar escape of Omicron BA.1 from Delta-elicited immunity across all samples tested, manifested as a 22.5-fold decrease (95% CI 14.4–35.0) in Omicron BA.1 neutralization compared to Delta virus neutralization (Fig. 3).

The large fold drop in Delta-infection-elicited neutralization capacity against Omicron BA.1 contrasts with the moderate and nonsignificant fold drops, or even fold increases, in neutralization of other variants relative to Omicron BA.1 in individuals infected with Omicron BA.1. However, in unvaccinated individuals, even though fold drops in neutralization were moderate and nonsignificant, the absolute levels of neutralization of the other variants, and of Omicron BA.1 itself, were low and on a similar scale to the cross-neutralization capacity against Omicron in Delta-infection-elicited immunity. This is consistent with other recently reported results20, and possibly indicates that Omicron is poorly immunogenic. In agreement with recent reports21,22, our observations show moderately and nonsignificantly lower neutralization of BA.2 by BA.1-elicited immunity. The results explain epidemiological observations showing that Omicron BA.2 reinfection is relatively rare soon after Omicron BA.1 infection23,24.

Our results may be supportive of a scenario in which hybrid immunity formed by Omicron infection combined with vaccination protects as well or better against reinfection with variants such as Delta relative to reinfection with Omicron itself. By contrast, unvaccinated participants infected with Omicron BA.1 only, have low neutralization capacity against the Omicron BA.2, Beta, Delta and ancestral viruses.

Limitations of this study include heterogeneity in participant immune history, including two vaccination types and one boost. On the basis of the high seroprevalence observed in South Africa25,26, some participants may also have had unreported previous infection. However, including two vaccine types did not mask the differences between vaccinated and unvaccinated participants, and the low levels of neutralization in unvaccinated participants against the ancestral, Beta and Delta viruses (the dominant strains in the preceding South African infection waves) support the notion that these participants were either not previously infected, or that immunity has waned completely. Participants were also mostly hospitalized, which may not be typical of Omicron infection13,27. Increased disease severity has been shown to lead to higher anti-SARS-CoV-2 antibody titres28. This should help in the detection of the neutralization response, but whether it would affect the trend we observed is unclear. Omicron infection is unlikely benign to the extent that hospitalization is an outlier outcome: in the USA, the number of individuals with coronavirus disease 2019 who died in the Omicron wave was similar to the number who died in the Delta wave27. Neutralizing immunity may have increased further in some participants had we sampled later: the neutralizing capacity did not plateau at the last time point in 8 of the 24 (33%) unvaccinated participants (participants 10, 14, 21, 27, 30, 31, 34 and 38; Extended Data Fig. 1) and 6 of the 15 (40%) vaccinated participants (participants 4, 6, 15, 16, 25 and 26). Therefore, the temporal dynamics give no clear indication that the immunity in the unvaccinated participants was delayed and would have reached similar levels to that of vaccinated participants if sampled later. However, the consequences of waning immunity several months post Omicron infection should be investigated.

The gap in immunity between unvaccinated individuals infected with Omicron BA.1 and vaccinated individuals with BA.1 breakthrough infection is concerning. Especially as immunity wanes, unvaccinated individuals post Omicron infection are likely to have poor cross-protection against existing and possibly emerging SARS-CoV-2 variants, despite acquiring some neutralizing immunity to the infecting Omicron BA.1 sub-lineage variant. The implication may be that Omicron BA.1 infection alone is not sufficient for protection and vaccination should be administered even in areas with a high prevalence of Omicron infection to protect against other variants.

## Methods

### Informed consent and ethics

Blood samples and the Delta isolate were obtained after written informed consent from adults with PCR-confirmed SARS-CoV-2 infection who were enrolled in a prospective cohort study at the Africa Health Research Institute approved by the Biomedical Research Ethics Committee at the University of KwaZulu-Natal (reference BREC/00001275/2020). Omicron BA.1 was isolated from a residual swab sample for SARS-CoV-2 where isolation from the sample was approved by the University of the Witwatersrand Human Research Ethics Committee (HREC; reference M210752). The sample to isolate Omicron BA.2 was collected after written informed consent as part of the study “COVID-19 transmission and natural history in KwaZulu-Natal, South Africa: Epidemiological Investigation to Guide Prevention and Clinical Care” of the Centre for the AIDS Programme of Research in South Africa (CAPRISA) and isolation from the sample approved by the Biomedical Research Ethics Committee at the University of KwaZulu-Natal (reference BREC/00001195/2020, BREC/00003106/2021).

### Reagent availability

Virus isolates and the cell line are available from the corresponding author. A Biosafety Level 3 facility is required for laboratories receiving live SARS-CoV-2.

### Cells

Vero E6 cells (originally ATCC CRL-1586, obtained from Cellonex in South Africa) were propagated in complete growth medium consisting of Dulbecco's modified Eagle medium with 10% fetal bovine serum (Hyclone) containing 10 mM HEPES, 1 mM sodium pyruvate, 2 mM l-glutamine and 0.1 mM nonessential amino acids (Sigma-Aldrich). Vero E6 cells were passaged every 3–4 days. The H1299-E3 cell line (H1299 originally from ATCC as CRL-5803) was propagated in growth medium consisting of complete Roswell Park Memorial Institute 1640 medium with 10% fetal bovine serum containing 10 mM HEPES, 1 mM sodium pyruvate, 2 mM l-glutamine and 0.1 mM nonessential amino acids. Cells were passaged every second day. The H1299-E3 (H1299-ACE2, clone E3) cell line was derived from H1299 as described in our previous work1,17. Cell lines were not authenticated. Cell lines tested negative for mycoplasma contamination.

### Virus expansion

All work with live virus was performed in Biosafety Level 3 containment using protocols for SARS-CoV-2 approved by the Africa Health Research Institute Biosafety Committee. ACE2-expressing H1299-E3 cells were seeded at 4.5 × 105 cells in a 6-well plate well and incubated for 18–20 h. After one Dulbecco's phosphate-buffered saline (PBS) wash, the subconfluent cell monolayer was inoculated with 500 μl universal transport medium diluted 1:1 with growth medium filtered through a 0.45-μm filter. Cells were incubated for 1 h. Wells were then filled with 3 ml complete growth medium. After 4 days of infection (completion of passage 1 (P1)), cells were trypsinized, centrifuged at 300 RCF for 3 min and resuspended in 4 ml growth medium. All cells from the P1 infection were added to Vero E6 cells that had been seeded at 2 × 105 cells ml−1, 20 ml total, 18–20 h earlier in a T75 flask for cell-to-cell infection. The coculture of ACE2-expressing H1299-E3 and Vero E6 cells was incubated for 1 h, and the flask was filled with 20 ml of complete growth medium and incubated for 4 days. The viral supernatant from this culture (passage 2 (P2) stock) was used for experiments.

### Live virus neutralization assay

H1299-E3 cells were plated in a 96-well plate (Corning) at 30,000 cells per well 1 day pre-infection. Plasma was separated from EDTA-anticoagulated blood by centrifugation at 500 RCF for 10 min and stored at −80 °C. Aliquots of plasma samples were heat-inactivated at 56 °C for 30 min and clarified by centrifugation at 10,000 RCF for 5 min. Virus stocks were used at approximately 50–100 focus-forming units per microwell and added to diluted plasma. Antibody–virus mixtures were incubated for 1 h at 37 °C, 5% CO2. Cells were infected with 100 μl of the virus–antibody mixtures for 1 h, and then 100 μl of a 1× Roswell Park Memorial Institute 1640 (Sigma-Aldrich, R6504), 1.5% carboxymethylcellulose (Sigma-Aldrich, C4888) overlay was added without removing the inoculum. Cells were fixed 18 h after infection using 4% paraformaldehyde (Sigma-Aldrich) for 20 min. Foci were stained with a rabbit anti-spike monoclonal antibody (BS-R2B12, GenScript A02058) at 0.5 μg ml−1 in a permeabilization buffer containing 0.1% saponin (Sigma-Aldrich), 0.1% BSA (Sigma-Aldrich) and 0.05% Tween-20 (Sigma-Aldrich) in PBS. Plates were incubated with primary antibody overnight at 4 °C, and then washed with wash buffer containing 0.05% Tween-20 in PBS. Secondary goat anti-rabbit antibody conjugated to horseradish peroxidase (Abcam ab205718) was added at 1 μg ml−1 and incubated for 2 h at room temperature with shaking. TrueBlue peroxidase substrate (SeraCare 5510-0030) was then added at 50 μl per well and incubated for 20 min at room temperature. Plates were imaged in an ImmunoSpot Ultra-V S6-02-6140 Analyzer ELISPOT instrument with BioSpot Professional built-in image analysis (Cellular Technology Limited).

### Statistics and fitting

Statistical methods were not used to predetermine sample size, and blinding and randomization were not used. All statistics and fitting were performed using custom code in MATLAB v.2019b. Neutralization data were fitted to:

$${\rm{Tx}}=1/1+(D/{{\rm{ID}}}_{50}).$$

Here Tx is the number of foci normalized to the number of foci in the absence of plasma on the same plate at dilution D, and ID50 is the plasma dilution giving 50% neutralization. FRNT50 = 1/ID50. Values of FRNT50 < 1 are set to 1 (undiluted), the lowest measurable value. We note that the most concentrated plasma dilution was 1:25 and therefore FRNT50 < 25 data were extrapolated. To calculate CIs, FRNT50 or fold change in FRNT50 per participant was log transformed and the arithmetic mean plus 2 s.d. and the arithmetic mean minus 2 s.d. were calculated for the log-transformed values. These were exponentiated to obtain the upper and lower 95% CIs on the geometric mean FRNT50 or the fold change in FRNT50 geometric means.

### Reporting summary

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