Despite delayed kinetics, people living with HIV achieve equivalent antibody function after SARS-CoV-2 infection or vaccination

The kinetics of Fc-mediated functions following SARS-CoV-2 infection or vaccination in people living with HIV (PLWH) are not known. We compared SARS-CoV-2 spike-specific Fc functions, binding, and neutralization in PLWH and people without HIV (PWOH) during acute infection (without prior vaccination) with either the D614G or Beta variants of SARS-CoV-2, or vaccination with ChAdOx1 nCoV-19. Antiretroviral treatment (ART)–naïve PLWH had significantly lower levels of IgG binding, neutralization, and antibody-dependent cellular phagocytosis (ADCP) compared with PLWH on ART. The magnitude of antibody-dependent cellular cytotoxicity (ADCC), complement deposition (ADCD), and cellular trogocytosis (ADCT) was differentially triggered by D614G and Beta. The kinetics of spike IgG-binding antibodies, ADCC, and ADCD were similar, irrespective of the infecting variant between PWOH and PLWH overall. However, compared with PWOH, PLWH infected with D614G had delayed neutralization and ADCP. Furthermore, Beta infection resulted in delayed ADCT, regardless of HIV status. Despite these delays, we observed improved coordination between binding and neutralizing responses and Fc functions in PLWH. In contrast to D614G infection, binding responses in PLWH following ChAdOx-1 nCoV-19 vaccination were delayed, while neutralization and ADCP had similar timing of onset, but lower magnitude, and ADCC was significantly higher than in PWOH. Overall, despite delayed and differential kinetics, PLWH on ART develop comparable responses to PWOH, supporting the prioritization of ART rollout and SARS-CoV-2 vaccination in PLWH.

Beyond neutralization, antibodies mediate cytotoxic functions through the interaction of the antibody Fc region with cellular Fc receptors or complement proteins. These Fc effector functions include antibody-dependent cellular cytotoxicity (ADCC), cellular phagocytosis (ADCP), cellular trogocytosis (ADCT), and complement deposition (ADCD). SARS-CoV-2-specific Fc effector functions have been associated with reduced COVID-19 mortality, are differentially imprinted by SARS-CoV-2 VOCs, are more durable than neutralization activity, and are required for optimal protection conferred by monoclonal antibodies (44-51). Furthermore, Fc effector functions elicited in non-human primates after vaccination correlate with protection (52)(53)(54)(55)(56)(57). Apart from a recent study showing similar response rates and levels of ADCP in PLWH and PWOH after infection, the kinetics and longevity of SARS-CoV-2-specific Fc effector function in PLWH are not known (23).
In this study, we investigated SARS-CoV-2 spike-specific Fc functions in hospitalized SARS-CoV-2 vaccine naïve PLWH and PWOH following infection with either D614G or Beta during the acute phase and compared these with ChAdOx1 nCoV-19 vaccinees. ART-naïve PLWH had impaired humoral responses after SARS-CoV-2 infection. Furthermore, subtle differences in Fc-mediated response level and kinetics depending on the infecting variant were observed, and both ADCP and neutralization responses were delayed in PLWH. In comparison with infection, ChAdOx1 nCoV-19 vaccination differed by eliciting delayed binding and higher ADCC in PLWH. Overall, despite early delays in some antibody functions following infection or vaccination, PLWH were able to elicit high levels of humoral immune responses similar to PWOH.  [3][4][5]. All samples were collected during the acute phase of COVID-19. Patients had PCR confirmed SARS-CoV-2 infection before blood collection, which was done at admission. A second sample was collected a median of 7 days after admission from participants (D614G, n = 37 and Beta, n = 14) that were not discharged early or deceased (D614G, n = 8 and Beta, n = 1). Full demographic and clinical characteristics of the Pretoria COVID-19 study participants are summarized in Supplementary Table S1.
Ethics approval was received from the University of Pretoria, Faculty of Health Sciences' Research Ethics Committee (247/2020).
For the ChAdOx1 nCoV-19 vaccine samples, eligible participants were adults living with (n = 13) or without HIV (n = 17) (Supplementary Table S2). Lack of prior infection in these individuals was confirmed by nucleocapsid enzyme-linked immunosorbent assay (ELISA) as previously described (59). Longitudinal sampling was conducted, and plasma was collected 28 days after the 1st dose, at day 42 (14 days after the 2nd dose) and then at day 182 (6 months post-vaccination). The cohort was established at the Chris Hani Baragwanath Hospital as part of the South African safety and efficacy ChAdOx1 nCov-19 randomized trial (28). This study was approved by the Human Research Ethics Committee of the University of the Witwatersrand (ethics reference number: M200501). Written informed consent was obtained from all participants.
For ELISA and Fc effector function assays, SARS-CoV-2 D614G and Beta variant full spike proteins were expressed in human embryo kidney (HEK) 293F suspension cells by transfecting the cells with the respective expression spike plasmids. The cells were incubated for a period of 6 days at 37°C, 70% humidity and 10% CO 2 . Proteins were purified using a nickel-charged resin followed by size-exclusion chromatography. The relevant fractions were collected and stored at −80°C until use.

Cell lines
HEK293F suspension cells were cultured in 293 Freestyle media (Gibco BRL Life Technologies, Ontario, CA) and incubated shaking at 37°C, 5% CO 2 and 70% humidity at 125 rpm. HEK293T cells were cultured at 37°C, 5% CO 2 , in DMEM (Gibco BRL Life Technologies, Ontario, CA) containing 10% heat-inactivated fetal bovine serum (FBS) and supplemented with 50 mg/ml Gentamicin. Cells were disrupted at confluence with 0.25% trypsin in 1 mM EDTA every 48h-72h. HEK293T/ACE2.MF cells were maintained as for HEK293T cells but were supplemented with 3 mg/ml Puromycin for selection of stably transduced cells. THP-1 cells were used for the ADCP assay and obtained from the AIDS Reagent Program, Division of AIDS, NIAID, NIH contributed by Dr. Li Wu and Vineet N. Kewal Ramani. Cells were cultured at 37°C, 5% CO 2 in RPMI (Gibco BRL Life Technologies, Ontario, CA) containing 10% heat-inactivated FBS with 1% Penicillin-Streptomycin (Pen/ Strep) and 2-mercaptoethanol to a final concentration of 0.05 mM and not allowed to exceed 4 × 10 5 cells/ml to prevent differentiation. Jurkat-Lucia ™ NFAT-CD16 cells (Invivogen, USA) were maintained in IMDM (Gibco BRL Life Technologies, Ontario, CA) media with 10% heat-inactivated FBS, 1% Pen/Strep, and 10 mg/ml of Blasticidin and 100 mg/ml of Zeocin were added to the growth medium every second passage to allow the selection of CD16 expressing cells.

SARS-CoV-2 spike enzyme-linked immunosorbent assay
Two microgram per mililiter of spike protein (D614G or Beta) was used to coat 96-well, high-binding plates and incubated overnight at 4°C. The plates were incubated in a blocking buffer consisting of 5% skimmed milk powder, 0.05% Tween 20, 1× phosphate buffered saline (PBS). Plasma samples were diluted to 1:100 starting dilution in a blocking buffer and added to the plates. IgG secondary antibody was diluted to 1:3000 in blocking buffer and added to the plates followed by TMB substrate (Thermo Fisher Scientific, USA). Upon stopping the reaction with 1 M H 2 SO 4 , absorbance was measured at a 450-nm wavelength. In all instances, mAbs CR3022 and BD23 (a differential control for Beta) were used as positive controls and Palivizumab was used as a negative control.

Pseudovirus neutralization assay
For all assays, including the neutralization assay, plasma samples were heat inactivated and clarified by centrifugation. IgG was isolated from the plasma samples of PLWH on ART, using Protein G (Pierce Biotechnology, USA), according to the manufacturer's instructions and confirmed by IgG ELISA. The plasma samples or IgG preparations from study participants were incubated with the SARS-CoV-2 pseudotyped virus for 1h at 37°C and 5% CO 2 . These samples were also tested against murine leukaemia virus pseudovirus to ensure that there was no interference from ART. Subsequently, 1 × 10 4 HEK293T cells engineered to overexpress ACE-2 (293T/ACE2.MF) [kindly provided by M. Farzan (Scripps Research)] were added and incubated at 37°C and 5% CO 2 for 72h upon which the luminescence of the luciferase gene was measured. Titers were calculated as the reciprocal plasma dilution (ID 50 ) causing 50% reduction of relative light units (RLUs). CB6 and CA1 were used as neutralization controls for D614G and Beta.

Antibody-dependent cellular phagocytosis (ADCP) assay
SARS-CoV-2 full spike proteins were biotinylated using the EZ-Link Sulfo-NHS-LC-Biotin kit (Thermo Scientific, USA) and coated onto fluorescent neutravidin beads as previously described (61). Briefly, beads were incubated for 2h with monoclonal antibodies at a starting concentration of 20 mg/ml or plasma at a single 1 in 100 dilution. Opsonized beads were incubated with the monocytic THP-1 cell line overnight, fixed and interrogated on the FACSAria II. Phagocytosis score was calculated as the percentage of THP-1 cells that engulfed fluorescent beads multiplied by the geometric mean fluorescence intensity of the population less the no antibody control. For this and all subsequent Fc effector assays, pooled plasma from five PCR-confirmed SARS-CoV-2-infected individuals and CR3022 were used as positive controls and plasma from five pre-pandemic healthy controls and Palivizumab were used as negative controls. 084-7D, a mAb isolated from a donor following SARS-CoV-2 infection, was used as a positive control for Beta. ADCP scores for different spikes were normalized to each other and between runs using CR3022.

FcgRIIIa reporter assay
The ability of plasma antibodies to cross-link spike SARS-CoV-2 protein and activate FcgRIIIa on Jurkat-Lucia ™ NFAT-CD16 cells (Invivogen, USA) was measured as a proxy for ADCC. Highbinding 96-well plates were coated with 1 mg/ml spike SARS-CoV-2 protein and incubated at 4°C overnight. Plates were then washed with PBS and blocked at room temperature for 1h with 2.5% bovine serum albumin (BSA)/PBS. After washing, heat-inactivated plasma (1:100 final dilution) or monoclonal antibodies (final concentration of 100 mg/ml) in RPMI media supplemented with 10% FBS 1% Pen/ Strep were added to the wells and incubated for 1h at 37°C. Jurkat-Lucia ™ NFAT-CD16 2 × 10 5 cells per well were added and incubated for 24h at 37°C and 5% CO 2 . Twenty microliter of supernatant was then transferred to a white 96-well plate with 50 ml of reconstituted QUANTI-Luc secreted luciferase and read immediately on a Victor 3 luminometer with 1 s integration time. The RLUs of a no antibody control were subtracted as background. Palivizumab was used as a negative control, while CR3022 was used as a positive control, and P2B-2F6 to differentiate the Beta from the D614G variant. To induce the transgene, 1× cell stimulation cocktail (Thermo Fisher Scientific, Oslo, Norway) and 2 mg/ml ionomycin in R10 was added. RLUs for spikes were normalized to each other and between runs using CR3022.
2.9 Antibody-dependent complement deposition (ADCD) assay ADCD was measured as previously described (62). Biotinylated spike protein was coated 1:1 onto red fluorescent 1 mM neutravidin beads (Molecular Probes Inc., USA) for 2h at 37°C. These were incubated with a single 1:10 plasma sample dilution or fivefold titration of mAb at a starting concentration of 100 mg/ml for 2h and guinea pig complement diluted 1 in 50 with gelatin/veronal buffer for 15 min at 37°C. Beads were washed in PBS and stained with anti-guinea pig C3b-FITC, fixed and interrogated on a FACSAria II. Complement deposition score was calculated as the percentage of C3b-FITC-positive beads multiplied by the geometric mean fluorescent intensity of FITC in this population less the no antibody or heat-inactivated controls. ADCD scores for D614G and Beta spikes were normalized to each other and between runs using CR3022.
2.10 Antibody-dependent cellular trogocytosis (ADCT) assay ADCT was performed as described in and modified from a previously described study (63). HEK293T cells transfected with a SARS-CoV-2-spike pcDNA vector as above were surface biotinylated with EZ-Link Sulfo-NHS-LC-Biotin as recommended by the manufacturer. Fifty-thousand cells per well were incubated with fivefold titration of mAb starting at 25 mg/ml or single 1 in 100 dilutions for 30 min. Following an RPMI media wash, these were then incubated with carboxyfluorescein succinimidyl ester (CFSE)stained THP-1 cells (5 × 10 4 cells per well) for 1h and washed with 15 mM EDTA/PBS followed by PBS. Cells were then stained for biotin using Streptavidin-PE and read on a FACSAria II. Trogocytosis score was determined as the proportion of CFSEpositive THP-1 cells also positive for streptavidin-PE less the no antibody control with infecting variants run head to head.

Dimeric Fc gamma receptor-binding ELISAs
High-binding 96-well ELISA plates were coated with 1 µg/ml spike protein in PBS overnight at 4°C. Three wells on each plate were directly coated with 5 µg/ml IgG, isolated from healthy donors, and signals from these wells were used to normalize the Fc receptor activity of the plasma samples. Plates were washed with PBS and blocked with PBS/1 mM EDTA/1% BSA for 1h at 37°C. Plates were then washed and incubated with 1:10 diluted plasma for 1h at 37°C and then with 0.2 µg/ml or 0.1 µg/ml of biotinylated FcgRIIa or FcgRIIIa dimer, respectively (constructs kindly provided by Prof. Mark Hogarth from the Burnet Institute, Australia), for 1h at 37°C (64). Subsequently, a 1:10000 dilution of Pierce high-sensitivity streptavidin-horseradish peroxidase (Thermo Scientific, USA) was added for a final incubation for 1h at 37°C. Last, TMB substrate (Sigma-Aldrich) was added, and color development was stopped with 1 M H 2 SO 4 and absorbance read at 450 nm.

Quantification and statistical analysis
Analyses were performed in Prism (v9; GraphPad Software Inc, San Diego, CA, USA). Non-parametric tests were used for all comparisons. Fisher's exact test was used to compare categorical variables between the study participants living with or without HIV, when assessing the clinical data. The Mann-Whitney test was used for unmatched samples. The Kruskal-Wallis test with Dunn's correction was conducted for multiple comparisons. All correlations reported are non-parametric Spearman's correlations. P-values less than 0.05 were considered to be statistically significant.

People living with HIV have similar levels of humoral responses irrespective of the SARS-CoV2 infecting variant
We first compared antibody responses in SARS-CoV-2infected PLWH with PWOH and assessed whether the infecting variant differentially triggered antibody functions. We used plasma from patients hospitalized during South Africa's first (PWOH, n = 39 and PLWH, n = 14) and second (PWOH, n = 13 and PLWH, n = 9) SARS-CoV-2 infection waves, which were dominated by the D614G and Beta variants, respectively. PWOH and PLWH were matched for age, sex assigned at birth, comorbidities, and SARS-CoV-2 disease severity for each of the infecting variants (Supplementary Table S1). Samples were collected at admission, which was a median of 2 days after a positive SARS-CoV-2 PCR test. During the acute infection, a second sample was taken a median of 7 days later, for a subset of patients who had not been discharged or were not deceased by this time (Supplementary Table S1).
As Fc effector function is modulated by Fc receptor binding, we examined the ability of antibodies from D614G and Beta infections to cross-link dimeric FcgRIIa or FcgRIIIa receptors (which modulate ADCP and ADCC, respectively) and the D614G or Beta spike protein by ELISA. As expected, Spearman's correlations > 0.5 p < 0.001 were noted between FcgRIIa binding and ADCP score, and between FcgRIIIa binding and ADCC reporter assay against D614G spike (Supplementary Figures S1A,  B). Similar to the functional readouts, there were no significant differences in FcgRIIa binding (Supplementary Figure  S1C) by HIV status. For the Beta infections, we observed lower FcgRIIa binding in comparison with the D614G infections (p = 0.021). However, no differences were noted between Beta-and D614G-triggered ADCP ( Figure 1C), suggesting the potential impact of variation in antigen presentation on functional beadbased and soluble protein assays. The FcgRIIIa cross-linking (Supplementary Figure S1D) and ADCC reporter assay, where FcgRIIIa is expressed on a target cell, both showed similar responses in hospitalized patients, regardless of HIV status, but the Beta variant compared with D614G resulted in significantly reduced FcgRIIIa binding (p = 0.010).

SARS-CoV-2 humoral responses are compromised in people living with HIV not on antiretroviral treatment
PLWH were stratified according to whether they were on ART (n = 18, all of whom had CD4 T-cell counts greater than 100 cells/ µl) or ART-naïve (n = 5, all with CD4 T-cell counts lower than 100 cells/µl). The ART-naïve PLWH had significantly lower IgG binding (p = 0.037), neutralization (p = 0.006), and ADCP (p = 0.012) activity than the PLWH on ART ( Figure 1G). Additionally, both ADCC and ADCD responses were lower in ART-naïve individuals, although not significantly so, while ADCT responses were comparable between the two groups. Overall, this indicates that PLWH who do not receive ART have impaired immune responses to SARS-CoV-2 infection, as has been previously reported (15-18).

People living with HIV have delayed neutralization and ADCP SARS-CoV-2 responses following D614G infection
We next investigated the kinetics of humoral responses between admission and 1 week of hospitalization in PLWH and PWOH following infection with D614G (PWOH, n = 27 and PLWH, n = 10) or Beta (PWOH, n = 9 and PLWH, n = 5). This analysis excluded patients that were either discharged early or deceased. Overall, humoral responses increased between admission and 1 week of hospitalization, as indicated by a fold change greater than 1 (week 1/admission responses), regardless of HIV status and the infecting variant ( Figure 2). There were no significant differences in the median fold changes of IgG binding, ADCC, ADCD, and ADCT by HIV status, indicating that these functional responses increased at the same rate in PLWH as in PWOH. However, during D614G infection, neutralization (p = 0.042) and ADCP responses (p = 0.032) had significantly lower fold changes in PLWH, indicating that these responses were delayed compared with PWOH. The Beta variant had slower ADCT responses in both PWOH (p = 0.001) and PLWH (p = 0.021), and in contrast to D614G, Beta rapidly elicited ADCP in PLWH ( Figure 2). These data suggest that some humoral responses are delayed in PLWH following acute SARS-CoV-2 infection and that different variants elicit altered levels of activity and also affect the kinetics of humoral responses.

People living with HIV show stronger coordination between SARS-CoV-2 IgG binding, neutralization, and Fc effector functions following infection
A coordinated antibody response against other diseases has been shown following vaccination and for HIV associated with improved protection (53,(65)(66)(67). We therefore assessed the correlations between antibody functions in PLWH (n = 10) and PWOH (n = and ADCT with FcgRIIa binding (r = 0.55 p < 0.01). Thus, antibody functional relationships differ between PLWH and PWOH, and PLWH in this study had a more coordinated functional response to SARS-CoV-2 infection, despite slower neutralization and ADCP kinetics in D614G infection.

ChAdOx1 nCoV-19 vaccination elicits higher IgG binding and ADCC over time in people living with HIV
The initial SARS-CoV-2 vaccines were based on the ancestral spike sequence, enabling us to examine IgG binding, neutralization, and Fc effector functions following ChAdOx1 nCoV-19 vaccination, to determine whether the kinetics are similar to D614G infection responses in both PLWH and PWOH. Thirty participants stratified by HIV status (PWOH, n = 17 and PLWH, n = 13) received two doses of the vaccine and plasma samples were collected at days 28 (after first dose), 42 (after second dose) and 182 post-vaccination ( Figure 4A). The PWOH and PLWH were A B FIGURE 3 Differential associations between antibody responses in hospitalized patients living with or without HIV. Spearman correlations for (A) PWOH (n = 27) and (B) PLWH (n = 10) 1-week post-admission during SARS-CoV-2 D614G infection. Color intensity from white to red is associated with increasing correlation, indicated within the blocks. Statistical significance: ***p < 0.001; **p < 0.01; *p < 0.05.  Table S2). Among the PLWH, the majority (77%) were on ART and had CD4 T-cell counts greater than 250 cells/µl. The onset of neutralization responses was similar regardless of HIV status, although this functional response trended lower in PLWH ( Figure 4B). At 28 days post-vaccination, the onset of ADCP responses was similar between the groups, but the maturation of ADCP following the second dose in PLWH was slowed, resulting in significantly lower activity at days 42 (p = 0.038) and 182 (p < 0.011) ( Figure 4C). In contrast to infection, IgG binding was the only response significantly delayed at day 28 (p = 0.003) in PLWH. However, following the second dose, PLWH achieved comparable levels and then exceeded the titers observed in PWOH at 182 days (p = 0.002) ( Figure 4D). In comparison with the other functions, ADCC was rapidly elicited in PLWH with significantly higher levels at day 28 (p = 0.035) than in the PWOH, and following the second dose, these higher levels persisted through to day 182 (p = 0.005) ( Figure 4E). Last, we observed similar antibody kinetics between ART-treated and ART-naïve PLWH, likely due to small participant numbers, following ChAdOx1 nCoV-19 vaccination (data not shown). In contrast to infection with D614G, there was an initial delay in IgG binding, which may have compromised the levels and kinetics of neutralization and ADCP responses in PLWH, although not significantly. Despite early delays in antibody functions, ChAdOx1 nCoV-19 vaccination was able to induce higher IgG binding and ADCC in PLWH than in PWOH.

Discussion
PLWH are at high risk of COVID-19 morbidity and mortality (5)(6)(7)(8)(9)(10), therefore investigating the kinetics of antibody responses is essential for understanding protection against SARS-CoV-2 and informing vaccine implementation. Here, we confirm that ARTnaïve individuals showed compromised SARS-CoV-2 immunity. We show that the infecting variants triggered nuanced differences in the levels and kinetics of Fc effector functions, as we have previously reported (44). Neutralization and ADCP responses were significantly delayed in PLWH, but despite these delays, a more coordinated antibody functional response was observed in comparison with PWOH. Following ChAdOx1 nCoV-19 vaccination, PLWH overall had slightly impaired neutralization and ADCP responses, and spike binding was delayed; however, ADCC was significantly higher. This study highlights that the kinetics of Fc-mediated functions differ by SARS-CoV-2 infection and vaccination in PLWH and despite initial delays, PLWH are able to reach an equivalent humoral immune response to PWOH.
Our observation that ART-naïve PLWH (CD4 counts less than 100 cells/µl) had significantly lower IgG binding, neutralization, and ADCP responses suggests impaired SARS-CoV-2 immunity in this immune-compromised population. This corroborates previous studies that showed lower IgG and pseudovirus neutralizing antibody titers in virally unsuppressed PLWH with low CD4 counts between 200 and 500 cells/ul (15)(16)(17)(18). In contrast, several studies reported unimpaired binding and neutralization in the context of ART-controlled HIV, who had even higher CD4 counts than participants in our study (19,20). Most PLWH in our cohort accessed ART, likely explaining why we saw no overall impairment in humoral responses (68)(69)(70).
Our study also investigated Fc effector functions ADCC, ADCD, and ADCT, which have not previously been measured in PLWH following SARS-CoV-2 infection. We found no differences in Fc effector functions compared with PWOH. This confirms a previous finding that ADCP in PLWH after SARS-CoV-2 infection showed no difference compared with PWOH (23). Thus, PLWH on ART were capable of eliciting a robust Fc-mediated response after SARS-CoV-2 infection, similar to PWOH. However, differences in the magnitude of Fc effector functions were observed depending on which infecting variant triggered the response. Beta induced significantly lower ADCC activity in PLWH and also trended lower in PWOH. In PLWH, ADCD trended higher and was shown to be significantly higher in PWOH following Beta infection. These observations were similar to our previous findings in a different cohort of PWOH, in which ADCC was reduced and ADCD was substantially higher in individuals infected with Beta (44). Here, Beta elicited higher levels of ADCT in both PWOH and PLWH, but in our previous study, responses were unaffected in comparison with D614G (44). This difference in studies suggests that other factors, such as disease severity, may also impact variant imprinting. Here, and in our previous study, we conclude that the sequence of infecting variants may alter antibody quality, perhaps by eliciting antibodies with varying glycosylation and/or isotype, a finding which requires further investigation.
While no quantitative differences in the level of responses were noted at enrollment, both neutralization and ADCP were delayed in PLWH following D614G infection. These delays were not observed for PLWH infected with Beta despite similarities in age, CD4 counts, and the proportion of individuals on ART, when compared with PLWH infected with D614G. This indicates that there is an impairment in SARS-CoV-2 responses in PLWH, perhaps due to the incomplete restoration of immune function following ART (37,38). Delayed responses could increase the risk of re-infections or vaccine breakthrough infections and have been associated with SARS-CoV-2 RNAemia and COVID-19 progression (71). However, these same delays were not noted following ChAdOx1 nCoV-19 vaccination. Vaccinated PLWH rapidly elicited robust ADCC responses exceeding the levels in PWOH after the first dose and these levels remained higher 6 months post-vaccination. ChAdOx1 nCoV-19-induced ADCC in PWOH showed greater cross-reactivity against Delta and lasted longer than binding and neutralizing antibodies as previously reported (72). This highlights that ADCC may also be an important mechanism for durable and broad vaccine-induced immunity in PLWH but requires further investigation into how this functional response is differentially regulated in these individuals. In contrast to comparable binding kinetics observed for D614G infection, binding responses were initially delayed in ChAdOx1 vaccinated PLWH. This could be due to lower antigen exposure in vaccination, as the only difference in spike immunogen sequences is the D614G mutation. However, these binding responses increased to equivalent levels in PWOH after the second vaccine and result in higher binding at 6 months post-vaccination. In an immunogenicity study for the ChAdOx1 vaccine, including larger numbers of PLWH, higher SARS-CoV-2 spike binding antibodies were reported with each vaccine dose in PLWH (28). The durability and higher binding is potentially due to the late onset of this response in PLWH compared with PWOH. While neutralization and ADCP were delayed in infection, after vaccination, these responses had similar kinetics but lower levels in PLWH, although not significantly so for neutralization. Overall, vaccination induces different effects compared with D614G infection despite a similar antibody-targeted immunogen.
In addition to delayed responses following infection, PLWH had a surprisingly better coordination of antibody functional responses. For SARS-CoV-2, early infection that triggers an enhanced coordinated Fc effector response has been associated with COVID-19 recovery (47,48). Here, our data suggest that PLWH have a unique trajectory in humoral responses against SARS-CoV-2. This is not restricted to SARS-CoV-2 responses, as we have previously reported on a distinct coordination of antibody responses in PLWH in comparison with PWOH after influenza vaccination (66). However, the mechanisms that drive this more coordinated response in PLWH require further investigation. Given that there was no difference in severity or death between PLWH and PWOH in this cohort, it is possible that a more balanced or synergistic immune response, despite delays in their development, may be required for protection in this population.
Limitations of this study include the lack of sequences for the infecting variant; however, this was mitigated by sampling when the circulating variant accounted for > 90% of infections in the population. For Beta-triggered neutralization responses, PWOH had higher responses; however, the lack of statistical significance when compared with responses in PLWH is possibly a result of small participant numbers. In several cases, due to small numbers of ARTnaïve PLWH, we pooled data from both waves, limiting our ability to assess the impact of the infecting variant. In addition, HIV viral loads were not available for all participants and, as a result, viral suppression could not be defined. Furthermore, sample sizes for the kinetics and coordination analysis were reduced due to the lack of follow-up samples from some patients. The SARS-CoV-2 infection humoral kinetics reported in this study are restricted to the first week following infection, which corresponds to the acute phase. Therefore, unlike in vaccination where longitudinal samples were obtained, a more complete analysis of the infection humoral kinetics was not possible.
In conclusion, our study shows that SARS-CoV-2 infection induces different kinetics of responses to vaccination between PLWH and PWOH. These data highlight the importance of rapid ART rollout and support the current SARS-CoV-2 vaccine implementation strategies in PLWH. Despite delayed kinetics, PLWH were able to elicit comparable responses to SARS-CoV-2 infection and vaccination. This study highlights the importance of investigating the kinetics of humoral immune responses against SARS-CoV-2 in PLWH, as these provide insights into the mechanisms required for immunity against severe disease and vaccine efficacy in this population.

Data availability statement
The original contributions presented in the study are included in the article/Supplementary Material. Further inquiries can be directed to the corresponding authors.

Ethics statement
The studies involving human participants were reviewed and approved by University of Pretoria, Faculty of Health Sciences' Research Ethics Committee (247/2020) and Human Research Ethics Committee of the University of the Witwatersrand (ethics reference number: M200501). The patients/participants provided their written informed consent to participate in this study.

Author contributions
BM, PM, and SR conceptualized the study. BM performed experiments, analysed data, generated the figures and wrote the manuscript. NM and SR performed Fc experiments and analysed data. HK and PK performed neutralization assays supervised by TH. FAy and ZM performed ELISA assays supervised by TM-G. MM processed samples and FAb, MB, VU, and TR established the Pretoria COVID-19 study, which provided participant samples from the Tshwane District Hospital. SM provided samples and clinical data for the ChAdOx1-nCoV19 vaccination trial. PM and SR assisted in data interpretation, reviewed and edited the manuscript and supervised the research. All authors contributed to the manuscript and approved the submitted version.