Broad spectrum SARS‐CoV‐2‐specific immunity in hospitalized First Nations peoples recovering from COVID‐19

Indigenous peoples globally are at increased risk of COVID‐19‐associated morbidity and mortality. However, data that describe immune responses to SARS‐CoV‐2 infection in Indigenous populations are lacking. We evaluated immune responses in Australian First Nations peoples hospitalized with COVID‐19. Our work comprehensively mapped out inflammatory, humoral and adaptive immune responses following SARS‐CoV‐2 infection. Patients were recruited early following the lifting of strict public health measures in the Northern Territory, Australia, between November 2021 and May 2022. Australian First Nations peoples recovering from COVID‐19 showed increased levels of MCP‐1 and IL‐8 cytokines, IgG‐antibodies against Delta‐RBD and memory SARS‐CoV‐2‐specific T cell responses prior to hospital discharge in comparison with hospital admission, with resolution of hyperactivated HLA‐DR+CD38+ T cells. SARS‐CoV‐2 infection elicited coordinated ASC, Tfh and CD8+ T cell responses in concert with CD4+ T cell responses. Delta and Omicron RBD‐IgG, as well as Ancestral N‐IgG antibodies, strongly correlated with Ancestral RBD‐IgG antibodies and Spike‐specific memory B cells. We provide evidence of broad and robust immune responses following SARS‐CoV‐2 infection in Indigenous peoples, resembling those of non‐Indigenous COVID‐19 hospitalized patients.


INTRODUCTION
Severe acute respiratory syndrome coronavirus 2 (SARS-CoV-2), the infectious agent that causes coronavirus disease 2019 (COVID- 19), has led to a substantial health burden globally. 1High-risk groups, including the elderly, 2 pregnant women, 3 individuals with underlying comorbidities 2 and Indigenous peoples globally 4 are more susceptible to SARS-CoV-2 infections and severe diseases.Higher COVID-19 cases and case fertility rates in Indigenous peoples have been reported in the USA and Peru (Ucayali region), but this was not the case for Australia, Canada and Ecuador during the early phase of the pandemic. 1During 2020-2021, COVID-19 cases were very low in Australian First Nations peoples due to early and strict public health measures, interstate border closures and a "zero COVID-19 policy" in Australia.However, even after the lifting of COVID-19 restrictions, based on nationwide high COVID-19 vaccination targets (> 80-90% state-to-state), Australian First Nations peoples still have a COVID-19 mortality rate 1.6 times higher than non-Indigenous people. 5ocioeconomic factors can contribute to the high risk of infectious diseases in Indigenous populations. 6Chronic conditions, including diabetes and renal diseases, which are highly prevalent in Indigenous populations, 7 increase the severity of viral infections. 2 However, it is not yet fully understood whether immunological factors associated with chronic comorbidities are playing a role in disease severity.Our recent study showed robust immune responses towards COVID-19 vaccination in Australian First Nations peoples, although reduced antibody responses were found in individuals with diabetes and renal diseases. 8Similarly, previously we highlighted that Australian First Nations peoples can mount robust immune responses towards seasonal influenza vaccination. 9Yet, Australian First Nations peoples are still disproportionately affected by severe influenza disease, as evidenced by the 2009 influenza pandemic and seasonal influenza epidemics. 10,11herefore, scant data exist on whether immunological factors contribute to disease susceptibility and severity towards SARS-CoV-2 infection or other respiratory viral infections in Indigenous peoples, but are urgently needed to inform rational strategical planning and to better protect Indigenous populations globally.
In this study, we recruited Australian First Nations peoples hospitalized with COVID-19 to assess their immune responses during acute illness and at the recovery phase.We performed in-depth immune analyses, including antibody responses towards the Ancestral SARS-CoV-2 strain and variants of concern strains, cytokines and chemokines, cellular activation, memory-Spike-specific B cell responses, activating Spikespecific T cell responses (activation-induced marker, AIM) and Spike-epitope-specific T cell responses ex vivo using peptide-human leukocyte antigen (p-HLA) class-I and class-II tetramers.Our study provides evidence for robust immune responses following SARS-CoV-2 infection in Indigenous peoples.During hospitalization and recovery of Australian First Nations COVID-19 patients, antibodies directed towards receptor-binding domain (RBD) and nucleocapsid (N), antibody-secreting cells (ASCs), T follicular helper (Tfh) cells, tetramer-specific T cells and peptide-stimulated AIM + T cell frequencies resembled those reported for non-Indigenous COVID-19 hospitalized patients.

LIFT cohort of Australian First Nations peoples hospitalized with COVID-19
To investigate immune responses towards SARS-CoV-2 infection in Indigenous peoples, we recruited 13 PCRconfirmed COVID-19 Australian First Nations peoples hospitalized at the Royal Darwin Hospital between 16 November 2021 and 13 May 2022 (Supplementary table 1).We collected blood samples at study enrolment (median 3 days post disease onset) and at (or just prior to) hospital discharge (median 11 days post disease onset) (Figure 1a).The median age was 53 years (range 33-73 years) and 31% of COVID-19 patients were female (Figure 1b).As the Northern Territory in Australia had a zero COVID-19 policy prior to opening its borders in December 2021, SARS-CoV-2 infections in our cohort reflected patients' first SARS-CoV-2 exposures, with Delta and Omicron variants circulating at the time.COVID-19 patients included nine ward patients (six requiring supplemental oxygen support) and four ICU patients, all requiring non-invasive oxygen support.Eleven patients (7/9 ward, 4/4 ICU) were previously vaccinated with the BNT162b2 COVID-19 vaccine with one ward patient receiving one dose, six receiving two doses (four ward, two ICU) and four receiving three doses (two ward, two ICU).Patients received a combination of treatments including antivirals (n = 8), monoclonal antibody (Sotrovimab, targeting the Spike protein; n = 2), antibiotics (n = 3), immunomodulators (n = 13) and/or bronchodilator (n = 1).No patients died in the study (Supplementary table 1).All patients had comorbidities with 10/13 exhibiting three or more (Figure 1c).12/13 patients had diabetes and/or renal disease, which we have recently described to be associated with lower humoral immune responses following BNT162b2 vaccination. 8

Cytokine and antibody responses to SARS-CoV-2 infection in Australian First Nations peoples
Heightened inflammatory responses such as IL-6 and IL-18 and an increase in IL-8 during hospitalization were observed in hospitalized severe COVID-19 patients prior to COVID-19 vaccination but not healthy, mildly acute or convalescent individuals. 12,13Here, in the COVID-vaccination era, plasma cytokine and chemokine levels were fairly stable in Indigenous peoples at enrolment (V1) and discharge (V2), except for MCP-1 (mean 1596 pg mL À1 , range 102.7-12 039 pg mL À1 ; involved in regulating migration and infiltration of monocytes/macrophages) and IL-8 (mean 2765 pg mL À1 , range 250.1-17 204 pg mL À1 ; an interferon-gamma inducing proinflammatory cytokine), which increased by 14-fold and 6-fold at hospital discharge, respectively (Figure 2a, b).Indeed, these peak MCP-1 and IL-8 levels were 1.5-7.2-foldhigher than what we have observed previously in our earlier hospitalized cohorts of unvaccinated non-Indigenous patients (mean MCP-1 level of 1047 and 222.3 pg mL À1 , mean IL-8 level of 1041 and 678.6 pg mL À1 ). 12,13Conversely, our patients had similar V2 but 0.2-fold V1 (mean 449.9 pg mL À1 ) IL-8 levels compared with baseline levels of Indigenous and non-Indigenous people with chronic comorbidities (mean 2263 pg mL À1 ), 8 perhaps due to the usage of IL-8 at acute infection.
Antibody responses were comparable when patients were stratified into having one dose, two doses, third dose booster or monoclonal antibody treatment (Sotrovimab) (Figure 2d).IgG titers specific for Ancestral RBD were highly correlated with IgG titers specific for Delta RBD, Omicron RBD and Ancestral N (Figure 2e).Ex vivo assessment of Spike probe-specific B cells revealed an increase in cell frequency from enrolment (V1) to discharge (V2) in 3/6 patients (Figure 2f; Supplementary figure 1a).Moreover, the frequency of Spike-specific B cells correlated positively with IgG antibody titers (Figure 2g).
We further defined SARS-CoV-2-specific CD4 + and CD8 + T cells in COVID-19 Australian First Nations patients using the AIM assay and stimulation with overlapping peptide pools spanning Ancestral Spike, Membrane and Nucleocapsid (Figure 4a; Supplementary figure 1c).Measuring the expression of CD69, CD134 and CD137 activation markers, SARS-CoV-2-reactive CD4 + and CD8 + T cells directed at Spike, Membrane and Nucleocapsid generally increased during hospital stay (Figure 4b), although this was more apparent for AIMspecific CD4 + T cells (Figure 4c).
(g) Activation of tetramer + CD4 + and CD8 + T cells based on the activation markers CD38, CD71, HLA-DR and PD-1.Only samples with ≥ 10 tetramer + -enriched events were included for phenotyping and activation analysis.Mean with SD is shown in stacked plots.Statistical significance was determined with the two-tailed Wilcoxon test.
Overall, our study demonstrated broad and concurrent anti-viral immune responses in Australian First Nations peoples hospitalized with SARS-CoV-2 infection.

DISCUSSION
Our study provides key insights into immune responses following SARS-CoV-2 infection in Indigenous peoples.It demonstrates that Australian First Nations peoples hospitalized with COVID-19, and predominantly vaccinated, elicited broad immunity, including RBD/N-antibodies towards the Ancestral, Delta and Omicron strains, ASCs, Tfh cells and tetramer-specific CD4 + /CD8 + T cell responses.
Age, male gender and chronic conditions are independent and co-risk factors for contributing to COVID-19 disease severity, 2 as well as to other infectious viral diseases such as influenza. 31,32This may explain why the majority of our First Nations hospitalized patients were middle-aged or older, male and with chronic conditions (8/13 or 61.5% having all three risk factors).Moreover, we have shown previously that individuals with diabetes and renal diseases have perturbed antibody responses towards COVID-19 vaccines. 8Given that 12 out of 13 First Nations COVID-19 patients had diabetes and/or renal diseases, our hospitalized cohort may be more vulnerable than healthy individuals even after COVID-19 vaccination.Indeed, the patients had a median log 10 titer of 3.623 against Ancestral RBD at enrolment (V1), which is lower than the median of 4.072 observed in First Nations individuals without comorbidities after the 2nd COVID-19 vaccine dose. 8revious studies showed that severe COVID-19 was associated with hypercytokinemia and hyperactivation of innate and adaptive immune cells, 13,[33][34][35][36] together with high titers of antibodies. 36In the current study, an increase in levels of MCP-1 and IL-8 from enrolment (V1) to discharge (V2) was still observed, whereas other cytokines and chemokines remained stable.Prolonged coexpression of HLA-DR and CD38 on CD8 + and CD4 + T cells can be detrimental leading to disease outcomes, as was the case for a cohort of H7N9 influenza A virusinfected patients in China in 2013. 37Co-expression of HLA-DR + CD38 + on T cells has been described previously to be largely contributed by non-specific bystander activation of T cells during acute influenza virus infection in a mouse model. 38However, in our study, activated CD38 + HLA-DR + CD4 + and CD8 + T cells were transiently elevated at enrolment (V1), with levels significantly decreased prior to discharge (V2), most likely reflecting the patients' recovery.][42][43] In terms of SARS-CoV-2-specific T cell responses, AIM + CD4 + and CD8 + T cell responses were elevated in multiple patients during hospitalization.The increase of AIM + T cells is similar to those reported in non-Indigenous COVID-19 patients, which occurred gradually over time at early timepoints post disease onset. 43A previous study also reported a negative correlation of AIM + T cells with viral loads, indicating viral clearance. 43he frequency of tetramer-specific T cells was stable during hospitalization, which has been reported to be stable up to 9 months post SARS-CoV-2 infection for the immunodominant B7/N 105 CD8 + T cell epitope. 24These T cells at enrolment were at frequencies comparable to Australian First Nations peoples after two COVID-19 vaccine doses and exhibited a T CM -like phenotype. 8imilar to our previous study in influenza-infected hospitalized patients, 17 epitope-specific CD4 + and CD8 + T cells mainly expressed CD38, CD71 and/or PD-1.
Our study has some limitations.We acknowledge that samples were collected from a small number of hospitalized patients.Thus, the results might not apply to patients who do not require hospitalization.However, given the very low case numbers in Darwin during the study recruitment (November 2021-May 2022), where interstate borders had started to lift, we were still able to capture a small cohort of patients for immune analyses including those critically ill in ICU.As discussed earlier, the patients had several known risk factors for severe COVID-19, which might predispose them to hospitalization.Further studies are needed to investigate immune responses in Indigenous peoples infected with SARS-CoV-2 that cover broader demographical ranges.
Despite the limitations, in the current study, we performed broad immune analyses in Australian First Nations peoples hospitalized who had recovered from COVID-19.During hospitalization, Australian First Nations patients with COVID-19 elicited prototypical antibodies directed towards RBD and N, ASCs, activated Tfh cells, tetramer-specific CD4 + /CD8 + T cells and peptide-stimulated AIM + T cells.As the majority of our COVID-19 patient cohort (85%) was vaccinated, we have not observed greatly exacerbated cytokine profiles (with the exception of increased IL-8 and MCP-1 levels) or hyperactivation of T cells depicted by HLA-DR and CD38 expression, both previously associated with severe COVID-19 in unvaccinated patients.

METHODS LIFT study participants and specimens
We enrolled 13 Australian First Nations patients hospitalized with COVID-19 in Darwin between November 2021 and May 2022.Samples were taken at Visit 1 (V1, enrolment, n = 13) and Visit 2 (V2, hospital discharge, n = 10).Heparinized peripheral blood and serum were collected for isolating peripheral blood monocular cells (PBMCs), plasma, sera and granulocytes. 9

Antibodies, memory B cells and T-cell assays
IgG ELISAs for Ancestral, Delta and Omicron RBD, as well as Ancestral N were performed as described. 30,40A volume of 200 lL of whole blood was used to measure ASCs (CD3 À CD19 + CD27 hi CD38 hi ), activated Tfh cells (CD3 + CD4 + CXCR5 + ICOS + PD-1 + ) and activated CD8 + and CD4 + T cells (HLA-DR + CD38 + ), as described previously. 16emory Spike-specific B cell responses were measured on thawed PBMCs or TAME-flow through fractions using Ancestral Spike recombinant probes, essentially as described including full staining panel. 30The SARS-CoV-2-specific AIM assay was performed essentially as described 15 using thawed PBMCs cultured at 37°C/5% CO 2 for 24 h with 10 lg mL À1 lots of Spike (181 peptides, 0.06 lg mL À1 per peptide), Membrane (31 peptides, 0.32 lg mL À1 per peptide) or Nucleocapsid (59 peptides, 0.17 lg mL À1 per peptide) peptide pools (all BEI Resources, Rockville, USA), or DMSO (Sigma-Aldrich, St Louis, USA) as a negative control.The full staining panel of antibodies has been described in detail. 15Ex vivo TAME was performed, as essentially described, 15,30  (TFEYVSQPFLMDLE).Monomers were generated by the Rossjohn Laboratory (Monash University, Melbourne, Australia) and conjugated to streptavidin-labeled PE or APC fluorochromes (both BD) to form tetramers.TAME was performed using anti-PE and/or anti-APC microbeads, LS columns and QuadroMACS TM Separator (all Miltenyi Biotec, Bergisch Gladbach, Germany) to enrich for tetramer + cells.Full antibody staining panel has been described in detail. 30Enriched tetramer + cells below 10 were not analyzed by phenotype (CD45RA, CD27, CD95) or activation status (CD71, CD38, HLA-DR, PD-1).The cells were analyzed on LSRII Fortessa (BD), and FCS files were analyzed using FlowJo v10 software (BD).Flow cytometry gating strategies are shown in Supplementary figure 1.

Statistical analyses
Statistical significance was assessed using the two-tailed Wilcoxon signed-rank test and Spearman's correlation coefficient (r s ) in Prism v10 (GraphPad, Dotmatics, Boston, USA) unless stated otherwise.

Figure 2 .
Figure 2. Cytokine and IgG responses towards RBD and N following SARS-CoV-2 infection in Australian First Nations peoples.(a) Concentration of 13 cytokines and chemokines at enrolment (V1) and hospital discharge (V2).(b) Concentration of MCP-1 and IL-8.Bars indicate median with interquartile range.(c) IgG titer against Ancestral, Delta and Omicron RBD and Ancestral N. Red horizontal dotted lines indicate seropositivity defined as mean + 2SD of unexposed/pre-vaccination samples. 14,15(d) IgG titer against Ancestral, Delta and Omicron RBD and Ancestral N grouped by dose of COVID-19 vaccination or monoclonal antibody (mAb) treatment.(e) Correlation between Ancestral RBD IgG titer and Delta RBD, Omicron RBD and Ancestral N. (f) Concatenated representative FACS plots and frequency (%) of Spike-specific IgD À B cells.(g) Correlation between Spike-specific IgD À B cells and Ancestral, Delta and Omicron RBD and Ancestral N. Correlation was determined with Spearman's correlation.Statistical significance was determined with the two-tailed Wilcoxon test.Exact P-values are shown unless P < 0.0001. activated

Figure 3 .
Figure 3. Prototypic cellular responses in Australian First Nations peoples infected with SARS-CoV-2.(a) Representative FACS plots and frequency (%) of CD27 + CD38 + antibody-secreting cells (ASC) and activated ICOS + PD-1 + CXCR5 + CD4 + T follicular helper cells (Tfh).(b) Representative FACS plots and % of activated CD38 + HLA-DR + CD4 + and CD8 + T cells.Red horizontal dotted lines indicate healthy cutoff with the median of % activated cells from healthy individuals.16(c) Correlation between % activated CD4 + T cells with % ASC, activated Tfh cells and activated CD8 + T cells.Correlation was determined with Spearman's correlation.Statistical significance was determined with the two-tailed Wilcoxon test.Exact P-values are shown.

971W
Zhang et al.SARS-CoV-2-specific immunity in hospitalized First Nations COVID-19 patients Bloods were shipped overnight from the Menzies School of Health Research to the University of Melbourne for immediate blood processing and whole blood analyses.Demographic, clinical and sampling information are described in Supplementary table 1.All donors consented to genetic analysis of HLA-I and HLA-II which was performed on genomic DNA isolated from granulocytes by the Victorian Transplant and Immunogenetics Service (Melbourne, Australia) and the results are shown in Supplementary table 2. Experiments conformed to the Declaration of Helsinki Principles and the Australian National Health and Medical Research Council Code of Practice.Written informed consent was obtained from all blood donors prior to the study.The study was approved by the Human Research Ethics Committee of the Northern Territory Department of Health and Menzies School of Health Research (#2012-1928, LIFT) and the University of Melbourne Human Research Ethics Committees (#21864).