Abstract
Aims
The HLA system has been implicated as an underlying determinant for modulating the immune response to SARS-CoV-2. In this study, we aimed to determine the association of patients’ HLA genetic profiles with the disease severity of COVID-19 infection.
Methods
Prospective study was conducted on COVID-19 patients (n = 40) admitted to hospitals in Saskatoon, Canada, between March and December 2020. Next-generation sequencing was performed on the patient samples to obtain high-resolution HLA typing profiles. The statistical association between HLA allelic frequency and disease severity was examined. The disease severity was categorized based on the length of hospital stay and intensive care needs or demise during the hospital stay.
Results
HLA allelic frequencies of the high and low-severity cohorts were normalized against corresponding background allelic frequencies. In the high-severity cohort, A*02:06 (11.8-fold), B*51:01 (2.4-fold), B*15:01(3.1-fold), C*01:02 (3.3-fold), DRB1*08:02 (31.2-fold), DQ*06:09 (11-fold), and DPB1*04:02(4-fold) were significantly overrepresented (p < 0.05) making these deleterious alleles. In the low-severity cohort, A*24:02 (2.8-fold), B*35:01 (2.8-fold), DRB1*04:07 (5.3-fold), and DRB1*08:11 (22-fold) were found to be significantly overrepresented (p < 0.05) making these protective alleles. These above alleles interact with NK cell antiviral activity via the killer immunoglobulin-like receptors (KIR). The high-severity cohort had a higher predilection for HLA alleles associated with KIR subgroups; Bw4-80I (1.1-fold), and C1 (1.6-fold) which promotes NK cell inhibition, while the low-severity cohort had a higher predilection for Bw4-80T (1.6-fold), and C2 (1.6-fold) which promote NK cell activation.
Conclusion
In this study, the HLA allelic repository with the distribution of deleterious and protective alleles was found to correlate with the severity of the clinical course in COVID-19. Moreover, the interaction of specific HLA alleles with the KIR-associated subfamily modulates the NK cell-mediated surveillance of SARS-CoV-2. Both deleterious HLA alleles and inhibitory KIR appear prominently in the severe COVID-19 group focusing on the importance of NK cells in the convalescence of COVID-19.
Keywords: HLA typing, COVID-19, Natural killer cells, Killer cell immunoglobulin-like receptors, KIR
1. Introduction
As of September 3rd, 2022, the Coronavirus disease 2019 (COVID-19) pandemic caused by the severe acute respiratory syndrome coronavirus 2 (SARS-CoV-2) has resulted in over 604 million cases and 6.4 million deaths [1]. This vast morbidity and mortality have imposed secondary pressures on the global public health system and propagated an economic crisis [2], [3], [4]. A peculiar spectrum of clinical severity amongst patients infected with SARS-CoV-2 was present. Some patients remained asymptomatic, while others presented with multiorgan failure and subsequent demise. The remainder of the patients presents within these two extremes. This variegated morbidity and mortality profile has been reported to be associated with several factors, including geographical, environmental, and biological (i.e., age, blood group, social norms, and vaccination policies) [5], [6], [7], genes involved in immunological antiviral defense and inflammatory cascades [8], [9].
It was reported that the human leukocyte antigen (HLA) genotype conferred differential viral susceptibility and, subsequently, a variegated clinical severity [10], [11]. HLA genes are encoded at the HLA super locus within chromosome 6p21, which codes for six classical proteins [12]. HLA presents peptides derived from foreign organic entities on their respective host cell surface to facilitate T-cell recognition. Many studies have shown that the polymorphisms of HLA genes modulated the clinical course in diseases caused by SARS-CoV-1, HIV, and Influenza H1N1 [13], [14], [15].
Furthermore, the amalgamation of the innate and adaptive host responses and the virion’s antigenic profile facilitates evasion of this host response, thereby dictating the clinical severity. Natural killer (NK) cells are rapid-acting innate lymphocytes involved in the early viral response. Therefore, studying the SARS-CoV-2 interaction with NK cells is prudent. SARS-CoV-2′s primary target is the respiratory epithelial cells. Upon virion interaction, these cells precipitate a localized inflammatory response via the pattern recognition receptors (PRRs) [16]. Coronaviridae family viruses are recognized by both immune and non-immune cells via pathogen recognition receptors (PRRs) such as toll-like receptors (TLR3, TLR7, TLR8) and retinoic acid-inducible gene I (RIG-I) [17], [18]. These PRRs upregulate the transcription of type I interferons that prime NK cells. One study showed that with anti-SARS coronavirus-specific antibodies, the NK cell expression of CD158b + correlated with the clinical severity [19]. A wide array of activating and inhibitory killer immunoglobulin-like receptors (KIRs) are expressed on the surface of NK cells which interacts with the HLA class I ligands on cells infected by SARS-CoV-2. Ligands such as KIR2DL1 recognize the allotype HLA-C group 2 (HLA-C2) with lysine at position 80, whereas KIR2DL2 and KIR2DL3 ligands recognize the allotype HLA-C group 1 (HLA-C1) with asparagine at position 80 [20]. Furthermore, the KIR3DL1 ligand has increased binding affinity for the HLA-B Bw4 molecules with isoleucine at position 80 (Bw4-80I) compared to Bw4-80 T has threonine at position 80 [21]. Lastly, the ligand KIR3DL2 recognizes the allotypes HLA-A3 and -A11 (A3/11) [22], [23]. Additionally, there are many other non-KIR binding ligands (NoL) that interact with HLA alleles with unknown clinical significance. Therefore, the intricate KIR-HLA interaction is one of the determinants of the antiviral response.
The rationale of this study is to identify specific protective and deleterious HLA alleles that can explain the low clinical severity and high clinical severity in the COVID-19 groups. The NK cells are the first response to viral infection. The ancillary aim was to understand the HLA allele interplay with NK cells’ KIR receptors subfamilies, which could explain the low clinical severity and high clinical severity in the COVID-19 groups.
2. Methods
2.1. Study participants and blood sample collection
This project was approved by the Biomedical Research Ethics Board (Bio ID 2513) and sponsored by the College of Medicine at the University of Saskatchewan. Prospective study on COVID-19 patients (n = 40) admitted in Saskatoon March-December 2020 during the prominent SARS-CoV-2 variant of concern B.1.1.7 (ALPHA strain). The study samples were categorized into two cohorts.1) Low severity COVID-19 cohort: Patients in this group consisted of individuals with a hospital stay of less than 7 days without intensive care unit transfer, need for intubation, or demise during the hospital stay; 2) High severity COVID-19 cohort: patients in this group were individuals that required intensive care unit transfer or ventilatory support via mechanical intubation or demise during the hospital stay. Residual EDTA blood samples were retrieved from the laboratory after completing the clinical testing.
2.2. Background allelic frequency
A library of HLA allele frequencies was compiled from January to March 2020 in 1079 participants. High-resolution HLA typing was performed on donors and recipient candidates that were involved with the hematopoietic stem cell and solid organ transplant services of the Saskatchewan health authority. We excluded matched HLA donors from the allele frequency to avoid bias in the calculation of allele frequencies. This allelic library was used as a background population allelic frequency.
2.3. DNA extraction
DNA extractions were prepared by the QIAGEN, BioRobot® EZ1 (Qiagen, Toronto, Canada) and EZ1 DNA Blood 350 μl Kit (Catalog 951054) using whole blood collected in EDTA tubes. DNA concentration and purity were quantified by a Nanodrop (ThermoScientific, Canada), and samples with a 260/280 ratio > 1.8 were processed.
2.4. HLA genotyping
HLA genotyping was performed using the One Lambda AllType FASTplex NGS kits (One Lambda, USA) and the Illumina MiniSeq platform (Catalog FC-420-1004) (Illumina, USA). The full genome sequence was done on HLA: A, B, C, DRB1, DRB3/4/5, DQA1, DQB1 DPA1, and DPB1. TypeStream Visual 2.1 NGS analysis software (One Lambda, USA) was used for HLA data analysis.
2.5. Statistical analysis
All HLA allele families and haplotype frequencies were calculated by direct counting, and these were compared between the two COVID-19 severity groups. Additionally, the allele frequencies in the low and high COVID-19 severity groups were compared against the background allelic frequencies to identify the over-represented and under-represented alleles in those cohorts. Due to our sample size (n = 40), we expected lower allele frequencies as such the Fisher’s exact test was utilized for comparisons instead of the chi-square test. All data were analyzed using SPSS V.28.0. A probability value with a P-value less than 0.05 was considered statistically significant. Furthermore, Bonferroni correction was applied such that the new P-value cut-off is representative of the total number of alleles compared to the nine HLA loci. For a power of 80% with the deleterious allele frequency of each protective and deleterious allele in our dataset compared to their corresponding background provincial frequency, we would require at least 400 cases per group for an alpha of 0.05. Since this is a qualitative comparative pilot study we intended to obtain 10% of the calculated sample size as suggested in the literature [24]. We acknowledge that this study is underpowered due to the smaller cohort size. Lastly, HLA alleles have different frequencies in different ethnic populations, and the patients in our COVID −19 severity cohorts as well as the background population cohort are multiethnic in Saskatchewan. A multiethnic population could further affect the power of this study with the presence of cryptic subpopulations forming confounding by locus due to linkage disequilibrium of specific alleles. Genomic control loci were not utilized in this pilot study and therefore limit the elimination of false positive results.
3. Results
3.1. Patients profile
The patient demographics, clinical and laboratory variables contributing to the hospitalization, and clinical course amongst the patients are summarized in Table 1 . The results showed a nearly equal distribution of patients within each COVID-19 clinical severity cohort. A higher predilection for males was found in both groups. The elderly population was predominant in the high-severity cohort, whereas the low-severity cohort contained a predominantly younger population (46 ± 6 years). Hypertension and type II diabetes mellitus were the most common comorbidities in both cohorts. In the high-severity cohort, the frequency of the presenting complaints with respiratory decompensation increased by 3-fold compared to the low-severity cohort. The presenting vitals showed higher oxygen desaturation and relative thrombocytopenia in the higher-severity group. The imaging findings with consolidation and/or glass ground opacities and length of stay in the high severity cohort were 2-fold and 10-fold higher, respectively.
Table 1.
Summary of patient demographics, presenting vitals, laboratory and imaging findings, and hospitalization course.
| Parameters |
COVID-19 Severity |
||
|---|---|---|---|
| Low | High | ||
| Demographics | n | 17 | 23 |
| Male: Female | 11:6 | 14:9 | |
| mean age (SD) | 46 (6) | 62 (4) | |
|
Comorbidity profile, n (%) |
ddCKD | 2 (11.7) | 1 (4.3) |
| T2DM | 6 (35.2) | 8 (34.8) | |
| HTN | 4 (23.5) | 9 (39.1) | |
| CAD | 0 (0) | 3 (13.0) | |
| Presenting complain, n (%) | Respiratory decompensation | 5 (29.4) | 15 (65.2) |
|
Presenting Vitals, n (%) |
SBP (mmHg) | 123 (23) | 125 (25) |
| DBP (mmHg) | 73 (15) | 71 (20) | |
| SaO2 (%) | 94.5 (4.6) | 89 (4.8) | |
| Tmax (⁰C) | 36.5 (0.4) | 37.1 (1.0) | |
|
Presenting Labs, n (%) |
WBC (109/L) | 8.5 (4.1) | 9.0 (4.8) |
| Hb (g/L) | 118.6 (20) | 114.9 (29) | |
| PLT (109/L) | 261 (136) | 237 (110) | |
| Imaging findings, n (%) | Consolidation and/or GGO | 6 (35.3) | 12 (54.5) |
|
Hospital course, n (%) |
Mean LOS. days (SD) | 3 (0.7) | 30 (7) |
| ICU admission | 0 (0) | 17 (73.9) | |
| Intubation | 0 (0) | 9 (39.1) | |
| Readmission post discharge | 13 (76.5) | 5 (21.7) | |
| Deceased | 0 (0) | 12 (52.2) | |
Abbreviations: Standard deviation (SD), dialysis dependent chronic kidney disease (ddCKD), type 2 diabetes mellitus (T2DM), hypertension (HTN), coronary artery disease (CAD), systolic blood pressure (SBP), diastolic blood pressure (DBP), arterial oxygen saturation (SaO2), maximum temperature (Tmax), White blood cell count (WBC), Hemoglobin (Hb), Platelet count (PLT), ground glass opacities (GGO), length of stay (LOS), Intensive care unit (ICU).
3.2. HLA alleles and clinical severity
We examined the distribution of HLA class I alleles in the low- and high-severity cohorts among all the COVID-19 infected patients (Supplementary Tables S1). HLA-A*11:01 allelic frequencies in low and high-severity COVID-19 infection were 2.9% (n = 1/34) and 8.7% (n = 4/46), respectively. Similarly, HLA-A*31:01 allelic frequencies in low- and high-severity COVID-19 infection were 2.9% (n = 1/34) and 8.7% (n = 4/46), respectively. In both A*11:01 and A*31:01, a 3-fold enrichment between the low- and the high-severity groups was observed. HLA-A*02:01 allelic frequencies in low- and high-severity COVID-19 infection were 20.5% (n = 7/34), and 26.1% (n = 12/46), respectively (Fig. 1 A). HLA-B*51:01 allelic frequencies in low- and high-severity COVID-19 infection were 5.9% (n = 2/34), and 15.2% (n = 7/46), respectively. Similarly, HLA-B*35:01 allelic frequencies in low- and high-severity COVID-19 infection are 17.6% (n = 6/34), and 6.5% (n = 3/46), respectively. Compared to the low-severity group, B*51:01 showed a 1.6-fold enrichment in the high-severity group, whereas B*35:01 showed a 0.37-fold enrichment in the high-severity group (Fig. 1B).HLA-C*04:01 allelic frequencies in low- and high-severity COVID-19 cohorts were 20.6% (n = 7/34), and 6.5% (n = 3/46), respectively. Similarly, HLA-C*05:01 and C* 12:03 allelic frequencies in low- and high-severity COVID-19 infection were 8.8% (n = 3/34) and 2.2% (n = 1/46), respectively. A 0.47-fold change in C*04:01 was found between the high- and low-severity groups. In both C*05:01 and C*12:03, a 0.25-fold enrichment between the low- and the high-severity groups (Fig. 1C).
Fig. 1.
Distribution of HLA class I alleles in low (green bars) and high (red bars) severity COVID-19 infection. A) allelic frequency distribution in the HLA-A group. B) allelic frequency distribution in the HLA-B group. C) allelic frequency distribution in the HLA-C group. Comparisons showing HLA class I protective alleles (overrepresented alleles in the low severity group) and deleterious alleles (overrepresented alleles in the high severity group). (For interpretation of the references to colour in this figure legend, the reader is referred to the web version of this article.)
We next examined the distribution of HLA class II alleles in the different COVID-19 severity groups as summarized in (Supplementary Table S2). DRB1*16:02 and DRB1*08:02 were exclusively seen in the low-severity (5.8%; n = 2/34) and high-severity (13%; n = 6/46) cohorts, respectively. Moreover, DRB1*03:01 allelic frequencies in low-, and high-severity cohorts were 2.9% (n = 1/34), and 6.5% (n = 3/46) respectively. A 2.2-fold enrichment between the high- and low-severity groups in the DRB1*03:01 frequencies was observed (Fig. 2 A). DRB5*02:02 were exclusively absent in the high-severity cohort and only seen in the low-severity (14.7%, n = 5/34). Both DRB3*03:01 and DRB3*01:01 allelic frequencies in low-, and high-severity cohorts were 5.9% (n = 2/34) and 13.6% (n = 6/46), respectively. DRB3* 03:01 and DRB3* 01:01 showed a 2.3-fold enrichment between the high-severity and low-severity groups (Fig. 2B). Both DQA1*01:04, DQB1*05:03, and DQB1*06:04 were exclusively absent in the high-severity cohort and only seen in the low-severity (5.9%, n = 2/34). Both DQA1*01:01 and DQB1*05:01 allelic frequencies in low-, and high-severity cohorts are 2.9% (n = 1/34), 10.9% (n = 5/46) and 2.9% (n = 1/34), 10.9% (n = 5/46); respectively. Similarly, both DQA1*01:01 and DQB1*05:01 showed a 3.8-fold enrichment between the high- and low-severity groups (Fig. 2C and 2D). Overall, DPA1*01:03 was the most common allele in the low (94.1%, n = 32/34) and high-severity (82.6%, n = 38/46) cohorts. DPB1*104:01 was exclusively seen in the low-severity cohort at a frequency of 5.8% (n = 2/34). DPB1*04:02 frequency in low- and high-severity cohorts was 17.6% (n = 6/34) and 34.8% (n = 16/46), respectively. There is a 2.7-fold enrichment between the high-severity and low-severity groups in the DPB1*04:02 frequencies (Fig. 2E).
Fig. 2.
Distribution of HLA class II alleles in low and high severity COVID-19 infection. A) allelic frequency distribution in the HLA-DRB1 group. B) allelic frequency distribution in the HLA-DRB235 group. C) allelic frequency distribution in the HLA-DQA1 group. 2D: allelic frequency distribution in the HLA-DQB1 group. E) allelic frequency distribution in the HLA-DPB1 group.
No statistically significant difference was seen between the low- and high-severity COVID-19 groups for any HLA class I and II alleles as shown in Tables S1-S2.
3.3. HLA homozygosity and clinical severity
The high-severity cohort had the highest frequency of homozygous genotypes (55.5%, n = 61/110) compared to the low severity (44.5%, n = 49/110) cohorts. Homozygosity was 1.2-fold higher in the high-severity group than in the low-severity group and did not meet the statistical significance (p = 0.31). In addition, homozygosity in the high-severity group showed a predilection for the HLA class II (72%, n = 44/61) (Fig. 3 and Supplementary table S3).
Fig. 3.
Cumulative frequency distribution of all homozygous genotypes from all 9 HLA-loci (HLA class I and II alleles) (n = 110) in low (green bar) and high (red bar) severity COVID-19 patients. (For interpretation of the references to colour in this figure legend, the reader is referred to the web version of this article.)
Overall, DPA1*01:03 was the most common homozygous genotype in the low (30.6%, n = 15/49) and high-severity (24.6%, n = 15/61) groups. DRB4*01:03 (9.8%, n = 6/61), and DPB1*04:02 (6.6%, n = 4/61) were found to be enriched in the high-severity group. The haplotype A*24:02 (6.1%, n = 3/49) was found to be enriched in the low-severity group. The frequency distribution of the homozygous haplotypes of the HLA class I and II are shown in Fig. 4 .
Fig. 4.
Summary of homozygous HLA genotypic frequency in COVID-19 low severity (LS) and high severity (HS) group. A) HLA class I genotypic frequencies and B) HLA class II genotypic frequencies.
3.4. KIR-HLA interaction and clinical severity
The distribution of HLA alleles associated with the KIR subgroups (A3/11, Bw4, Bw6, C1, and C2) was further analyzed in low and high-severity cohorts. As shown in Fig. 5 , the high-severity cohort had a higher predilection for HLA alleles associated with the A3/11 (6.4% vs 4.9%), Bw4-80I (24.4% vs 21.3%), and C1 (21.8% vs 13.1%) KIR subgroups. In contrast, the low-severity cohort had a higher predilection for HLA alleles associated with the Bw4-80T (9.8% vs 6.4%), Bw6 (24.6% vs 23.1%), and C2 (26.2% vs 17.9%) KIR subgroups.
Fig. 5.
Summary of HLA class I allele frequencies associated with killer Ig-like receptors genes (KIR) subgroups (A3/11, Bw4, Bw6, C1, and C2) in low severity (n = 61) and high severity (n = 78) COVID-19 patients.
The deleterious HLA alleles with their respective killer Ig-like receptors genes (KIR subtypes; A3/11, Bw4-80I, Bw4-80T, Bw6, C1, and C2) associations in the high-severity group are summarized in Table 2 . All displayed HLA frequencies in the high-severity cohort were normalized to their corresponding background allelic frequencies in the Saskatchewan population (Supplementary table S4). No deleterious HLA alleles were found in the low-severity cohort that attained statistical significance (p < 0.05). Of the statistically significant HLA alleles, the deleterious alleles had a predilection for HLA class II (n = 53) compared to HLA class I (n = 34). Notable enriched deleterious HLA class I alleles associated with KIR subtypes consisted of Bw4-80I (B*51:01), Bw6 (B*15:01, B*15:10), and C1 (C*01:02). As shown in Table 2, multiple other class I and II alleles pertaining to the non‐KIR‐binding molecules (NoL) were also enriched. Enriched alleles associated with KIR and NoL that did not meet statistical significance were illustrated in Supplementary Table S5 (HLA class I) and S6 (HLA class II).
Table 2.
Enrichment of deleterious HLA alleles in the high severity groups normalized to their background allelic frequency.
| HLA alleles | Associated KIR subtypes | Allele frequency n (%) | Fold change compared to Background allelic frequency | Fisher exact test statistic | Bonferroni correction | ||
|---|---|---|---|---|---|---|---|
| High Severity Cohort | HLA class I | B*51:01 | Bw4-80I | 7 (15.2) | 2.3x | 0.033 | NS |
| B*15:01 | Bw6 | 8 (17.4) | 2.9x | 0.006 | NS | ||
| C*01:02 | C1 | 5 (10.9) | 3.1x | 0.025 | NS | ||
| C*04:04 | NoL | 2 (4.3) | 17.5x | 0.009 | NS | ||
| A*02:06 | NoL | 6 (10.8) | 11.8x | 0.000038 | <0.001* | ||
| A*30:02 | NoL | 3 (6.5) | 7.2x | 0.014 | NS | ||
| B*53:01 | NoL | 2 (4.3) | 12.8x | 0.015 | NS | ||
| HLA class II | DRB1*01:02 | NoL | 2 (4.4) | 6.6x | 0.045 | NS | |
| DRB1*08:02 | NoL | 6 (13.3) | 30.7x | 0.00000032 | <0.003* | ||
| DRB1*08:11 | NoL | 3 (6.7) | 12.6x | 0.0029 | <0.003* | ||
| DQB1*03:02 | NoL | 11 (23.9) | 2.6x | 0.0028 | <0.005* | ||
| DQB1*04:02 | NoL | 8 (17.4) | 5.5x | 0.00013 | <0.005* | ||
| DQB1*06:09 | NoL | 4 (8.7) | 10.4x | 0.00096 | <0.005* | ||
| DPB1*04:02 | NoL | 16 (34.8) | 2.7x | 0.00015 | <0.006* | ||
Abbreviations: non‐KIR‐binding molecules (NoL), not statistically significant (NS), statistical significance (*).
The protective HLA alleles associated with the KIR subtypes and NoL in the high and low-severity groups are summarized in Table 3 . All displayed HLA frequencies in the high and low-severity cohorts were normalized to their corresponding background allelic frequencies in the Saskatchewan population. All protective HLA alleles listed in Table 3 were statistically significant. In the high-severity cohort, two protective alleles were found associated with the KIR subtype A3/11 (A*03:01) and NoL (A*31:01). Notable enriched protective HLA class I alleles associated with KIR subtypes consisted of Bw4-80I (A*24:02), and Bw6 (B*35:01). Table 3 also shows multiple other enriched class I and II alleles pertaining to the NoL group. Enriched protective alleles associated with KIR and NoL that did not meet statistical significance were illustrated in Supplementary Table S7 (HLA class I) and S8 (HLA class II). In addition, no background data on population-based allelic frequency was available for DRB3/4/5, DQA1, and DPA. Therefore, no corresponding harmful or protective alleles for DRB3/4/5, DQA1, and DPA1 were not included.
Table 3.
Enrichment of protective alleles in the high and low severity groups normalized to their background allelic frequency.
| HLA alleles | Associated KIR subtypes |
Allele frequency n (%) |
Fold change compared to Background allelic frequency | Fisher exact test statistic | Bonferroni correction | ||
|---|---|---|---|---|---|---|---|
| Low Severity |
HLA class I |
A*24:02 | Bw4-80I | 9 (26.5) | 2.8x | 0.005 | NS |
| B*35:01 | Bw6 | 6 (17.6) | 2.8x | 0.018 | NS | ||
| A*02:05 | NoL | 2 (5.8) | 6.0x | 0.05 | NS | ||
| A*02:06 | NoL | 3 (8.8) | 7.5x | 0.011 | NS | ||
| C*03:04 | NoL | 6 (17.6) | 2.6x | 0.029 | NS | ||
| HLA class II | DRB1*08:11 | NoL | 4 (11.8) | 22.1x | 0.000068 | < 0.003* | |
| DRB1*16:02 | NoL | 2 (5.9) | 20.2x | 0.007 | NS | ||
| DQB1*03:02 | NoL | 8 (23.5) | 2.6x | 0.011 | NS | ||
| DQB1*04:02 | NoL | 5 (14.7) | 4.7x | 0.005 | NS |
Abbreviations: non‐KIR‐binding molecules (NoL), not statistically significant (NS), statistical significance (*).
4. Discussion
The scientific literature on the association of HLA allelic repertoire with COVID-19 clinical severity has rapidly expanded in the past two years. Many studies showed accordant or discordant results pertaining to HLA alleles that were protective or resistant to SARS-CoV-2 infection. Furthermore, these studies often described the allelic repository within a unique cultural cohort to minimize genotypic heterogeneity. However, to our knowledge, this study was the first to describe the HLA allelic repository in Canada within a multicultural and multiethnic population stratified based on clinical severity [25].
Many concordant trends in the HLA allelic frequency changes with COVID-19 severity are seen in our multicultural population and the literature. Within the HLA class I, A*24:02 was protective in our population, but it was reported in the literature that A*24:02 was equally prominent within the low and high-severity cohorts in the Russian and Japanese populations respectively [26], [27]. Additionally, A*11:01, A*31:01, and A*02:01 were prominently enriched in the high-severity cohort. Notably, allelic frequency and COVID-19 severity studies from Greece, Turkey, Japan, Spain, and the USA showed that A*11 has consistently been enriched in patients with severe COVID-19 [28], [29], [30], [31], [32]. Many high-resolution typing studies have shown HLA-A*11:01 as a predisposing factor for severe COVID-19 disease courses [30], [32], [33], [34]. The HLA-A loci, A*02:05, A*29:02, and A*68:01 were enriched in our low-severity COVID-19 groups. The literature also shows that A*68:01 and A*02:05 were found to be protective alleles in the Mexican and Sardinian populations, respectively [35], [36]. B51 was enriched in the high-severity cohort, while B35 was seen as a protective allele. Both these HLA-B alleles show similar trends in the south Asian population [33], [37], [38]. Interestingly, the specific antigens coded by B*35 present a higher peptide loading capacity than all other HLA-B proteins, which may translate to the observed protective immune surveillance. Furthermore, in the HLA-C loci, C*04:01, C*05:01, and C*12:03 appear to be protective in our low-severity data. In contrast, C*01:02, C*06:02, and C*15:02 appeared to be harmful in our high-severity data. Consistent with our findings, one study analyzing the repertoire of SARS-CoV-2 epitopes involved in the T-cell-based viral mitigation showed that C*04:01 was the most elevated antigen-presenting allele, and C*01:02 was their worse presenting allele which corresponds to its protective and harmful status respectively [9].
Amongst the HLA allelic studies published in the literature, only a few studies focused on the HLA class II allele repertoire contribution to the COVID-19 severity. In the HLA-DRB1 group, DRB1*07:01, DRB1*04:04, DRB1*08:11, and DRB1*16:02 were protective alleles in our low-severity dataset. However, DRB1*08:02, DRB1*03:01, and DRB1*13:02 were deleterious alleles in our high-severity dataset. One study showed significant negative associations with spike antibody titers post single dose of the BNT162b2 (Pfizer- BioNTech) vaccine, and DRB1*04:04 and DRB1*07:01 were observed among single dose vaccinated individuals with prior SARS-CoV-2 infection. Conversely, DRB1*03:01 patients showed a significantly elevated titer of anti-Spike titers [39]. The DRB1*08:02 was reported to associate with symptomatic COVID-19 disease in the Hispanic population [40]. Many of the identified protective and deleterious HLA alleles in this study were also seen in the literature as ethnically specific alleles. The appearance of many different ethnically specific alleles from the literature in our population could be indicative of our multicultural population. It is important to note that the specific contribution of each of these alleles unique to the ethnic population depends on their census contribution to our multicultural dataset.
Another concept explored in our study was the role of HLA heterozygosity advantage in COVID-19 immune surveillance. An increase in the HLA gene homozygosity can theoretically reduce the range of the HLA alleles that can detect viral antigens and therefore leaves these homozygous individuals more susceptible to COVID-19. Our data shows a minuscule predilection of increasing homozygosity by 1.2-fold in the high-severity cohort compared to the low-severity cohort. However, this data was not statistically significant and may indicate that the specific homozygous genotype can contribute more to the increasing susceptibility rather than the sheer number of homozygous genotypes. Alternatively, it could be the composite effect from the overall allelic repository.
In this study, the interaction of various HLA alleles with their associated KIR subfamilies was explored to predict the NK cell antiviral activity status. When the alleles in the high-severity and low-severity were normalized to the background frequencies in Saskatchewan, there was an enrichment of specific protective and deleterious alleles. The high-severity cohort has a higher predilection for HLA alleles associated with the KIR subgroups; A3/11, Bw4-80I, and C1. The inhibitory KIR2DL2 often binds to its cognate HLA alleles associated with the A3/11 [41]. Additionally, the inhibitory KIR2DL2 and KIR2DL3 receptors bind to the C1 associated HLA alleles, wherein the amino acid asparagine is positioned at 80 [41]. The Bw4 subtype further splits into Bw4-80I and Bw4-80 T based on whether isoleucine or threonine is positioned at 80. Isoleucine at position 80 has a higher affinity for the inhibitory receptor to KIR3DL1 compared to threonine [41]. All the above KIR receptors and their cognate HLA shows increased engagement with the inhibitory receptors compared to the cross-reactive activating KIR receptors. This shared feature in the high-severity cohort could explain the severe clinical course due to overall inhibition of the activation of NK cell-mediated lysis of SARS-CoV-2 infected cells. The low-severity cohort has a higher predilection for HLA alleles associated with Bw4-80T, Bw6, and C2. These subgroups have shown either a weaker affinity for inactivating KIR receptors or strongly activating NK cell activity.
There were several limitations of this study. Firstly, our study has a small sample size. Based on the allelic frequency proportions, the over-represented HLA alleles in the low-severity and the high-severity groups were identified as protective and deleterious alleles, respectively. Due to our smaller sample size, many of these over-represented protective and deleterious alleles did not meet the statistical significance when the Bonferroni correction was applied. This lack of statistical significance was also seen in many publications analyzing HLA allelic repertoire in COVID-19 clinical severity in the literature. Secondly, HLA alleles have different frequencies in different ethnic populations, which could affect the power of this study with the presence of cryptic subpopulations forming heterogeneous cohorts (e.g. COVID19-hg GWAS meta-analyses) and this could lead to false positives. The next iteration of this study could mitigate the contribution of these cryptic subpopulations by documenting the ethnic background of the patients and implementing tagged SNPs to evaluate the significance of known strong linkage disequilibrium considered in HLA alleles to the corresponding ethnic backgrounds.
Additionally, only limited sets of HLA alleles were studied, mainly focusing on HLA class I. There is an inability to assess the relative contribution of the HLA alleles as protective or deleterious alleles without accounting for confounding factors such as clinical comorbidities, host factors, etc. Lastly, enriching specific HLA alleles and their interaction with the KIR molecules for NK cell immune surveillance could be further explored by incorporating a KIR typing and activation assay.
5. Conclusion
Our study identified the HLA repertoire and COVID-19 severity associations and provided insights into the pathogenesis of SARS-CoV-2. These findings of protective and deleterious HLA genes could facilitate classifying individuals for vaccination booster prioritization and provide early optimized management of patients with COVID-19 by identifying those at the greatest risk of severe morbidity and mortality.
Financial Support
This research is funded by the Respiratory Research Center/Office of the Vice Dean Research (OVDR), College of Medicine, University of Saskatchewan (2020 Rapid Response COVID-19 Research Award).
Declaration of Competing Interest
The authors declare that they have no known competing financial interests or personal relationships that could have appeared to influence the work reported in this paper.
Footnotes
Supplementary data to this article can be found online at https://doi.org/10.1016/j.humimm.2023.01.003.
Appendix A. Supplementary data
The following are the Supplementary data to this article:
References
- 1.J.H.C.R. Center, Coronavirus COVID-19 Global Cases by the Center for Systems Science and Engineering (CSSE) at Johns Hopkins University (JHU). 2020 [cited 2020 March 22]; Available from: https://coronavirus.jhu.edu/map.html.
- 2.Pareek M., et al. Ethnicity and COVID-19: an urgent public health research priority. Lancet. 2020;395(10234):1421–1422. doi: 10.1016/S0140-6736(20)30922-3. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 3.Abou Ghayda R., et al. The global case fatality rate of coronavirus disease 2019 by continents and national income: A meta-analysis. J. Med. Virol. 2022;94(6):2402–2413. doi: 10.1002/jmv.27610. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 4.Kakodkar P., Kaka N., Baig M.N. A Comprehensive Literature Review on the Clinical Presentation, and Management of the Pandemic Coronavirus Disease 2019 (COVID-19) Cureus. 2020;12(4):e7560–e. doi: 10.7759/cureus.7560. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 5.Chen S., et al. Climate and the spread of COVID-19. Sci. Rep. 2021;11(1):9042. doi: 10.1038/s41598-021-87692-z. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 6.S. Abdul-Fattah, et al., History and Recent Advances in Coronavirus Discovery, in In Silico Modeling of Drugs Against Coronaviruses: Computational Tools and Protocols, K. Roy, Editor. 2021, Springer US: New York, NY. p. 3-24.
- 7.Yap C., et al. Comprehensive literature review on COVID-19 vaccines and role of SARS-CoV-2 variants in the pandemic. Ther. Adv. Vaccines. Immunother. 2021;9 doi: 10.1177/25151355211059791. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 8.Genomewide Association Study of Severe Covid-19 with Respiratory Failure. New England J. Med., 2020. 383(16): p. 1522-1534. [DOI] [PMC free article] [PubMed]
- 9.Nguyen A., et al. Human Leukocyte Antigen Susceptibility Map for Severe Acute Respiratory Syndrome Coronavirus 2. J. Virol. 2020;94(13) doi: 10.1128/JVI.00510-20. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 10.Dendrou C.A., et al. HLA variation and disease. Nat. Rev. Immunol. 2018;18(5):325–339. doi: 10.1038/nri.2017.143. [DOI] [PubMed] [Google Scholar]
- 11.Wang W., et al. Distribution of HLA allele frequencies in 82 Chinese individuals with coronavirus disease-2019 (COVID-19) Hla. 2020;96(2):194–196. doi: 10.1111/tan.13941. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 12.Shiina T., et al. The HLA genomic loci map: expression, interaction, diversity and disease. J. Hum. Genet. 2009;54(1):15–39. doi: 10.1038/jhg.2008.5. [DOI] [PubMed] [Google Scholar]
- 13.Lin M., et al. Association of HLA class I with severe acute respiratory syndrome coronavirus infection. BMC. Med. Genet. 2003;4:9. doi: 10.1186/1471-2350-4-9. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 14.Dutta M., et al. Polymorphism of HLA class I and class II alleles in influenza A(H1N1)pdm09 virus infected population of Assam, Northeast India. J. Med. Virol. 2018;90(5):854–860. doi: 10.1002/jmv.25018. [DOI] [PubMed] [Google Scholar]
- 15.Jiang Y., et al. KIR3DS1/L1 and HLA-Bw4-80I are associated with HIV disease progression among HIV typical progressors and long-term nonprogressors. BMC. Infect. Dis. 2013;13:405. doi: 10.1186/1471-2334-13-405. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 16.Janeway C.A., Jr., Medzhitov R. Innate immune recognition. Annu. Rev. Immunol. 2002;20:197–216. doi: 10.1146/annurev.immunol.20.083001.084359. [DOI] [PubMed] [Google Scholar]
- 17.N. Kaka, et al., The Origin, Transmission, and Clinical Therapies in the Management of Coronavirus Diseases, in: In Silico Modeling of Drugs Against Coronaviruses: Computational Tools and Protocols, K. Roy, Editor. 2021, Springer US: New York, NY. p. 25-44.
- 18.Kikkert M. Innate Immune Evasion by Human Respiratory RNA Viruses. J. Innate. Immun. 2020;12(1):4–20. doi: 10.1159/000503030. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 19.The involvement of natural killer cells in the pathogenesis of severe acute respiratory syndrome. Am. J. Clin. Pathol., 2004. 121(4): p. 507-11. [DOI] [PMC free article] [PubMed]
- 20.Colonna M., et al. HLA-C is the inhibitory ligand that determines dominant resistance to lysis by NK1- and NK2-specific natural killer cells. Proc. Natl. Acad. Sci. USA. 1993;90(24):12000–12004. doi: 10.1073/pnas.90.24.12000. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 21.Cella M., et al. NK3-specific natural killer cells are selectively inhibited by Bw4-positive HLA alleles with isoleucine 80. J. Exp. Med. 1994;180(4):1235–1242. doi: 10.1084/jem.180.4.1235. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 22.Döhring C., et al. A human killer inhibitory receptor specific for HLA-A1,2. J. Immunol. 1996;156(9):3098–3101. [PubMed] [Google Scholar]
- 23.Hansasuta P., et al. Recognition of HLA-A3 and HLA-A11 by KIR3DL2 is peptide-specific. Eur. J. Immunol. 2004;34(6):1673–1679. doi: 10.1002/eji.200425089. [DOI] [PubMed] [Google Scholar]
- 24.Connelly L.M. Pilot studies. Medsurg. Nurs. 2008;17(6):411–412. [PubMed] [Google Scholar]
- 25.Canada, S. Census Profile, 2016 Census - Saskatchewan Canada. 2016 [cited 2022; Available from: https://www12.statcan.gc.ca/census-recensement/2016/dp-pd/prof/details/page.cfm?Lang=E&Geo1=PR&Code1=47&Geo2=PR&Code2=01&SearchText=Canada&SearchType=Begins&SearchPR=01&B1=Ethnic%20origin&type=0.
- 26.Shkurnikov M., et al. Association of HLA Class I Genotypes With Severity of Coronavirus Disease-19. Front. Immunol. 2021;12 doi: 10.3389/fimmu.2021.641900. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 27.Pretti M.A.M., et al. Class I HLA Allele Predicted Restricted Antigenic Coverages for Spike and Nucleocapsid Proteins Are Associated With Deaths Related to COVID-19. Front. Immunol. 2020;11 doi: 10.3389/fimmu.2020.565730. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 28.Detsika M.G., et al. Demographic, Clinical and Immunogenetic Profiles of a Greek Cohort of COVID-19 Patients. Life. (Basel) 2021;11(10) doi: 10.3390/life11101017. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 29.Ertosun M.G., et al. The relationship between COVID-19 and HLA in kidney transplant recipients, an evaluation of predictive and prognostic factors. Clin. Transplant. 2022;36(2):e14525. doi: 10.1111/ctr.14525. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 30.Khor S.S., et al. HLA-A*11:01:01:01, HLA-C*12:02:02:01-HLA-B*52:01:02:02, Age and Sex Are Associated With Severity of Japanese COVID-19 With Respiratory Failure. Front. Immunol. 2021;12 doi: 10.3389/fimmu.2021.658570. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 31.Lorente L., et al. HLA genetic polymorphisms and prognosis of patients with COVID-19. Med. Intensiva. (Engl. Ed) 2021;45(2):96–103. doi: 10.1016/j.medin.2020.08.004. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 32.Warren R.L., Birol I. HLA alleles measured from COVID-19 patient transcriptomes reveal associations with disease prognosis in a New York cohort. PeerJ. 2021;9:e12368. doi: 10.7717/peerj.12368. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 33.Wang F., et al. Initial whole-genome sequencing and analysis of the host genetic contribution to COVID-19 severity and susceptibility. CellDiscov. 2020;6(1):83. doi: 10.1038/s41421-020-00231-4. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 34.Toyoshima Y., et al. SARS-CoV-2 genomic variations associated with mortality rate of COVID-19. J. Hum. Genet. 2020;65(12):1075–1082. doi: 10.1038/s10038-020-0808-9. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 35.Hernández-Doño S., et al. Protective HLA alleles against severe COVID-19: HLA-A*68 as an ancestral protection allele in Tapachula-Chiapas, Mexico. Clin. Immunol. 2022;238 doi: 10.1016/j.clim.2022.108990. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 36.Littera R., et al. Human Leukocyte Antigen Complex and Other Immunogenetic and Clinical Factors Influence Susceptibility or Protection to SARS-CoV-2 Infection and Severity of the Disease Course. The Sardinian Experience. Front. Immunol. 2020:11. doi: 10.3389/fimmu.2020.605688. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 37.Correale P., et al. HLA-B*44 and C*01 Prevalence Correlates with Covid19 Spreading across Italy. Int. J. Mol. Sci. 2020;21(15) doi: 10.3390/ijms21155205. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 38.Naemi F.M.A., et al. Association between the HLA genotype and the severity of COVID-19 infection among South Asians. J. Med. Virol. 2021;93(7):4430–4437. doi: 10.1002/jmv.27003. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 39.Astbury S., et al. HLA-DR polymorphism in SARS-CoV-2 infection and susceptibility to symptomatic COVID-19. Immunology. 2022;166(1):68–77. doi: 10.1111/imm.13450. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 40.Schindler E., et al. HLA genetic polymorphism in patients with Coronavirus Disease 2019 in Midwestern United States. Hla. 2021;98(4):370–379. doi: 10.1111/tan.14387. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 41.Maruthamuthu S., et al. Individualized Constellation of Killer Cell Immunoglobulin-Like Receptors and Cognate HLA Class I Ligands that Controls Natural Killer Cell Antiviral Immunity Predisposes COVID-19. Front. Genet. 2022;13 doi: 10.3389/fgene.2022.845474. [DOI] [PMC free article] [PubMed] [Google Scholar]
Associated Data
This section collects any data citations, data availability statements, or supplementary materials included in this article.