HIV Viremia Is Associated With APOL1 Variants and Reduced JC-Viruria

Variants in the Apolipoprotein L1 (APOL1) gene (G1-rs60910145, rs73885319, G2-rs71785313) are common in Africans and in individuals of recent African ancestry and are associated with an increased risk of non-diabetic chronic kidney disease (CKD) and in particular of HIV associated nephropathy (HIVAN). In light of the significantly increased risk of HIVAN in carriers of two APOL1 risk alleles, a role in HIV infectivity has been postulated in the mechanism of APOL1 associated kidney disease. Herein, we aim to explore the association between HIV viremia and APOL1 genotype. In addition, we investigated interaction between BK and JC viruria, CKD and HIV viremia. A total of 199 persons living with HIV/AIDS (comprising 82 CKD cases and 117 controls) from among the participants in the ongoing Human Heredity and Health in Africa (H3Africa) Kidney Disease Research Network case control study have been recruited. The two APOL1 renal risk alleles (RRA) genotypes were associated with a higher risk of CKD (OR 12.6, 95% CI 3.89–40.8, p < 0.0001). Even a single APOL1 RRA was associated with CKD risk (OR 4.42, 95% CI 1.49–13.15, p = 0.007). The 2 APOL1 RRA genotypes were associated with an increased probability of having HIV viremia (OR 2.37 95% CI 1.0–5.63, p = 0.05). HIV viremia was associated with increased CKD risk (OR 7.45, 95% CI 1.66–33.35, P = 0.009) and with a significant reduction of JC virus urine shedding (OR 0.35, 95% CI 0.12–0.98, p = 0.046). In contrast to prior studies, JC viruria was not associated with CKD but was restricted in patients with HIV viremia, regardless of CKD status. These findings suggest a role of APOL1 variants in HIV infectivity and emphasize that JC viruria can serve as biomarker for innate immune system activation.

Variants in the Apolipoprotein L1 (APOL1) gene (G1-rs60910145, rs73885319, G2-rs71785313) are common in Africans and in individuals of recent African ancestry and are associated with an increased risk of non-diabetic chronic kidney disease (CKD) and in particular of HIV associated nephropathy (HIVAN). In light of the significantly increased risk of HIVAN in carriers of two APOL1 risk alleles, a role in HIV infectivity has been postulated in the mechanism of APOL1 associated kidney disease. Herein, we aim to explore the association between HIV viremia and APOL1 genotype. In addition, we investigated interaction between BK and JC viruria, CKD and HIV viremia. A total of 199 persons living with HIV/AIDS (comprising 82 CKD cases and 117 controls) from among the participants in the ongoing Human Heredity and Health in Africa (H3Africa) Kidney Disease Research Network case control study have been recruited. The two APOL1 renal risk alleles (RRA) genotypes were associated with a higher risk of CKD (OR 12.6, 95% CI 3.89-40.8, p < 0.0001). Even a single APOL1 RRA was associated with CKD risk (OR 4.42, 95% CI 1.49-13.15, p = 0.007). The 2 APOL1 RRA genotypes were associated with an increased probability of having HIV viremia (OR 2.37 95% CI 1.0-5.63, p = 0.05). HIV viremia was associated with increased CKD risk (OR 7.45, 95% CI 1.66-33. 35, P = 0.009) and with a significant reduction of JC virus urine shedding (OR 0. 35, 95% INTRODUCTION Two variants at the APOL1 gene, encoding Apolipoprotein L1, account for more than 70% of the increased risk for non-diabetic chronic kidney disease (CKD) in individuals of African ancestry (1)(2)(3). The ancestral non-renal risk allele has been designated G0 and the two risk alleles (renal risk alleles, RRA) have been designated as G1 (encoding S342G and I384M substitutions) and G2 (encoding the combined N388 and Y389 deletions), respectively (1,3). These variants have risen to high allele frequency due to a dominant selective advantage that restores trypanolytic activity against T. b. rhodesiense (G2) and confers protection from active illness caused by T. b. gambiense (G1) (1,(4)(5)(6). Unexpectedly, the G2 genotype is associated with active T. b. gambiense infection (5). It is an enigma why some of the highest frequencies of G2 are also found in regions of West Africa where T. b. gambiense is endemic, despite the observed G2 association with symptomatic chronic sleeping sickness. This raises the likelihood of a broader role in pathogen resistance by the RRA, extending beyond trypanosomes (4)(5)(6). Therefore, the protective role of APOL1 variants in various infectious diseases is under continued extensive investigation. HIV infection deserves special consideration in regard to modulation of HIV infectivity by APOL1 variants. In light of the significantly increased risk of HIV associated nephropathy (HIVAN) in carriers of two APOL1 RRA (2, 7), a role in HIV infectivity has been postulated. In vitro studies have demonstrated that G0 APOL1 restricts HIV replication in macrophages and differentiated monocytes (8). APOL1 was also identified as one of the HIV restriction proteins, by using genome-wide scans for human genes sharing molecular and evolutionary signatures of known restriction factors. These findings were also validated using in vitro studies demonstrating anti-HIV-activity (9). In addition, the presence of RRA but not G0 APOL1 led to persistence of HIV infection in human podocytes in synergy with IL-1beta (10). Recently, a genetic association study (ALIVE -AIDS Link to the Intravenous Experience cohort) revealed no evident association of APOL1 RRA with HIV infection acquisition, viral load or disease progression, in antiretroviral therapy (ART) naïve patients (11). That study examined HIV natural history cohorts enrolling African Americans prior to the ART era. The large number of treatment-naïve seroconverters made it a choice cohort for unbiased exploration of HIV-related outcomes. Herein, we explored the association of HIV viremia and CKD with APOL1 variants in participants recruited from HIV clinics in Africa. All the participants were attending HIV clinics were likely to have received ART, but we cannot be certain of adherence. We also examined the association between JC and BK viruria in all participants, in light of previous findings demonstrating reduced JC viruria in CKD patients and the expected increased BK urine shedding in immunocompromised HIV infected patients (12)(13)(14)(15)(16)(17).

Ethics Statement
Ethical approval for the study was obtained from all the Institutional Review Boards of each participating institution as well as the lead clinical center at the University of Ghana (Ghana Health Service Ethics Review Committee) who approved the study protocols. Written informed consent was obtained from all study participants.

Study Participants
Age and sex matched Africans from Ghana and Nigeria with HIV infection were recruited as part of the H3Africa Kidney Disease Research Network, from 2013 to 2018.

HIV Viral Load
Plasma HIVRNA quantitative analysis was assayed as previously reported by Tung (19). The test quantitates HIV RNA over the range of 20-10,000,000 copies/mL.

Nucleic Acid Extraction and Quantitative PCR for JCV and BKV
Frozen urine (200 mL) samples from the participants were thawed, and total nucleic acid was extracted. The extraction procedure used the Magna Pure LC, version 2.0, apparatus and Magna Pure LC total Nucleic Acid Isolation Kit Reagents

TaqMan Real Time PCR
The primers and probes were derived from nucleotide sequences of the N terminus of the viral T antigen of JCV and BKV: The PCR reaction was performed in a total volume of 25 ul containing Absolute Blue quantitative PCR mix (Thermo Scientific, UK) in the presence of 10 mL target DNA, 300 nM of each primer and 200 nM of the probe. PCR was performed using the rotor gene 6000/Q instrument (Corbett Research/Qiagen, Hilden, Germany) under the following conditions: 15 min at 95 • C and 45 cycles of 15 s at 95 • C and 60 s at 60 • C. For quantitative results analysis, an average standard curve was constructed using quantified JCV and BKV DNA (Advanced Biotechnology Industry, MD). The results are reported as the number of JCV/BKV genome copies per 1 mL of urine. The lowest detection level was 100 genomic copies per 1 mL.

Statistical Analysis
Univariate analysis (Chi-squared test, t-test) was used to compare case and control groups. Multivariate logistic regression with stepwise, forward and backward model selection was used to predict the risks for HIV viral load, JC and BK viruria and CKD. Each covariate was corrected for all other variables. The covariates that were included for CKD were: age, gender, APOL1 RRA, BK viruria, JC viruria and HIV viremia. The interaction between APOL1 RRA and HIV viremia was added to the CKD risk association model. For CKD risk, we adjusted sequentially to four models: model 1: for APOL1 variants, age and sex, model 2: further adjusted for BK and JC viruria, model 3: as for model 2 plus HIV viremia and model 4: as for model 3 plus the interaction of APOL1 RRA and HIV viremia. For models predicting HIV viremia, JC and BK viruria, the following covariates were included in the logistic regression: age, gender, CKD, urine albumin/creatinine, APOL1 RRA and the other two viruses, respectively.

Association With CKD
The demographic and clinical characteristics for participants in this cohort, stratified by case-control status (n = 82, n = 117, respectively) are presented in Tables 1, 2. The CKD patient group and the control group did not differ significantly by age (pvalue = 0.06) or sex (p-value = 0.154) ( Table 1). We compared the distribution of G1 and G2 RRAs in HIV seroprevalent subjects with and without CKD. The following covariates were included: age, gender, APOL1 RRA, HIV viremia, BK viruria, JC viruria and the interaction 0f HIV * APOL1 RRA. Multivariate logistic regression with a stepwise, forward and backward model selection showed that even a single APOL1 RRA was associated with CKD, OR 4.429 (95% CI 1.49-13.15, p = 0.007). As expected, 2 APOL1 RRA were significantly associated with CKD, OR 12.6 (95% CI 3.89-40.80, p < 0.0001). HIV viremia was also associated with CKD, OR 7.45 (95% CI 1.66-33.35, p = 0.009). The interaction between APOL1 RRA and HIV was significant (p = 0.011), with OR 0.081 and 0.073 with 1 and 2 APOL1 RRA, respectively. This reverse association is perplexing and might be explained by Simpson's paradox, and does not seem to confer biological significance (21-23) ( Table 3). Interestingly, adjusted models for age, sex, BK, JC and HIV viremia (models 2 and 3, respectively) did not change the association of 2 APOL RRA with CKD while adjustment for APOL1 RRA * HIV

Prevalence of HIV Viremia
Participants with CKD (cases) and those without CKD (controls) were evaluated for HIV viral load, JCV and BKV viruria. All HIV-positive patients recruited from HIV clinics were likely to have been initiated on antiretroviral therapy (ART) before recruitment into the study. To determine whether APOL1 G1 or G2 RRAs affect host susceptibility to HIV viremia, multivariate logistic regression with a stepwise, forward and backward model selection was used to predict HIV viral load. Variables included in the model were age, gender, JC viruria, BK viruria, APOL1 RRA (2 vs. 1 vs. 0), CKD, e GFR and urine albumin to creatinine ratio. The two APOL1 RRA genotypes were associated with an increased probability of having HIV viremia OR 2.37 (95% CI 1.0-5.63, p = 0.05) ( Table 5). Moreover, a higher viral load (more than 40,000

Association of BK and JC Viruria
HIV viremia was associated with a significant effect on the probability of JC virus urine shedding, reducing the probability of JC viruria compared to individuals without HIV viremia OR 0.35 (95% CI 0.12-0.98, P = 0.046; Table 5). JC viruria did not differ between patients with CKD and controls (p-value = 0.67).
In addition, subjects with APOL1 RRA who had JC viruria did not display a lower prevalence of CKD compared to individuals who did not have JC viruria (p-value = 0.692). BK viruria was not associated with CKD and HIV viremia (p-value 0.65 and 0.503, respectively). In light of previous publications about the inhibitory interaction between BK and JC viruria (25), we assessed this possibility. As we previously reported (14), there was no correlation between BK and JC viruria (Pearson correlation coefficient: 0.43, p=0.23).

DISCUSSION
In this genetic epidemiological study of HIV infected individuals living in Ghana and Nigeria (26), we observed an association between 2 APOL1 RRA and HIV viremia and CKD. Notably, there was a trend for increased CKD risk with one APOL1 RRA that was significantly increased after adjustment for APOL1 RRA * HIV viremia interaction. This increased risk is in keeping with the reported marginal association of HIVAN with 1 risk allele, in individuals of black African ancestry from South Africa (7). Similarly, previous studies have also reported a marginal effect of one APOL1 RRA (1, 2, 27, 28), suggesting that the possibility that an environment (HIV)-gene interaction may enhance the risk of CKD, in carriers of one APOL1 RRA.
Therefore, HIV is a robust risk factor for kidney disease, that may potentiate the genetic risk, even in individuals with only one risk allele, into manifest kidney disease. Innate immune responses to viruses can drive APOL1 kidney disease in patients with APOL1 high-risk genotypes as was demonstrated in case series of collapsing glomerulopathy caused by therapeutic administration of interferon products and in patients infected with COVID-19 (29,30). Interestingly, one APOL1 RRA in kidney allograft was associated with collapsing glomerulopathy in a patient infected with COVID-19 (30), suggesting that one APOL1 risk allele may confer a pathogenic role in diseases that share enhanced activation of the innate immune system. For completeness, we point out that in a combined interaction analysis, the interaction of HIV viremia and 2 APOL1 RRA was associated with decreased CKD risk. We consider this not to reflect a biological effect but rather a statistical effect that can occur in interaction analyses of separately contributory associations, termed Simpson's paradox, but still make note in case future biological or clinical research is warranted to explain this result (21)(22)(23).
In addition, we demonstrate a complex interaction of HIV viremia with polyoma virus urine shedding. The increased HIV viral load was associated with reduced JC viruria, while BK viruria was not associated with CKD or HIV viremia. We did not demonstrate the protective association of JC viruria with preserved kidney function, as previously reported in patients who did not have HIV infection and had non-diabetic as well as diabetic kidney disease (12)(13)(14). We reasoned that in the presence of HIV viremia, the secretion of inflammatory cytokines, such as Interferon itself, might lead to restriction of other viruses, including suppression of JC viruria and thereby mask any such association.
The role of APOL1 as a member of the innate immune system is only partially understood. APOL1 has a wellcharacterized protective role that restores trypanolysis activity against Trypanosoma brucei rhodesiense (G2) and confers protection from active illness caused by Trypanosoma brucei gambiense (G1) (1,(4)(5)(6). It is not clear if APOL1 has broader protective roles as an innate immunity protein.
Most interestingly, a recent analysis of the African-American participants of the Reasons for Geographic and Racial Differences in Stroke (REGARDS) study, demonstrated that APOL1 RRA were associated with sepsis risk under dominant [Hazard Ratio (HR) 1.55, 95% confidence interval (CI), 1.13-2.11] and additive (HR per variant allele copy 1.25, 95% CI, 1.02-1.53) genetic models adjusted for covariates and ancestry, suggesting a complex interaction between APOL1 and general risk for severe infection (31). In contrast to the increased sepsis risk, some in vitro studies have demonstrated that APOL1 restricts HIV infectivity and protects human keratinocytes from the lethal effect of the Staphylococcus aureus α toxin (8,9,32). Specifically, impaired HIV restriction in podocytes by the APOL1 encoded by RRA, has been suggested (10). However, an epidemiological study did not detect an association between HIV viral load and APOL1 variants. The ALIVE cohort revealed no evidence of association of APOL1 RRA with HIV infection acquisition, viral load or disease progression, in ART naïve patients (11).
In contrast, the current study of ART treated patients points to  We demonstrate complex interactions between HIV viremia, CKD and polyoma viruria. Previous studies reported reduced JC viruria in CKD patients (12)(13)(14). Herein, reduced JC viruria was driven by HIV viremia and not CKD. We postulated that the reduced JC viruria in kidney disease and the presence of HIV viremia share similar signaling pathways that enhance the activation of the innate immune system. Enhanced inflammatory signaling (e.g., via the type I interferon signaling cascade that abrogates JC replication) restricts JCV growth and leads to decreased shedding of JCV (14,36). Indeed, in vitro studies have demonstrated divergent interaction between IFN-β and the two polyoma viruses. IFN-β controls JCV replication but fails to restrict BKV (36), thereby, explaining the specific restriction of JC virus shedding in urine and not BK in patients with HIV viremia. We hypothesize that absence of JC viruria is an epiphenomenon, serving as a biomarker that reflects innate immune system activation. In contrast to JC viruria, only BK viruria is enhanced in immunocompromised patients with HIV infection and correlates inversely with the CD4+ cell count (15)(16)(17). In the current cohort, BK viruria was not associated with CKD and HIV viremia. Most BK virus associated nephropathy occurs in renal allograft patients after kidney transplantation. However, some case reports have described BK virus-associated nephropathy in the native kidney, particularly in patients with human immunodeficiency virus infection (37,38). Further studies are needed to explore the role of BK viruria in CKD patients infected with HIV.
The present study has limitations: This cohort has a relatively small sample size. HIV viral load was measured only at enrollment, so we could not determine if the enhanced viral load was transient or persistent. Similarly, we do not have information about CD4 cells count, ART compliance and longitudinal clinical outcomes of kidney function were missing. Therefore, we could not ascertain if the increased HIV viremia will translate to faster CKD progression, as was previously demonstrated (39). In addition, JC and BK viruria may fluctuate and was measured at a single time point over time in the current study.
In conclusion, this population genetic study demonstrates an increased risk for HIV viremia and CKD in patients with one and two APOL1 renal risk alleles. The association of one APOL1 risk allele with CKD suggests that if G0 has a protective role, it does not have complete suppressive penetrance; alternatively, this finding is also consistent with a gain of injury mechanism mediating kidney injury in individuals with a robust environmental insult, such as HIV infection. Future studies are warranted to explore the complex interaction of APOL1 RRA, CKD risk, HIV viremia and urine shedding of polyoma viruses. Confirmation in future studies of an association between two APOL1 risk alleles with HIV viremia and CKD would have significant therapeutic implications regarding APOL1 inhibition in patients with two risk alleles, who do not reside in Trypanosoma endemic areas.

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 author/s.

ETHICS STATEMENT
The studies involving human participants were reviewed and approved by University of Ghana (Ghana Health Service Ethics Review Committee). The patients/participants provided their written informed consent to participate in this study.