PARV4 prevalence, phylogeny, immunology and coinfection with HIV, HBV and HCV in a multicentre African cohort

Abstract

Background: The seroprevalence of human parvovirus-4 (PARV4) varies considerably by region. In sub-Saharan Africa, seroprevalence is high in the general population, but little is known about the transmission routes or the prevalence of coinfection with blood-borne viruses, HBV, HCV and HIV. Methods: To further explore the characteristics of PARV4 in this setting, with a particular focus on the prevalence and significance of coinfection, we screened a cohort of 695 individuals recruited from Durban and Kimberley (South Africa) and Gaborone (Botswana) for PARV4 IgG and DNA, as well as documenting HIV, HBV and HCV status. Results: Within these cohorts, 69% of subjects were HIV-positive. We identified no cases of HCV by PCR, but 7.4% were positive for HBsAg. PARV4 IgG was positive in 42%; seroprevalence was higher in adults (69%) compared to children (21%) (p<0.0001) and in HIV-positive (52%) compared to HIV-negative individuals (24%) (p<0.0001), but there was no association with HBsAg status. We developed an on-line tool to allow visualization of coinfection data (https://purl.oclc.org/coinfection-viz). We identified five subjects who were PCR-positive for PARV4 genotype-3. Ex vivo CD8+ T cell responses spanned the entire PARV4 proteome and we propose a novel HLA-B*57:03-restricted epitope within the NS protein. Conclusions: This characterisation of PARV4 infection provides enhanced insights into the epidemiology of infection and co-infection in African cohorts, and provides the foundations for planning further focused studies to elucidate transmission pathways, immune responses, and the clinical significance of this organism.

Keywords: Africa; HBV; HCV; HIV; PARV4; co-infection; epidemiology; parvovirus.

Figure 1.

Relationship between PARV4 IgG status and age ( A– C), HIV status ( D– G), and HBV status ( H– K). Boxes show median, 25 ^th and 75 ^th centiles; whiskers show 5 ^th–95 ^th centiles. P-values by Fisher’s exact test (bar charts), linear regression (scatter plot), and Mann Whitney U test (box and whisker plots). Denominator stated on each panel varies based on availability of relevant data.

Figure 2.. Phylogeny of PARV4 sequences detected in serum from five individuals from South Africa.

Phylogeny inferred using maximum likelihood trees from partial VP1 ( A), complete NS ( B) and complete VP1 ( C) nucleotide sequences (equivalent to nucleotides 3067-3310, 283-2271 and 2378-5035, respectively, of the PARV4 reference sequence NC007018). In each case, the new sequences derived from South Africa are highlighted (lavender bars). Bootstrap support of branches (500 replications) is indicated.

Figure 3.. IFN-gamma CD8+ T cell responses to PARV4 peptides determined by ELISpot assays.

Data in ( A) and ( B) are derived by screening 14 subjects positive for PARV4 IgG recruited from the Kimberley (n=7, children) and Thames Valley (n=7, adults) cohorts. None of these subjects was PCR positive for PARV4 from serum. Raw data showing the responses made by each individual subject can be viewed in Supplementary data 4. ( A) Proportion of 14 screened subjects who made each individual response. ( B) Mean magnitude (box) and range of response (whiskers); the dashed horizontal line allows visualization of peptides for which the mean response is >1000 SFCs/10 ⁶ PBMC. Responses to NS peptides are shown in grey, to VP peptides in black, and to ARF in hatched bars. ( C) Prediction of a novel HLA-B*5703-restricted CD8+ T cell epitope in PARV4 NS protein. Cryopreserved PBMCs from PARV4 IgG-positive adult subject N087 (HIV-positive adult recruited via the Thames Valley cohort, HLA class I genotype HLA-A*0301/-A*3001/-B*5703/-B*5801/-C*0602/-C*1801) were screened by IFN-gamma ELISpot for responses to peptide truncations from PARV4 NS protein (sequences within OLPs 9.6 and 9.7) at different concentrations. Plots and error bars show mean and SEM of assays performed in triplicate. On the basis of the HLA-B*5703 binding motif and the greatest magnitude responses, the putative optimal epitope is HLA-B*5703-QF9 (QTRITMFQF).