Highly replicating hepatitis C virus variants emerge in immunosuppressed patients causing severe disease

Abstract

Hepatitis C virus (HCV) exists as a heterogenous quasispecies, but the phenotypic consequences of viral variability are widely unexplored. Here we identify a replication enhancing domain (ReED) in non-structural protein 5A conferring high replication fitness to clinical isolates. Accumulation of mutations in the ReED mediates high genome replication capacity. In a cohort of liver transplant patients, high replicator variants are exclusively found in individuals with severe disease outcome, suggesting that high viral replication fitness is associated with increased viral pathogenesis. Analysis of large sequence cohorts reveals that overall only 10% of viral genomes show genetic signatures of high replicators, which are enriched in recipients of liver transplantations, patients developing hepatocellular carcinoma and in HIV coinfected individuals. Overall, our data suggests that low replication fitness is a hallmark of HCV, contributing to establishment of persistence, whereas high replicators appear to have an advantage under conditions of immune suppression, thereby enforcing pathogenesis.

Fig. 1. HCV evolution in the GLT1 patient.

A Schematic illustrating time points of collection and viral load of longitudinal serum samples from the GLT1 patient. Note that the patient passed away 6 weeks post LTX2. B Schematic illustrating the workflow to retrieve bulk sequence information about HCV from patient sera. We extracted viral RNA from the serum, converted it into cDNA, amplified the HCV genome via PCR and performed next generation sequencing (NGS). For technical reasons, the HCV genome was amplified in two fragments spanning the entire coding sequence of the viral genome. C Amino acid alignments of the consensus sequences at the indicated time points with the post LTX2 ( = GLT1) sequence, differences are indicated by black dashes. D Number of DNA variants (QD > 1) in the HCV population of each time point. E Schematic of generation of sequencing data on the level of individual viral clones (left panel) and resulting phylogenetic tree for the second PCR fragment encompassing most of the viral replicase (right panel). F SGRs of the indicated constructs were electroporated into Huh7-Lunet SEC14L2 cells (upper panel), luciferase activity in cell lysates (RLU) was quantified as a correlate of RNA replication efficiency at the given time points and normalised to 4 h to account for differences in transfection efficiency. Con1 ∆GDD served as a replication deficient negative control. Data are from three independent biological replicates measured in technical duplicates. Each dot depicts the result of one replicate and the bar indicates the mean of all replicates. Schematics in A, B, E, F were created in BioRender. Lohmann, V. (2025) https://BioRender.com/upia2pp.

Fig. 2. Viral determinants of high replication fitness.

A, B, D SGRs of the indicated constructs were electroporated into Huh7-Lunet SEC14L2 cells, luciferase activity in cell lysates (RLU) was quantified as a correlate of RNA replication efficiency at the given time points and normalised to 4 h to account for differences in transfection efficiency. Con1 ∆GDD served as a replication deficient negative control. Data are from three independent biological replicates measured in technical duplicates. Each dot depicts the result of one replicate, and the bar indicates the mean of all replicates. C Amino acid alignment of LCS1D2 between GLT1 and the respective subspecies pre LTX1, dots indicate an amino acid being identical to GLT1.

Fig. 3. Genetic and phenotypic analysis of the FCH cohort.

A Amino acid deviations from the gt1b consensus sequence were counted for each patient sequence post LTX, an insertion was counted as one mutation, irrespective of its length. The analysed regions are depicted in a schematic (upper panel). Each dot depicts the result of one patient, and the bar indicates the mean of all patients. 14 FCH patients were compared to 9 non-FCH patients. B Con1 based chimeric SGRs harbouring the ReED of a patient (upper panel) were electroporated into Huh7-Lunet SEC14L2 cells, luciferase activity in cell lysates (RLU) was quantified as a correlate of RNA replication efficiency at the given time point and normalised to 4 h to account for differences in transfection efficiency. Data are from three independent biological replicates measured in technical duplicates. Each dot represents the mean RLU of one patient and the bar indicates the mean of all non-FCH/FCH patients. 14 FCH patients were compared to 9 non-FCH patients. A, B Statistical significance was determined with a two-sided Mann–Whitney-U-test. ns = not significant, *** = p < 0.001, **** = p < 0.0001. Exact p-values: p = 0.00004 (A, ReED), p = 0.00001 (A, ISDR), p = 0.116 (A, C-term) and p = 0.0005 (B). C–D SGRs of the indicated constructs were electroporated into Huh7-Lunet SEC14L2 cells, luciferase activity in cell lysates (RLU) was quantified as a correlate of RNA replication efficiency at the given time points and normalised to 4 h to account for differences in transfection efficiency. Con1 ∆GDD served as a replication deficient negative control. Data are from three independent biological replicates measured in technical duplicates. Each dot depicts the result of one replicate and the bar indicates the mean of all replicates. D Replication 96 h after electroporation is depicted, treatment with the indicated concentration of Pibrentasvir was performed 24 h after electroporation. E Sections of human livers from post-transplant patients either developing FCH or not were stained via IHC with an HCV NS5A targeting antibody. Representative image from one non-FCH (left panel, patient non-FCH12) and one FCH patient (middle panel, patient FCH14), scale bar represents 100 µm. Quantification of HCV positive cells per 500 × 500 µm region of interest (ROI) (right panel). Six regions per patient were analysed. Each dot depicts the mean result of one patient and the bar indicates the mean of all non-FCH/FCH patients. 6 FCH patients were compared to 10 non-FCH patients. Statistical significance was determined with a two-sided Student’s t-test. ** = p < 0.01. Exact p-value: p = 0.0014.

Fig. 4. Detailed sequence determinants of elevated replication fitness.

A Con1 based SGRs harbouring either patient derived ReEDs or chimeric ReEDs combining a patient derived ISDR with the gt1b consensus C-term. B, C Chimeric Con1 SGRs containing the gt1b consensus ReED and the indicated point mutations in the ISDR. D Chimeric Con1 SGRs containing the indicated patient derived ReED or a variant of that ReED with residue 2209 mutated back to the WT proline. E Chimeric Con1 SGRs containing the gt1b consensus ReED and the indicated insertion in the ISDR. The ReEDs from which the respective insertions were derived served as reference. A–E SGRs of the indicated constructs were electroporated into Huh7-Lunet SEC14L2 cells, luciferase activity in cell lysates (RLU) was quantified as a correlate of RNA replication efficiency at the given time points and normalised to 4 h to account for differences in transfection efficiency. Con1 ∆GDD served as a replication deficient negative control. Data are from three independent biological replicates measured in technical duplicates. Each dot depicts the result of one replicate, and the bar indicates the mean of all replicates.

Fig. 5. The ReED can regulate replication fitness in all major genotypes.

H77 (A) or S52 (C) SGRs harbouring patient derived ReEDs either from FCH patients^, or from LTX patients from the HCV Research UK cohort. B, D Chimeric H77 SGRs containing the gt1a consensus ReED and the indicated point mutations in the ISDR (B) or S52 SGRs containing the gt3a consensus ReED and the indicated point mutations in the ISDR (D) were analysed. A–D SGRs of the indicated constructs were electroporated into Huh7-Lunet SEC14L2 cells, luciferase activity in cell lysates (RLU) was quantified as a correlate of RNA replication efficiency at the given time points and normalised to 4 h to account for differences in transfection efficiency and subsequently normalised to the values for the gt specific consensus of the respective replicate. H77 ∆GDD (A) or S52 ∆GDD (C) served as a replication deficient negative control. Data are from three independent biological replicates measured in technical duplicates. Each dot depicts the result of one replicate, and the bar indicates the mean of all replicates. Statistical significance was determined with a two-sided Student’s t-test. ns = not significant, * = p < 0.05, ** = p < 0.01. Exact p-values: p = 0.03 (A, FCH1a_2), p = 0.01 (A, 1a_LTX_3), p = 0.001 (C, 3a_LTX7), p = 0.02 (C, 3a_LTX11).

Fig. 6. Sequence signatures of high RF in various clinical contexts.

For each patient’s ISDR sequence, the number of amino acid differences compared to the gt specific consensus sequence was determined. If an ISDR had 3 or more mutations, the patient was considered to be infected with a potential high replicator. A–C, E–H Data based on the HCV Research UK cohort. A Fraction of potential high replicators stratified by gt. B–I Data for gts 1a, 1b and 3a was pooled, for all analyses the results for the individual gt had the same trend as the combined analysis. B Fraction of potential high replicators in patients who ever received an LTX depending on whether samples for HCV sequencing were collected pre or post LTX. C In the patients depicted in panel B, number of mutations in the replicase (excluding the ReED) compared to the gt specific consensus. B, C Comparison of 94 pre LTX and 86 post LTX patients. D–I Post LTX patients were excluded from the analysis to prevent confounding the results. D Sequences from the HITS-p cohort combined with 28 patients participating in the Montreal Hepatitis C cohort (HEPCO) sequenced for this study, informing about the fraction of potential high replicators in acute patients who either developed a chronic HCV infection or cleared the virus without antiviral treatment. Comparison of 71 patients developing a chronic infection and 31 clearing the infection. E Relative abundance of potential high replicators in the context of very high serum titers ( > 10,000,000 IU/ml). Comparison of 1407 low titer and 112 high titer patients. F Correlation of the presence of potential high replicators and HCV serum titers. 193 potential high replicators were compared to 1295 potential low replicators. The most recent titer measurement was used and only patients with an available titer measurement within a year of sequencing were included. G Fraction of high replicators in patients with an HCV-HIV coinfection. Comparison of 1980 patients without coinfection and 95 patients with coinfection. H Fraction of high replicators in patients who ever had an HCC diagnosis depending on whether samples for HCV sequencing were collected pre or post HCC diagnosis. Comparison of 156 patients sequenced before HCC diagnosis and 46 patients sequenced after HCC diagnosis. I Combined data from patients of the HCV Research UK, STOPHCV and BOSON cohorts where host genomic sequence information was available, showing the fraction of potential high replicators depending on the host’s IFNL4 rs12979860 genotype. Comparison of 944 non-CC and 537 CC patients. C,F The centre line signifies the median, the box the 25^th and 75^th percentile, and the whiskers the 1.5 interquartile range. Statistical significance was determined using a two-sided Student’s t-test (C,F) or a Fisher’s exact test (B, D, E, G–I). ns = not significant, * = p < 0.05, ** = p < 0.01, **** = p < 0.0001. Exact p-values: p = 1.478e-05 (B), p = 0.005 (G), p = 0.04 (H), p = 0.01 (I).