Hepatitis C

Clearance of hepatitis C virus is associated with early and potent but narrowly-directed, Envelope-specific antibodies

Generation of infectious autologous T/F viruses from subjects with varied disease outcomes

Longitudinally collected samples from 14 newly viremic, but still seronegative, subjects were selected for this study. The estimated days post infection (DPI) of the initial infection time point ranged from 4 to 45 days (median 30). Of the 14 subjects, seven naturally cleared the primary infection (termed here clearers) and seven developed chronic infection (termed here chronic progressors) (Table 1). The median time to natural clearance was 178 DPI (range 124 487).

Table 1 Subject characteristics and time point analysis.

The T/F viruses were estimated from the distribution of variants at the earliest sampling time point as previously described13,34,35,36,37,38. To determine the number of T/F viruses that established infection, Poison Fitter and phylogenetic analyses were performed on the haplotypes generated across the Core-E1E2 region to determine if there was 1 or >1 T/F virus (Supplementary Fig. 1). Twelve of the 14 subjects had a single T/F virus, and the E1E2 was cloned for all these variants (Table 1). Subject 023_Ch (the_Ch suffix refers to the chronic infection outcome for this subject), as previously published, had two T/F viruses identified13, and subject 256_Ch, had an unresolved number of variants, estimated to be between 1 and 10 (Supplementary Fig. 1). Both T/F E1E2 viruses were cloned for subject 023_Ch, and subject 256_Ch had three E1E2 variants selected for cloning, with the most frequent haplotype selected from each of the three main phylogenetic clusters. In total, 16 T/F variants were considered for HCVpp production.

The 16 T/F HCVpp generated from the 14 subjects were tested for infectivity. A total of 9 out of 16 (56%) T/F HCVpp were infectious in vitro, defined as greater than five-fold over background (Supplementary Fig. 2). Those that were non-infectious, included three T/F HCVpp from seven clearer T/F viruses (43%; non-infectious clones from 306_Cl, 277_Cl and 4087_Cl) and four T/F HCVpp from nine chronic progressor T/F viruses (44%; non-infectious clones from 240_Ch T/F, 023_Ch T/F2, HOK T/F and 256_Ch V3). Correctly folded E1E2 was confirmed for all T/F HCVpp by binding with a panel of 12 well-characterized mAbs, with the exception of 277_Cl and HOK_Ch (Supplementary Table 1). For HOK_Ch, an infectious variant from a sample collected 30 DPI was used for HCVpp, which only differed from the T/F at residue R446K (Supplementary Fig. 8).

Clearers have an earlier and potent nAb response

In order to determine if the timing of nAb development against the autologous T/F virus(es) was associated with disease outcome, nAb activity in plasma was tested longitudinally against the autologous T/F HCVpp. For the five subjects with non-infectious HCVpp (277_Cl, 240_Ch, 306_Cl, 4087_Cl and 4059_Ch), representative HCVpps were chosen based on sequence similarity (Supplementary Table 2). All samples were also tested for neutralization against control pseudo-particle VSV-G to confirm specificity.

All 14 subjects demonstrated HCV-specific neutralizing activity, as indicated by a 50% neutralization capacity at a 1/40 dilution for at least one-time point during infection (Fig. 1 and Supplementary Figs 3 and 4). No nAb activity against VSV-G was detected in any subject or sample (Fig. 1).

Figure 1

HCV and VSV-G neutralizing activity and HCV RNA levels (IU/ml) were examined longitudinally for clearers and chronic progressors. Panels(A,B) show representative subjects who cleared the infection. Panels (C,D) show representative subjects who developed chronic HCV infection. The shaded area represents the longitudinal HCV RNA levels (IU/ml). The maroon line represents neutralising antibody (nAb) ID50 titer with squares representing time points tested on autologous virus and circles representing time points tested on heterologous virus. The red line represents neutralization of control pseudo-particle VSV-G. Neutralization results were generated from quadruplicates using two-fold dilutions from 1/40 to 1/2560.

Analysis of the longitudinal data indicated that HCV-specific nAbs first emerged (defined as an ID50 < 1/40) significantly earlier in clearer subjects at a median of 71 DPI (range 44 95) compared to a median of 425 DPI in chronic progressors (range 74 945, p = 0.0012, Fig. 2A). The timing of the peak in magnitude of the nAb response was also earlier in clearer subjects at a median of 89 DPI (range 56 144) when compared to chronic progressors, who had a median of 538 DPI (range 197 945, p = 0.0006, Fig. 2B). In order to examine if the earlier nAb responses were more potent in the clearers, the data was stratified into time windows ( < 50, 51 65, 66 80, 81 95, 96 110, 111 150, 151 300, 301 450, 451 600). Each subject was only represented once per time window with the largest value selected per window. Loess regression analysis of the inhibitory-dose (ID)50 data partitioned into time windows illustrated rapid emergence of highly potent nAbs in clearers (Fig. 2C). A repeated measures ANOVA was applied for groupwise, and then ‘protected’ timepoint comparisons until 300 DPI. The logID50 values changed significantly over time F(2.39 14.29) = 6.88, p = 0.006) and differed by outcome group with significantly higher ID50 values in clearers than chronic progressors between 81 and 110 DPI (81 95 window p = 0.024, 96 110 window p = 0.002) (Fig. 2D).

Figure 2

Timing and potency of nAb responses in clearers and chronic progressors. (A) Kaplan-Meier survival analysis was performed to compare the timing of the first nAb responses, defined here as a 50% reduction of HCVpp infectivity at the highest serum concentration (i.e. lowest dilution, 1/40) between clearers (blue) and chronic progressors (red). Clearers had significantly earlier nAb responses with a median of 71 days post-infection (DPI) (range 44 95) compared to a median of 425 DPI in chronic progressors (range 74 945) (B). The timepoint where the peak ID50 responses recorded was significantly earlier in clearers (C). The logID50 values were stratified into time windows, with each subject only represented once per time window, and curves were fitted with loess regression (grey shading represents 90% CI) for clearers (blue) and chronic (red) subjects. (D) A repeated measures ANOVA was applied to the nAb responses groupwise, and then ‘protected’ timepoint comparisons until 300 DPI. The logID50 values changed significantly over time F(2.39 14.29) = 6.88, p = 0.006) and differed by outcome group with significantly higher ID50 values in clearers than chronic progressors between 81 110 DPI (*p = 0.05, **p = 0.005).

Early nAb responses are narrowly directed

In order to confirm that the nAb activity was truly delayed in the chronic progressors and not influenced by neutralization sensitivity of the T/F, plasma samples from each subject collected closest in time to 71 DPI (the median time to nAb emergence in clearers) was examined for nAb activity against a panel of 7 HCVpp bearing E1E2 representing seven subtypes (1a, 1b, 2a, 2b, 3a, 4a and 6a) (Fig. 3, and Supplementary Fig. S5). In the samples from the 7 chronic progressors only one had weak neutralization of one of the HCVpp at 51%. For comparison, plasma from four of the 7 clearers demonstrated neutralization of one or more HCVpp from the panel, but the nAb activity was still narrow, with a maximum of only two HCVpp neutralized per subject which was not significantly different to breadth of the delayed onset responses in the chronic progressors (p = 0.1189).

Figure 3

Neutralization breadth of clearers and chronic progressors. nAb breadth was calculated longitudinally for both clearers and chronic progressors. All samples were tested against HCVpp subtypes: 1a, 1b, 2a, 2b, 3a, 4a and 6a. The number of HCVpp neutralized by each sample was calculated and compared. Comparisons using a Wilcoxon rank test were made between clearers and chronic progressor samples collected at timepoint matched samples closest to 71 DPI (median time to nAb emergence in clearers), and also at the peak nAb response.

The breadth of neutralization at the time of the peak ID50 response detected longitudinally was also compared between the infection outcome groups, with a non-significant trend towards greater breadth in chronic progressors (p = 0.3998). It should be noted that sample selection for this analysis was not time matched as the nAb responses in chronic progressors emerged much later (clearers range: 58 144 DPI, chronics range: 197 945 DPI) than clearers, and likely accounts for the previously described increase in chronically infected subjects21. It was interesting to note that of the 8 subjects with neutralization of at least 1 of the HCVpp panel, 7 subjects only neutralized viruses that were of the same genotype as their infection. In combination, these data indicate that early nAb responses are narrowly directed.

As there are instances in HIV infection where the delayed nAbs are specific for variants which emerge longitudinally, rather than the T/F variant, this was tested for two subjects, THD_Ch and THG_Ch, who shared the same T/F (believed to have been infected from the same source) and had three longitudinal variants with good infectivity in the HCVpp system. The variants, THD_72DPI and THG_58DPI and THG_184DPI (named for the timepoint of isolation) were all the dominant variant at that timepoint. The HCVpp for THD were tested against plasma collected across five time points (85, 109, 198, 394, 583 DPI) and the HCVpp for THG were tested across two time points (184 and 380 DPI) with plasma at a 1/40 dilution (Supplementary Fig. S6). Interestingly, responses against the THG_58DPI variant revealed a neutralization profile identical to those against the T/F variant in plasma from both of the timepoints tested. The other two variants, THD_72DPI and THG_184DPI, were more sensitive to neutralization than the T/F variant with plasma from all timepoints tested, but the 1/40 dilution was only neutralizing at greater than 50% (the defined cutoff for nAb activity) at the same time point as the T/F variant (394 and 380 DPI for THD and THG, respectively). This finding, along with the delayed E2 seroconversion (Table 1), supports the observation that nAb responses were delayed in these two chronic progressors, and raises the possibility that the nAb response may have been induced by the later variant which then became cross-reactive against the T/F variant as the response broadened, rather than being induced by the T/F. It is also possible T/F-induced responses developed earlier but were below the limit of detection of these assay systems.

nAb activity is delayed in chronic progressors until multiple known epitopes are targeted

Longitudinal epitope mapping was performed to determine which epitopes on the HCV Envelope proteins were being targeted, and therefore could be associated with nAb activity. Plasma samples that had a minimal signal-to-noise ratio (OD: cut-off ratio >1.8) in an E1E2 IgG ELISA were further characterized in a competition-based ELISA using the non-autologous genotype 1a H77 E1E2. A total of 43 plasma samples from 13 of the 14 subjects were characterized (the remaining subject, 4032_Cl, did not show binding to E1E2). MAbs targeting nine distinct antigenic regions (AR1, AR2, AR3/Domain B, AR4, AR5, Domain E, Domain D, Domain A and Domain C) were selected to determine which epitopes were being targeted longitudinally. A competition matrix was performed to ensure the epitope targeting by these mAbs could be distinguished from each other (Supplementary Fig. S9). Some crossover was apparent between Domain D (HC84.26) and AR3, and between AR2A and CBH7, but most epitopes could be distinguished.

In all seven chronic progressors, binding against multiple epitopes were identified in the competition binding assay in plasma samples taken as nAb activity emerged (Fig. 4 and Supplementary Fig. 7). In general, the amount of competition for the BnAbs could be seen to increase proportionately over time. However, in some subjects delayed emergence for some epitope binding was observed. Generally, nAb activity was delayed until high binding with multiple well-known neutralizing epitopes was evident. In summary, in chronic progressors the emergence of nAbs correlated with mapping to known nAb epitopes, and the number of epitopes targeted and the level of competition increased over time.

Figure 4

Epitopes targeted longitudinally in representative clearers and chronic progressors. Nine mAbs were used in competition binding assays to determine which epitopes (see key) were targeted throughout the infection. The maroon arrow indicates first nAb response (see key) and grey shading represents the viral load (see key). Representative chronic progressor subjects: 023_Ch (A) and THG_Ch (B), and clearer subject: 277_Cl (C), are shown. Responses in subjects 023_Ch, THG_Ch, and 277_Cl were mapped against non-autologous H77 E1E2.

Surprisingly, limited competition in the binding assay with non-autologous E1E2 was observed in plasma samples collected from the clearers. Low competition, <30%, was observed for subjects 686_Cl, 306_Cl, 168_Cl and 4087_Cl (Supplementary Fig. 7). Subject 277_Cl, was the only clearer subject where competition was evident>30%, with an AR3-like dominant response (Fig. 4C). As most of the recognized HCV neutralizing mAbs act via inhibition of binding of HCV to CD81, a CD81 binding competition assay was also performed for the seven clearers, with only the plasma from the same subject, 277_Cl inhibiting CD81 binding (data not shown).

Due to the very narrow breadth of the binding activity in clearer samples, it was considered possible that lack of reactivity to the H77 E1E2 might have spuriously prevented competition against known nAb epitopes. Therefore, competition mapping against autologous E1E2 was attempted. However, for five of the seven clearers, despite good binding to rE2 in ELISA, poor binding to the lectin-bound autologous E1E2 was also observed (OD: cut-off ratio <1.8) (data not shown).

Irrespective of infection outcome the majority of T/Fs utilize CD81, but as antibodies targeting well-characterized epitopes are generated, the E1E2-CD81 interaction is modulated

As HCV viremia continued in chronic progressors despite nAb activity, longitudinal sequence evolution from the T/F virus was sought in the next generation sequencing dataset of the longitudinally collected samples of clearers and chronic progressors. These data were examined for single amino acid polymorphisms (SAPs) occurring in >10% of the viral population within nAb epitopes, or CD81 binding sites (Supplementary Fig. 8). Limited viral evolution in antigenic regions was observed in clearers, with only two of seven subjects (4087_Cl and 306_Cl) having any SAPs identified.

In contrast, all seven chronic progressors had SAPs occurring in antigenic regions (Supplementary Fig. 8). A significant proportion of chronic progressors had SAPS occurring within AR3 (n = 7, p = 0.041), Domain B (n = 5, p = 0.041) and Epitope II (n = 6, p = 0.029) (Supplementary Table 3). Interestingly, mutations within these regions have been implicated in disruption of CD81 binding25,27,39. Four of the seven chronic progressors (023_Ch, 240_Ch, THD_Ch and THG_Ch) developed mutations in CD81 binding sites, equating to just over half of this group (57%). Notably, of these four subjects, all developed a mutation at residue Y443, which has been reported to be critical for E2 binding to CD8139,40. The viral population in all four subjects mutated away from the highly conserved tyrosine (Y), to a more polar, bulky, positively charged histidine (H) (THD_Ch, THG_Ch or 240_Ch), or to a bulky hydrophobic, isoleucine (I) (023_Ch).

To understand the functional impact that these mutations had on CD81 binding, HCVpp representative of the longitudinal chronic variants (n = 12) were produced, and a CD81 binding assay was performed on T/F variants from both clearers and chronic progressors, as well as the matched 12 longitudinal variants from chronic progressors. Correctly folded E1E2 for each HCVpp was confirmed with a panel of well-characterized mAbs (Supplementary Table 1). Table S1 indicated that at least within subjects the level of E1E2 detected via ELISA with a panel of mAbs was consistent across the longitudinal variants, though some variation between subjects was evident, likely a natural phenomenon of the neutralization profile. Unfortunately none of the well-characterised BnAbs were broad enough to enable proper quantification of the E1E2s.

To determine whether CD81 binding of T/F variants influenced disease outcome or epitope recognition by nAbs, CD81 binding (Bmax) and affinity (Kd) were compared between clearers and chronic progressors (Fig. 5). No significant difference was observed between the T/F variants of clearer and chronic progressor groups for Bmax (Fig. 5B), or for Kd (affinity) (Fig. 5C). Examination of the later variants that emerged after nAb development in chronic infection revealed that both Bmax and affinity to CD81 decreased in samples collected after the emergence of nAbs. The decrease in the CD81 Bmax, but not affinity, was significantly different in post-nAb variants, when compared to pre-nAb variants (p = 0.0405, Fig. 5D). Unfortunately, the lack of a broad binding mAb made it impossible to confirm that all E1E2 were added at comparable amounts, however as outlined in the methods, E1E2 was added to saturation and Table S1 indicates that at least within individual subjects, the variants had comparable mAb binding for the relevant mAbs. Loess regression curves were fitted for the CD81 Kd values from the representative variants isolated from the seven chronic progressors, in conjunction with the log ID50 scores (Fig. 5E) revealing the temporal trends towards decreased CD81 binding as the nAb ID50 increased.

Figure 5

CD81 binding of viral variants by ELISA. Binding measurements were performed with HCVpp lysates incubated with a range of recombinant CD81 concentrations (0.4 300 μg/mL) (A). The Y-axis shows the mean OD values for bound CD81, and the dotted line represents background cut-off. Error bars represent SD. The saturation binding curves were fitted by non-linear regression. HCVpp incorporating E1E2 from T/F viruses from clearers (n = 7) are represented in warmer colors and from chronic progressors are represented in cooler colors (n = 9). All variants are labelled with subject number and clone identifier. Clearer and chronic progressor HCVpp T/F variants were compared for CD81 binding as Bmax (B) and Kd (C). Variants were assessed for CD81 binding before and after emergence of anti-CD81 antibodies was observed using the epitope mapping ELISA (D). Loess regression curves illustrate the reciprocal relationship between the emergence of nAb activity (ID50) and the decline in Kd of the circulating viral variants over time from 6 of the subjects that developed chronic infection (E).

Higher diversity in viral population prior to nAb emergence predicts chronic progression

To test the hypothesis that delayed emergence of nAbs may advantage the virus by enabling a greater repertoire of emergent variants and hence potential for immune escape from nAb selection, the Shannon Entropy (SE) across the E1 and E2 regions was calculated from viral sequences and analyzed over the course of infection (Fig. 6). Firstly, in a time-matched comparison, the timepoint identified to have the greatest SE prior to 51 DPI was selected for each of the 15 subjects (Core/NS seroconversion in the commercial anti-HCV assay was defined as occurring at 51 DPI, see Methods for explanation). Comparison of the SE within E1 and E2 over this period showed no significant difference between clearers and chronic progressors (E1: clearers SE = 0.02989 chronic progressors SE = 0.03525, p = 0.8048. E2: clearers SE = 0.02758, chronic progressors SE = 0.03475, p = 0.6200) (Fig. 6A,B). In contrast, when the last time point prior to nAb emergence was compared, chronic progressors had significantly higher SE in both E1 and E2 when compared to clearers (E1: clearers SE = 0.02502, chronic progressors SE = 0.04318, p = 0.0262. E2: clearers SE = 0.01879, chronic progressors SE = 0.0361, p = 0.0175) (Fig. 6C,D).

Figure 6

Shannon entropy across E1 and E2 regions. Shannon entropy (SE) across the E1 and E2 region was calculated and compared for clearers and chronic progressors over the first 51 days post-infection (DPI) (pre-Core/NS seroconversion) for both E1 (A) and E2 (B). Comparison of SE at the last sampling point prior to nAb emergence was also compared for clearers and chronic progressors across E1 (C) and E2 (D).


This blog is for information purposes only. The content is not intended as medical advice, diagnosis, or treatment. Should you have a medical or dermatological problem, please consult with your physician. None of the information or recommendations on this website should be interpreted as medical advice.

All product reviews, recommendations, and references are based on the author’s personal experience and impressions using the products. All views and opinions are the author’s own.

This blog post may contain affiliate links. An affiliate link means we may earn a commission if you click on a link and make a purchase, without any extra cost to you.

Please see our Disclaimer for more information.

Source Link


diseases, diagnosis and treatment methods, drugs and their side effects on this site. online diseases, diagnosis and treatment methods

Related Articles

Back to top button
%d bloggers like this: