Quasispecies composition appears to be important in the IFN resistance mechanisms , and genetic variability has been studied mainly in E2 and NS5A regions of the HCV genome [30–35]. Additional investigation of NS5A region variability may be relevant as heterogeneity in this region before treatment has been related to IFN-monotherapy responsiveness. However, studies have presented controversial data and the mechanism underlying how NS5A interferes with responsiveness to treatment has yet to be elucidated.
In the present study, the genetic variability of the complete NS5A region in patients infected with HCV genotype 1 was analyzed. Altogether, 690 sequences of the entire NS5A from 11 patients’ samples were analyzed. Samples showed lower genetic diversity of quasispecies composition at baseline for SVR patients than NR and patients who relapsed after the end of therapy (ETR). Those significant differences among genetic distance of before treatment were not observed in our previous work when we exclusively analyzed baseline samples . However, the group of patients and sequences analyzed were not identical, being only part of sequences of previous work used to these analyses.
Our findings are consistent with previous observations made on patients undergoing anti-HCV therapy that less complexity and lower diversity of HCV quasispecies at the baseline is associated with viral clearance [19, 21, 37–39]. Therefore, high genetic variability of HCV could be related to the low efficiency of anti-HCV treatment . Jain et al. observed that the initial anti-viral effect of interferon is influenced by the quasispecies composition at the time of treatment initiation, and patients who demonstrated high viral diversity were less likely to respond to treatment . In agreement with the genetic distance and entropy results, the number of viral strains at baseline was lower for SVR patients than NR and ETR patients, showing that quasispecies composition was more homogeneous than other groups of responses. A recent study evaluating the clinical and virological parameters that could be associated with or predictive of therapy outcome at baseline demonstrated that a higher number of quasispecies variants in the E1/E2 region was significantly associated with treatment failure. However, it is inferred that independent factors cannot provide a consistent prediction of therapy response .
In order to investigate whether therapy caused a significant change in the population over time, we compared variants heterogeneity of baseline, during therapy and follow-up by analyzing the mean of genetic distance and Shannon entropy among each time-point. No specific profile of quasispecies diversity for all patients from each response type could be identified over time. However, significant differences were observed when baseline and post-therapy variants of NRs and ETRs were compared. In either case post-therapy population were significant less heterogeneous than baseline showing a decrease of diversity over time.
We also calculated the mean of genetic distances between each two time points and could observe that when compared to the baseline, post-therapy populations of ETR patients presented significantly higher diversity from baseline variants than NR patients.
Those findings correlate with data obtained by tracking the number of strains. Analyzing ETR, it was evident that quasispecies composition was more homogeneous in post-therapy samples. This homogeneity was most evident in patients P37 and P03, who presented with homogeneous quasispecies composition from the rebound and sustained it. As with ETR patients, the HCV RNA is undetectable by the end of therapy, suggesting that the selective pressure of treatment acts on quasispecies diversity during it until HCV becomes undetectable in samples. One explanation is that during or after the end of therapy predominant quasispecies rise in frequency, while others decrease progressively in frequency or are eliminated. The persistence of this predominant variant could be favorable for sustaining the infection, causing it to become chronic again.
Farci at al. evaluated the quasispecies composition of the E1/E2 region in samples from patients who were untreated or treated with conventional IFN combined with ribavirin. Analyzing samples of relapse patients, they observed greater viral diversity at baseline, and an emergence at relapse of a new dominant strain, or the emergence of a minor dominant population . They inferred that the emergence of a new viral strain after rebound implies that the majority of baseline variants were sensitive to IFN and that the origin of these new strains is uncertain. The authors concluded that it is probable that very low levels of virus replication continued to occur despite the disappearance of viremia .
A study focusing on breakthrough response pattern, defined by patients who have an initial response followed by reactivation whilst receiving IFN therapy, did not identify quasispecies variants from breakthrough in baseline samples . This study assumed that HCV variants sampled at the time of breakthrough represent drug-resistant quasispecies, and if present at baseline, they must be present very low frequency. According to the authors viral breakthrough could be attributed to the selection of pre-existing drug-resistant variants or the emergence of different quasispecies with reduced sensitivity to IFN. However, the study indicated that selection is mostly responsible for appearance of drug resistant quasispecies at breakthrough . As breakthrough and relapse indicate rebound of infection after undetectable HCV RNA levels, is reasonable that the results presented herein are in agreement when considering the quasispecies composition and diversity at breakthrough and relapse time.
The data collected from non-responder samples revealed a very diverse composition of viral strains over the time course of the investigation. As described previously, the high variability of the quasispecies population in these patient samples might represent a continuous process of adaptation , as these variants were continuously eliminated during and after the end of treatment in most cases. Alternatively, it could offer the virus some advantage in sustaining the infection as a large number of quasispecies may indicate a better opportunity for virus persistence [44, 45].
The analysis of the number of effective populations among time-point analyzed showed different profiles for the types of response. A strong decrease of effective populations was observed after therapy for ETR patients whereas NRs presented a continuous variation or stability of its populations over time. Those data are in agreement to nucleotide diversity results where ETRs demonstrated a decrease of population number and heterogeneity presenting more distinct post-therapy variants. In general, populations of those patients experience a bottleneck phenomenon after therapy. On the other hand, profiles found to NRs assent to the continuous high diverse population and a lower diversity between pre and post-therapy variants probably due to a minor effect of treatment in variants selection. Pawlotsky et al. compared pre- and post-therapy NS5A amino acid sequences from non-responder patients and demonstrated that most variants from after-treatment samples were not detected before treatment, whereas most pre-treatment variants were no longer evident after treatment . The data presented here agree with their findings, with the exception of samples from patient P08. In the P08 samples, a variant was identified at baseline (13% of NS5A amino acid sequences) that was detected after 12 weeks of treatment and two months after therapy was completed. Several studies have showed that variants resistant to therapy may be present at baseline, and it is proposed that the persistence of these variants in patients who fail to respond suggests the existence of virus strains with inherent resistance to IFN [41, 43].
Consistent with previous observations, phylogenetic analysis of virus sequences obtained from patients failed to show any clustering associated with specific response [34, 41, 46]. However, the data suggested that pre- and post-therapy isolates from samples of ETR patients tended to group together in different branches. For NR samples, before-treatment quasispecies variants tended to group with quasispecies found after 12 weeks of treatment, and post-therapy sequences grouped together in another branch. The distinctive clustering of pre/post-therapy sequences had been described previously [22, 42, 47], demonstrating the evolutionary process occurred over time. In some cases, isolates identified in after-treatment samples had the longest distances from the main nodes in phylogenetic analysis. This may suggest that changes can improve the fitness of quasispecies resulting in a persistent infection after treatment selection pressure.
To investigate whether quasispecies were under differential selective pressure over time, we performed analysis with the set of sequences of each NR and ETR patient. In general, the results showed that a purifying selection is driving the evolution of quasispecies over time. In some cases, the purifying selection was relaxed (clades selected for patients P11 and P47). In such cases, two scenarios could be envisaged. Variants undetected in the baseline sample due its low frequency could increase in frequency under relaxed selective pressure and consequently be detected in the after therapy samples for some time. Alternatively, those variants could have arisen by mutation during treatment increasing in frequency and being detected then. Relaxed purifying selection may also be indicative of some codons of the entire protein being under positive selection. However, we failed to identify sites under positive selection for those patients. Codons under positive selective pressure observed in this work could not be related to the therapy outcome and were not present in preferential position in the NS5A protein. A study analyzing PKR binding domain and the V3 domain in the NS5A region identified positively selected sites but also failed to detect a pattern mechanism for the inefficient response to anti-viral treatment . Therefore, further analyses are necessary.
Genetic analyses identified nonsense mutations in the NS5A sequence of some HCV variants. Nonsense mutations in samples from HCV-infected patients have been described [34, 36, 49], but the structural and functional implications of such mutations detected in vivo are unclear. Most of the nonsense mutations detected were located in the N-terminal region of NS5A (Alpha Helix, Domain I or Low Complexity Sequence I regions of the NS5A); only one was located in the C-terminal region (domain III). It is known that N-terminal mutants of NS5A are preferentially located in the nucleus and are reported to function as transcriptional regulators. A recent study using a HCV replicating cell system revealed that during the life cycle a variety of N-terminally truncated NS5A fragments are generated . Tests on several of these truncated NS5A fragments demonstrated that they were preferentially located in the nucleus or equally distributed between the cytoplasm and nucleus. The full length NS5A (1–449) was located in the extra-nuclear compartment as previously described. However, truncated forms impaired HCV replication. In contrast, domain III of the C-terminal NS5A region can be deleted with no or minimal effect on RNA replication [51, 52], but the C-terminal region residing between amino acid residues 2404 and 2435 of domain III is crucial for virus production . To elucidate the effects of nonsense mutations identified in the present study, further analysis with HCV replicating cells is necessary. Truncated NS5A forms observed in previous studies show drawbacks in terms of replication or assembly, but these effects could be overcome by the diversity of quasispecies composition.