[PROVISIONAL] Resistance mutations of NS3 and NS5b in treatment-naïve patients infected with Hepatitis C virus in Santa Catarina and Rio Grande do Sul States, Brazil.

Hepatitis C virus (HCV) infection is a worldwide health problem. Nowadays, direct-acting antiviral agents (DAAs) are the main treatment; however, the high level of variability of HCV lead to development of resistance-associated variants (RAVs). Thus, look into the RAVs among infected patients is an important tool for monitoring the efficacy of the therapy. The aim of our study was to investigate the presence of naturally occurring resistance mutations in HCV NS3 and NS5 regions in treatment-naïve patients. Ninety-six anti-HCV positive serum samples from blood donors at Center of Hematology and Hemotherapy of Santa-Catarina State (HEMOSC) were collected retrospectively in 2013, and evaluated in this study. HCV subtypes 1a (37.9%), 1b (25.3%) and 3a (36.8%) were found. The frequency of patients with RAVs in our study was 6.9%. The HCV NS5b sequencing reveled 1 sample with L320F mutation and 4 samples with the polymorphism C316N/R. The analysis of the NS3 region revealed mutations D168A/G/T (3.45%), S122G (1.15%) and V55A (2.3%). All samples from genotype 3a (36.8%) presented the non-synonymous mutation V170 I/V. In conclusion, we have shown that mutation in NS3 and NS5b genes are present in Brazilian isolates from therapy-naïve patients.


Introduction
Hepatitis C virus (HCV) infection is a worldwide health problem. According to the Global Hepatitis Report, from the World Health Organization (WHO), approximately 71 million people have chronic HCV infection, and nearly 399,000 people die each year, mostly due to cirrhosis or hepatocellular carcinoma (WHO, 2017). HCV has a high genetic heterogeneity and is classified into seven genotypes (1 to 7) and 67 subtypes (Smith et al., 2014). The genotype distribution depends on geographical location and risk groups (Cantaloube et al., 2005). Genotype 1 is the most frequent in Brazil, followed by genotypes 3 and 2 (Campiotto et al., 2005;Lampe et al., 2013).
There is no vaccine available for preventing HCV infections. The main antiviral treatment until 2011, was PEGylated interferon-alfa (aPeg-IFN) alone or in combination with ribavirin, leading to a sustained virological response (SVR) in 50% of treated patients, depending on the virus genotype causing the HCV infection (Peres-da-Silva et al., 2012;Paolucci et al., 2013;Gross et al., 2018). Nowadays, direct-acting antiviral agents (DAAs) have been approved for HCV infection treatment, with an average SVR above 95%, at least for genotypes 1 and 4 (Leuw and Stephan, 2018). In Brazil, DAAs were incorporated by the Ministry of Health for the treatment of hepatitis C under the Unified Health System (SUS) since 2015 (Ministério da Saúde, 2018). Unfortunately, there is little data about the efficacy of DAAs in Brazil, with some information found in the study by Sette Jr et al. (2017).
The primary targets of DAAs are nonstructural proteins essentials for HCV replication, which include the NS3 protease, NS5B polymerase, and NS5A protein (Paolucci et al., 2013;Lontok et al., 2015). However, a challenge in HCV treatment is the emergence of viral resistance mutations that reduces susceptibility of the virus to DAA therapies (Hoffmann et al., 2015;Gededzha et al., 2017). The development of resistance-associated variants (RAVs) is due to the high level of virus variability, from the combination of the virus' high replication rate, low RNA polymerase fidelity rate, and selective pressure for drug or immunomediated treatment (Peres-da-Silva et al., 2012;Paolucci et al., 2013;Gededzha et al., 2017).
The presence of RAVs in patients not yet under treatment has been reported previously in different countries (Peres-da Silva et al., 2010;Paolucci et al., 2013;Zeminian et al., 2013;Gededzha et al., 2017). In addition, a systematic review regarding HCV resistance-associated substitutions and their clinical relevance was published recently (Sorbo et al., 2018). Therefore, assessing RAVs in infected patients is important for monitoring the efficacy of therapy (Loggi et al., 2017) and the epidemiology of HCV in Brazil. Thus, the aim of our study was to investigate the presence of naturally occurring resistance mutations in HCV NS3 and NS5 regions in treatment-naïve patients.

Materials and Methods
The Center of Hematology and Hemotherapy of Santa Catarina State (HEMOSC) is currently responsible for the nucleic acid testing for HIV, HCV, and HBV in samples from blood donors from Santa Catarina and Rio Grande do Sul states. Annually, HEMOSC receives around 300 thousand blood donations. A total of 96 samples that were positive for HCV in 2013 were used for this study, retrospectively.
HCV RNA was extracted from plasma previously conserved at -80°C using a molecular biology workstation (BioRobot MDx, Qiagen), with the Qiamp one-for-all nucleic acid kit (Qiagen), according to the manufacturer's in-structions. Plasma HCV RNA was quantified using COBAS/Taqman HCV Test v2.0 (Roche).
Genotyping/subtyping was performed by amplifying and sequencing a 339-bp amplicon of the NS5b region, according to Cantaloube et al. (2005). The nucleotide sequences obtained were analyzed in the Geno2pheno [HCV] (Kalaghatgi et al., 2016) for genotypes and subtypes, and possible resistance against licensed DAAs.
The amplification of the entire NS3 region of the HCV genome, followed by a second PCR was performed as described previously (Peres-da-Silva et al., 2010), using primers specific to subtypes 1a, 1b, and 3a. The nucleotide sequences obtained from each subtype were analyzed for drug resistance in the Geno2pheno [HCV] (Kalaghatgi et al., 2016).
The statistical program SPSS (IBM SPSS Statistics Base 22.0) was used. Multivariate analysis of variance (ANOVA) was applied to compare means of continuous variables with normal distribution (p<0.05).

Results and Discussion
From the 96 HCV-positive samples collected, nine did not have enough material to perform the assays and were excluded from the analysis. Eighty-seven samples were used for genotyping and analysis of NS3 and NS5b regions.
(36.8%) were of genotype 3a. Genotype 1 (1a plus 1b) was the most frequent, followed by genotype 3, a result that is in agreement with what was previously reported in Brazil (Campiotto et al., 2005;Lampe et al., 2013;Nishiya et al., 2014). We did not find genotypes 2, 4, and 5, known to be less frequent in Brazil. Sixty-three samples were successfully geno-and subtyped by the NS3 and NS5b region, and there was no disagreement between the HCV genotypes in both regions. Twenty-four samples had no amplification of the NS5b region, even with an alternative protocol (Sandres-Sauné et al., 2003), and were thus genotyped according to Peres-da-Silva (2010) protocol (NS3 region). This amplification divergence has already been discussed in Larrat et al. (2013), who reported a failure of some quantitative RT-PCR assays to detect or amplify correctly the NS5b region in some strains of HCV, even when using three sets of primers covering two different regions. This could be explained by the great variety of viruses, the use of primers not suitable for these peculiar strains, or by a mixed infection in the plasma sample.
The mean viral loads were 5.31 log IU/mL for genotype 1a, 5.18 log IU/mL for 1b, and 5.38 log IU/mL for 3a. There was no difference in viral load between the genotypes (p=0.6). The detected HCV genotypes and viral loads are both important predictors for therapeutic outcomes. It has been reported that patients infected with genotype 1 are more likely to have higher viral loads than those infected with genotype 2 and 3 (Scott et al., 2007;Soriano et al., 2008;Nishiya et al., 2014). In contrast to our results, in a study carried out with blood donors from São Paulo, the viral load from genotype 3a (5.22 log10 IU/mL) had a lower log mean than genotype 1a (5.99 log10 IU/ mL) (p=0.0002) and genotype 1b (6.35 log10 IU/mL) (Nishiya et al., 2014), in agreement with a previous report.
The frequency of patients with RAVs in our study (6.9%) was intermediate when compared with other Brazilian studies among HCV chronic carriers not treated with protease inhibitors (3.2% -18.9%) (Hoffmann et al., 2013;Nishiya et al., 2014). Figure 1 present the frequency of specific NS3 and NS5b resistance-associated variants found in this study by HCV subtype.
The HCV NS5b sequencing from 63 samples was analyzed. The L320F mutation was present in only one sample (1.59%) of genotype 1a. L320F is known to confer low resistance to sofosbuvir and sofosbuvir associated with mericitabine (Paolucci et al., 2013), which are associated to treatment failure in clinical trials (Constantino et al., 2015). In a previous study, L320F single mutation had no significant impact on the 50% effective concentration (EC50) and EC90 values for mericitabine (£2.7 fold) (Tong et al., 2014). To our knowledge, this is the first time that mutation L320F is reported as naturally occurring in DAA treatment-naïve patients and it should be monitored due to treatment failures reported previously in clinical trials.
Although it was not a goal of our study, we also observed the presence of D244N, Q309R, and A333E mutations conferring resistance to ribavirin and interferon in 42 samples (57.14%). Twenty samples (26.98%) presented Q309R, and three (1.58%) A333E. Seventeen were Q309R and D244N, two were Q309R and A333E, and two were triple positive. In a previous work, the most frequent mutation observed in Brazil was Q309R, present in all HCV subtypes (Castilho et al., 2011); in our study, it was present in 38 samples. No double mutations in the NS5b region conferring resistance to DAAs was observed in our samples. The emergence of double or triple-sites RAVs in the clinics is threatening the effectiveness of anti-HCV therapies, as published previously (Gane et al., 2016).
No mutations were found that confer resistance to glecaprevir and voxilaprevir, drugs known to present a high Resistance mutations of NS3 and NS5b barrier to resistance (Sorbo et al., 2018). However, this study found samples with mutations that decrease the susceptibility of HCV to these drugs, which reinforces the importance of monitoring HCV RAVs.
Samples from genotype 3a presented no mutations that confer or diminish resistance to glecaprevir and voxilaprevir, drugs recommended for treatment of patients infected with this genotype. This means that the standard protocol for treatment of patients with genotype 3 should be effective in Santa Catarina and Rio Grande do Sul. However, we found the non-synonymous mutation V170 I/V in all 32 samples of this genotype. In agreement with our results, Peres-da-Silva (2010) found that 100% (32/32) of the HCV 3a sequences contained the V170I substitution. Few data is available on effects of V170I substitution. The conservative substitution at this site was detected in up to 45% of patients infected with HCV genotype 1 (López-Labrador et al., 2008) A different pattern of resistance associated with NS3 protease domain in therapy-naïve patients was previously reported in Brazil. V36L mutation was found in genotype 1a at a frequency of 5.6%, in 1b at 100% (Peres-da-Silva et al., 2010), and in genotypes 2, 3, 4, and 5 V36L mutation was found as a genetic signature with frequency of 99% (Vidal et al., 2016); in another work, V36L was found at a frequency of 4% in genotype 1a (Nishiya et al., 2014). T54S mutations were found in 4.1% of genotype 1a (Peres-da-Silva et al., 2010) and 100% in genotype 2 (Vidal et al., 2016). The samples investigated by our study presented none of these mutations. The Q80K, a common mutation in the USA (40%) (Bartels et al., 2013) that confers resistance to simeprevir, was not found in our study, but has previously been reported at prevalence ranging from 0.4% to 2.7% in Brazil (Nishiya et al., 2014;Vidal et al., 2015;Moreira et al., 2018). There is a strong geographic correlation regarding the frequency of the Q80K substitution (Moreira et al., 2018), and for this reason, studies from different geographic regions are of great importance, especially in a large country as Brazil.
Of all the samples evaluated, only one sample of genotype 1b showed mutation in the genes NS3 and NS5b, conferring resistance to sofosbuvir (C316N) and decreased susceptibility to gazoprevir (Y56F). This shows the importance of studying both NS3 and NS5 proteins when evaluating or choosing the therapy strategy for HCV-positive patients. Patients carrying combinations of resistance mutations are of particular interest, since they may increase the possibility of failure in the treatment with DAAs.
The frequency of resistance mutations and genotypes was twice as high among patients with subtype 1a compared to those with subtype 1b. A similar result was found in a study with blood donor's samples from São Paulo (Nishiya et al., 2014). In addition, a higher frequency of virological failure for subtype 1a compared to 1b has been reported (Pawlotsky et al., 2011).
At last, several polymorphisms not associated with resistance to DAAs were observed in our study (Table 1), and previously reported by others (Constantino et al., 2015). Polymorphisms, prior to therapy, are part of the quasispecies population in infected individuals, and may not alter viral fitness (Peres-da-Silva et al., 2012;Paolucci et al., 2013;Nishiya et al., 2014).
In conclusion, we have shown that mutations in NS3 and NS5b domains are present in Brazilian isolates from therapy-naïve patients, in this case, blood donors with unknown HCV infection. Monitoring the presence of RAVs is important for predicting the response to antiviral therapy, and regional discernment can help determine local policies for treatment. The results presented here will help ensure therapy strategies that are more successful for HCV-infected patients in Santa Catarina and Rio Grande do Sul states in Brazil. *Some samples had more than one variant.