Genetic diversity of NS5A protein from hepatitis C virus genotype 3a and its relationship to therapy response

The viral load data show that all patients presented high viral loads, as expected for pre-treatment samples, ranging from log 5.76 to log 7.03. There was no correlation between viral load and treatment response (Table 1).

This study generated 15 contig sequences of the full NS5A region from 11 patients; 14 sequences from patient RF80 could be sequenced, totaling 179 contigs.

Quasispecies analysis revealed that the HCV NS5A region was highly variable, as only three patients showed two identical nucleotide sequences (Figure 1A). The amino acid data showed more identical sequences owing to synonymous substitution, as can be seen when the nucleotide and amino acid patterns of patient RF15 are compared. All nucleotide sequences from this patient were different; however, the amino acid sequences showed one of the highest degrees of conservation (Figure 1B).

Nucleotide substitution analysis showed that the sustained virological response (SVR) group had the highest means of nucleotide and amino acid substitutions, except for the amino acid sequences of regions NLS and V3, where the end-of-treatment response (ETR) group showed the highest values (Table 2). However, none of the differences between groups were statistically significant.

To identify the mutation sites in the sequences used in this study, they were represented graphically according to the reference sequence NZL1 (GeneBank D17763) (Figure 2). This representation indicates that no specific mutation could be associated with any kind of treatment response. Figure 2 also shows that some of the nucleotide mutations resulted in stop codons. Of the nine stop codons found in the 179 sequences generated in this study, the site was the same in two or more clones in seven cases. The same stop codon sites were found in NS5A amino acids 4 (RF31 and RF145), 166 (RF60 - 2 clones and RF145) and 447 (RF07 and RF109). Also, the sites where the translation stop codon was observed showed no other mutation, except for one clone from patient RF145, which showed a mutation in aa 447.

For the within-sample (ws) data, the mean rate of synonymous substitution per synonymous site (dS) was higher in the SVR group (Table 2). The rate of non-synonymous substitution per non-synonymous site (dN) was also higher in the SVR group, except for regions CRS and PKR-Bd. These data showed no statistically significant differences.

Between-sample (bs) analysis also showed a higher dS mean in the SVR group. SVR had the highest mean dN values, except for regions NLS and V3. All the bs data showed statistical significance, except for the dS means in CRS between SVR and ETR and between NR and ETR.

The mean genetic distances for the complete NS5A region and the other regions studied were higher in the SVR group in either ws or bs analysis (Table 2). No statistically significant difference was found for genetic distance.

Phylogenetic analysis was carried out using the 179 1356-bp sequences generated in this study, the reference sequence for genotype 3, NZL1 (GenBank accession number D17763), the six full-length NS5A sequences from genotype 3a, with country information, available in GenBank (Accession numbers: AY956467; DQ430819; DQ430820; DQ437509; X76918; GQ300882.1), and 20 Brazilian NS5A sequences of 1308 bp (Accession numbers: EF207999.1 to EF208018.1). The resulting phylogenetic tree is presented in Figure 3. All sequences from clones from the same patient are grouped in a monophyletic branch, with high bootstrap values (95-100%). There was no clustering in a monophyletic branch of the sequences from patients in the same treatment response group.

Plasma samples were collected from 12 HCV genotype 3-positive patients, seen at the Hemocenter of São José do Rio Preto, State of São Paulo, Brazil. All samples were collected before treatment, and a 6-months follow-up provided treatment response data. Four patients were sustained virological responders (SVR), i.e the virus was not detected after treatment or during the 6-months follow-up; four were non-responders (NR), since they showed no virological response; and four were end-of-treatment responders (ETR), i.e. a virological response was detected, but at the 6-months follow-up there was a rebound. All patients were infected with HCV genotype 3a. Treatment consisted of INF-α and ribavirin administration for 24 weeks. Patients with a history of alcoholism or infection with another agent that could cause liver damage were excluded. This study was approved by the ethics committee of the Base Hospital of São José do Rio Preto, and all participants signed an informed consent.

The viral load was quantified by Cobas TaqMan HCV Test.

Viral RNA was extracted from blood serum samples using a QIAamp Viral RNA Mini Kit (QIAgen), and cDNA was synthesized using a High-Capacity cDNA Archive Kit (Applied Biosystems) according to the manufacturer's instructions. The NS5A region was amplified using a set of primers specific for genotype 3, designed for this study. For the PCR reaction, two sets of forward and reverse primers were designed. The forward primers were H.NS5AP-F (5' GAGCGGTACAGTGGATGAAC 3' - nucleotide [nt] 6089 to 6108 in genotype 3 reference sequence NZL1 - GenBank D17763G) and H.NS5AP-F2 (5' GGTACAGTGGATGAACAGG 3' - nt 6093 to 6111 in NZL1). The reverse primers were H.NS5AP-R (5' CCTCCTTTAATGCAGTCTTG 3' - nt 7821 to 7840 in NZL1) and H.NS5AP-R2 (5' ACGACGTTGAATAGACTAGG 3' nt 7734 to 7753 in NZL1). Two sets of forward and reverse primers were also designed for the nested-PCR reaction. The forward primers were H.NS5AN-F (5' CGCATTGCTGAGTTCTCTAAC 3' from nt 6192 to 6212 in NZL1) and H.NS5AN-F2 (5' CTCTAACTGTCACAAGTCTGC 3' nt 6206 to 6226 in NZL1). The reverse primers were H.NS5AN-R (5' CAACAAGGAGTTGCTGAGTG 3' nt 7703 to 7722 in NZL1) and H.NS5AN-R2 (5' CAGCACTACATGGTGTTATC 3' nt 7659 to 7678 in NZL1). For the amplification reaction, 1 U of a proofreading polymerase was used (Elongase^® Enzyme Mix; Invitrogen) along with 10 μl of buffer B [300 mM Tris-SO₄, (pH 9.1 at 25°C), 90 mM (NH₄)₂SO₄ and 10 mM MgSO₄], 10 μl of dNTP, 30 pmol of sense and anti-sense primers, 10 μl of cDNA for PCR reaction, and 5 μl of PCR product for the nested-PCR reaction, plus Milli-Q autoclaved water to a final volume of 50 μl. The amplified products were analyzed on a 1% agarose gel.

Cloning was performed using a TOPO XL Cloning TM Kit (Invitrogen). Fragments from 15 clones from each patient were purified using a PureLink Mini-prep Plasmid Purification Kit (Invitrogen). The entire NS5A region was sequenced using eight primers: the vector primers M13F and M13R, and six inner primers, three sense and three anti-sense, designed for this study. The forward primers used in the sequencing reaction were: H.NS5AI-F1 (5' TGGCTGCGTATCATCTGGGA 3' - nt 6283 to 6302 in NZL1), H.NS5AI-F2 (5' ACCTCGATGTTGAGAGACCC 3' - nt 6871 to 6890 in NZL1) and H.NS5AI-F3 (5' TATCCTCCAGCCCTTCCTAT 3' - nt 7198 to 7217 in NZL1). The reverse primers used in the sequencing reaction were: H.NS5AI-R1 (5' CACGGACACTTGAGCTCATC 3' - nt 6679 to 6698 in NZL1), H.NS5AI-R2 (5' TTCTTGAAACACTCTGCAGC 3' - nt 7168 to 7187 in NZL1) and H.NS5AI-R3 (5' GTGGACCAAGAGTCGCAACT 3' - nt 7573 to 7592 in NZL1). The sequencing reaction was performed with Dyenamic ET Terminator (GE) and the products were sequenced in an ABI Prism 377 sequencer (Applied Biosystems). The reaction mixture consisted of 1 μl of Milli-Q autoclaved water, 1 μl of primer (5 pmol/μl), 2 μl of sequencing reagent mix, plus 2 μl of sample. Occasionally, when a good quality sequence could not be obtained, the sequencing reaction had to be doubled and a "hot start" had to be performed for 10 min at 95°C for better results. Cycling was carried out according to the manufacturer's instructions.

The sequences obtained were subjected to BioMol - Electropherogram quality analysis http://​adenina.​biomol.​unb.​br/​phph/​[46], a phred phrap [47, 48] analysis site, for quality check and contig construction. The contigs obtained for each clone were aligned, along with the reference sequence NZL1 for genotype 3 (GenBank accession number D17763), using Clustal X 1.81 software [49]. All sequences were edited on Bio Edit 7.0.5.3 [50] to remove the vector fragments, leaving only the complete sequence of the NS5A region.

Quasispecies analysis was carried out using LOCSPEQ 1.0 software [51], specially designed for our group for this kind of analysis.

The number of mutations and the genetic distances were calculated using MEGA 4.0 software [52]. The rates of synonymous substitution per synonymous site (dS) and non-synonymous substitution per non-synonymous site (dN), as well as the dN/dS ratio, were obtained at SNAP - Synonymous Non-synonymous Analysis Program -http://​www.​hiv.​lanl.​gov[53, 54].

Phylogenetic analysis was performed using PAUP* version 4 software [55]. A neighbor-joining phylogenetic tree was constructed with Tamura-Nei's substitution model including invariant sites (I) and Gamma distribution shape (G) parameter (TRN+I+G), determined by hierarchical likelihood ratio test score criteria using Modeltest 3.06 [56]. One thousand replicates were used to test the reliability of the tree topology, and bootstrap values >70 were considered significant [57].

Statistical analysis was performed by ANOVA. Comparisons between groups were made using Tukey's method for multiple comparisons. Values of P < 0.05 were considered significant. Standard error of the mean (SEM) values were calculated in Minitab 15.

In this work, we chose to analyze the results obtained by two different approaches:

The nucleotide sequence data reported here have been submitted to the GenBank nucleotide sequence database with accession numbers from EU826174 to EU826352.