Intergenotypic 2k/1b hepatitis C virus recombinants in the East Macedonia and Thrace region of Greece

Introduction

Hepatitis C virus (HCV) infection presents a major global health burden. Chronic HCV infection may lead to severe clinical outcomes, such as cirrhosis and hepatocellular carcinoma [1]. HCV is a positive sense RNA virus belonging to the Flaviviridae family. The viral genome encodes for a polyprotein cleaved by host and viral proteases to produce the individual viral proteins. HCV is characterized by an extensive genetic variability that has an impact on disease progress and treatment outcome. Genotypes 1, 3 and 4 are characterized by a higher incidence of hepatocellular carcinoma and low rates of sustained virological response (SVR) following pegylated interferon (peg-IFN)/ribavirin therapy [2]. On the other hand, genotype 2 is more likely to be cleared spontaneously, while it shows more than 80% response to peg-IFN/ribavirin therapy [3]. The development of new, direct-acting antivirals (DAAs) against HCV has increased SVR rates to more than 95%; however, their high cost has limited their availability [4]. The first HCV recombinant to be identified was the CRF01 strain from St. Petersburg in Russia, comprising a rare 2k 5’ moiety followed by a 1b 3’ moiety [5]. Since then, 2k/1b recombinant strains have been identified in countries bordering Russia and other European countries, including Ireland, France and Portugal [6-13]. In the present study we identified for the first time three 2k/1b recombinant clinical isolates in the Greek region of East Macedonia and Thrace.

Patients and methods

HCV genotyping and phylogenetic analysis

To identify a possible reason for the interferon resistance, we sequenced the core/E1 region of strains to reevaluate genotyping. Serum samples from the three HCV type 2 infected patients were used in this study. Serum from patients 1, 2 and 3 were obtained in 2014, 2015 and 2015, respectively. HCV RNA was extracted using the NucleoSpin Dx Virus kit (Macherey-Nagel) and reverse-transcribed to complementary DNA using M-MLV Reverse Transcriptase (Promega). Core_F and Core_R primers were used to amplify a 1028 bp 5’ UTR-Core-E1 polymerase chain reaction (PCR) product, while NS2_F and NS2_R primers were used for amplification of the NS2 region. Amplification of NS5A (520 bp) and NS5B (450 bp) regions was performed using primer pairs NS5A_F/NS5A_R and NS5B_F/NS5B_R, respectively (Table 1). The PCR products were purified using NucleoSpin gel and PCR clean-up purification kit (Macherey-Nagel), and then the HCV sequences were determined by cycle sequencing. The sequencing reactions were carried out using a BigDye® Terminator v3.1 Cycle Sequencing kit (Applied Biosystems) with the automated capillary DNA Sequencer ABI 3730 XL Analyzer (Applied Biosystems), according to the manufacturer’s instructions. GenBank accession numbers of the sequences obtained in this study are MH883229-MH883231 for the Core-E1 sequences, MH883226-MH883228 for the NS2 sequences and MH883232-MH883234 for the NS5A sequences.

Table 1 Oligonucleotides used in the study

The phylogenetic analysis was based on the core/E1 region of 2k/1b strains identified in Europe from patients originating from Azerbaijan (FJ435462, FJ435480, FJ435490, FJ435497, FJ435505, FJ435514), Georgia (EU684686, EU684728, JF949903, JF949904, JF949905, JF949906, JF949907, FJ821465), Russia (JF949908, AB327018, AY070214, AY070215, AB327015, AB327011, AB327010, AB327012, AY587845, AB327013, AB327014), and Uzbekistan (AB327016, AB327122). Evolutionary analysis in the Core/E1 and NS5A regions was conducted using MEGA7 [14] by applying the neighbor-joining method [15]. The evolutionary distances were computed using the Maximum Composite Likelihood method [16,17] and are in the units of the number of base substitutions per site. Genotyping of rs12979860 was performed by PCR using oligos IFNL3F and IFNL3R, resulting in the amplification of a 242 fragment and RFLP with restriction endonuclease BstU1. Genotyping of rs368234815 was performed by PCR using oligos IFNL4F and IFNL4R, resulting in the amplification of a 203 fragment and RFLP with restriction endonuclease Eco0109I (Table 1).

Results

Sequence analysis of the core-Ε1 region revealed that the three patients were infected with HCV strains phylogenetically related to the Russia-derived CRF01 2k/1b strain. The core/E1 region of all three strains, originally genotyped as 2a or 2a/2c, clustered together with the 2k core/E1 region of the 2k/1b intergenotypic recombinants (Fig. 1). Following partial sequencing of the NS5A and NS5B regions we verified the HCV 1b origin of the regions in all three patients (Fig. 2). Sequencing analysis of the NS2 region, where the recombination was anticipated, revealed that all three strains encompassed the same recombination site as the original RF1 2k/1b strain from Russia at genomic position 3175 (Fig. 3) [10]. The identification of the particular recombination site established the RF1 2k/1b origin of the strains as every recombinant HCV strain identified up to date has a unique recombination site position.

Figure 1 Evolutionary relationships of taxa in the core/E1 region. The percentage of replicate trees in which the associated taxa clustered together in the bootstrap test (1000 replicates) are shown next to the branches [21]. The tree is drawn to scale, with branch lengths in the same units as those of the evolutionary distances used to infer the phylogenetic tree. The analysis involved 40 nucleotide sequences of the core/E1 region. Codon positions included were 1st+2nd+3rd+noncoding. All positions containing gaps and missing data were eliminated. There were a total of 263 positions in the final dataset

Figure 2 Evolutionary relationships of taxa in the NS5A region. The percentage of replicate trees in which the associated taxa clustered together in the bootstrap test (1000 replicates) are shown next to the branches [21]. The tree is drawn to scale, with branch lengths in the same units as those of the evolutionary distances used to infer the phylogenetic tree. The analysis involved 20 nucleotide sequences of the NS5A region. Codon positions included were 1st+2nd+3rd+noncoding. All positions containing gaps and missing data were eliminated. There were a total of 404 positions in the final dataset

Figure 3 (A) Alignment of Patient 1 NS2 region recombination site with sequences of 2k and 1b origin (Grey = similarity with 2k, black = similarity with 1b). (B) Alignment of NS2 recombination site of all three patients with RF1_2k/1b strain

In parallel with the reassessment of virus genotype, we performed analysis of the genetic background of the patients as it has been linked to the treatment outcome. Thus, we identified the IFNL3 rs12979860 and the IFNL4 rs368234815 genotypes/variants of these patients. Patient 1 was genotyped as ΔG/ΔG for rs368234815 and TT for rs12979860, both favoring a poor outcome of treatment. Patient 3 was genotyped as TT/ΔG for rs368234815 and CC for rs12979860, favoring poor outcome and good outcome respectively.

Discussion

Intergenotypic recombinant HCV strains emerge rarely during coinfection of the same individual with two HCV genotypes. For reasons that have not been adequately addressed, the 5’ moiety is derived mostly from HCV genotype 2, while the 3’ part varies [18]. All HCV intergenotypic recombinants identified to date are restricted to a single individual [18]. The only strain that has been shown to circulate is the RF1 2k/1b strain, which was initially identified in Russia and other neighboring countries [5,7,10,19]. However, during the last decade a number of publications have shown the spread of RF1 2k/1b in other European countries, from individuals of Russian or Georgian origin [6-10]. This is the first study to report the circulation of recombinant 2k/1b strains in Greece or the Balkan Peninsula. All patients had lived in Georgia, suggesting an imported infection. The ability of this recombinant strain to spread in the population, especially among intravenous drug users (IDU) underlines the high possibility of the virus spreading into the indigenous population.

There are contradicting indications regarding the response of RF1 strain to IFN therapy. A recent report on 2k/1b strains from Georgia showed lower ratios of SVR (<50%) to pure HCV 1b genotype, possibly due to the 24-week rather than 48-week regimen followed in genotype 1b patients [7]. Besides, the study of Susser et al suggested that genotype 1-based DAAs are very effective in eradication of 2k/1b virus, compared to the standard sofosbuvir plus ribavirin treatment for HCV genotype 2 [20]. Our results indicate that 2k/1b strains show mild resistance to IFN treatment, as both patients 1 and 3 failed to develop SVR after IFN administration. The latter patient managed to develop SVR on his second treatment, which may be associated with the favorable IFNL3 rs12979860 CC background. On the other hand, patient 1 had a genetic background favoring a poor outcome of treatment, according to IFNL3 rs12979860 and IFNL4 rs368234815 genotypes/variants. Therefore, the coexistence of a nonstructural region of 1b genotype, which is more resistant to IFN than genotype 2, and a non-favorable SNP would rather suggest a cumulative effect on treatment outcome. Patient 2, who stopped treatment after 8 weeks, has now agreed to receive DAAs. Thus, given the wide distribution of RF1 2k/1b in certain areas, it is important to check all HCV genotype 2 isolates for the presence of recombination.

In summary, this is the first identification of 2k/1b recombinants in Greece, possibly imported from Georgia. Further analysis of the spread of the strain in the indigenous population of the area of East Macedonia and Thrace, where all three current incidences were reported, will be required, especially within the IDU community.