Hepatitis C Virus Genotyping Based on 5′ Noncoding Sequence Analysis (Trugene)

Abstract

Hepatitis C virus (HCV) genotyping of samples from 184 patients with chronic HCV infection by the Trugene 5′NC genotyping kit, based on sequence analysis of the 5′ noncoding region (5′ NCR), and the InnoLiPA assay was evaluated. In addition to these methods, the 184 samples were also analyzed by sequencing of part of the NS5B of the HCV genome after in-house PCR amplification, as a means of validating results obtained with the 5′ NCR. The distribution of the genotypes typed by NS5B sequence analysis was as follows: 1a, 41 samples; 1b, 58 samples; 1d, 1 sample; 2a, 5 samples; 2b, 2 samples; 2c, 7 samples; 3a, 46 samples; 4a, 7 samples; 4c, 1 samples; 4e, 9 samples; 5a, 6 samples; 6a, 1 sample. The Trugene and InnoLiPA assays gave concordant results within HCV types in 100% of cases. The ability to discriminate at the subtype level was 76 and 74% for the Trugene and the InnoLiPA assays, respectively.

Hepatitis C virus (HCV) is considered the major cause of posttransfusion non-A, non-B hepatitis. The viral genome, a positive-sense single-stranded RNA of about 9,400 nucleotides (5), is characterized by a high genetic heterogeneity like other RNA viruses. HCV isolates show four levels of genetic variability: types, subtypes, isolates, and quasispecies (3). An HCV genotype is therefore used with the histological results from liver biopsy and viral load for counseling individual patients about the risk-benefit ratio of therapy (17, 21, 22). HCV genotypes are distributed differently depending on geography and the etiology of infection (15, 25). For the purpose of nomenclature, it has been proposed that HCV be classified into types, corresponding to the main branches in the phylogenetic tree, and subtypes, corresponding to the more related sequences within the major groups (23, 24). HCV genotypes can be established by methods based on PCR typing and/or serological typing (2, 4, 16). The high degree of conservation in the 5′ noncoding regions (5′ NCR) has made it the target of choice for reverse transcriptase PCR-based detection assays. Moreover, several PCR typing methods, such as reverse dot blot (26), restriction fragment length polymorphism (10), cleavase fragment length polymorphism, dideoxy fingerprinting, heteroduplex mobility analysis (28), and hybridization to genotype-specific probes (18), exploit sequence-based differences and/or differences in secondary structure of the 5′ NCR for HCV genotyping (9). Nucleotide sequence analysis is the reference method for identifying different genotypes of HCV (9, 13). However, because this method is expensive and time-consuming and requires special equipment for sequencing, it has been restricted to the research setting and considered impractical for large clinical studies. A standardized sequencing assay has recently been developed for routine determination of HCV genotypes. In order to determine whether direct sequencing could be a routine tool for the determination of hepatitis C virus genotype, we assessed this first commercial type of HCV genotyping method (Trugene 5′NC HCV genotyping kit; Visible Genetics, Toronto, Canada), based on sequence analysis of the 5′ NCR, with samples from 184 patients with chronic HCV infection, and we compared the data with data from the most widely used genotyping method based on the reverse dot blot method, the InnoLiPA assay. In addition to these methods, the 184 patient samples were also subjected to amplification and sequence analysis of a region within the NS5B gene as a means of validating results obtained with the 5′ NCR.

(Part of this work was presented in abstract form at the AASLD symposium in Dallas in 1999.)

MATERIALS AND METHODS

The serum specimens analyzed in this study were obtained from 184 consecutive patients with histologically proven chronic HCV infection, monitored in two hepatology units in Marseilles and Bordeaux, France. All the patients were HCV RNA positive as determined by PCR with a commercial detection kit (Amplicor HCV; Roche Molecular Systems, Neuilly, France) (29). HCV RNA was quantified in sera with the nonisotopic branched-DNA signal amplification method (version 2.0; Bayer Corporation, Emeryville, Calif.). Patient viral load ranged from 0.5 × 10⁶ to 50 × 10⁶ eq/ml. In an effort to minimize sample degradation from multiple freeze-thaw cycles, all samples were thawed on receipt, aliquoted, and stored at −70°C prior to testing (11).

HCV genotypes were determined by two HCV typing methods using the 5′ NCR of the genome: the Trugene 5′NC HCV genotyping kit method and the InnoLiPA test. Both tests were carried out after RNA extraction, reverse transcription, and amplification using the Amplicor HCV RNA assay (Roche Diagnostic Systems, Neuilly, France) according to the manufacturer's instructions. The amplified product was analyzed using a reverse dot blot assay (InnoLiPA) according to the manufacturer's instructions (26) and by the Trugene 5′NC genotyping kit. The minimum viral load needed for all typing tests to work ranges from 100 to 1,000 copies/ml.

Trugene 5′NC assay.

Bidirectional DNA sequencing of the amplification products was performed using CLIP, a sensitive sequencing method developed by Visible Genetics Inc. Each sequencing reaction was loaded on a long read tower (Visible Genetics Inc.), an automated DNA sequencer. The resulting sequence for each sample was then compared to a database containing known HCV isolates using the Clustal W method (8). A phylogenetic analysis of a 196-bp segment (nucleotides −256 to −70) of the 5′ NCR was performed on all of the sequences generated in this study (19). Prototype sequences obtained from GenBank were included in this analysis.

All of the samples were also analyzed by sequence analysis of the NS5B gene after home brew PCR amplification. Serum RNA was extracted by the acid-guanidinium-phenol-chloroform method. Briefly, 50 μl of serum was homogenized with 150 μl of RNAzol (Leedo Laboratories, Houston, Tex.) and was extracted with chloroform. The RNA was then isopropanol precipitated. Next, the RNA was washed with 75% ethanol, dissolved in 10 μl of diethyl pyrocarbonate-treated water and stored at −80°C until use. cDNA synthesis was carried out using random hexamer oligonucleotides. Then cDNA was amplified in a single reaction with primers thought to be highly conserved among different isolates of HCV (12). A second PCR was carried out using multiple primer sets, published by Simmonds et al. (23). The resulting PCR product of 222 bp was subsequently purified by phenol-chloroform extraction and precipitated by ethanol. The purified product of 222 bp was subsequently purified and bidirectionally sequenced by the dideoxynucleotide method with the automated Clipper sequencer (Visible Genetics). Amplification of negative samples with PCR in the NS5B region produced no sequence. To avoid contamination, all analyses were performed following Kwok and Higuchi's recommendations (14).

RESULTS

The NS5B sequence analysis was used as a reference test for the accuracy of the genotyping. The distribution of the genotypes typed by NS5B sequence analysis was as follows: 1a, 41 samples; 1b, 58 samples; 1d, 1 sample; 2a, 5 samples; 2b, 2 samples; 2c, 7 samples; 3a, 46 samples; 4a, 7 samples; 4c, 1 sample; 4e, 9 samples; 5a, 6 samples; and 6a, 1 sample. All the samples were successfully typed by the HCV Trugene 5′NC genotyping kit and with the InnoLiPA assay.

Accuracy was defined as the number of correct genotype samples from the Trugene 5′NC genotyping kit and InnoLiPA assays divided by the total number of genotypes determined by the NS5B sequence analysis. The overall accuracies of the two tests regardless of the genotype were 76% for the Trugene 5′NC HCV and 74% for the InnoLiPA assays.

Among the samples classified as genotype 1a by NS5B sequence analysis, 31 of 41 (76%) were in agreement with the two 5′ NCR typing assays, and 10 of 41 (24%) were discordant or incompletely identified by these two assays.

Among the samples classified as genotype 1b by NS5B sequence analysis, 55 of 58 (95%) were in agreement with the Trugene 5′NC genotyping kit, versus 53 of 58 (91%) with the InnoLiPA assay; 3 of 58 (5%) and 5 of 58 (9%) were discordant or incompletely identified by, respectively, the Trugene 5′NC genotyping kit and the InnoLiPA assays. The only sample typed 1d by NS5B sequence analysis was typed 1a by the two 5′ NCR typing assays.

Among the samples classified as genotype 2a, 2b, or 2c by NS5B sequence analysis, 1 of 5 (20%), 1 of 2 (50%), and 1 of 7 (15%), respectively, were in agreement with the Trugene 5′NC genotyping kit, while 4 of 5 (80%), 1 of 2 (50%), and 6 of 7 (85%) were discordant or incompletely identified by the Trugene 5′NC genotyping kit. One sample identified as type 2b by NS5B sequence analysis was correctly identified using the InnoLiPA assay, while 13 of 14 (91%) were discordant or incompletely identified.

Among the 46 samples classified as genotype 3a by NS5B sequence analysis, 44 (96%) were in agreement with the Trugene 5′NC genotyping kit and the InnoLiPA assay, while 2 (4%) were discordant or incompletely identified by Trugene 5′NC genotyping kit and the InnoLiPA assays.

For the 17 samples classified as genotype 4a (seven samples), 4c (one sample), and 4e (nine samples) by NS5B sequence analysis, neither of the two assays was able to correctly identify the subtype.

For the six samples classified as genotype 5a and the only sample classified 6a by NS5B sequence analysis, both tests were able to identify the correct subtype.

Lower accuracy was observed with the two assays due to the highly conserved nature of the 5′ NCR, particularly for subtypes 2 and 4. The InnoLiPA test was less discriminate than the Trugene assay in discriminating subtype 2 isolates. However, this difference was not statistically significant.

DISCUSSION

The Trugene assay is the first generation of direct sequencing tests which provide complete sequence information to characterize HCV genotypes. However, because these methods are expensive and time-consuming and require special equipment for sequencing, they have been restricted to the research setting and considered impractical for clinical diagnosis. The time required for the Trugene assay is comparable (less than 5 h, including only 30 min of sequence analysis) to that for the InnoLiPA assay, which is one of the more commonly used genotyping assays worldwide. The sequencer is more compact than other currently available sequencers, and the cost of the genotyping assay (including the sequencer) is generally comparable. The Trugene assay may be used in a routine clinical laboratory, like the InnoLiPA assay, after amplification with the Amplicor assay to genotype all the HCV isolates. The Trugene assay does not require an additional specimen-processing step and utilizes products obtained from a single, nonnested amplification reaction, thus eliminating delays and the risks of carryover contamination but without incorporation of dUTPs and uracyl-N glycosidase. Other methods, such as restriction fragment length polymorphism analysis, are able to characterize genotypes in 93% of the cases (10).

As determined by NS5B sequence analysis, which is the reference method for discriminating nucleotide sequence variation, the accuracies of the Trugene and InnoLiPA assays were, respectively, 74 and 76%. Our data on sequence analysis of the 5′ NCR were confirmed by Germer et al. (9). In their study, 89.4% of the specimens were successfully classified after computer-assisted analysis of the sequence data.

The limitations of the Trugene 5′NC assay are related to the low discriminating power of 5′ NCR for determination of particular types or subtypes (27). The high level of conservation found in this region cannot discriminate subtypes, as in the case with subtypes 2a and 2c. The performance of the direct sequencing method is similar to that of the InnoLiPA Test, with relative failure to subtype the type 2 genotype (27). There are also examples in which only one or two minor nucleotide changes distinguish unique subtypes from each other. Two examples of this situation are illustrated by the minor differences seen in the 5′ NCR sequences of subtypes 1a and 1b: a single base change at position −99 (adenine to guanine) is the only change, and the thymine-to-cytosine polymorphism at position −94 describes genotypes 1a and 1b.

A crucial assumption of all genotyping assays is that the region analyzed (5′ NCR, core, E1, NS4, and NS5B) is representative of the genome as a whole. This assumption would break down if recombination between HCV genotypes occurred during replication, producing hybrid viruses containing contributions from different genotypes in different parts of the genome (3). This phenomenon, well established for human immunodeficiency virus, has yet to be demonstrated for HCV.

The direct sequencing of amplification products provides more detailed sequence information than genotyping assays based on hybridization, heteroduplex mobility analysis, single-strand conformation polymorphism, or restriction analysis. This additional information could prove to be quite useful in the detection of new viral types or to demonstrate nosocomial, sporadic, or interfamilial transmission by comparing the sequences (1, 6, 7, 20).

In conclusion, considering the importance of genotyping in hepatitis C treatment management, the Trugene HCV assay may be considered reliable for differentiating between all HCV types and could be used as a routine tool for the determination of HCV genotypes. This test provided a sensitive and efficient means of HCV genotyping in a clinical setting, particularly in light of studies which show that the clinically relevant distinction is only between genotypes.