Detection of hepatitis B virus YMDD variants using mass spectrometric analysis of oligonucleotide fragments

Journal of Hepatology 40 (2004) 837–844 www.elsevier.com/locate/jhep

Detection of hepatitis B virus YMDD variants using mass spectrometric analysis of oligonucleotide fragments Sun Pyo Hong1, Nam Keun Kim2, Seong Gyu Hwang2,3,*, Hyun Jae Chung1, Sukjoon Kim1, Jin Hee Han2, Hyung Tae Kim3, Kyu Sung Rim2,3, Myung Seo Kang4, Wangdon Yoo1, Soo-Ok Kim1 1 GeneMatrix Inc., Cancer Research Institute, Seoul National University College of Medicine, Seoul, South Korea Institute for Clinical Research, Pundang CHA General Hospital, College of Medicine, Pochon CHA University, Seongnam, South Korea 3 Department of Internal Medicine, Pundang CHA General Hospital, College of Medicine, Pochon CHA University, Seongnam, South Korea 4 Department of Laboratory Medicine, Pundang CHA General Hospital, College of Medicine, Pochon CHA University, Seongnam, South Korea 2

Background/Aims: Mutations in hepatitis B virus (HBV) to lamivudine resistance that arise during prolonged treatment frequently cause amino acid substitutions in the YMDD motif of HBV DNA polymerase. Current methods of detecting such variants are time-consuming, labor intensive, and unsuitable for screening large numbers of samples. Here, we describe the development of a matrix-assisted laser desorption/ionization time of flight mass spectrometry (MALDI-TOF MS) genotyping assay suitable for detecting HBV variants in a sensitive and specific manner. Methods: The assay is based on PCR amplification and mass measurement of oligonucleotides containing sites of mutation of the YMDD motif. Results: The MALDI-TOF MS-based genotyping assay is sufficiently sensitive to detect as few as 100 copies of HBV genome per millilitre of serum, with superior specificity for determining mixtures of wild-type and variant viruses. When sera from 40 patients were analyzed, the MALDI-TOF MS-based assay correctly identified known viral variants and additional viral quasi-species not detected by previous methods, as well as their relative abundance. Conclusions: The sensitivity, accuracy and amenability to high-throughput analysis makes the MALDI-TOF MS-based assay suitable for mass screening of HBV infected patients receiving lamivudine, and can help provide further understanding of disease progression and response to therapy. q 2004 European Association for the Study of the Liver. Published by Elsevier B.V. All rights reserved. Keywords: Hepatitis B virus; Lamivudine; YMDD variant; Genotyping; Mass spectrometry 1. Introduction Lamivudine [(2 )-b-L -20 ,30 -dideoxy thiacytidine, 3TC] has revolutionized the treatment of chronic hepatitis B and opened new options for the management of patients with decompensated cirrhosis or recurrent hepatitis B, post liver transplantation [1]. Unfortunately, long-term lamivudine therapy promotes the selection of HBV mutants with altered DNA polymerases resistant to therapy. The altered DNA Received 16 September 2003; received in revised form 14 November 2003; accepted 14 January 2004 * Corresponding author. Address: Department of Internal Medicine, Division of Gastroenterology – Hepatology, Pundang CHA General Hospital, College of Medicine, Pochon CHA University 351, Yatapdong, Pundang-gu, Sungnam-si, Kyonggi-do 463-712, South Korea. Tel.: þ 82-31-780-5213; fax: þ82-31-780-5219. E-mail address: [email protected] (S. Gyu Hwang).

polymerases frequently exhibit changes in a highly conserved tyrosine, methionine, aspartate, and aspartate (YMDD) motif, and so are named YMDD mutants. The most commonly described mutations cause the substitution of valine or isoleucine for methionine at residue 204 (revised nomenclature according to Stuyver et al.) [2 – 4]. Recent studies have shown that in some instances YMDD mutants may be responsible for increased liver damage in 10 – 25% of patients, which can be lifethreatening [5,6]. An additional concern is that individuals infected by HBV YMDD mutants pre-transplant may have an increased risk of recurrent infection post-transplant with reappearance of high level of serum HBV [7,8]. Because the DNA polymerase mutations can be detected before viral breakthrough, advances in antiviral therapy like lamivudine have brought about an increasing need for sensitive and early detection of emerging drug-resistant mutant [9 – 14].

0168-8278/$30.00 q 2004 European Association for the Study of the Liver. Published by Elsevier B.V. All rights reserved. doi:10.1016/j.jhep.2004.01.006

838

S. Pyo Hong et al. / Journal of Hepatology 40 (2004) 837–844

In this study, we established a MALDI-TOF MS-based HBV genotyping assay that exploits differences in molecular weights between HBV wild type and variant oligonucleotides encoding the YMDD motifs in the viral DNA polymerase. We have compared the results of the MALDIMS genotyping assay with DNA sequencing and RFLP assays, and measured the time course of takeover of resistant variants to determine whether they could be detected earlier with the MALDI-TOF MS-based genotyping assay than by RFLP and DNA sequencing assays.

2. Materials and methods 2.1. Patients and sera Sera were collected from 40 hepatitis B patients who received lamivudine (Glaxo-Wellcome, Greenford, UK) therapy for more than 3 months at the Liver Clinic of Pundang CHA General Hospital in Korea between March 1998 and November 2001 (Table 2). None of the patients was positive for either anti-hepatitis C virus antibody or anti-human immunodeficiency virus antibody. Informed consent was obtained form each patient and experimental protocol conformed to the ethical guidelines of the 1975 Declaration of Helsinki as reflected in a priori approval by the Pundang CHA General Hospital human research committee. Hepatitis B surface antigen and hepatitis B e antigen/antibody were determined by enzyme immunoassay (Abbott Diagnostics, Chicago, IL). HBV DNA was measured by the Digene hybrid capture assay (Digene Diagnostics, Beltsville, MD) with a lower limit of 0.5 pg/ml or Cobas Amplicor HBV Monitor Test (Roche Diagnostics, Branchburg, NJ) with a lower limit of 200 copies/ml.

2.2. HBV DNA extraction and PCR amplification HBV DNA was extracted from 200 ml of serum using the QIAamp blood kit (Qiagen, Chatworth, CA) according to the manufacturer’s instructions Two micro litre of the viral DNA was used for the PCR reaction. For MALDI-TOF MS-based genotyping, PCR was performed in 18 ml reaction mixture containing 20 mM Tris –HCl (pH 8.4), 50 mM KCl, 0.2 mM of each dNTP, 10 pmol of each primer, and 0.4 units of Platinumw Taq DNA polymerase (Invitrogen, Carlsbad, CA). The amplification conditions included initial denaturation at 94 8C for 2 min, 10 cycles of denaturation at 94 8C for 15 s, annealing at 50 8C for 15 s and extension at 72 8C for 30 s, followed by 35 cycles of denaturation at 94 8C for 15 s, annealing at 55 8C for 15 s, and extension at 72 8C for 30 s. The sequences of forward and reverse primers used in the PCR were, respectively: 50 TTCCCCCACTGTTTGGCTGGATGTCAGTTAT-30 (nucleotide number 712 – 738) and 50 -TACAGACTTGGCCCCCAATACCACATGA-30 (nucleotide number 771–744). To insert a new Fok I site or to eliminate the naturally occurring Fok I site in the products, sequences underlined in each primer were modified as shown in Fig. 1. To amplify the HBV polymerase gene encoding the YMDD motif for cloning or sequencing analysis, PCR was performed as described in Chayama et al. [15]. Nucleotide sequence positions were numbered according to Ono et al. [16].

Fig. 1. Schematic representation of MALDI-TOF MS-based genotyping strategy. Nucleotide sequences of the PCR-amplified region including the YMDD motif of HBV DNA polymerase gene and PCR primers are shown. A five-nucleotide sequence (GGATG) was inserted in the forward primer (nucleotides 712 –738) to introduce a Fok I site; one nucleotide (T) that is indicated by the blank box was not included to improve hybridization of the primer. To erase the naturally occurring Fok I site (nucleotides 741–745), the backward primer (nucleotides 771 –744) was designed to harbor a single base mismatch (underlined G) at the second position from the 30 -terminal. Cleavage sites of Fok I and BstF5I, an isoschizomer for Fok I, are indicated by filled arrows and blank arrows, respectively, and recognition sites of both restriction endonucleases by shaded box. After completion of the enzyme digestion, the central part of the PCR products should produce 7mer fragment (50 -AGTTATN-30 ) and 13-mer fragment (30 -AGTCAATANANCT-50 ) where the nucleotides denoted by N would represent the variation at codon 204 of the samples in MALDI-TOF MS. An example of the 7-mer (50 -AGTTATA-30 ) and 13-mer (30 -AGTCAATATACCT50 ) resulting from the wild-type virus is shown with the sequence of codon 204 in bold italic. MALDI-TOF MS (Bruker Daltonics Biflex IV, Billerica, MA) workstation in a positive ion, delayed extraction mode.

2.4. Validation of genotyping by MALDI-TOF MS The MALDI-TOF MS assay results were compared with results from a conventional RFLP assay [15]. For confirmation of viral quasispecies, we cloned the PCR products into the pCR-Script Amp cloning vector (Stratagene, La Jolla, CA), for sequence analysis. Sequence analysis was performed by ABI PRISM 310 Genetic Analyzer (Applied Biosystems, New York, NY).

3. Results 3.1. Genotyping strategy

2.3. Genotyping by MALDI-TOF MS Restriction enzyme digestion of PCR products was performed by mixing the PCR reaction mixture with 10 ml of buffer containing 50 mM potassium acetate, 20 mM Tris –acetate, 10 mM magnesium acetate, 1 mM dithiothreitol and 1 unit of Fok I. The reaction mixture was incubated at 37 8C for 2 h and further incubated at 45 8C for 2 h with Bst F5I. The resulting digest was desalted by vacuum filtration through a 384-well sample preparation plate containing 5 mg of polymeric solvent (Waters, Miliford, MA) per well. The desalted reaction mixtures were resuspended with matrix solution containing 50 mg/ml 3-hydroxy picolinic acid, 0.05 M ammonium citrate, and 30% acetonitrile, and were spotted in 3 ml volumes on a polished anchorchip plate. Mass spectra were acquired on a linear

The MALDI-TOF MS assay is based on mass spectrometric analysis of small DNA fragments containing sites of variation (Fig. 1). The first step requires PCR amplification using primers flanking the altered bases. The forward primer (nucleotide 712– 738) was designed to introduce a Fok I site (an isoschizomer of Bst F5I) in the amplified product by substituting the restriction recognition sequence GGATG for T (nucleotide 730). The backward primer (nucleotides 744– 771) was designed to have the second C (nucleotide

S. Pyo Hong et al. / Journal of Hepatology 40 (2004) 837–844

745) from its 30 end replaced by G to erase a naturally occurring Fok I site in HBV virus. Both Fok I and Bst F5I are type IIS restriction enzymes that cleave DNA outside the recognition sequence. The Fok I enzyme cleaves DNA nine bases 30 to the recognition site on one strand and 13 bases from the recognition site on the other strand, leaving a four base overhang protruding 50 end. Bst F5I cleaves DNA 2 bases 30 to the recognition site on one strand and immediately 30 to the recognition site on the opposite strand, leaving a two base overhang. As outlined in Fig. 1, the 7-mer fragment contains the polymorphic base at nucleotide 739 and the 13-mer fragment contains the polymorphic bases at nucleotide 739– 741. These fragments are then analyzed by mass spectrometry to identify the bases at codon 204 (nucleotides 739 –741). 3.2. Sensitive and specific detection of YMDD motif variants We assessed the sensitivity and validity of the MALDITOF MS-based genotyping assay using the DNA polymerase region (nucleotide 411 –886) of wild-type and variant viruses in plasmids whose sequences at codon 204 were known. As shown in Fig. 3A, wild-type virus DNA released two fragments with molecular masses of 2219.6 and 3998.0 which are identical to the expected masses presented in Table 1. Also in case of variants, the YIDD variant produced two peaks corresponding to molecular masses of 2219.6 and 4021.8 while the YVDD variant produced two peaks at mass positions 2215.9 and 3982.9 as predicted (Fig. 3A and Table 1). When assayed with serially diluted plasmids or viruses of known concentration, to determine the lower limit of detection, the MALDI-TOF MS-based genotyping assay was able to detect extremely low amounts of DNA corresponding to 100 copies of HBV genome per millilitre, regardless of whether it was wild type or variant DNA (data not shown). To evaluate the ability to distinguish wild-type and variant DNA in mixed populations, the MALDI-TOF MS-based genotyping assay was performed with samples composed of different ratios of wild type and variant sequences. Various copies of five-fold diluted plasmids containing the nucleotide sequence of the YIDD variant were mixed with plasmids containing the wild-type sequence in the following ratios (1:1, 1:5, 1:52, 1:53, 1:54, 1:55). A total

839

of 102 copies of YIDD DNA were detected in the mixture of 1.25 £ 104 copies of the wild-type DNA (Fig. 2). 3.3. Detection of variant HBV genomes with YMDD motif by MALDI-TOF MS in serum samples HBV DNAs encoding the YMDD motif of 40 chronic hepatitis patients, who showed an HBV DNA breakthrough after prolonged lamivudine treatment, were analyzed by conventional RFLP and the MALDI-TOF MS assay. The results were compared and validated by clonal sequencing analysis (Table 2). Consistent results were obtained in 38 (95%) out of 40 samples when only the predominant genotype identified by each method was analyzed. When the detection of mixed genotypes was considered, the MALDITOF MS and RFLP assays had 100 and 85% agreement with clonal sequencing results, respectively. At the onset of viral breakthrough, YMDD mutations were identified in 36 (90%), and YMDD wild type in 4 (10%) of the 40 patients (Fig. 3A, upper left). As demonstrated by the representative mass spectra (Fig. 3A), 24 (60%) out of the 40 subjects exhibited the YIDD mutation (upper right), eight (20%) of the subjects exhibited the YVDD mutation (lower left), and four (10%) subjects exhibited the mixed YIDD/YVDD mutation (lower right). In six of 40 samples, the MALDITOF MS assay was superior to RFLP for distinguishing mixed populations of viruses or identifying uncharacterized variants. For example, by RFLP assay patients 4 and 21 were shown to harbor only YIDD variant by the RFLP assay, but were shown to have YIDD co-existing with YVDD virus by MALDI-TOF MS assay and clonal sequencing. In four patients out of 40, the MALDI-TOF MS method detected additional viral genotypes of ATC (Patient 7 and 26) or ATA (Patient 11 and 17) mixing with predominant ATT, which could not be detected by RFLP assay. Existence of these genotypes was verified by the clonal analysis using direct sequencing of multiple clones representing PCR products spanning sequences encoding the YMDD motif. The clinical course of a cirrhotic patient who exhibited HBV DNA breakthrough and flare-up of hepatitis with emergence of YMDD variants 10 months after lamivudine treatment is shown in Fig. 4. The lamivudine-resistant HBV variant YIDD (ATT and ATA) along with the wild type

Table 1 Expected and observed masses of oligonucleotides resulting from restriction enzyme digestion of PCR products spanning YMDD motif Genotype

YMDD YVDD YIDD YIDD YIDD a

Codon 204

aTg gTg aTt aTc aTa

Expected fragmentsa

Expected mass (Da)

Observed mass (Da)

7mer

13mer

7mer

13mer

7mer

13mer

AGTTATa AGTTATg AGTTATa AGTTATa AGTTATa

TCcAtATAACTGA TCcAcATAACTGA TCaAtATAACTGA TCgAtATAACTGA TCtAtATAACTGA

2199.4 2215.4 2199.4 2199.4 2199.4

3997.6 3982.6 4021.6 4037.6 4012.6

2199.6 2215.9 2199.6 2199.6 2199.6

3998.0 3982.9 4021.8 4037.6 4012.6

The first and third bases of codon 204 in YMDD motif are noted by a small letter.

840

S. Pyo Hong et al. / Journal of Hepatology 40 (2004) 837–844

Fig. 2. Specific and sensitive detection of HBV YMDD variants by MALDI-TOF MS method. MALDI-TOF MS results from mixed populations of HBV wild-type and YIDD variant plasmids are shown. One hundred copies of plasmid with the nucleotide sequence of YIDD variant were serially mixed with plasmid containing the wild-type sequence at various ratios as following; (A) 1:1, (B) 1:5, (C) 1:52, (D) 1:53, (E) 1:54, (F) 1:55. The wild-type plasmid released two fragments with molecular masses of 2199.6 Da and 3998.0 Da in mass spectrum and YIDD variant plasmid produced two peaks corresponding to molecular masses of 2199.6 and 4021.8 Da. Because of the same mass of the smaller fragment, only the mass spectra of the larger fragment are shown here. AI and m=z represent actual peak intensity and mass to charge ratio, respectively.

virus were first detected by the MALDI-TOF MS method 4 months after initiation of therapy, which is 6 months ahead of viral breakthrough; mutant virus was detected by RFLP and direct sequencing in samples taken at 2 months before breakthrough (Fig. 3B, upper left). At 5 months post therapy, the wild-type virus disappeared and only the YIDD variants (ATT and ATA) were detected (Fig. 3B, upper right). One month later, the YVDD variant was transiently detected (Fig. 3B, lower left). Thereafter, the YIDD variants persisted even though HBV DNA remained below the detection limit of the commercial HBV DNA quantification assay (Cobas Amplicor HBV Monitor Test; Roche Diagnostics, Branchburg, NJ) during up to 7 months of therapy. The YIDD variant (ATT) became predominant and was detected at 8 months post therapy, 2 months before breakthrough, until 12 months post therapy when the patient stopped taking medication. One month after cessation of therapy, a wild-type virus started to re-emerge and a mixed YMDD/YIDD genotype was detected (Fig. 3B, lower right). Two months after cessation of therapy, only the wild-type virus remained without detectable YMDD variants. These changes in the viral populations were also confirmed by clonal analysis of the PCR product. Using RFLP only the

YIDD variant (ATT) was detected 2 months before breakthrough (Fig. 4). It did not show the spasmodic emergence of the YVDD and other YIDD (ATC, ATA) variants during the period of low-level viral load before breakthrough, nor the trailing variant virus overlapped with re-emerging wild-type virus 2 months after medication stopped.

4. Discussion HBV variants in the YMDD motif have been associated with reduced susceptibility to lamivudine [17,18] Particularly for patients at high risk for disease progression, it is important to detect these variants early and precisely before the emergence of viral breakthrough when variant viruses represent only a minor fraction of the total viral population. Presently, the detection of HBV variants is largely performed by sequencing analysis, RFLP and hybridization-based assays [15,19,20]. Sequencing can give information on the majority species present in the viral populations, but generally cannot detect species comprising fewer than 15 –50% of a viral population [21]. Thus it is

S. Pyo Hong et al. / Journal of Hepatology 40 (2004) 837–844

841

Table 2 Genotyping results of YMDD motif in chronic hepatitis B patients with an HBV DNA breakthrough Patient

1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40

Breakthrough (months)

7 18 10 11 16 16 13 11 12 16 14 24 6 18 6 6 7 18 24 12 5 18 11 11 16 16 13 11 12 16 14 24 6 18 6 6 7 18 24 12

HBV-DNA (pg/ml)

105 120 50 720 68 25 1173 263 464 3294 620 732 2305 574 180 195 62 309 221 638 150 230 400 105 304 250 240 140 20 45 3030 650 55 120 340 180 550 230 460 55

ALT (IU/l)

34 273 220 37 108 51 55 102 21 36 137 50 28 40 245 19 57 143 92 154 65 85 100 40 120 108 95 80 133 36 450 328 28 35 240 32 110 156 85 95

HBeAg

þ þ þ þ þ þ þ þ þ þ þ þ þ þ þ þ þ þ þ þ þ þ þ þ þ þ þ þ þ þ þ þ þ þ þ þ þ þ þ þ

HBeAb

– – – – – – – – – – – – – – – – – – – – – – – – – – – – – – – – – – – – – – – –

Codon 204 in YMDD motif MALDI-TOF MSa

RFLP

Clonal sequencingb

I V I I, V I I I (ATT, ATC) I I V I (ATT, ATA) I V I, V I I V M I M I, V I I V I I (ATT, ATC) I (ATT, ATA) M I V V I M I, V I I V I I I

I V I I I I I I I V I I V I, V I I V M I M I, V I I V I I I M I V V I M I, V I I V I I I

ATT GTG ATT ATT, ATT ATT ATT, ATT ATT GTG ATT, ATT GTG ATT, ATT ATT GTG ATG ATT ATG ATT, ATT ATT GTG ATT ATT, ATT, ATG ATT GTG GTG ATT ATG ATT, ATT ATT GTG ATT ATT ATT

GTG (17, 3)

ATC (17, 3)

ATA (15, 5)

GTG (12, 8)

GTG (18, 2)

ATC (18, 2) ATA (15, 5)

GTG (13, 7)

a

Isoleucine quasispecies observed in codon 204 of YMDD motif by MALDI-TOF MS method is presented in parentheses. Number of clones with corresponding genotypes is shown in parentheses and no parenthesis indicates single genotype results obtained from 20 independent clones. b

necessary to analyze multiple clones representing viral quasispecies for determining the heterogeneity of a population. In particular, an rtM204V/rtM204I variant mixture could be incorrectly scored as rtM204V variant/wild type mixture by direct sequencing even after visual inspection. This is due to the inability of the sequencing software to determine the correct variants present in the sample when the nucleotide mixture at position 1 (R ¼ A or G) and position 3 (K ¼ T or G) of codon 204 are present [22]. Assays based on RFLP and hybridization, have been contributing to the understanding of the occurrence of the mutant HBV strains. However, such assays are time consuming, labor intensive, and are not suitable for screening

large number of samples, as they require multiple DNA amplifications and enzyme digestions or complex hybridization steps. The MALDI-TOF MS genotyping method we describe here provides extra levels of accuracy for YMDD motif genotyping: Firstly, a 13-mer fragment with the correct mass can only be generated by specific PCR amplification around a region of the HBV DNA polymerase gene including YMDD motif since the 13-mer fragment is a 5 base-extended product spanning the codon 204 from the end nucleotide of a forward PCR primer. Secondly, a 7-mer fragment can only be simultaneously generated with a 13-mer by restriction digestion. In addition to being very

842

S. Pyo Hong et al. / Journal of Hepatology 40 (2004) 837–844

Fig. 3. Representative MALDI-TOF MS results of YMDD motif in patient samples. Panel A: MALDI-TOF MS results from various types of YMDD motifs. Upper left, mass spectrum of wild-type virus with the nucleotide sequence of ATG for amino acid Met in the YMDD motif from patient 18. Upper right, mass spectrum of YIDD variant with the nucleotide sequence of ATT for amino acid Ile from patient 1. Lower left, mass spectrum of YVDD variant with the nucleotide sequence of GTG for amino acid Val from patient 9. Lower right, mass spectrum of mixed population of YIDD and YVDD variants from patient 2. Panel B: MALDI-TOF MS results from follow-up samples from patient 3. Upper left, mass spectrum after 4 months treatment of lamivudine showing the wild-type virus along with the two types of YIDD variants containing the nucleotide sequence of ATT or ATA for amino acid Ile. Upper right, detection of the two types of YIDD variants in the absence of the wild-type virus after 5 months treatment. Lower left, emergence of another types of YIDD variant with the nucleotide sequence of ATC for amino acid Ile and YVDD variant with the nucleotide sequence of GTG for amino acid Val after 6 months treatment. Lower right, re-emergence of the wild-type YMDD motif after 1 month from the cessation of lamivudine treatment at 12 months post-therapy. AI and m=z represent actual peak intensity and mass to charge ratio, respectively.

accurate the MALDI-TOF MS genotyping system is also sensitive, being able to detect as few as 100 HBV copies/ml and is more specific than DNA sequencing and hybridization-based assays in determining the ratios of wild-type to variant virus (Fig. 2). A clear correlation was observed

between peak ratios and relative genotype concentration of different YMDD motif-containing plasmids in mixed populations. The performance of the MALDI-TOF MS genotyping method for detecting YMDD variants was compared with

S. Pyo Hong et al. / Journal of Hepatology 40 (2004) 837–844

Fig. 4. Longitudinal data of serum ALT levels and HBV DNA viral load with the genotyping results of HBV YMDD motif in a patient with lamivudine-resistant HBV infection. Clinical course of patient 3 who showed an HBV DNA breakthrough and flare-up of the hepatitis with emergence of the YMDD variants 10 months after lamivudine treatment. Lamivudine treatment was stopped at 12 months post therapy. The duration of the lamivudine treatment is shown with blank box and genotyping results by both MALDI-TOF MS, RFLP, and direct sequencing assays on each tme points were shown with codon 204 in YMDD motif. (2) indicates no result.

that of RFLP and clonal sequencing analysis, using samples of 40 chronically infected patients at differing stages of HBV breakthrough after prolonged lamivudine therapy (Table 2). Consistent results were obtained by the MALDITOF MS genotyping method and RFLP for YMDD motif genotypes at codon 204 in 38 out of 40 (95%) samples. In six samples (15%), mixed viral genotypes, some of which contained ATA or ATC genotype encoding isoleucine, were successfully detected by the MALDI-TOF MS, indicating that MALDI-TOF MS genotyping method is more sensitive than RFLP in detecting the complex quasispecies of HBV during lamivudine treatment. The ATC or ATA variants have been previously reported by Cane et al. [23] in a liver transplanted patient with lamivudine resistance and by Stuyver et al. [20] in clonal analysis of patients with viral breakthrough. Although ATC or ATA variations lead to conversion of tryptophan to serine or to a stop codon at codon 196 in the overlapping reading frame of surface antigen, respectively, clinical differences between these variants are currently unknown. It is interesting to note that in none of the patient samples in our study does an ATC or an ATA variant other than ATT appear as the single major variant, even though the ATT variant was the predominant type and was detected as a single major species in 24 of 40 patient samples. A possible explanation for this observation could be that ATC and/or ATA variants are defective and

843

thus require co-infection with other types of virus that can provide more intact surface antigen. In a longitudinal study of a patient during early stages of lamivudine therapy from 0 to 15 months, we could detect quasispecies of YMDD variants mixed with wild-type virus even though HBV DNA remained below the detection limit of the commercial HBV DNA quantification assay during the 120-day window period before viral breakthrough (Fig. 4). As lamivudine treatment continued, the major species of virus were replaced by YMDD variants and wild type virus could not be detected. At the end of the study, the single YMDD variant (YIDD) became the predominant and almost sole virus two months before viral breakthrough, when viremia started to arise sharply (Fig. 4). In most patients examined, the YMDD variant predominance, defined by 25-times higher amount of variant compared to wild type, seems to be linked with phenotypic resistance and pure variants without the presence of wild type virus preceded viral breakthrough by several months. Taken together with previous finding that patients with HBV DNA breakthrough had higher percentages of YMDD variants without wild type presence compared with patients without HBV DNA breakthrough [24], the predominance of variant virus need to be further tested as a predictable marker for viral breakthrough. This should be facilitated by the capacity of the MALDI-TOF MS assay to show relative abundance among viral species. In conclusion, the MALDI-TOF MS genotyping method utilizing mass difference of oligonucleotides requires simple steps of single PCR amplification and restriction enzyme digestion, and is amenable to high-throughput system. The MALDI-TOF MS method has practical advantages over both RFLP assay and direct DNA sequencing, including the ability to more sensitively and specifically detect mixed populations of the wild-type and variant viruses as well as being a simpler procedure without need for cloning or multiple PCR steps, enabling earlier detection of variant virus. The MALDI-TOF MS genotyping assay is easily adaptable for the detection of other mutations or polymorphisms. Therefore, we believe that the MALDI-TOF MS genotyping assay will be most useful to evaluate the kinetics of emergence of variant viruses, and to correlate them to clinical consequences of response to drug therapy. With the development of new antiviral options, we expect our genotyping assays along with viral quantification will be clearly beneficial for optimal monitoring of antiviral therapy of chronic hepatitis B.

Acknowledgements We gratefully acknowledge the valuable assistance of Dr William Folk at the University of Missouri (Missouri, USA) for revising the manuscript.

844

S. Pyo Hong et al. / Journal of Hepatology 40 (2004) 837–844

References [1] Merle P, Trepo C. Therapeutic management of hepatitis B-related cirrhosis. J Viral Hepat 2001;8:391 –399. [2] Allen MI, Deslauriers M, Andrews CW, Tipples GA, Walters KA, Tyrrell DL, et al. Identification and characterization of mutations in hepatitis B virus resistant to lamivudine. Hepatology 1998;27: 1670–1677. [3] Han JH, Hwang SG, Chung KW, Oh SW, Hong SP, Park PW, et al. YMDD motif mutants in hepatitis B virus polymerase during lamivudine therapy. Korean J Genet 2002;24:219– 226. [4] Stuyver LJ, Locarnini SA, Lok A, Richman DD, Carman WF, Dienstag JL, et al. Nomenclature for antiviral-resistant human hepatitis B virus mutations in the polymerase region. Hepatology 2001;33:751 –757. [5] Liaw YF, Chien RN, Yeh CT, Tsai SL, Chu CM. Acute exacerbation and hepatitis B virus clearance after emergence of YMDD motif mutation during lamivudine therapy. Hepatology 1999;30:567–572. [6] Wang JH, Lu SN, Lee CM, Lee JF, Chou YP. Fatal hepatic failure after emergence of the hepatitis B virus mutant during lamivudine therapy in a patient with liver cirrhosis. Scand J Gastroenterol 2002; 37:366–369. [7] Bartholomew MM, Jansen RW, Jeffers LJ, Reddy KR, Johnson LC, Bunzendahl H, et al. Hepatitis-B-virus resistance to lamivudine given for recurrent infection after orthotopic liver transplantation. Lancet 1997;349:20 –22. [8] Chen RY, Edwards R, Shaw T, Colledge D, Delaney 4th WE, Isom H, et al. Effect of the G1896A precore mutation on drug sensitivity and replication yield of lamivudine-resistant HBV in vitro. Hepatology 2003;37:27 –35. [9] Papatheodoridis GV, Dimou E, Papadimitropoulos V. Nucleoside analogues for chronic hepatitis B: antiviral efficacy and viral resistance. Am J Gastroenterol 2002;97:1618 –1628. [10] Niesters HG, de Man RA, Pas SD, Fries E, Osterhaus AD. Identification of a new variant in the YMDD motif of the hepatitis B virus polymerase gene selected during lamivudine therapy. J Med Microbiol 2002;51:695–699. [11] Zhang M, Gong Y, Osiowy C, Minuk GY. Rapid detection of hepatitis B virus mutations using real-time PCR and melting curve analysis. Hepatology 2002;36:723– 728. [12] Lok AS, Zoulim F, Locarnini S, Mangia A, Niro G, Decraemer H, et al. Monitoring drug resistance in chronic hepatitis B virus (HBV)infected patients during lamivudine therapy: evaluation of performance of INNO-LiPA HBV DR assay. J Clin Microbiol 2002; 3729–3734. [13] Kirishima T, Okanoue T, Daimon Y, Itoh Y, Nakamura H, Morita A, et al. Detection of YMDD mutant using a novel sensitive method in

[14]

[15]

[16]

[17]

[18]

[19]

[20]

[21]

[22]

[23]

[24]

chronic liver disease type B patients before and during lamivudine treatment. J Hepatol 2002;37:259–265. Paik YH, Chung HY, Ryu WS, Lee KS, Lee JS, Kim JH, et al. Emergence of YMDD motif mutant of hepatitis B virus during shortterm lamivudine therapy in South Korea. J Hepatol 2001;35:92 –98. Chayama K, Suzuki Y, Kobayashi M, Kobayashi M, Tsubota A, Hashimoto M, et al. Emergence and takeover of YMDD motif mutant hepatitis B virus during long-term lamivudine therapy and re-takeover by wild type after cessation of therapy. Hepatology 1998;27: 1711–1716. Ono Y, Onda H, Sasada R, Igarashi K, Sugino Y, Nishioka K. The complete nucleotide sequences of the cloned hepatitis B virus DNA; subtype adr and adw. Nucleic Acids Res 1983;1747–1757. Ahmed SS, Tavan D, Pichoud C, Berby F, Stuyver L, Johnson M, et al. Early detection of viral resistance by determination of hepatitis B virus polymerase mutations in patients treated by lamivudine for chronic hepatitis B. Hepatology 2000;32:1078–1088. Yeh CT, Chien RN, Chu CM, Liaw YF. Clearance of the original hepatitis B virus YMDD-motif mutants with emergence of distinct lamivudine-resistant mutants during prolonged lamivudine therapy. Hepatology 2000;31:1318 –1326. Allen MI, Gauthier J, DesLauriers M, Bourne EJ, Carrick KM, Baldanti F, et al. Two sensitive PCR-based methods for detection of hepatitis B virus variants associated with reduced susceptibility to lamivudine. J Clin Microbiol 1999;37:3338–3347. Stuyver L, Van Geyt C, De Gendt S, Van Reybroeck G, Zoulim F, Leroux-Roels G, et al. Line probe assay for monitoring drug resistance in hepatitis B virus-infected patients during antiviral therapy. J Clin Microbiol 2000;38:702–707. Pas SD, de Man RA, Fries E, Osterhaus AD, Niesters HG. The dynamics of mutations in the YMDD motif of the hepatitis B virus polymerase gene during and after lamivudine treatment as determined by reverse hybridisation. J Clin Virol 2002;25:63–71. Aberle SW, Kletzmayr J, Watschinger B, Schmied B, Vetter N, Puchhammer-Stockl E. Comparison of sequence analysis and the INNO-LiPA HBV DR line probe assay for detection of lamivudineresistant hepatitis B virus strains in patients under various clinical conditions. J Clin Microbiol 2001;39:1972–1974. Cane PA, Mutimer D, Ratcliffe D, Cook P, Beards G, Elias E, et al. Analysis of hepatitis B virus quasispecies changes during emergence and reversion of lamivudine resistance in liver plantation. Antiviral Ther 1999;4:7–14. Yuen MF, Sablon E, Hui CK, Yuan HJ, Decraemer H, Lai CL. Factors associated with hepatitis B virus DNA breakthrough in patients receiving prolonged lamivudine therapy. Hepatology 2001;34: 785 –791.