Detection of Hepatitis C Virus by Polymerase Chain Reaction and Response to Interferon-cx Therapy: Relationship to Genotypes of Hepatitis C Virus KENTAE~O YOSHIOKA,'SHINICHI KAKUMu,' TAKAJI WAKITA,'TETSWAISHIKAWA,' YUJI IT OH,^ HI GAS HI,^ MOTOHIROSHIBATA' AND TSUNEOMORISHIMA'
MASAHIRO TAKAYANAGI,~ YASUYUKI
'Third Department of Internal Medicine and 2Department of Pediatrics, Nagoya University School of Medicine, Nagoya 466, Japan
To investigate the relationshipbetween genotypes of hepatitis C virus and response to interferon-atherapy, hepatitisC virus RNA was assayed by polymerase chain reaction with three sets of primers and probes in 70 patients with non-A, non-B chronic hepatitis who received interferon-a.Twenty-four patients sustained long-term remissions (complete responders). Polymerase chain reaction for 5'-terminal noncoding region detected hepatitis C virus RNA in 94.3% (66 of 70) of the patients. Polymerase chain reaction for nonstructural region 3, in which primers and a probe were synthesized to be identical to hepatitis C virus-J, detected hepatitis C virus RNA in 40 patients. Polymerase chain reaction for nonstructural region 5-in which sequences of primers and a probe were derived from hepatitis C virus-K2, a genotype different from hepatitis C virus4 - detected hepatitis C virus RNA in 17 patients. Only one patient was positive on both nonstructural region 3 and nonstructural region 5 polymerase chain reaction. Nucleotide sequence of clones obtained from 5' terminal noncoding region polymerase chain reaction products of two patients positive on polymerase chain reaction for nonstructural region 3 and negative on polymerase chain reaction for nonstructural region 5 (group 1) corresponded to that of the hepatitis C virus-J group, and those of clones from two patients negative on polymerase chain reaction for nonstructural region 3 and positive on polymerase chain reaction for nonstructurd region 5 (group 2) corresponded to that of hepatitis C virus-K!Z. A clone from a patient negative on polymerase chain reaction for nonstructural region 3 and polymerase chain reaction for nonstructural region 6 ( w u p 3) showed low nucleotide sequence homology with the hepatitis C virus4 and hepatitis C virus-K2groups. The complete response rates of group 2 (10 of 16 [62.5%]) and group 3 (6 of 10 [60.0%1) were signiscantly higher than that of group 1 (5 of 39 [ 12.8%]) (p < 0.01 for both). Logarithms of hepatitis C virus RNA concentrations (copies per milliliter) were Received October 21, 1991; accepted March 16, 1992. A d k reprint requests to: Kentaro Yoshioka, M.D., Third Department of Internal Medicine, Nagoya University School of Medicine, 65 Tsuruma-cho, Showa-ku, Nagoya 466, Japan.
31/1/38110
significantly higher in group 1 (5.0 f 1.1) than in ?up 2 (3.8 f 1.1) or group 3 (3.2 f 1.1) (p < 0.01 for either). These results indicate that detection of hepatitis C virus RNA by polymerase chain reactions with different sets of primers and probes may be valuable in classifying hepatitis C virus into genotypes, and that amount of hepatitis C virus RNA in sera and response to interferon-amay vary among different genotypes of
HCV. (HEPATOLOGY
1992;16:293-299.)
Hepatitis C virus (HCV), a small, positive-stranded RNA virus, is considered a major causative agent of blood-borne non-A, non-B (NANB)hepatitis (1, 2). Recently, partial and entire sequence data of various HCV isolates have been reported, suggesting several genotypes of HCV (3-9). In addition, it has become possible to detect HCV RNA by the polymerase chain reaction (PCR)with various sets of primers derived from these sequence data (10-13). Although interferons (IFNs) have been shown to be effective in NANB hepatitis (14-161,they are not equally efficacious for all patients with the disease. So far no clinical, biochemical, serological or histological features have been identified that predict response to IFN-a therapy. We have investigated the relationship between response to IFN-a therapy and genotypes of HCV, and we present here data suggesting that detection of HCV RNA by PCR might be useful in classifying HCV into genotypes and that the response to IFN-a is closely associated with genotype of HCV. MATERLALS AND METHODS Patients and ZFN--rr Therapy. Seventy consecutive patients with NANB hepatitis who received recombinant human 1FN-aza(IFN-a) (provided by Nippon Roche Inc.) from June 1987 to December 1990 in our hospital were investigated retrospectively. NANB hepatitis was dmgnosed after exclusion of infection with known viruses, including hepatitis A and B, cytomegalovirus, Epstein-Barr virus and human immunodeficiencyvirus. Patients classified as having NANB hepatitis did not have autoimmune hepatitis, drug-induced hepatitis or alcoholic liver disease. Diagnosis of hepatitis was based on appropriate biochemical and histological criteria (17). Liver biopsies were performed in all but two patients; specimens
293
YOSHIOKA ET AL.
294
I
2
3
HEPATOLOGY
4
5x104 5x103 5x102 5x10
5
5
copies
FIG.1.Quantitation of HCV RNA by competitive PCR. Agarose-gel electrophoresis of "nested" PCR products showed 159-bpbands (products from competitor DNA with a deletion) in lanes 1to 3 and 193-bp bands (products from normal HCV cDNA) in lanes 4 and 5. The amount of competitor DNA added to the samples was 5 x lo4 copies for lane 1 , 5 x lo3 copies for lane 2,5 x lo2 copies for lane 3 , 5 x 10 copies for lane 4 and 5 copies for lane 5. Because the gel shows that the amount of HCV cDNA is smaller than that of competitor DNA in lane 3 (5 x 10' copies) and larger in lane 4 (5 x 10 copies), the amount of HCV cDNA in the sample is considered to be approximately 100 copies. TABLE1. Characteristics of the patients before treatment Interferon dose 2 million
Characteristics
Age (Yr)" Sex (FM)
Disease duration
units (n = 13)
50 & 12 2111 7.3 f 5.1
W"
Blood transfusion
815
4 million units
8 million
(n = 20)
units (n = 37)
51 ? 11 8/12 5.3 t 6.5
48 t 14 15/22 4.2 f 4.7
1614
23/14
(-I+)
Serum levelsb ALT (IUIL)" y-GTP (IUL)" y-Globulin (gmldl)" Anti-HCV ( +I - 1" HAIc
305 ? 203 186 ? 299 1.5 ? 0.4 1310 11.1 ? 2.0
214 ? 151 69 t 48 1.7 f 0.5 1515 10.4 2 3.4
274 ? 126 88 ? 64 1.6 2 0.5 2918 9.8 f 3.7
y-GTP = y-transpeptidase. "Data expressed as mean f S.D. bNormal values: 0 to 45 IUIL for ALT, 0 to 40 IUL for y-GTP and 0.76 to 1.69 gmldl for y-globulin. 'Liver biopsy specimens were not available for two patients. were assessed by two of us (K.Y. and S.K.) in double-blind fashion and graded with the histological activity index (HAI) described by Knodell and coworkers (18). Patients received intramuscular injections of IFN-a in various doses 3 timeslwk for 16 wk: 13 patients received 2 million unitslday, 20 received 4 million unitslday and 37 received 6 million unitslday (Table 1). All patients gave informed consent, and the trials were approved by the hospital committee on human ethics. Patients whose ALT levels normalized during therapy or within a month after therapy and remained so for more than 6 mo after therapy were regarded as complete responders. Patients whose ALT levels remained more than 1.5 times the upper limit of normal range during IFN-a administration were regarded as nonresponders. Those whose ALT levels fell below 1.5 times the upper limit of normal range during therapy but did not remain normal for more than 6 mo were considered partial responders. Antibody to HCV. Antihody to (anti-HCV) was assayed in pretreatment serum by ELISA (Ortho Diagnostic Systems, Raritan, NJ) (2). H C V R N . . HCV RNA was detected in pretreatment sera by PCR. Serum RNA was extracted, reverse-transcribed, am-
plified and hybridized with probes. Three sets of oligonucleotide primers and probes were synthesized (Table 2). One pair of primers (Al, A2) and a probe (probe A) were derived from the 5'-terminal noncoding region of HC-Jl (4) cloned in Japan and used to amplify a 221-bp sequence or to hybridize with it (5'-NC-PCR).The second pair of primers (Bl, B2) and a probe (probe B) were derived from the putative nonstrudural region 3 of HCV-J (3) and used to amplify a 458-bp sequence or to hybridize with it (NS3-PCR). The third pair of primers (Cl, C2) and a probe (probe C) were derived from the putative nonstructural region 5 of HCV-K2 (8),which is a genotype different from HCV prototype (6) or HCV-J, and used to amplify a 231-bp sequence or to hybridize with it (NS5-PCR). RNA was prepared from 100 ~1 serum by treating it with guanidium isothiocyanate buffer containing 4.0 mom guanidium thiocyanate, 0.1 m o m Tris-HC1 (pH7.5), 1% 2-mercaptoethanol and 0.5% sodium lauryl sarcosinate. This was centrifuged through a 5.7 m o m CsCVO.01 m o m EDTA cushion at 140,000 g for 16 hr in an SW 41 rotor (Beckman Instruments, Palo Alto, CA). RNA pellets were resuspended in water, and PCR was then performed. Complementary DNA (cDNA) was synthesized with avian myeloblastosis virus reverse transcriptase (Bethesda Research Laboratories, Gaithersburg, MD) with an antisense primer. After a sense primer and Taq polymerase (2.5 U; Perkin-Elmer Cetus, Norwalk, CT) were added, 40 cycles of PCR were performed as follows: denaturation for 1minute at 94" C, annealing for 2 min at 55" C and extension for 3 min at 72" C. A 1 0 - ~ aliquot 1 of PCR products was subjected to agarose-gelelectrophoresis. The gels were stained with ethidium bromide and examined for the presence of bands of the expected size. Then the gels were blotted onto Nytran (Schleicher & Schuell, Dassel, Germany), and the blots were hybridized with a probe conjugated with alkaline p h o s p h a t e in 5 x standard saline citrate (1x standard saline citrate = 0.15 moVL NaCl and 0.015 m o m sodium citrate) containing 5 x Denhardt's solution and 0.5% SDS a t 50" C. Conjugationwas performed with an E-LINK oligonucleotide labeling kit (Cambridge Research Biochemicals, Cheshire, UK). The blots were washed in 1x standard saline citrate and 0.5% SDS at 50" C, sprayed with Lumi-Phos substrate (CambridgeResearch Biochemicals) and exposed to autoradiograph films. When the band of expected size hybridized to the probe, it was considered positive. Sensitivity of our PCR method was examined with a plasmid containing a partial sequence of the 5'-NC region. To avoid false positivity, we applied the contamination-prevention measures of Kwok and Higuchi (19). All PCRs were run and analyzed in a blinded fashion. Qurrntitation of HCVRNA by Competitive PCR. HCV RNA concentrations were measured by competitive PCR in 63 of
Vol. 16,No. 2, 1992
295
HCV GENOTYPES AND IFN-aTHERAPY TABLE2. Synthetic oligonucleotidesequences
Primem and probe
5'-NC-PCR A1 A2 Probe A
N S3-PCR B1 PTb B2 PT Probe B PT N S5-P CR c1 PT c2 PT Probe C
PT
Origin of Bequence
Sequence (6'-3')
C A C A
T A C C
G C G C
T A G G
G A G G A A C T A C T G T C T T C T A C T C G G C T A G C A G T G A G A G C C A T A G T G G T C T G C G G A G
T G T C A C T C A G A C G G T C G A T T T C A G C _ _ . _ C_ __ . . . _ A _ _ _ _ _ _ _ A C T T C C A C A T T T G A T C C C A C G A T G G _ _ _ _ _ _ _ _ _ _ C - - G_ . . _ . . _ _ C T G T G C T T G G T A T G A G C T C A .
.
.
.
T C C G -A C A G G - -
A C -
.
_
T A - C A T - -
.
.
.
.
.
_
_
.
.
_
.
.
Product size
HC-J1(4) -297-- -278 221 bp - 9 6 - -77 -213%- -183
HCV-J(3) 4371--4395 _
_
.
_
_
_
_
458 bp
_
4804-- 4828 4572--4591
_
T A T C A G G C C T G C T C C C T G C - A T G T - - T G A - - - C G T G A G G G T G C A A C T A T C C C T G - T - C A G T - C T G G -G - - - -
HCV-K2(8) 7972--7996
231 bp
- - C -
G C C C T T G C T G T T G A A C A T G G C - - - C - T G A A - - - G T A - G A ~~
_
Nucleotide positions"
~~~~
817%- 8202 805%- 8072
~~
"Nucleotide positions are numbered according to HCV prototype sequence (6). *PT-HCVprototype sequence. Al,A2 and probe-A sequences are identical to HCV prototype. B1 and B2 sequences differ from HCV prototype a t two nucleotide positions as shown. Probe B sequence is identical to HCV prototype. C1,C2 and probe C sequences differ from HCV prototype at 13,12 and 11 nucleotide positions, respectively. C1,C2 and probe C sequences also differ from two strains of HCV-J group (5,7)at 12 to 15 nucleotide positions.
TABLE3. Characteristicsof the patients before treatment 66 5'-NC-PCR-positive patients (20). After the usual reverse transcription with primer A2, 1:10 volumes of the according to their response to IFN-a cDNA were mixed with serial dilutions of a competitor Complete Partial DNA that differed from the cDNA of interest by a small responders responders Noxiresponders Characteristics (n = 24) (n = 32) (n = 14) deletion (34 bases). Samples were amplified by "nested" PCRS with two sets of primers (A1 and A2 for outer Age (yr)" 45 t 14' 50 -t 12 54 t 9 primers; D1, 5'-CGCAGAAAGCGTCTAGCCAT-3', and D2, Sex ( F M ) 7/17 14/18 4/10 5'-ACTCGGCTAGCAGTCTCGCG-3', for inner primers) and Disease duration 4.8 -t 6.6 4.2 k 4.0' 7.5 t 5.4' the same reaction mixture as described above. Competitor (yr)" DNA with deletion was produced from a cloned DNA from Blood transfusion 14/10 22/10 1113 5'-NC-PCR of one patient in group 2. For both PCRs, 35 ( i t ) cycles of PCR were performed as follows: denaturation for Serum levelsd 1 min at 94" C, annealing for 1 min at 55" C and extension ALT (IUL)" 274 t 151 232 t 129 313 t 189 for 1 min at 72" C. Both PCRs were performed in 50-p.1 y-GTP (IUL)" 73 t 43 79 t 66 201 t 282 mixtures, and 1 k.1 of the first-PCR product was transferred y-Globulin (gmfdl)" 1.6 -t 0.4 1.6 2 0.5 1.7 2 0.5 to the second-PCR mixture. "Nested" PCR detected as few anti-HCV ( + I - ) 2913 1113 1717 as five copies of cloned HCV DNA. HCV RNA concentrations 12.0 t2 9 9.3 t 3.0' 10.2 t 3.5 HAI score", were determined from the results of agarose gel electrophoresis of PCR products (Fig. 1). y-GTP = y-glutamyltranspeptidase. Cloning and Sequencing. Nucleotide sequences of a part of "values are mean t SD. 5'-NC region between primer A1 and A2 (nucleotide -277 to hThemean age of complete responders was significantly lower than around nucleotide -97) were determined in six patients (two that of nonresponders (p < 0.01). patients from group 1,two from group 2, one from group 3 and 'Duration of disease in partial responders was significantly shorter the NS3-PCR and NSB-PCR-positive patient). 5'-NC-PCR than that in nonresponders (p < 0.01) products were cloned into the SmaI site of plasmid dNormalvalues: 0 to 45 I U L for ALT, 0 to 40 IUL for y-GTP and pGEM7Zf( + )(Promega Biotec, Madwon, WI).These clones 0.76to 1.69gm/dl for y-globulin. were sequenced by the dideoxy chain-termination method. 'Liver biopsy specimens were scored for histological activity by Statistical Analysis. Data were analyzed by the x2 test and Knodell's index (HAI).Specimens were not available for two patients. Student's t test. fThe mean HAI score of complete responders was significantly
RESULTS ZFN-ru Thempy. The response to IFN-a was related to the dose. Thus 16 of 37 (43.2%)patients treated with 6 million units responded, compared with 7 of 20 (35.0%) treated with 4 million units and 1 of 13 (7.7%)treated with 2 million units. When pretreatment characteristics were compared
lower than that of nonresponders (p < 0.05).
among the patient groups, age and HAI scores of complete responders (45 k 14 yr, 9.3 & 3.0; mean 2 S.D.) were significantly lower than those of nonresponders (54 +- 9 yr, 12.0 -+ 2.9; [p < 0.01, age, and p < 0.05, MI) The . duration of disease in partial
YOSHIOKA ET AL.
296
HEPATOLOGY
-PllTlENT 1
PATIENT 2
FIG.2. Detection of HCV RNA by PCRs in three patients. Patient 1 was positive on 5'-NC-PCR and NS3-PCR and negative on NS5-PCR. Patient 2 was positive on 5'-NC-PCR and NS5-PCR and negative on NS3-PCR. Patient 3 was positive on all three PCRs.
TABLE 4. Pattern of results of PCRs and response to IFN-u therapy Group
Characteristics
5'-NC-PCR response NS3-PCR response NS5-PCR response No. of patients (%) No. of complete responders (rate of complete responders) No. of partial responders No. of nonresponders ~
_____
1
2
3
+ +
-
+
+ -
+ + +
-
l(1.4) 1 (100) 0 0
39 (55.7) 5 (12.8) 25 9
+
-
16 (22.9) 10" (62.5) 4
10 (14.3) 6' (60.0)
2
2
2
4 (5.7) 2 (50.0) 1 1
_______~
+ = double infection with two genotypes; - = disease caused by different agent from HCV. "Group 2 had significantly more complete responders than did group 1 (p < 0.01). bGroup 3 had significantly more complete responders than did group 1 (p < 0.01).
responders (4.2 2 4.0 yr) was significantly shorter than that in nonresponders (7.5 2 5.4 yr) (p < 0.01) (Table 3). However, no factors differentiated complete responders from partial responders. HCVRNA. Anti-HCV was present in 81.4% (57 of 70) of the patients. HCV RNA was present in 94.3% (66 of 70) as assayed by 5'-NC-PCR. 5'-NC-PCR detected as few as 500 copies of a cloned sequence by ethidium bromide staining, and as few as five copies were detected by hybridization with probe A. All 5'-NC-PCR-negative patients were also negative on NS3-PCR or NS5-PCR, and all but one were also negative for anti-HCV. HCV RNA was present in 57.1% (40 of 70) of the patients assayed by NS3-PCR and in 24.3% (17 of 70) assayed by NS5-PCR. In three patients who had no band with ethidium bromide staining in NS5-PCR, a specific 231-bp band was found after hybridization with probe C. Only one patient was positive on both NS3-PCR and NS5-PCR, and four were negative on all three PCRs. Thus patients could be divided into five groups according to the PCR results (Table 4, Fig. 2). The three main groups were 39 NS3-PCR-positive and NS5-PCRnegative patients (group l ) , 16 NS3-PCR-negative and NS5-PCR-positive patients (group 2) and 10 NS3-PCRnegative, NS5-PCR-negative, 5'-NC-PCR-positive patients (group 3). When pretreatment characteristics were compared among the three groups, the number of
patients with histories of blood transfusion was significantly lower in group 2 (6.3%, 1of 16) than in group 1 (38.5%, 15 of 39) (p < 0.05) and in group 3 (50.0%, 5 of 10) (p < 0.05). HCV RNA and Responses to Therapy. Response rate to IFN-a varied according to PCR group. The best response was noted in group 2 (10 of 16,63%),whereas group 3 patients had a slightly lower response (6 of 10, 60%). The response in both of these groups was significantly better than that in group 1 (5 of 39, 13%) (p < 0.01) (Table 4). It may be more appropriate to exclude patients treated by the lower dose in evaluating the response to treatment. Therefore we also analyzed 57 patients treated with 4 million or 6 million units. The response in group 2 (10 of 14 [71.4%1)and in group 3 (5 of 8 [62.5%])was significantly better than that in group 1 ( 5 of 30 [16.7%1) (p < 0.011. These findings were identical to those for the complete set of 70 patients. Quantitation of HCV RNA by Competitive PCR. The range of HCV RNA concentrations in serum of 63 patients before therapy was lo2 to lo8 copies per milliliter. Mean k S.D. of logarithms of HCV RNA concentrations was 4.5 f 1.3. Logarithms of HCV RNA concentrations were significantly higher in group 1 (5.0 & 1.1) than in group 2 (3.8 2 1.1)and group 3 (3.2 f 1.1)(p < 0.01 for both) (Fig. 3). Logarithms of HCV RNA concentrations were significantly lower in
-
HCV GENOTYPES AND IFN-CX THERAPY
Vol. 16, No. 2, 1992
complete responders (3.8 2 1.4) than in partial responders (4.7 & 1.2) and nonresponders (5.0 ? 1.2) (p < 0.05 for both). Cloning and Sequencing. Sequences of 181 nucleotides (nucleotide - 277 to nucleotide - 97) were analyzed (Fig. 4). Nucleotide sequences of two clones from two patients in group 1 (patient 1 and patient 2) corresponded completely with that of HC-J4 (4) and differed from that of HCV prototype at only one nucleotide. Nucleotide sequence of the clone from a patient of group 2 (Patient 3) showed 97.8% homology with HCV-K2a, 92.8% with HCV-J4 and 93.4% with HCVprototype. The clone from another patient in group 2 (Patient 4) corresponded completely with that of HCV-K2a. The clone froin a patient of group 3 (Patient 5 ) showed 93.4% homology with HCV-J4, 93.9% homology with HCV prototype and 95.0% homology with HCV-K2a. The clone from the NS3-PCR-positive and NS5-PCR-positive patient (Patient 6) showed 92.3% homology with HCV-J4, 92.8% homology with HCV prototype and 96.1% homology with HCV-K2a. DISCUSSION
HCV RNA was detectable in 94.3% (66 of 70) of the patients included in this study on 5‘-NC-PCR. The sequence of the 5’-NC region is considered the most conserved among the different strains (4). Therefore it is naturally expected that 5’-NC-PCR will detect HCV RNA at a higher rate than will the other two PCRs. On the other hand, considerable sequence variability exists in NS3 and NS5 reeons among the different strains. It is probably the reason why NS3-PCR and NS5-PCR detected HCV-RNA less frequently than did 5’-NC-PCR. Houghton et al. (21) classified HCV into three genotypes: HCV-I (HCV prototype group), HCV-I1 (HCV-J group) and HCV-I11 (HCV-K2 group). Because the sequences of primers and a probe used for NS3-PCR are derived from HCV-J, NS3-PCR will detect both HCV-J and HCV prototype, whose nucleotide sequence is similar to that of HCV-J. In fact, the nucleotide sequences of clones from two patients in group 1 (NS3-PCR-positive and NS5-PCR-negative patients) were shown to be identical to HC-J4 (HCV-J group). The sequences of primers and a probe used for NS5-PCR are derived from HCV-K2 and differ considerably from HCV prototype or HCV-J. Therefore NS5-PCR will detect HCV-K2 but not HCV prototype or HCV-J. In fact, the nucleotide sequences of clones from two patients in group 2 (NS3-PCR-negative and NS5PCR-positive patients) were shown to be identical or similar to those in the HCV-K2a (HCV-K2) group. It seems that HCV-K2 significantly differs from the sequences of primers and a probe used in NS3-PCR and is not detected by our NS3-PCR. In practice, only one patient was positive for both NS3-PCR and NS5-PCR. These results indicate that group 1 patients may be infected with the HCV prototype group or the HCVJ group and group 2 patients may be infected with the HCV-m group. A clone from a patient in group 3 showed low
297
**
**
m
o complete responder 0 partial rsspondsr x non-responder
X
.... ..:.... ....
f
oxn
000
onx 0..
0.
X
00
oonnnn
1.00 on n
T
.O
f
.on
.O
000
0
000
group 1
group 2
group 3
(37)
(15)
(9)
FIG. 3. HCV RNA concentrations in the three groups, group 1 (NS3-PCR positive and NS5-PCR-negative patients), group 2 (NS3PCR-negative and NSS-PCR-positive patients) and group 3 (5‘-NCPCR-positive, NS3-PCR-negative, NSS-PCR-negative patients). Logarithms of HCV RNA concentrations were significantly higher in group 1 (5.0t 1.1) than in group 2 (3.8 k 1.1) and in group 3 (3.2 -+ 1.1) (p < 0.01 for both). Logarithms of HCV RNA concentrations were sigdicantly lower in complete responders (3.6 2 1.4) than in partial responders (4.7 t 1.2) and in nonresponders (5.0f 1.2) ( p < 0.05 for both).
nucleotide homology with either HCV prototype, H C J 4 (HCV-J group) or HCV-ma. Therefore HCV of group 3 probably differs from that of any of HCV prototype, HCV-J or HCV-K2. Nucleotide sequence of a clone from the NS3-PCR- and NS5-PCR-positive patient resembled that of HCV-K2a but differed from that of HCV prototype or HCV-J. It is probable that the other NS3-PCR-positive strain was also present in the serum of this patient. Garson et al. reported that their “nested” PCR method detected as few as 0.5 copies of HCV sequence (12). Our 5’-NC-PCR detected as few as five copies of HCV sequence. In addition, high specificity was obtained by hybridization with nonradioactive probes. PCR with primers from NS3 and NS5 detected HCV RNA in serum samples at concentrations of lo2 copies/ml (10 copiesltest sample). Therefore it is believed that their sensitivity was as high as that of 5’-NC-PCR. Our “nested” PCR also detected as few as five copies of HCV sequence.
298
YOSHIOKA ET AL.
HEPATOLOGY
FIG.4. Sequences of 181nucleotides in 5’-NC region (nucleotide - 277 to nucleotide - 97) of six patients. Nucleotide sequences of two clones from Patient 1 and Patient 2 corresponded completely with that of HC-J4 (HCV-J group) and differed from that of HCV prototype only a t one nucleotide. Nucleotide sequence of the clone from Patient 3 showed 97.8% homology with HCV-K2a, 92.8% homology with H C J 4 and 93.4% homologywith HCVprototype. The clone from Patient 4 corresponded completelywith that of HCV-Wa. The clone from Patient 5 showed 93.4% homology with H C J 4 , 93.9% homology with HCV prototype and 95.0% homology with HCV-K2a. The clone from Patient 6 showed 92.3% homology with HC-J4, 92.8% homology with HCV prototype and 96.1% homology with HCV-K2a.
So far no clinical, biochemical, serological or histological features have been identified to predict response to IFN treatment (13-15).This study also demonstrated that there were no clinical or biochemical features to differentiate complete responders from partial responders or nonresponders reliably before treatment, although younger age, lower HAI score and shorter duration of disease were the factors that predicted better response to IFN-a therapy. We showed that group 2 (putative HCV-K!2 group) yielded significantly more complete responders than did group 1 (putative HCV prototype and HCV-J group). This finding indicates that the HCV prototype group or HCV-J group is more resistant to IFN-a and that the HCV-K!2 group is more responsive. At the same time, HCV RNA concentrations in group 2 were significantly lower than those in group 1. This suggests that the HCV-K2 group is less able to replicate in humans and thus may be more responsive to IFN-a therapy. In fact, pretreatment HCV RNA concentrations in complete responders were significantly lower than in the others. The different HCV RNA concentrations among genotypes may be attributed to the difference in susceptibility to immunological response of the host to the virus or in replicative capacity of the virus by itself. From these results, we could suppose that the genotype resistant to IFN-a has stronger virulence and thus causes more severe hepatitis. A large-scale study on the relationship between severity of disease and genotypes is under way in our laboratory. Alternatively, it was shown that group 2 patients had lower rates of history of blood transfusion than did group 1 or 3 patients. Further study is necessary to clarify the relationship between genotypes and routes of transmission.
It was also shown that group 3 had a significantly larger number of complete responders than did group 1 and that HCV RNA concentrations in group 3 patients were significantly lower than those in group 1patients. This may suggest that another HCV genotype (different from any HCV prototype, HCV-J or HCV-K2) is also responsive to IFN-a and has a lower replicative capacity. There may be still the possibility that these findings are due to differential sensitivity in the PCRs using different primers and probes. To rule out that possibility, it will be necessary to test the sensitivity of the assay with all 3 sets of primers and also to determine the sequences of amplified NS3 and NS5 regions. Sequencing is under way in our laboratory. In conclusion, our study suggests that PCR is valuable in defining the genotypes of HCV and predicting the response to IFN-a therapy and that the amount of HCV RNA in sera and response to IFN-a therapy may vary among different genotypes of HCV. REFERENCES 1. Choo QL, Kuo G, Weiner AJ, Overby LR, Bradley DW, Houghton M. Isolation of a cDNA clone derived from a blood-borne non-A, non-B viral hepatitis genome. Science 1989;244:1-3. 2. Kuo G, Choo QL, Alter HJ, Gitnick GL, Redeker AG, Purcell RH, Miyamura T, et al. An assay for circulating antibodies to a major etiologic virus of human non-A, non-B hepatitis. Science 1989; 244:362-364. 3. Kubo Y, Takeuchi K, Boonmar S, Katayama T, Choo Q-L,Kuo G, Weiner AJ, et al. A cDNA fragment of hepatitis C virus isolated from an implicated donor of post-transfusion non-A, non-B hepatitis in Japan. Nucleic Acids Res 1989;17:10367-10372. 4. Okamoto H, Okada S, Sugiyama Y, Yotsumoto Y, Tanaka T, Yoshizawa H, Tsuda F, et al. The 5’-terminal sequence of the hepatitis C virus genome. Jpn J Exp Med 1990;60:167-177. 5. Kato N, Hijikata M, Ootsuyama Y, Nakagawa M, Ohkoshi S,
Vol. 16, No. 2, 1992
H C V G E N O T W E S AND IFN-n THERAPY
Sugyama T , Shimotohno K, e t a]. Molecular cloning of the human hepatitis C virus genome from dapanese patients with non-A. non-B hepatitis. Proc Natl Acad Sci LISA 1990:87:9524-9526. 6. Choo QL, Richman KH. Han JII. Berger I
MASAHIRO TAKAYANAGI,~ YASUYUKI
'Third Department of Internal Medicine and 2Department of Pediatrics, Nagoya University School of Medicine, Nagoya 466, Japan
To investigate the relationshipbetween genotypes of hepatitis C virus and response to interferon-atherapy, hepatitisC virus RNA was assayed by polymerase chain reaction with three sets of primers and probes in 70 patients with non-A, non-B chronic hepatitis who received interferon-a.Twenty-four patients sustained long-term remissions (complete responders). Polymerase chain reaction for 5'-terminal noncoding region detected hepatitis C virus RNA in 94.3% (66 of 70) of the patients. Polymerase chain reaction for nonstructural region 3, in which primers and a probe were synthesized to be identical to hepatitis C virus-J, detected hepatitis C virus RNA in 40 patients. Polymerase chain reaction for nonstructural region 5-in which sequences of primers and a probe were derived from hepatitis C virus-K2, a genotype different from hepatitis C virus4 - detected hepatitis C virus RNA in 17 patients. Only one patient was positive on both nonstructural region 3 and nonstructural region 5 polymerase chain reaction. Nucleotide sequence of clones obtained from 5' terminal noncoding region polymerase chain reaction products of two patients positive on polymerase chain reaction for nonstructural region 3 and negative on polymerase chain reaction for nonstructural region 5 (group 1) corresponded to that of the hepatitis C virus-J group, and those of clones from two patients negative on polymerase chain reaction for nonstructural region 3 and positive on polymerase chain reaction for nonstructurd region 5 (group 2) corresponded to that of hepatitis C virus-K!Z. A clone from a patient negative on polymerase chain reaction for nonstructural region 3 and polymerase chain reaction for nonstructural region 6 ( w u p 3) showed low nucleotide sequence homology with the hepatitis C virus4 and hepatitis C virus-K2groups. The complete response rates of group 2 (10 of 16 [62.5%]) and group 3 (6 of 10 [60.0%1) were signiscantly higher than that of group 1 (5 of 39 [ 12.8%]) (p < 0.01 for both). Logarithms of hepatitis C virus RNA concentrations (copies per milliliter) were Received October 21, 1991; accepted March 16, 1992. A d k reprint requests to: Kentaro Yoshioka, M.D., Third Department of Internal Medicine, Nagoya University School of Medicine, 65 Tsuruma-cho, Showa-ku, Nagoya 466, Japan.
31/1/38110
significantly higher in group 1 (5.0 f 1.1) than in ?up 2 (3.8 f 1.1) or group 3 (3.2 f 1.1) (p < 0.01 for either). These results indicate that detection of hepatitis C virus RNA by polymerase chain reactions with different sets of primers and probes may be valuable in classifying hepatitis C virus into genotypes, and that amount of hepatitis C virus RNA in sera and response to interferon-amay vary among different genotypes of
HCV. (HEPATOLOGY
1992;16:293-299.)
Hepatitis C virus (HCV), a small, positive-stranded RNA virus, is considered a major causative agent of blood-borne non-A, non-B (NANB)hepatitis (1, 2). Recently, partial and entire sequence data of various HCV isolates have been reported, suggesting several genotypes of HCV (3-9). In addition, it has become possible to detect HCV RNA by the polymerase chain reaction (PCR)with various sets of primers derived from these sequence data (10-13). Although interferons (IFNs) have been shown to be effective in NANB hepatitis (14-161,they are not equally efficacious for all patients with the disease. So far no clinical, biochemical, serological or histological features have been identified that predict response to IFN-a therapy. We have investigated the relationship between response to IFN-a therapy and genotypes of HCV, and we present here data suggesting that detection of HCV RNA by PCR might be useful in classifying HCV into genotypes and that the response to IFN-a is closely associated with genotype of HCV. MATERLALS AND METHODS Patients and ZFN--rr Therapy. Seventy consecutive patients with NANB hepatitis who received recombinant human 1FN-aza(IFN-a) (provided by Nippon Roche Inc.) from June 1987 to December 1990 in our hospital were investigated retrospectively. NANB hepatitis was dmgnosed after exclusion of infection with known viruses, including hepatitis A and B, cytomegalovirus, Epstein-Barr virus and human immunodeficiencyvirus. Patients classified as having NANB hepatitis did not have autoimmune hepatitis, drug-induced hepatitis or alcoholic liver disease. Diagnosis of hepatitis was based on appropriate biochemical and histological criteria (17). Liver biopsies were performed in all but two patients; specimens
293
YOSHIOKA ET AL.
294
I
2
3
HEPATOLOGY
4
5x104 5x103 5x102 5x10
5
5
copies
FIG.1.Quantitation of HCV RNA by competitive PCR. Agarose-gel electrophoresis of "nested" PCR products showed 159-bpbands (products from competitor DNA with a deletion) in lanes 1to 3 and 193-bp bands (products from normal HCV cDNA) in lanes 4 and 5. The amount of competitor DNA added to the samples was 5 x lo4 copies for lane 1 , 5 x lo3 copies for lane 2,5 x lo2 copies for lane 3 , 5 x 10 copies for lane 4 and 5 copies for lane 5. Because the gel shows that the amount of HCV cDNA is smaller than that of competitor DNA in lane 3 (5 x 10' copies) and larger in lane 4 (5 x 10 copies), the amount of HCV cDNA in the sample is considered to be approximately 100 copies. TABLE1. Characteristics of the patients before treatment Interferon dose 2 million
Characteristics
Age (Yr)" Sex (FM)
Disease duration
units (n = 13)
50 & 12 2111 7.3 f 5.1
W"
Blood transfusion
815
4 million units
8 million
(n = 20)
units (n = 37)
51 ? 11 8/12 5.3 t 6.5
48 t 14 15/22 4.2 f 4.7
1614
23/14
(-I+)
Serum levelsb ALT (IUIL)" y-GTP (IUL)" y-Globulin (gmldl)" Anti-HCV ( +I - 1" HAIc
305 ? 203 186 ? 299 1.5 ? 0.4 1310 11.1 ? 2.0
214 ? 151 69 t 48 1.7 f 0.5 1515 10.4 2 3.4
274 ? 126 88 ? 64 1.6 2 0.5 2918 9.8 f 3.7
y-GTP = y-transpeptidase. "Data expressed as mean f S.D. bNormal values: 0 to 45 IUIL for ALT, 0 to 40 IUL for y-GTP and 0.76 to 1.69 gmldl for y-globulin. 'Liver biopsy specimens were not available for two patients. were assessed by two of us (K.Y. and S.K.) in double-blind fashion and graded with the histological activity index (HAI) described by Knodell and coworkers (18). Patients received intramuscular injections of IFN-a in various doses 3 timeslwk for 16 wk: 13 patients received 2 million unitslday, 20 received 4 million unitslday and 37 received 6 million unitslday (Table 1). All patients gave informed consent, and the trials were approved by the hospital committee on human ethics. Patients whose ALT levels normalized during therapy or within a month after therapy and remained so for more than 6 mo after therapy were regarded as complete responders. Patients whose ALT levels remained more than 1.5 times the upper limit of normal range during IFN-a administration were regarded as nonresponders. Those whose ALT levels fell below 1.5 times the upper limit of normal range during therapy but did not remain normal for more than 6 mo were considered partial responders. Antibody to HCV. Antihody to (anti-HCV) was assayed in pretreatment serum by ELISA (Ortho Diagnostic Systems, Raritan, NJ) (2). H C V R N . . HCV RNA was detected in pretreatment sera by PCR. Serum RNA was extracted, reverse-transcribed, am-
plified and hybridized with probes. Three sets of oligonucleotide primers and probes were synthesized (Table 2). One pair of primers (Al, A2) and a probe (probe A) were derived from the 5'-terminal noncoding region of HC-Jl (4) cloned in Japan and used to amplify a 221-bp sequence or to hybridize with it (5'-NC-PCR).The second pair of primers (Bl, B2) and a probe (probe B) were derived from the putative nonstrudural region 3 of HCV-J (3) and used to amplify a 458-bp sequence or to hybridize with it (NS3-PCR). The third pair of primers (Cl, C2) and a probe (probe C) were derived from the putative nonstructural region 5 of HCV-K2 (8),which is a genotype different from HCV prototype (6) or HCV-J, and used to amplify a 231-bp sequence or to hybridize with it (NS5-PCR). RNA was prepared from 100 ~1 serum by treating it with guanidium isothiocyanate buffer containing 4.0 mom guanidium thiocyanate, 0.1 m o m Tris-HC1 (pH7.5), 1% 2-mercaptoethanol and 0.5% sodium lauryl sarcosinate. This was centrifuged through a 5.7 m o m CsCVO.01 m o m EDTA cushion at 140,000 g for 16 hr in an SW 41 rotor (Beckman Instruments, Palo Alto, CA). RNA pellets were resuspended in water, and PCR was then performed. Complementary DNA (cDNA) was synthesized with avian myeloblastosis virus reverse transcriptase (Bethesda Research Laboratories, Gaithersburg, MD) with an antisense primer. After a sense primer and Taq polymerase (2.5 U; Perkin-Elmer Cetus, Norwalk, CT) were added, 40 cycles of PCR were performed as follows: denaturation for 1minute at 94" C, annealing for 2 min at 55" C and extension for 3 min at 72" C. A 1 0 - ~ aliquot 1 of PCR products was subjected to agarose-gelelectrophoresis. The gels were stained with ethidium bromide and examined for the presence of bands of the expected size. Then the gels were blotted onto Nytran (Schleicher & Schuell, Dassel, Germany), and the blots were hybridized with a probe conjugated with alkaline p h o s p h a t e in 5 x standard saline citrate (1x standard saline citrate = 0.15 moVL NaCl and 0.015 m o m sodium citrate) containing 5 x Denhardt's solution and 0.5% SDS a t 50" C. Conjugationwas performed with an E-LINK oligonucleotide labeling kit (Cambridge Research Biochemicals, Cheshire, UK). The blots were washed in 1x standard saline citrate and 0.5% SDS at 50" C, sprayed with Lumi-Phos substrate (CambridgeResearch Biochemicals) and exposed to autoradiograph films. When the band of expected size hybridized to the probe, it was considered positive. Sensitivity of our PCR method was examined with a plasmid containing a partial sequence of the 5'-NC region. To avoid false positivity, we applied the contamination-prevention measures of Kwok and Higuchi (19). All PCRs were run and analyzed in a blinded fashion. Qurrntitation of HCVRNA by Competitive PCR. HCV RNA concentrations were measured by competitive PCR in 63 of
Vol. 16,No. 2, 1992
295
HCV GENOTYPES AND IFN-aTHERAPY TABLE2. Synthetic oligonucleotidesequences
Primem and probe
5'-NC-PCR A1 A2 Probe A
N S3-PCR B1 PTb B2 PT Probe B PT N S5-P CR c1 PT c2 PT Probe C
PT
Origin of Bequence
Sequence (6'-3')
C A C A
T A C C
G C G C
T A G G
G A G G A A C T A C T G T C T T C T A C T C G G C T A G C A G T G A G A G C C A T A G T G G T C T G C G G A G
T G T C A C T C A G A C G G T C G A T T T C A G C _ _ . _ C_ __ . . . _ A _ _ _ _ _ _ _ A C T T C C A C A T T T G A T C C C A C G A T G G _ _ _ _ _ _ _ _ _ _ C - - G_ . . _ . . _ _ C T G T G C T T G G T A T G A G C T C A .
.
.
.
T C C G -A C A G G - -
A C -
.
_
T A - C A T - -
.
.
.
.
.
_
_
.
.
_
.
.
Product size
HC-J1(4) -297-- -278 221 bp - 9 6 - -77 -213%- -183
HCV-J(3) 4371--4395 _
_
.
_
_
_
_
458 bp
_
4804-- 4828 4572--4591
_
T A T C A G G C C T G C T C C C T G C - A T G T - - T G A - - - C G T G A G G G T G C A A C T A T C C C T G - T - C A G T - C T G G -G - - - -
HCV-K2(8) 7972--7996
231 bp
- - C -
G C C C T T G C T G T T G A A C A T G G C - - - C - T G A A - - - G T A - G A ~~
_
Nucleotide positions"
~~~~
817%- 8202 805%- 8072
~~
"Nucleotide positions are numbered according to HCV prototype sequence (6). *PT-HCVprototype sequence. Al,A2 and probe-A sequences are identical to HCV prototype. B1 and B2 sequences differ from HCV prototype a t two nucleotide positions as shown. Probe B sequence is identical to HCV prototype. C1,C2 and probe C sequences differ from HCV prototype at 13,12 and 11 nucleotide positions, respectively. C1,C2 and probe C sequences also differ from two strains of HCV-J group (5,7)at 12 to 15 nucleotide positions.
TABLE3. Characteristicsof the patients before treatment 66 5'-NC-PCR-positive patients (20). After the usual reverse transcription with primer A2, 1:10 volumes of the according to their response to IFN-a cDNA were mixed with serial dilutions of a competitor Complete Partial DNA that differed from the cDNA of interest by a small responders responders Noxiresponders Characteristics (n = 24) (n = 32) (n = 14) deletion (34 bases). Samples were amplified by "nested" PCRS with two sets of primers (A1 and A2 for outer Age (yr)" 45 t 14' 50 -t 12 54 t 9 primers; D1, 5'-CGCAGAAAGCGTCTAGCCAT-3', and D2, Sex ( F M ) 7/17 14/18 4/10 5'-ACTCGGCTAGCAGTCTCGCG-3', for inner primers) and Disease duration 4.8 -t 6.6 4.2 k 4.0' 7.5 t 5.4' the same reaction mixture as described above. Competitor (yr)" DNA with deletion was produced from a cloned DNA from Blood transfusion 14/10 22/10 1113 5'-NC-PCR of one patient in group 2. For both PCRs, 35 ( i t ) cycles of PCR were performed as follows: denaturation for Serum levelsd 1 min at 94" C, annealing for 1 min at 55" C and extension ALT (IUL)" 274 t 151 232 t 129 313 t 189 for 1 min at 72" C. Both PCRs were performed in 50-p.1 y-GTP (IUL)" 73 t 43 79 t 66 201 t 282 mixtures, and 1 k.1 of the first-PCR product was transferred y-Globulin (gmfdl)" 1.6 -t 0.4 1.6 2 0.5 1.7 2 0.5 to the second-PCR mixture. "Nested" PCR detected as few anti-HCV ( + I - ) 2913 1113 1717 as five copies of cloned HCV DNA. HCV RNA concentrations 12.0 t2 9 9.3 t 3.0' 10.2 t 3.5 HAI score", were determined from the results of agarose gel electrophoresis of PCR products (Fig. 1). y-GTP = y-glutamyltranspeptidase. Cloning and Sequencing. Nucleotide sequences of a part of "values are mean t SD. 5'-NC region between primer A1 and A2 (nucleotide -277 to hThemean age of complete responders was significantly lower than around nucleotide -97) were determined in six patients (two that of nonresponders (p < 0.01). patients from group 1,two from group 2, one from group 3 and 'Duration of disease in partial responders was significantly shorter the NS3-PCR and NSB-PCR-positive patient). 5'-NC-PCR than that in nonresponders (p < 0.01) products were cloned into the SmaI site of plasmid dNormalvalues: 0 to 45 I U L for ALT, 0 to 40 IUL for y-GTP and pGEM7Zf( + )(Promega Biotec, Madwon, WI).These clones 0.76to 1.69gm/dl for y-globulin. were sequenced by the dideoxy chain-termination method. 'Liver biopsy specimens were scored for histological activity by Statistical Analysis. Data were analyzed by the x2 test and Knodell's index (HAI).Specimens were not available for two patients. Student's t test. fThe mean HAI score of complete responders was significantly
RESULTS ZFN-ru Thempy. The response to IFN-a was related to the dose. Thus 16 of 37 (43.2%)patients treated with 6 million units responded, compared with 7 of 20 (35.0%) treated with 4 million units and 1 of 13 (7.7%)treated with 2 million units. When pretreatment characteristics were compared
lower than that of nonresponders (p < 0.05).
among the patient groups, age and HAI scores of complete responders (45 k 14 yr, 9.3 & 3.0; mean 2 S.D.) were significantly lower than those of nonresponders (54 +- 9 yr, 12.0 -+ 2.9; [p < 0.01, age, and p < 0.05, MI) The . duration of disease in partial
YOSHIOKA ET AL.
296
HEPATOLOGY
-PllTlENT 1
PATIENT 2
FIG.2. Detection of HCV RNA by PCRs in three patients. Patient 1 was positive on 5'-NC-PCR and NS3-PCR and negative on NS5-PCR. Patient 2 was positive on 5'-NC-PCR and NS5-PCR and negative on NS3-PCR. Patient 3 was positive on all three PCRs.
TABLE 4. Pattern of results of PCRs and response to IFN-u therapy Group
Characteristics
5'-NC-PCR response NS3-PCR response NS5-PCR response No. of patients (%) No. of complete responders (rate of complete responders) No. of partial responders No. of nonresponders ~
_____
1
2
3
+ +
-
+
+ -
+ + +
-
l(1.4) 1 (100) 0 0
39 (55.7) 5 (12.8) 25 9
+
-
16 (22.9) 10" (62.5) 4
10 (14.3) 6' (60.0)
2
2
2
4 (5.7) 2 (50.0) 1 1
_______~
+ = double infection with two genotypes; - = disease caused by different agent from HCV. "Group 2 had significantly more complete responders than did group 1 (p < 0.01). bGroup 3 had significantly more complete responders than did group 1 (p < 0.01).
responders (4.2 2 4.0 yr) was significantly shorter than that in nonresponders (7.5 2 5.4 yr) (p < 0.01) (Table 3). However, no factors differentiated complete responders from partial responders. HCVRNA. Anti-HCV was present in 81.4% (57 of 70) of the patients. HCV RNA was present in 94.3% (66 of 70) as assayed by 5'-NC-PCR. 5'-NC-PCR detected as few as 500 copies of a cloned sequence by ethidium bromide staining, and as few as five copies were detected by hybridization with probe A. All 5'-NC-PCR-negative patients were also negative on NS3-PCR or NS5-PCR, and all but one were also negative for anti-HCV. HCV RNA was present in 57.1% (40 of 70) of the patients assayed by NS3-PCR and in 24.3% (17 of 70) assayed by NS5-PCR. In three patients who had no band with ethidium bromide staining in NS5-PCR, a specific 231-bp band was found after hybridization with probe C. Only one patient was positive on both NS3-PCR and NS5-PCR, and four were negative on all three PCRs. Thus patients could be divided into five groups according to the PCR results (Table 4, Fig. 2). The three main groups were 39 NS3-PCR-positive and NS5-PCRnegative patients (group l ) , 16 NS3-PCR-negative and NS5-PCR-positive patients (group 2) and 10 NS3-PCRnegative, NS5-PCR-negative, 5'-NC-PCR-positive patients (group 3). When pretreatment characteristics were compared among the three groups, the number of
patients with histories of blood transfusion was significantly lower in group 2 (6.3%, 1of 16) than in group 1 (38.5%, 15 of 39) (p < 0.05) and in group 3 (50.0%, 5 of 10) (p < 0.05). HCV RNA and Responses to Therapy. Response rate to IFN-a varied according to PCR group. The best response was noted in group 2 (10 of 16,63%),whereas group 3 patients had a slightly lower response (6 of 10, 60%). The response in both of these groups was significantly better than that in group 1 (5 of 39, 13%) (p < 0.01) (Table 4). It may be more appropriate to exclude patients treated by the lower dose in evaluating the response to treatment. Therefore we also analyzed 57 patients treated with 4 million or 6 million units. The response in group 2 (10 of 14 [71.4%1)and in group 3 (5 of 8 [62.5%])was significantly better than that in group 1 ( 5 of 30 [16.7%1) (p < 0.011. These findings were identical to those for the complete set of 70 patients. Quantitation of HCV RNA by Competitive PCR. The range of HCV RNA concentrations in serum of 63 patients before therapy was lo2 to lo8 copies per milliliter. Mean k S.D. of logarithms of HCV RNA concentrations was 4.5 f 1.3. Logarithms of HCV RNA concentrations were significantly higher in group 1 (5.0 & 1.1) than in group 2 (3.8 2 1.1)and group 3 (3.2 f 1.1)(p < 0.01 for both) (Fig. 3). Logarithms of HCV RNA concentrations were significantly lower in
-
HCV GENOTYPES AND IFN-CX THERAPY
Vol. 16, No. 2, 1992
complete responders (3.8 2 1.4) than in partial responders (4.7 & 1.2) and nonresponders (5.0 ? 1.2) (p < 0.05 for both). Cloning and Sequencing. Sequences of 181 nucleotides (nucleotide - 277 to nucleotide - 97) were analyzed (Fig. 4). Nucleotide sequences of two clones from two patients in group 1 (patient 1 and patient 2) corresponded completely with that of HC-J4 (4) and differed from that of HCV prototype at only one nucleotide. Nucleotide sequence of the clone from a patient of group 2 (Patient 3) showed 97.8% homology with HCV-K2a, 92.8% with HCV-J4 and 93.4% with HCVprototype. The clone from another patient in group 2 (Patient 4) corresponded completely with that of HCV-K2a. The clone froin a patient of group 3 (Patient 5 ) showed 93.4% homology with HCV-J4, 93.9% homology with HCV prototype and 95.0% homology with HCV-K2a. The clone from the NS3-PCR-positive and NS5-PCR-positive patient (Patient 6) showed 92.3% homology with HCV-J4, 92.8% homology with HCV prototype and 96.1% homology with HCV-K2a. DISCUSSION
HCV RNA was detectable in 94.3% (66 of 70) of the patients included in this study on 5‘-NC-PCR. The sequence of the 5’-NC region is considered the most conserved among the different strains (4). Therefore it is naturally expected that 5’-NC-PCR will detect HCV RNA at a higher rate than will the other two PCRs. On the other hand, considerable sequence variability exists in NS3 and NS5 reeons among the different strains. It is probably the reason why NS3-PCR and NS5-PCR detected HCV-RNA less frequently than did 5’-NC-PCR. Houghton et al. (21) classified HCV into three genotypes: HCV-I (HCV prototype group), HCV-I1 (HCV-J group) and HCV-I11 (HCV-K2 group). Because the sequences of primers and a probe used for NS3-PCR are derived from HCV-J, NS3-PCR will detect both HCV-J and HCV prototype, whose nucleotide sequence is similar to that of HCV-J. In fact, the nucleotide sequences of clones from two patients in group 1 (NS3-PCR-positive and NS5-PCR-negative patients) were shown to be identical to HC-J4 (HCV-J group). The sequences of primers and a probe used for NS5-PCR are derived from HCV-K2 and differ considerably from HCV prototype or HCV-J. Therefore NS5-PCR will detect HCV-K2 but not HCV prototype or HCV-J. In fact, the nucleotide sequences of clones from two patients in group 2 (NS3-PCR-negative and NS5PCR-positive patients) were shown to be identical or similar to those in the HCV-K2a (HCV-K2) group. It seems that HCV-K2 significantly differs from the sequences of primers and a probe used in NS3-PCR and is not detected by our NS3-PCR. In practice, only one patient was positive for both NS3-PCR and NS5-PCR. These results indicate that group 1 patients may be infected with the HCV prototype group or the HCVJ group and group 2 patients may be infected with the HCV-m group. A clone from a patient in group 3 showed low
297
**
**
m
o complete responder 0 partial rsspondsr x non-responder
X
.... ..:.... ....
f
oxn
000
onx 0..
0.
X
00
oonnnn
1.00 on n
T
.O
f
.on
.O
000
0
000
group 1
group 2
group 3
(37)
(15)
(9)
FIG. 3. HCV RNA concentrations in the three groups, group 1 (NS3-PCR positive and NS5-PCR-negative patients), group 2 (NS3PCR-negative and NSS-PCR-positive patients) and group 3 (5‘-NCPCR-positive, NS3-PCR-negative, NSS-PCR-negative patients). Logarithms of HCV RNA concentrations were significantly higher in group 1 (5.0t 1.1) than in group 2 (3.8 k 1.1) and in group 3 (3.2 -+ 1.1) (p < 0.01 for both). Logarithms of HCV RNA concentrations were sigdicantly lower in complete responders (3.6 2 1.4) than in partial responders (4.7 t 1.2) and in nonresponders (5.0f 1.2) ( p < 0.05 for both).
nucleotide homology with either HCV prototype, H C J 4 (HCV-J group) or HCV-ma. Therefore HCV of group 3 probably differs from that of any of HCV prototype, HCV-J or HCV-K2. Nucleotide sequence of a clone from the NS3-PCR- and NS5-PCR-positive patient resembled that of HCV-K2a but differed from that of HCV prototype or HCV-J. It is probable that the other NS3-PCR-positive strain was also present in the serum of this patient. Garson et al. reported that their “nested” PCR method detected as few as 0.5 copies of HCV sequence (12). Our 5’-NC-PCR detected as few as five copies of HCV sequence. In addition, high specificity was obtained by hybridization with nonradioactive probes. PCR with primers from NS3 and NS5 detected HCV RNA in serum samples at concentrations of lo2 copies/ml (10 copiesltest sample). Therefore it is believed that their sensitivity was as high as that of 5’-NC-PCR. Our “nested” PCR also detected as few as five copies of HCV sequence.
298
YOSHIOKA ET AL.
HEPATOLOGY
FIG.4. Sequences of 181nucleotides in 5’-NC region (nucleotide - 277 to nucleotide - 97) of six patients. Nucleotide sequences of two clones from Patient 1 and Patient 2 corresponded completely with that of HC-J4 (HCV-J group) and differed from that of HCV prototype only a t one nucleotide. Nucleotide sequence of the clone from Patient 3 showed 97.8% homology with HCV-K2a, 92.8% homology with H C J 4 and 93.4% homologywith HCVprototype. The clone from Patient 4 corresponded completelywith that of HCV-Wa. The clone from Patient 5 showed 93.4% homology with H C J 4 , 93.9% homology with HCV prototype and 95.0% homology with HCV-K2a. The clone from Patient 6 showed 92.3% homology with HC-J4, 92.8% homology with HCV prototype and 96.1% homology with HCV-K2a.
So far no clinical, biochemical, serological or histological features have been identified to predict response to IFN treatment (13-15).This study also demonstrated that there were no clinical or biochemical features to differentiate complete responders from partial responders or nonresponders reliably before treatment, although younger age, lower HAI score and shorter duration of disease were the factors that predicted better response to IFN-a therapy. We showed that group 2 (putative HCV-K!2 group) yielded significantly more complete responders than did group 1 (putative HCV prototype and HCV-J group). This finding indicates that the HCV prototype group or HCV-J group is more resistant to IFN-a and that the HCV-K!2 group is more responsive. At the same time, HCV RNA concentrations in group 2 were significantly lower than those in group 1. This suggests that the HCV-K2 group is less able to replicate in humans and thus may be more responsive to IFN-a therapy. In fact, pretreatment HCV RNA concentrations in complete responders were significantly lower than in the others. The different HCV RNA concentrations among genotypes may be attributed to the difference in susceptibility to immunological response of the host to the virus or in replicative capacity of the virus by itself. From these results, we could suppose that the genotype resistant to IFN-a has stronger virulence and thus causes more severe hepatitis. A large-scale study on the relationship between severity of disease and genotypes is under way in our laboratory. Alternatively, it was shown that group 2 patients had lower rates of history of blood transfusion than did group 1 or 3 patients. Further study is necessary to clarify the relationship between genotypes and routes of transmission.
It was also shown that group 3 had a significantly larger number of complete responders than did group 1 and that HCV RNA concentrations in group 3 patients were significantly lower than those in group 1patients. This may suggest that another HCV genotype (different from any HCV prototype, HCV-J or HCV-K2) is also responsive to IFN-a and has a lower replicative capacity. There may be still the possibility that these findings are due to differential sensitivity in the PCRs using different primers and probes. To rule out that possibility, it will be necessary to test the sensitivity of the assay with all 3 sets of primers and also to determine the sequences of amplified NS3 and NS5 regions. Sequencing is under way in our laboratory. In conclusion, our study suggests that PCR is valuable in defining the genotypes of HCV and predicting the response to IFN-a therapy and that the amount of HCV RNA in sera and response to IFN-a therapy may vary among different genotypes of HCV. REFERENCES 1. Choo QL, Kuo G, Weiner AJ, Overby LR, Bradley DW, Houghton M. Isolation of a cDNA clone derived from a blood-borne non-A, non-B viral hepatitis genome. Science 1989;244:1-3. 2. Kuo G, Choo QL, Alter HJ, Gitnick GL, Redeker AG, Purcell RH, Miyamura T, et al. An assay for circulating antibodies to a major etiologic virus of human non-A, non-B hepatitis. Science 1989; 244:362-364. 3. Kubo Y, Takeuchi K, Boonmar S, Katayama T, Choo Q-L,Kuo G, Weiner AJ, et al. A cDNA fragment of hepatitis C virus isolated from an implicated donor of post-transfusion non-A, non-B hepatitis in Japan. Nucleic Acids Res 1989;17:10367-10372. 4. Okamoto H, Okada S, Sugiyama Y, Yotsumoto Y, Tanaka T, Yoshizawa H, Tsuda F, et al. The 5’-terminal sequence of the hepatitis C virus genome. Jpn J Exp Med 1990;60:167-177. 5. Kato N, Hijikata M, Ootsuyama Y, Nakagawa M, Ohkoshi S,
Vol. 16, No. 2, 1992
H C V G E N O T W E S AND IFN-n THERAPY
Sugyama T , Shimotohno K, e t a]. Molecular cloning of the human hepatitis C virus genome from dapanese patients with non-A. non-B hepatitis. Proc Natl Acad Sci LISA 1990:87:9524-9526. 6. Choo QL, Richman KH. Han JII. Berger I
