Journal of Medical Virology 36259-264 (1992)
Association of Hepatitis C Virus Carrier State With the Occurrence of Hepatitis C Virus Core Antibodies Hendrik Claeys, Andre Volckaerts, Hans De Beenhouwer, and Carl Vermylen Belgian Red Cross Blood Transfusion Centre, Leuuen, Belgium An enzyme immunoassay (EIA) was developed for the determination of antibodies against the "putative" core protein of hepatitis C virus (HCV). Antigens used were recombinant fragments (amino acids f3-77 or 6-143) of the HCV core protein, produced in Escherichia coli with truncated hepatitis B core (HBc) as fusion protein. Evaluation of 385 sera positive for HCV antibodies by first generation EIA, revealed 98 (25.4%)with HCV core antibodies. HCV-RNA, determined by the polymerase chain reaction (PCR), was exclusively found in the sera positive for HCV core antibodies (89 PCR positives). In random screening of 3,708 sera, 3 sera with HCV core antibodies were found PCR positive. Only 2 of these sera were positive in the first generation EIA. It is concluded that HCV core antibody determination is a reliable test for identifying HCV carriers among blood donors.
KEY WORDS: non-A,no n-B hepatit is, blood screening, HCV core expression, truncated HBc
INTRODUCTION Transfusion-associated hepatitis non-A, non-B (NANBH) is a serious complication of blood transfusion. Analysis of transmission of NANBH revealed a correlation between increased serum alanine aminotransferase (ALT) levels or the presence of antibodies to hepatitis B virus core (HBc) in donor blood and the occurrence of NANBH in transfusion recipients [Wick et al., 19851. These observations prompted blood banks to implement surrogate tests for NANBH. The recent discovery of hepatitis C virus (HCV) improved considerably knowledge about NANBH [Choo et al., 19891. HCV has been shown to be a major etiological agent of NANBH [Kuo et al., 19891. Commercial enzyme immunoassays (EIA) became available for the detection of antibodies to HCV in serum. These first generation assays are based on a non-structural viral protein. However an only moderate sensitivity and specificity was achieved [Van der Poel et al., 19911. The 0 1992 WILEY-LISS, INC.
discovery of the 5'-terminal sequence of the HCV genome, postulated to contain two structural genes [Okamot0 et al., 19901, opened the possibility of improving sensitivity by using structural proteins for antibody detection. The nucleocapsid protein of hepatitis B virus (HBV) has been shown to be a generic epitope carrier [Stahl and Murray, 19891. Fusion proteins are produced in high yield in Escherichia coli, assemble into core-like particles which can easily be purified, and were found to be good immunogens. In the present study we constructed fusion genes in which coding regions for HCV structural proteins were linked to the 3' end of the truncated HBc-gene (the first 144 amino acids [AAI). Recombinant proteins were expressed in E . coli, purified by fractional ultracentrifugation and gel filtration, and evaluated as a n antigen in EIA for HCV core antibodies (HCV core Ab EIA). The combination with HBc made it also possible to evaluate anti-HBc surrogate testing.
MATERIALS AND METHODS Molecular Cloning Technology HBV DNA was isolated from the serum of a n HBV chronic carrier. The coding sequences of the entire HBc antigen, AA 1-183 and of a truncated form, AA 1-144, were polymerised by the polymerase chain reaction (PCR). HCV-RNA was isolated from a patient with documented NANBH, reverse transcribed, and the coding sequences of the putative core protein (AA 6-77 and AA 6-143) polymerised by the PCR. All primers contained restriction sites to facilitate subsequent cloning. As expression vector a commercial plasmid (pBTac 2, Boehringer, Mannheim, Germany) was used. Restriction endonucleotide digestions, ligations, and transformations were performed by standard procedures [Sambrook et al., 19891. Plasmid Constructions The line drawing of Figure 1 shows the plasmid (p) constructs made: Accepted for publication October 3,1991 Address reprint requests to H. Claeys, Blood Transfusion Centre, 0.L.Vrouwstraat 42,3000 Leuven, Belgium.
Claeys et al.
260
In pilot experiments, cultures were induced by the (IPTG) addition of isopropyl-P-D-thiogalactopyranoside to 1mM.
E B H
pBTac2-
-
E
p 333
HBc 1 - 1 8 3
HBc 1 - 1 4 4
B H a
E
HBc 1 - 1 4 4
B HCV 6 - 7 7
HBc 1 - 1 4 4
B
I
E
p 601
I
E p 405
p 520
B H
I
I
I
EIA 'To evaluate the recombinant fusion proteins as antiH
HCV 6 - 1 4 3
H
___t_
Fig. 1. Line drawing indicating the structure of pBTac2 and of the different recombinant plasmids. Restriction sites for EcoRl (El, BamHl (B),and Hind111 (H) are indicated.
p 333: the entire coding sequence of HBc ligated between the EcoR1-BamH1 sites of pBTac2. p 405: the coding sequence of truncated HBc ligated between the EcoR1-BamH1 sites of pBTac2. p 520: the coding sequence of HCV core AA 6-77 ligated between the BamH1-Hind111 sites of p 405. p 601: the coding sequence of HCV core AA 6-143 ligated between the BamH1-Hind111 sites of p 405.
Expression and Purification of Recombinant Proteins Cultures of transformed E. coli (DH5a competent cells, Gibco-BRL, Paisly, Scotland) were grown overnight at 37°C in 1liter volumes under continuous shaking. Cells were harvested by centrifugation and lysed by ultrasonification. Fusion protein particles were purified by ultracentrifugation and gel filtration on Sephacryl S-500 (Pharmacia, Uppsala, Sweden) essentially a s described [Stahl and Murray, 19891. Purified proteins were characterised by sodium dodecyl sulfate polyacrylamide gel electrophoresis (SDS-PAGE).
gens (Ag) in EIA, protein particles (4 mg/ml) were incubated with 0.1% SDS (60 min, 37°C) and subsequently diluted in 10 mM Na bicarbonate buffer pH 9.6. Polyvinylchloride (PVC) assay plates (Costar, Cambridge, MA, USA) were coated overnight at 4°C with the SDStreated proteins at a final concentration of 1 pg/ml. Plates were washed with 10 mM Tris buffer pH 7.5, air dried, and stored at 4°C until use. As conjugate a goat anti-human immunoglobulin was used, labelled with 1,250 units horseradish peroxidase per 10 mg IgG [Tijssen and Kurstak, 19841. EIA was performed by standard procedures on 1/40 or 1/100 dilutions of serum. As positive control, serum from a donor positive in PCR for HCV [De Beenhouwer et al., 19921 was used. The cutoff ((3.0.) was arbitrary calculated from the optical densities (O.D.) a s follows: O.D. negative control
+ 1/10 O.D. positive control.
All results were expressed as the ratio between the O.D. of the sample to the O.D. of the C.O. A ratio of > 1.0 was considered positive. To discriminate between HBc and HCV antibodies, initially reactive sera were reassayed after HBc neutralisation by the addition of 10 pg/ml HBc antigen (Ag333) to the dilution buffer. This treatment may also exclude false positives caused by anti-E. coli.
Fig. 2. SDS-PAGE of recombinant protein after reduction. Samples in the tracks from left to right are: lane 1: molecular weight (MW) markers (MW 110,000; 84,000; 47,000; 33,000; 24,000; 16,000);lane 2 Ag333; lane 3: Ag405; lane 4: Ag520; lane 5 Ag601; lane 6 MW markers.
HCV Core Antibodies and Carrier State
261
TABLE I. EIA for HBc and HCV Core Antibodies* Antigens Sera Negative HBc Ab positive HCV Ab positive HCV and HBc Ab positive HBc Ab positive (HBV carrier)
0.118 8.950
Ag333 HBc neutralisation 0.112 0.696
0.115 6.630
Ag520 HBc neutralisation 0.128 0.688
0.110 0.423
Ag601 HBc neutralisation 0.119 0.110
Evaluation Negative Negative
0.143
0.132
9.298
9.407
6.210
6.830
Positive
6.448
0.608
6.211
5.089
2.580
2.110
Positive
7.892
6.208
7.478
5.419
4.078
2.551
I.D.
*Five representative sera were assayed by EIA using different antigens, with and without HBc neutralisation. Results were expressed as ratios. Evaluation as positive, negative or indeterminate (1.D.)was related to the presence of HCV antibodies. Sera with HBc Ab were also positive for HBe Ab.
Other HCV Assays First and second generation EIA (Ortho, Beerse, Belgium, or Abbott Wiesbaden, Germany) and Immunoblots (4-RIBA, Chiron-Ortho, Emeryville, CA, USA; INNO-LIA, Innogenetics, Antwerp, Belgium) were performed according to the directions provided by the manufacturers. Diagnostic PCR was performed on the noncoding region of the HCV genome using nested primers [De Beenhouwer et al., 19921.
TABLE 11. HCV Core Antibody in First Generation EIA Positive Sera* Ag520 EIA/Ag333 EIA Initial screening Assay with HBc neutralisation Total number of sera Ag601 EIA/Ag333 EIA
RESULTS Expression, Purification, and Characterisation of Recombinant Proteins After transformation of E. coli with the different plasmid constructs, the recombinant proteins were spontaneously produced, without benefit of IPTG-induction. On the Sephacryl-S500 they behaved a s HBclike particles and eluted in the higher molecular weight region of the column immediately after the void volume fraction. Yields of purified recombinant proteins differed appreciably: A333, Ag405, and Ag520 were obtained in yields of about 15 mglliter culture, while Ag601 yielded only about 1mg/liter. SDS-PAGE revealed the purified reduced proteins as doublets with average molecular weights of respectively 20,000, 15,000, 24,000, and 32,500 (Fig. 2). For the first 3 antigens a high degree of purity was obtained, but for Ag601 there was still some contamination with E. coli proteins.
EIA for HBc and HCV Core Antibodies For optimal coating of the PVC assay plates, a pretreatment of the recombinant HBc-like particles with 0.1% SDS was found beneficial. To evaluate the different antigens, representative sera were analysed. Results are shown in Table I. As expected Ag333 only measured antibodies to HBc and Ag520 measured antibodies to both HBc and HCV. Ag601 also reacted strongly with HCV antibodies, but showed only a minor reactivity for HBc antibodies.
Initial screening Assay with HBc neutralisation Total number of sera
HCV core antibody Positive I.D. Negative 91 14 280 J.
7 98
4
1 1
.L
6 286
HCV core antibody Positive I.D. Negative
2 7 98
1 1
0
286
*Three hundred eighty-five first generation EIA positive sera were assayed by Ag520 and Ag601 EIA. Positive sera were analysed on Ag333 for HBc antibodies and if present HBc neutralisation was performed. In the Table the number of sera with a positive, negative, or indeterminate (I.D.) result is given.
Discrimination between HBc and HCV antibodies could be made by HBc neutralisation, except in sera with very high titers of HBc antibodies as found in HBV carriers. In such cases a n indeterminate result (1.D.)had to be accepted.
EIA for HBc and HCV Core Antibodies in Sera Received for HCV Confirmation In our blood bank we have a collection of 385 sera found positive in first generation HCV-EIA [De Beenhouwer et al., 19911. When assayed on Ag520,104 sera were initially reactive with a n average ratio of 6.22 1.41 and 281 were negative with a n average ratio of 0.33 ? 0.14 (Table 11).Of the 105 initially reactive sera only 14 were also positive on Ag333 and therefore remained provisionally indeterminate. To discriminate further, these 14 sera were reassayed after HBc neutralisation. For 7 sera the signal was neutralised on Ag333 while on Ag520 it remained essentially un-
*
Claeys et al.
262 TABLE 111. Correlation of Several HCV Assays With HCV Core Antibody* HCV core antibody 98 Positive 287 Negative Second generation EIA I positive \ /98\ 4-RIBA INNO-LIA 4-RIBA INNO-LIA Immunoblot I( L k! L “reactive” 95 98 JL 0 0 87 89 PCR positive
I
H
t
*On 385 first generation EIA positive sera, second generation EIA (Ortho and Abbott) and PCR were performed. The 179 (98 + 81) sera positive in both second generation EIA were further characterised by immunoblotting (CRIBA and INNO-LIA). For both the HCV core Ab positive and negative group the number of positive or “reactive” sera in different assays is given.
changed, for 6 sera i t was neutralised on bot,h antigens, and for one serum no definite neutralisation was observed. These sera could therefore be allocated respectively as positive, negative, and indeterminate for HCV core antibodies. When assayed on Ag601, a similar distribution of sera was obtained but the 6 sera t h a t were only positive for HBc antibodies, were negative already in the initial screening.
Significance of HCV Core Antibodies All 385 sera were also assayed in 2 commercial second generation EIA (Abbott and Ortho) and PCR. The 179 sera positive in both second generation EIA were characterised further by immunoblotting (4-RIBA and INNO-LIA) [De Beenhouwer et al., 19921. The number of positive or “reactive” (Immunoblot)sera in the different assays was compared in the group of HCV core Ab positives and negatives (Table 111). All 98 HCV core Ab positive sera were positive in both second generation EIA and “reactive” by INNOLIA immunoblotting. On 4-RIBA immunoblotting 95 (96.9%) were “reactive” and 3 were “C22-3-I.D.” (only reaction with the c22-3 band). All 89 (90.8%)PCR positive sera belonged to this group. The band pattern of the remaining 9 PCR negative sera could not be distinguished from that of the positives. Of the 287 HCV core Ab negative sera 81 (28.9%) were positive in both second generation EIA. Only 2 (2.5%) of these 81 sera were “reactive” on 4-RIBA immunoblotting and 3 (3.7%) on INNO-LIA. None was “reactive” on both. In addition colour intensity was weak and only reactions against non-structural proteins were observed. All 287 HCV core Ab negative sera were PCR negative. Evaluation of Ag520 E I A as Screening Assay A series of 3,708 sera were assayed for HCV antibodies with the Ortho first generation EIA and the Ag520 EIA in parallel. The results are summarized in Table IV. On initial screening with Ag520 EIA, 131(3.5%) sera were positive. Only 2 sera were non-reactive on Ag333 and could therefore be immediately interpreted HCV core Ab positive. HBc neutralisation on the 129 sera reactive on both antigens, revealed another 3 positive
-
TABLE IV. Evaluation of Ag520 EIA as Screeninn Assav* Ag520 EIA/Ag333 EIA Initial screening Assav with HBc neutralisation Total number Ortho first generation EIA
Positive 2
3
I.D. 129
Negative 3,577
5
3 3
123 3,700
21
-
3,687
*Three thousand seven hundred eight randomly collected sera were screened by Ortho first generation EIA and by Ag520 EIA-Ag333 EIA. Indeterminate (I.D.) sera were reassayed after HBc neutralisation. The number of positive, I.D., or negative sera is shown.
sera, 123 negative sera, and 3 indeterminates. Thus 5 (0.13%) sera were positive in the HCV core Ab EIA against 21 (0.57%) in the Ortho first generation EIA. Only 2 sera were positive in both assays. Confirmation of positive sera was performed by second generation EIA, immunoblotting, and PCR (Table V). Only 3 of the 24 initially positive sera could be serologically confirmed. All 3 were reactive on Ag520, but only 2 were detected by the Ortho first generation EIA. All 3 sera were PCR positive.
DISCUSSION HBc- and HBc-HCV plasmid constructs were made in the pBTac2 expression vector and recombinant HBcHCV proteins were produced in transformed E . coli. Unexpectedly, recombinant proteins were spontaneously produced and no influence of stimulation by IPTG was observed. Yields of recombinant proteins averaged 15 mg/liter culture except for the larger protein construct (Ag601).These data confirm earlier observations t h a t the nucleocapsid protein of HBC can be used as a carrier to express foreign polypeptides in E . coli [Stahl and Murray, 19891. Recombinant proteins apparently assembled into HBV core-like particles and could therefore be easily purified by sequential ultracentrifugation and gel filtration. In this respect Sephacryl S-500 resulted in a considerably better purification of the particles than Sepharose 4B-CL. HBc fusion proteins were shown to be good immunogens CStahl and Murray, 19891. We explored the possibility of using them a s coating material for EIA. Intact
HCV Core Antibodies and Carrier State
Positive in screening
263
TABLE V. Confirmation of HCV Status in EIA Positive Sera* Ortho first generation EIA
Y 1 2 / i 51 ! f
Ag520 EIA
19
Positive in both second generation EIA “Reactive” in both immunoblots PCR positive
0
2 .L
2
.L 1
-1
1 L 1
*The HCV s t a t u s of t h e 24 sera, reactive in t h e Ortho first generation EIA or t h e Ag520 EIA,w a s evaluated by additional serologic assays and by PCR. The number of positive sera in each assay is shown.
HBc-like particles had a low affinity for PVC, but the conformational changes provoked by SDS treatment [Sallberg e t al., 19901 resolved this problem. As expected the HBc fusion proteins had a n affinity for both HBc and HCV antibodies. However compared to the unsubstituted HBc protein (Ag333), the af€inity of the fusion proteins for HBc antibodies decreased with the length of the substituent. Indeed, while Ag520 with a substituent of 72 AA still measured HBc antibodies quite effectively, Ag601 with a substituent of 138 AA only detected high levels (Table I). By the addition of HBc antigen (Ag333)to the dilution buffer the majority of HBc reactions were neutralised. Thus depending on the set-up HBc fusion proteins can be used to measure both HBc and substituent or only substituent antibodies. In NANBH, testing for both HCV and HBc antibodies could potentially be valuable. Of the 385 first generation EIA positive sera only 98 126.4%) were positive for HCV core antibodies (Table 111).This discrepancy appeared to be mainly due to the poor specificity of the first generation EIA in a population of blood donors, where the frequency of HCV positivity is low [De Beenhouwer et al., 19921. The present HCV core Ab EIA appeared to be extremely specific in distinguishing false from true positives. Of the 287 HCV core Ab negative sera only 81 were found by both second generation EIA and from those none was “reactive” on both immunoblots. In contrast all 98 HCV core Ab positive sera were confirmed in both second generation EIA and 95 were “reactive” in both blots (Table 111).Moreover the 89 PCR positive sera were all found in this group and the remaining 9 could serologically by no means be further distinguished. Thus only these 98 donors should be considered as HCV carriers. To further evaluate the sensitivity and specificity of HCV core Ab testing we screened 3,708 sera with both Ag520 EIA and a first generation EIA (Ortho). The data obtained (Table V) essentially confirm the reported sensitivity: 3 sera were “reactive” by immunoblotting and positive by PCR and all 3 were detected by the HCV core Ab EIA while the first generation EIA, based on non-structural antigens, only found 2 of these sera. Specificity of the Ag520 EIA was slightly hampered by the interference of HBc antibodies resulting in
2 false positive and 3 indeterminate sera. Specificity was still considerably better than that of the first generation EIA. The present data also allowed the evaluation of HBc testing as a surrogate assay for NANBH. Of the 98 positive sera in the confirmation study 7 were also positive for HBc antibodies. In the screening study 129 HBc Ab positive sera were found of which one was HCV Ab positive. Thus in our population of blood donors a poor association was observed between HCV infection and the presence of HBc antibodies. In conclusion, confirmation of HCV first generation EIA positive donors, by HCV core Ab determination was reliable in identifying viral carriers. Screening for HCV core Ab resulted in a higher sensitivity and specificity than screening with first generation EIA.
ACKNOWLEDGMENTS The authors thank Ortho Diagnostics, Abbott Diagnostic, and Innogenetics for supplying test kits. Primers were synthesized by Prof. F. Van Leuven (K.U. Leuven-Genetics). We are grateful to Mrs. W. Mertens, E. Noe, H. Verhaert, and G. Wuyts for technical support and to Dr. K. Gautama for data management. REFERENCES Choo QL, Kuo G, Weiner AJ, Overby LR, Bradley DW, Houghton M (1989):Isolation of a cDNA clone derived from blood borne Non-A, Non-B viral hepatitis genome. Science 244:359-362. De Beenhouwer H, Verhaert H, Claeys H, Vermylen C (1992):Confirmation of HCV positive blood donors by immunoblotting and PCR. Vox Sanguinis (in press). Kuo G, Choo QL, Alter HJ, Gitnick G1, Redeker AG, Purcell RH, Miyamura T, Dienstag JL, Alter MJ, Stevens CE, Tegtmeier GE, Bonino F, Colombo M, Lee WS, Kuo C, Berger K, Shuster JR, Overby LR, Bradley DW, Houghton M (1989):An assay for circulating antibodies to a major etiologic virus of human non-A, non-B hepatitis. Science 244:362-364. Okamoto H, Okada S, Sugiyama Y, Yotsumoto S, Tanaka T, Yoshizawa H, Tsuda F, Miyakawa Y, Mayumi M (1990): The 5’terminal sequence of the hepatitis C virus genome. Japanese Journal of Experimental Medicine 60:167-177. Sallberg M, Norder H, Magnius LO (1990):Comparisons of class and subclass distribution of antibodies to the hepatitis €3 core and B e antigens in chronic hepatitis B. Journal of Medical Virology 3O:l-6.
264 Sambrook J, Fritsch EF, Maniatis T (1989): Molecular Cloning, A Laboratory Manual. Cold Spring Harbor: Cold Spring Harbor Laboratory Press. Stahl SJ, Murray K (1989):Immunogenicity of peptide fusions to hepatitis B virus core antigen. Proceedings ofthe National Academy of Science USA 86:62834287. Tijssen P, Kurstak E (1984):Highly efficient and simple methods for the preparation of peroxidase and active peroxidase-antibody conjugates for enzyme immunoassays. Analytical Biochemistry 136:451-457.
Claeys et al. Van der Poel CL, Cuypers HTM, Reesink HW, Weiner AJ, Quan S, di Nello R, Van Boven J J P , Winkel I, Mulder-Folkerts D, ExelOehlers PJ, Schaasberg W, Leentvaar-Kuypers A, Polito A, Houghton M, Lelie PN (1991):Confirmation of hepatitis C virus infection by new 4-antigen recombinant immunoblot assay. Lancet 337:317-319. Wick MR, Moore S, Taswell H F (1985):Non-A, non-B hepatitis associated with blood transfusion. Transfusion 25:93-101.
Association of Hepatitis C Virus Carrier State With the Occurrence of Hepatitis C Virus Core Antibodies Hendrik Claeys, Andre Volckaerts, Hans De Beenhouwer, and Carl Vermylen Belgian Red Cross Blood Transfusion Centre, Leuuen, Belgium An enzyme immunoassay (EIA) was developed for the determination of antibodies against the "putative" core protein of hepatitis C virus (HCV). Antigens used were recombinant fragments (amino acids f3-77 or 6-143) of the HCV core protein, produced in Escherichia coli with truncated hepatitis B core (HBc) as fusion protein. Evaluation of 385 sera positive for HCV antibodies by first generation EIA, revealed 98 (25.4%)with HCV core antibodies. HCV-RNA, determined by the polymerase chain reaction (PCR), was exclusively found in the sera positive for HCV core antibodies (89 PCR positives). In random screening of 3,708 sera, 3 sera with HCV core antibodies were found PCR positive. Only 2 of these sera were positive in the first generation EIA. It is concluded that HCV core antibody determination is a reliable test for identifying HCV carriers among blood donors.
KEY WORDS: non-A,no n-B hepatit is, blood screening, HCV core expression, truncated HBc
INTRODUCTION Transfusion-associated hepatitis non-A, non-B (NANBH) is a serious complication of blood transfusion. Analysis of transmission of NANBH revealed a correlation between increased serum alanine aminotransferase (ALT) levels or the presence of antibodies to hepatitis B virus core (HBc) in donor blood and the occurrence of NANBH in transfusion recipients [Wick et al., 19851. These observations prompted blood banks to implement surrogate tests for NANBH. The recent discovery of hepatitis C virus (HCV) improved considerably knowledge about NANBH [Choo et al., 19891. HCV has been shown to be a major etiological agent of NANBH [Kuo et al., 19891. Commercial enzyme immunoassays (EIA) became available for the detection of antibodies to HCV in serum. These first generation assays are based on a non-structural viral protein. However an only moderate sensitivity and specificity was achieved [Van der Poel et al., 19911. The 0 1992 WILEY-LISS, INC.
discovery of the 5'-terminal sequence of the HCV genome, postulated to contain two structural genes [Okamot0 et al., 19901, opened the possibility of improving sensitivity by using structural proteins for antibody detection. The nucleocapsid protein of hepatitis B virus (HBV) has been shown to be a generic epitope carrier [Stahl and Murray, 19891. Fusion proteins are produced in high yield in Escherichia coli, assemble into core-like particles which can easily be purified, and were found to be good immunogens. In the present study we constructed fusion genes in which coding regions for HCV structural proteins were linked to the 3' end of the truncated HBc-gene (the first 144 amino acids [AAI). Recombinant proteins were expressed in E . coli, purified by fractional ultracentrifugation and gel filtration, and evaluated as a n antigen in EIA for HCV core antibodies (HCV core Ab EIA). The combination with HBc made it also possible to evaluate anti-HBc surrogate testing.
MATERIALS AND METHODS Molecular Cloning Technology HBV DNA was isolated from the serum of a n HBV chronic carrier. The coding sequences of the entire HBc antigen, AA 1-183 and of a truncated form, AA 1-144, were polymerised by the polymerase chain reaction (PCR). HCV-RNA was isolated from a patient with documented NANBH, reverse transcribed, and the coding sequences of the putative core protein (AA 6-77 and AA 6-143) polymerised by the PCR. All primers contained restriction sites to facilitate subsequent cloning. As expression vector a commercial plasmid (pBTac 2, Boehringer, Mannheim, Germany) was used. Restriction endonucleotide digestions, ligations, and transformations were performed by standard procedures [Sambrook et al., 19891. Plasmid Constructions The line drawing of Figure 1 shows the plasmid (p) constructs made: Accepted for publication October 3,1991 Address reprint requests to H. Claeys, Blood Transfusion Centre, 0.L.Vrouwstraat 42,3000 Leuven, Belgium.
Claeys et al.
260
In pilot experiments, cultures were induced by the (IPTG) addition of isopropyl-P-D-thiogalactopyranoside to 1mM.
E B H
pBTac2-
-
E
p 333
HBc 1 - 1 8 3
HBc 1 - 1 4 4
B H a
E
HBc 1 - 1 4 4
B HCV 6 - 7 7
HBc 1 - 1 4 4
B
I
E
p 601
I
E p 405
p 520
B H
I
I
I
EIA 'To evaluate the recombinant fusion proteins as antiH
HCV 6 - 1 4 3
H
___t_
Fig. 1. Line drawing indicating the structure of pBTac2 and of the different recombinant plasmids. Restriction sites for EcoRl (El, BamHl (B),and Hind111 (H) are indicated.
p 333: the entire coding sequence of HBc ligated between the EcoR1-BamH1 sites of pBTac2. p 405: the coding sequence of truncated HBc ligated between the EcoR1-BamH1 sites of pBTac2. p 520: the coding sequence of HCV core AA 6-77 ligated between the BamH1-Hind111 sites of p 405. p 601: the coding sequence of HCV core AA 6-143 ligated between the BamH1-Hind111 sites of p 405.
Expression and Purification of Recombinant Proteins Cultures of transformed E. coli (DH5a competent cells, Gibco-BRL, Paisly, Scotland) were grown overnight at 37°C in 1liter volumes under continuous shaking. Cells were harvested by centrifugation and lysed by ultrasonification. Fusion protein particles were purified by ultracentrifugation and gel filtration on Sephacryl S-500 (Pharmacia, Uppsala, Sweden) essentially a s described [Stahl and Murray, 19891. Purified proteins were characterised by sodium dodecyl sulfate polyacrylamide gel electrophoresis (SDS-PAGE).
gens (Ag) in EIA, protein particles (4 mg/ml) were incubated with 0.1% SDS (60 min, 37°C) and subsequently diluted in 10 mM Na bicarbonate buffer pH 9.6. Polyvinylchloride (PVC) assay plates (Costar, Cambridge, MA, USA) were coated overnight at 4°C with the SDStreated proteins at a final concentration of 1 pg/ml. Plates were washed with 10 mM Tris buffer pH 7.5, air dried, and stored at 4°C until use. As conjugate a goat anti-human immunoglobulin was used, labelled with 1,250 units horseradish peroxidase per 10 mg IgG [Tijssen and Kurstak, 19841. EIA was performed by standard procedures on 1/40 or 1/100 dilutions of serum. As positive control, serum from a donor positive in PCR for HCV [De Beenhouwer et al., 19921 was used. The cutoff ((3.0.) was arbitrary calculated from the optical densities (O.D.) a s follows: O.D. negative control
+ 1/10 O.D. positive control.
All results were expressed as the ratio between the O.D. of the sample to the O.D. of the C.O. A ratio of > 1.0 was considered positive. To discriminate between HBc and HCV antibodies, initially reactive sera were reassayed after HBc neutralisation by the addition of 10 pg/ml HBc antigen (Ag333) to the dilution buffer. This treatment may also exclude false positives caused by anti-E. coli.
Fig. 2. SDS-PAGE of recombinant protein after reduction. Samples in the tracks from left to right are: lane 1: molecular weight (MW) markers (MW 110,000; 84,000; 47,000; 33,000; 24,000; 16,000);lane 2 Ag333; lane 3: Ag405; lane 4: Ag520; lane 5 Ag601; lane 6 MW markers.
HCV Core Antibodies and Carrier State
261
TABLE I. EIA for HBc and HCV Core Antibodies* Antigens Sera Negative HBc Ab positive HCV Ab positive HCV and HBc Ab positive HBc Ab positive (HBV carrier)
0.118 8.950
Ag333 HBc neutralisation 0.112 0.696
0.115 6.630
Ag520 HBc neutralisation 0.128 0.688
0.110 0.423
Ag601 HBc neutralisation 0.119 0.110
Evaluation Negative Negative
0.143
0.132
9.298
9.407
6.210
6.830
Positive
6.448
0.608
6.211
5.089
2.580
2.110
Positive
7.892
6.208
7.478
5.419
4.078
2.551
I.D.
*Five representative sera were assayed by EIA using different antigens, with and without HBc neutralisation. Results were expressed as ratios. Evaluation as positive, negative or indeterminate (1.D.)was related to the presence of HCV antibodies. Sera with HBc Ab were also positive for HBe Ab.
Other HCV Assays First and second generation EIA (Ortho, Beerse, Belgium, or Abbott Wiesbaden, Germany) and Immunoblots (4-RIBA, Chiron-Ortho, Emeryville, CA, USA; INNO-LIA, Innogenetics, Antwerp, Belgium) were performed according to the directions provided by the manufacturers. Diagnostic PCR was performed on the noncoding region of the HCV genome using nested primers [De Beenhouwer et al., 19921.
TABLE 11. HCV Core Antibody in First Generation EIA Positive Sera* Ag520 EIA/Ag333 EIA Initial screening Assay with HBc neutralisation Total number of sera Ag601 EIA/Ag333 EIA
RESULTS Expression, Purification, and Characterisation of Recombinant Proteins After transformation of E. coli with the different plasmid constructs, the recombinant proteins were spontaneously produced, without benefit of IPTG-induction. On the Sephacryl-S500 they behaved a s HBclike particles and eluted in the higher molecular weight region of the column immediately after the void volume fraction. Yields of purified recombinant proteins differed appreciably: A333, Ag405, and Ag520 were obtained in yields of about 15 mglliter culture, while Ag601 yielded only about 1mg/liter. SDS-PAGE revealed the purified reduced proteins as doublets with average molecular weights of respectively 20,000, 15,000, 24,000, and 32,500 (Fig. 2). For the first 3 antigens a high degree of purity was obtained, but for Ag601 there was still some contamination with E. coli proteins.
EIA for HBc and HCV Core Antibodies For optimal coating of the PVC assay plates, a pretreatment of the recombinant HBc-like particles with 0.1% SDS was found beneficial. To evaluate the different antigens, representative sera were analysed. Results are shown in Table I. As expected Ag333 only measured antibodies to HBc and Ag520 measured antibodies to both HBc and HCV. Ag601 also reacted strongly with HCV antibodies, but showed only a minor reactivity for HBc antibodies.
Initial screening Assay with HBc neutralisation Total number of sera
HCV core antibody Positive I.D. Negative 91 14 280 J.
7 98
4
1 1
.L
6 286
HCV core antibody Positive I.D. Negative
2 7 98
1 1
0
286
*Three hundred eighty-five first generation EIA positive sera were assayed by Ag520 and Ag601 EIA. Positive sera were analysed on Ag333 for HBc antibodies and if present HBc neutralisation was performed. In the Table the number of sera with a positive, negative, or indeterminate (I.D.) result is given.
Discrimination between HBc and HCV antibodies could be made by HBc neutralisation, except in sera with very high titers of HBc antibodies as found in HBV carriers. In such cases a n indeterminate result (1.D.)had to be accepted.
EIA for HBc and HCV Core Antibodies in Sera Received for HCV Confirmation In our blood bank we have a collection of 385 sera found positive in first generation HCV-EIA [De Beenhouwer et al., 19911. When assayed on Ag520,104 sera were initially reactive with a n average ratio of 6.22 1.41 and 281 were negative with a n average ratio of 0.33 ? 0.14 (Table 11).Of the 105 initially reactive sera only 14 were also positive on Ag333 and therefore remained provisionally indeterminate. To discriminate further, these 14 sera were reassayed after HBc neutralisation. For 7 sera the signal was neutralised on Ag333 while on Ag520 it remained essentially un-
*
Claeys et al.
262 TABLE 111. Correlation of Several HCV Assays With HCV Core Antibody* HCV core antibody 98 Positive 287 Negative Second generation EIA I positive \ /98\ 4-RIBA INNO-LIA 4-RIBA INNO-LIA Immunoblot I( L k! L “reactive” 95 98 JL 0 0 87 89 PCR positive
I
H
t
*On 385 first generation EIA positive sera, second generation EIA (Ortho and Abbott) and PCR were performed. The 179 (98 + 81) sera positive in both second generation EIA were further characterised by immunoblotting (CRIBA and INNO-LIA). For both the HCV core Ab positive and negative group the number of positive or “reactive” sera in different assays is given.
changed, for 6 sera i t was neutralised on bot,h antigens, and for one serum no definite neutralisation was observed. These sera could therefore be allocated respectively as positive, negative, and indeterminate for HCV core antibodies. When assayed on Ag601, a similar distribution of sera was obtained but the 6 sera t h a t were only positive for HBc antibodies, were negative already in the initial screening.
Significance of HCV Core Antibodies All 385 sera were also assayed in 2 commercial second generation EIA (Abbott and Ortho) and PCR. The 179 sera positive in both second generation EIA were characterised further by immunoblotting (4-RIBA and INNO-LIA) [De Beenhouwer et al., 19921. The number of positive or “reactive” (Immunoblot)sera in the different assays was compared in the group of HCV core Ab positives and negatives (Table 111). All 98 HCV core Ab positive sera were positive in both second generation EIA and “reactive” by INNOLIA immunoblotting. On 4-RIBA immunoblotting 95 (96.9%) were “reactive” and 3 were “C22-3-I.D.” (only reaction with the c22-3 band). All 89 (90.8%)PCR positive sera belonged to this group. The band pattern of the remaining 9 PCR negative sera could not be distinguished from that of the positives. Of the 287 HCV core Ab negative sera 81 (28.9%) were positive in both second generation EIA. Only 2 (2.5%) of these 81 sera were “reactive” on 4-RIBA immunoblotting and 3 (3.7%) on INNO-LIA. None was “reactive” on both. In addition colour intensity was weak and only reactions against non-structural proteins were observed. All 287 HCV core Ab negative sera were PCR negative. Evaluation of Ag520 E I A as Screening Assay A series of 3,708 sera were assayed for HCV antibodies with the Ortho first generation EIA and the Ag520 EIA in parallel. The results are summarized in Table IV. On initial screening with Ag520 EIA, 131(3.5%) sera were positive. Only 2 sera were non-reactive on Ag333 and could therefore be immediately interpreted HCV core Ab positive. HBc neutralisation on the 129 sera reactive on both antigens, revealed another 3 positive
-
TABLE IV. Evaluation of Ag520 EIA as Screeninn Assav* Ag520 EIA/Ag333 EIA Initial screening Assav with HBc neutralisation Total number Ortho first generation EIA
Positive 2
3
I.D. 129
Negative 3,577
5
3 3
123 3,700
21
-
3,687
*Three thousand seven hundred eight randomly collected sera were screened by Ortho first generation EIA and by Ag520 EIA-Ag333 EIA. Indeterminate (I.D.) sera were reassayed after HBc neutralisation. The number of positive, I.D., or negative sera is shown.
sera, 123 negative sera, and 3 indeterminates. Thus 5 (0.13%) sera were positive in the HCV core Ab EIA against 21 (0.57%) in the Ortho first generation EIA. Only 2 sera were positive in both assays. Confirmation of positive sera was performed by second generation EIA, immunoblotting, and PCR (Table V). Only 3 of the 24 initially positive sera could be serologically confirmed. All 3 were reactive on Ag520, but only 2 were detected by the Ortho first generation EIA. All 3 sera were PCR positive.
DISCUSSION HBc- and HBc-HCV plasmid constructs were made in the pBTac2 expression vector and recombinant HBcHCV proteins were produced in transformed E . coli. Unexpectedly, recombinant proteins were spontaneously produced and no influence of stimulation by IPTG was observed. Yields of recombinant proteins averaged 15 mg/liter culture except for the larger protein construct (Ag601).These data confirm earlier observations t h a t the nucleocapsid protein of HBC can be used as a carrier to express foreign polypeptides in E . coli [Stahl and Murray, 19891. Recombinant proteins apparently assembled into HBV core-like particles and could therefore be easily purified by sequential ultracentrifugation and gel filtration. In this respect Sephacryl S-500 resulted in a considerably better purification of the particles than Sepharose 4B-CL. HBc fusion proteins were shown to be good immunogens CStahl and Murray, 19891. We explored the possibility of using them a s coating material for EIA. Intact
HCV Core Antibodies and Carrier State
Positive in screening
263
TABLE V. Confirmation of HCV Status in EIA Positive Sera* Ortho first generation EIA
Y 1 2 / i 51 ! f
Ag520 EIA
19
Positive in both second generation EIA “Reactive” in both immunoblots PCR positive
0
2 .L
2
.L 1
-1
1 L 1
*The HCV s t a t u s of t h e 24 sera, reactive in t h e Ortho first generation EIA or t h e Ag520 EIA,w a s evaluated by additional serologic assays and by PCR. The number of positive sera in each assay is shown.
HBc-like particles had a low affinity for PVC, but the conformational changes provoked by SDS treatment [Sallberg e t al., 19901 resolved this problem. As expected the HBc fusion proteins had a n affinity for both HBc and HCV antibodies. However compared to the unsubstituted HBc protein (Ag333), the af€inity of the fusion proteins for HBc antibodies decreased with the length of the substituent. Indeed, while Ag520 with a substituent of 72 AA still measured HBc antibodies quite effectively, Ag601 with a substituent of 138 AA only detected high levels (Table I). By the addition of HBc antigen (Ag333)to the dilution buffer the majority of HBc reactions were neutralised. Thus depending on the set-up HBc fusion proteins can be used to measure both HBc and substituent or only substituent antibodies. In NANBH, testing for both HCV and HBc antibodies could potentially be valuable. Of the 385 first generation EIA positive sera only 98 126.4%) were positive for HCV core antibodies (Table 111).This discrepancy appeared to be mainly due to the poor specificity of the first generation EIA in a population of blood donors, where the frequency of HCV positivity is low [De Beenhouwer et al., 19921. The present HCV core Ab EIA appeared to be extremely specific in distinguishing false from true positives. Of the 287 HCV core Ab negative sera only 81 were found by both second generation EIA and from those none was “reactive” on both immunoblots. In contrast all 98 HCV core Ab positive sera were confirmed in both second generation EIA and 95 were “reactive” in both blots (Table 111).Moreover the 89 PCR positive sera were all found in this group and the remaining 9 could serologically by no means be further distinguished. Thus only these 98 donors should be considered as HCV carriers. To further evaluate the sensitivity and specificity of HCV core Ab testing we screened 3,708 sera with both Ag520 EIA and a first generation EIA (Ortho). The data obtained (Table V) essentially confirm the reported sensitivity: 3 sera were “reactive” by immunoblotting and positive by PCR and all 3 were detected by the HCV core Ab EIA while the first generation EIA, based on non-structural antigens, only found 2 of these sera. Specificity of the Ag520 EIA was slightly hampered by the interference of HBc antibodies resulting in
2 false positive and 3 indeterminate sera. Specificity was still considerably better than that of the first generation EIA. The present data also allowed the evaluation of HBc testing as a surrogate assay for NANBH. Of the 98 positive sera in the confirmation study 7 were also positive for HBc antibodies. In the screening study 129 HBc Ab positive sera were found of which one was HCV Ab positive. Thus in our population of blood donors a poor association was observed between HCV infection and the presence of HBc antibodies. In conclusion, confirmation of HCV first generation EIA positive donors, by HCV core Ab determination was reliable in identifying viral carriers. Screening for HCV core Ab resulted in a higher sensitivity and specificity than screening with first generation EIA.
ACKNOWLEDGMENTS The authors thank Ortho Diagnostics, Abbott Diagnostic, and Innogenetics for supplying test kits. Primers were synthesized by Prof. F. Van Leuven (K.U. Leuven-Genetics). We are grateful to Mrs. W. Mertens, E. Noe, H. Verhaert, and G. Wuyts for technical support and to Dr. K. Gautama for data management. REFERENCES Choo QL, Kuo G, Weiner AJ, Overby LR, Bradley DW, Houghton M (1989):Isolation of a cDNA clone derived from blood borne Non-A, Non-B viral hepatitis genome. Science 244:359-362. De Beenhouwer H, Verhaert H, Claeys H, Vermylen C (1992):Confirmation of HCV positive blood donors by immunoblotting and PCR. Vox Sanguinis (in press). Kuo G, Choo QL, Alter HJ, Gitnick G1, Redeker AG, Purcell RH, Miyamura T, Dienstag JL, Alter MJ, Stevens CE, Tegtmeier GE, Bonino F, Colombo M, Lee WS, Kuo C, Berger K, Shuster JR, Overby LR, Bradley DW, Houghton M (1989):An assay for circulating antibodies to a major etiologic virus of human non-A, non-B hepatitis. Science 244:362-364. Okamoto H, Okada S, Sugiyama Y, Yotsumoto S, Tanaka T, Yoshizawa H, Tsuda F, Miyakawa Y, Mayumi M (1990): The 5’terminal sequence of the hepatitis C virus genome. Japanese Journal of Experimental Medicine 60:167-177. Sallberg M, Norder H, Magnius LO (1990):Comparisons of class and subclass distribution of antibodies to the hepatitis €3 core and B e antigens in chronic hepatitis B. Journal of Medical Virology 3O:l-6.
264 Sambrook J, Fritsch EF, Maniatis T (1989): Molecular Cloning, A Laboratory Manual. Cold Spring Harbor: Cold Spring Harbor Laboratory Press. Stahl SJ, Murray K (1989):Immunogenicity of peptide fusions to hepatitis B virus core antigen. Proceedings ofthe National Academy of Science USA 86:62834287. Tijssen P, Kurstak E (1984):Highly efficient and simple methods for the preparation of peroxidase and active peroxidase-antibody conjugates for enzyme immunoassays. Analytical Biochemistry 136:451-457.
Claeys et al. Van der Poel CL, Cuypers HTM, Reesink HW, Weiner AJ, Quan S, di Nello R, Van Boven J J P , Winkel I, Mulder-Folkerts D, ExelOehlers PJ, Schaasberg W, Leentvaar-Kuypers A, Polito A, Houghton M, Lelie PN (1991):Confirmation of hepatitis C virus infection by new 4-antigen recombinant immunoblot assay. Lancet 337:317-319. Wick MR, Moore S, Taswell H F (1985):Non-A, non-B hepatitis associated with blood transfusion. Transfusion 25:93-101.
