Molecular and Cellular Probes (1990) 4, 43-51

Amplification of hepatitis delta virus RNA sequences by polymerase chain reaction : a tool for viral detection and cloning

A . L . Zignego, 1 ' 3 P. Deny,1 ' 5 C. Feray, l A. Ponzetto, 2 P. Gentilini ; 1 '4 P. Tiollaisl and C. Bréchot 'Unité de Recombinaison et Expression Génétique (Inserm U 163/CNRS UA 271) Institut Pasteur, 25-28 rue du Dr . Roux 75724, Paris Cedex 15, France, 3 2 Division of Gastroenterology, Molinette Hospital, Turin, Italy, lstituto di Clinica Medica 11, University of Florence, School of Medicine, Florence, Italy, 4 lnserm U 75 Chu Necker and Unité d'Hépatologie, Hôpital Laennec, Paris, France and 'Service de Bactériologie-Virologie, Hôpital Avicenne, 93009-Bobigny Cedex, France (Received 28 August 1989, Accepted 14 September 1989)

In this investigation we have evaluated the feasibility of using the polymerase chain reaction (PCR) for hepatitis delta virus (HDV) RNA detection, cloning and sequencing . Total RNA from HDV-infected liver and serum samples was purified and Moloney murine leukaemia virus (M-MLV) reverse transcribed . HDV cDNA was then directly amplified with Taq polymerase using three pairs, of specific primers . It was possible to amplify a region of about 1200 by in three partially overlapping fragments including the whole HDAg-ORF . A DNA fragment of the expected size was repeatedly obtained from an initial sample of less than 0 .1 pg of liver RNA and from 10 pl of infected serum . An amplified fragment of 359 by obtained by PCR from an infected woodchucks' liver was sequenced . The sequence was 91 . 8% and 98 . 6% identical to previously published HDV sequences. In addition, amplified and ' 2 P-radiolabelled HDV sequences were shown to hybridize specifically to HDV RNA extracted from HDV-infected liver and serum . In conclusion this technique promises to be of great value in the appraisal of HDV infection, rapid synthesis of HDV probes and analysis of the genetic variability of the virus .

KEYWORDS : HDV RNA reverse transcriptase polymerase chain-reaction, HDV RNA

cloning and sequencing, HDV probes synthesis .

Address correspondence to : Anna Linda Zignego, M .D . PhD ., Institut Pasteur, Unité de recombinaison et expression génétique, 28 rue du Dr. Roux, 75015 Paris Cedex, France .

0890-8508/90/010043 + 09 $03 .00/0

43

© 1990 Academic Press Limited



44

A. L. Zignego et al.

INTRODUCTION Hepatitis delta virus (HDV) is a defective RNA virus dependent on hepatitis B virus (HBV) for infection . The virus particle has a diameter of about 36 nm and a chimeric structure composed of the specific viral antigen (HDAg) and nucleic acid (HDV RNA) together with molecules of the helper hepadnavirus envelope proteins in the outer coat .' The viral genome (HDV RNA) is circular" and about 1700 bases in length (1679-1683) . 2 HDV sequences have been determined 2,3,5-e and one of the open reading frames (ORF) identified on the antigenomic strand (+) has been assigned to the gene coding for the delta antigen (HDAg) .2 Infection with HDV can either occur simultaneously with acute HBV infection (coinfection) or may be superimposed on chronic HBV infection (superinfection) . In cases of coinfection the acute disease is usually self-limiting although a fulminant course may also occur . Superinfection by delta virus on a HBV chronic carrier state is generally associated with severe liver disease . Recently, direct detection of HDV by molecular hybridization techniques has been shown to be a powerful tool in the appraisal of HDV infections. In particular, HDV RNA detection in serum by the 'slot-hybridization' method has been used as a diagnostic test in acute infections . Indeed, none of the three conventional HDV markers (anti-HD, IgM anti-HD and HDAg) can be systematically detected in acute infections . Similarly, RNA hybridization is a useful technique in immunocompromised patients, where the pattern of serological HDV markers may be modified (Zignego et al ., in press) . However, standard hybridization procedures so far used for HDV RNA determination may be not sufficiently sensitive to detect very low levels of HDV replication . This might be important in the first phase of acute infection or in the late stages of chronic infection . As a consequence a rather large percentage of subjects with chronic hepatitis have been found to be HDV RNA-negative in serum although liver HDAg was positive . 9-" However, several studies in different systems indicate that genomic amplification by polymerase chain reaction (PCR) may allow a very sensitive identification of viral DNA and RNA sequences . ' 2- 13 Furthermore, the difficulty in HDV RNA cloning from infected samples renders approaches aiming to obtain different HDV probes or to compare HDV sequences from different isolates laborious . We have therefore evaluated the feasibility of using the PCR for HDV RNA detection, HDV probe synthesis and to facilitate cloning .

MATERIALS AND METHODS Samples Serum and liver samples from two patients with chronic active delta hepatitis (IgM anti-HD and liver HDAg positive), were used as well as those from a chimpanzee and a woodchuck with acute HDV infections .

HDV RNA purification Total RNA was extracted from liver samples as previously described ." Serum samples were digested for 12 h at 37 ° C in 2 volumes of a lysis buffer containing 0 . 4 M



Amplification of HDV RNA sequences by PCR

45

NaCl, 40 mm EDTA, 3% SDS and 500 µg ml - ' proteinase K . RNA was then extracted with phenol-chloroform, precipitated with 1 volume of propan-2-ol and redissolved in 50 tl of diethyl pyrocarbonate-treated water (DEPC H 2 O) (Zignego et al., in press) .

Slot-blot analysis RNA was denatured in formaldehyde at 65°C for 15 min then applied to a nylon membrane by suction using a hybrislot apparatus . A single-stranded RNA (ssRNA) probe of antigenomic polarity was obtained from a cDNA clone of HDV RNA (pDelta 1) (Fig . 1) of 391 by (kindly provided by Dr Kos, TNO Centre, Rotterdam, The Netherlands)' subcloned into the Pstl site of the Bluescribe plasmid (Stratagene, San Diego) . Radiolabelling of the ssRNA was carried out as previously described . 14 The ssRNA probe used to hybridize to the RNA on the filters had a specific activity of 1-2 109 cpm µg -1 (Zignego et al., in press). Hybridization was performed at 55°C as previously described . 14

Fig . 1 . HDV genetic map with the different primers and probes . S 2-5 and S3 : oligonucleotide probes . P-delta 1 : cDNA probe. 2A, 25; 3A, 3S and 5A, 5S: HDV specific primers. R2, R3 and R5 : HDV RNA regions amplified . HDV sequence was from Makino et al .'

46

A. L. Zignego et ai.

HDV RNA reverse transcriptase-polymerase chain reaction (HDV RNA RT-PCR) Oligonucleotide primers Oligonucleotide primers were synthesized on a 308 A DNA synthesizer (Applied Biosystems) by the methoxyphosphoramidite method . For HDV RNA reverse transcriptase PCR (HDV RNA RT-PCR), three sets of oligonucleotide primers were used . Their sequences were selected according to available HDV sequences''' and are as follows : 2A 2S 3A 3S 5A 5S

= 5'-TAT TCT TCT TTC CCT TCT-3' = 5'-AGA AGT TAG AGG AAC TGC-3' = 5'-CGA CCT GGG CAT CCC AAG GAG GAC G-3' = 5'-GGA GAC CCG AAG CGA GGA GGA AAG TA-3' = 5'-AGA AAA GAG TAA GAG TAC TCA GGA C-3' = 5'-CAA ACC TGT GAG TGG AAA CCC GCT T-3'

(S = sense strand and A = antisense strand polarity) (Fig . 1) .

RT-PCR After denaturation at 100°C for 3 min total RNA samples were cooled in dry ice . HDV cDNA-RNA hybrids were synthesized by adding to these samples the following components in a final volume of 20µl : a downstream primer consisting of an antigenomic delta specific sequence (2A, 3A or 5A=50 pmol) ; the Moloney murine leukaemia virus reverse transcriptase (M-MLV RT) (BRL) (200 U jig - ' of liver RNA or 200 ltl of serum) ; 1 mm dNTPs ; 4 gl of 5 reverse transcriptase buffer (50 mm Tris-HCI, pH 8 . 3, 75 mm KCI, 3 mm MgCl 2 , 10 mm DTT and 100 jig ml -1 BSA, pH 8) and RNAsin (1 U µl -1 ) . The sample was incubated at 45°C for 30 min and then heated to 95°C for 5 min to inactivate reverse transcriptase . Finally, PCR amplified HDV cDNA was generated 12 as described after addition of a genomic delta oligonucleotide primer as the upstream primer (2S, 3S or 5S = 50 pmol) and cloned Taq polymerase (Perkin Elmer Cetus) (1 U jig - ' of liver RNA or 200 gl of serum) in a final volume of 100 µl. A Perkin-Elmer Cetus automatic thermal cycler was used . The first denaturation was performed at 94°C for 5 min and for 1 min in subsequent cycles and annealing and elongation at 55°C and at 72°C for 1 min respectively ; the 40th cycle had a 5 min elongation step to fully extend incomplete DNA fragments . To eliminate possible contamination, all reagents were prepared with the use of disposable pipettes (Microman, Gilson), aliquoted and stored- in new disposable containers . In addition all reagents were tested to verify the absence of HDV RNA or previously amplified HDVcDNA .

Analysis of amplified cDNA 10 gl of the PCR mix was subjected to electrophoresis on a 1 . 5% agarose gel in Trisborate buffer (TBE) . The bands were visualized by staining with ethidium bromide



Amplification of HDV RNA sequences by PCR

47

(PCR-EB) and by Southern blotting (PCR-SB) on a 'Gene Screen plus' (NEN) nylon membrane . The filters were prehybridized in 3 sodium chloride citrate (SCC, 1SOrnrno!! -, sodium chloride, 15nnmno!! -, sodium citrate), 5 Denhardt's solution (0-02% ficoU,O'O2% po!yviny!pyrro!idone, O'0296 bovine serum albumin), 0-5 0/o SDS and 30% formamide . The filters were incubated at 42C overnight with one of two "P-labelled oligonucleoticle probes specific for the amplified fragments as follows (Fig . 1) : 52'S=5'-TTG TCG GTG AATC[TCCC ClC AGAGGC [TCTTC CIA GGT C-3' S3~~5'-CCACAG AAA AGTCG[TCTCCC TTG CCCATCCCA GTG CA-3' Filters were then washed in 2 SS{, 1To SDS for 15 min at room temperature and in 0'2S6[ ' 0196SDS6or1Onoinat5G"[ . The autoradiograph was exposed for 3 h and 16 h with a single intensification screen .

Cloning and sequencing of amplified products in M13 vector To facilitate the cloning of the RNA, restriction sites (Hind UI and Pstl) were included at the 5'-ends of the primers 2S and 2A . After amplification, the HDVcDNA was purified from an agarose gel digested with 12 units of each Pstl and Hind U![ The resulting Hind Ill-Pstl fragment was cloned, and its nucleoticle sequence was obtained by the dideoxy termination method using a commercially-available kit (Sequenase, 1,1S . Diochennicu! Corporation, Cleveland) .

Synthesis o f HDV probes from amplified products To prepare HDV cDNA probes, amplified products were separated on agarose gek, was radiolabelled by nick translation or the random primer labelling method using commercially available kits (Amershamh The specific activities of HDV probes was 3 x 1W-1 x 109 cpm pg - ' . in order to obtain larger amounts of cDNA for use as probes, amplification products were generally reamplified two to three times .

With 1gg of total RNA from infected liver or 30 gI of serum samples, amplification of different HDV RNA sequences was obtained . indeed three partially overlapping HDVkNA regions named region 2(3S9bp), region 3(478bu)and region (716bo) (Fig . 1) were obtained with the different sets of primers . About 1200 bases pairs of the HDV genome, including the whole HDAg ORF, were identified . It is clear, therefore, that the secondary structures of HDV RNA did not prevent the synthesis of cDNA and thus the amplification experiment . Specificity of HDV RI PCR was shown by : (1) the negative results observed with controls including RNA extracted from normal human liver and serum (Fig . 2, lanes c-) and baker's yeast (Fig . 2, lanes ry) and DNA from lambda bacteriophage and various plasmids (data not shown) ; (2) the negative results obtained with the



48

A. L . Zignego et ai.

Fig. 2. Amplification of HDV RNA sequences after ethidium bromide staining (A) and hybridization with 32P-labelled oligonucleotide probes (B). Panel 1 : amplification with primers 5A, 5S . Panel 2 : amplification with primers 3A, 3S. Lanes M : reference molecular weight markers (123 by ladder) ; lanes c- : negative control (normal liver RNA) ; lanes F : HDV+ woodchucks' liver; lanes S: serum of a patient with delta chronic active hepatitis ; lanes Ss: HDV+ chimpanzee's serum ; lanes -rt : the same reaction mix used for S without reverse transcriptase; lanes ry : baker's yeast RNA; lane -R: distilled water.

reaction mixture without the reverse transcriptase (Fig . 2, lanes -rt) ; (3) the absence plification using distilled water rather than RNA (Fig . 2, lane -R) ; (4) the size of of t e fragments amplified which corresponded to that expected from the genetic map . An example is shown in Fig . 2, where the products of amplification Al and A2, from region 5 and 3 respectively, have the expected size of 716 and 478 by respectively . Similarly in Fig . 3 the amplification products of region 2 were, as predicted, 359 base pairs long ; (5) the specificity of the reaction was further confirmed by hybridization of 32 P-labelled amplified fragments to a 1 .7 kb band on Northern blots from HDV-infected liver or serum (data not shown) . The sensitivity of RT PCR for the detection of HDV sequences was tested in experiments using serial dilutions of RNA extracted from serum and liver samples . A band in SB PCR analysis has been obtained with 01 pg of infected liver or 8 pl of infected serum (Fig. 3, panel B) . Parallel analysis of the same samples by the slot blot technique (Fi2 . 3 . Danel C) showed that sensitivity of HDV RT PCR, was 10 4-10 6



Amplification of HDV RNA sequences by PCR

49

greater than that of slot blot analysis . In fact slot blot analysis was not able to clearly detect HDV sequences using less than 1 ng of liver RNA and 8 gl of serum . In the different amplification experiments from serial dilutions of liver RNA the amount of amplification product regularly increased with increasing concentrations of RNA, across the range 0 . 1 pg-1 gg in reaction mix . By contrast, the efficiency of HDV RT PCR was lower with initial serum volumes above 100 Id but reliable PCR results were obtained when the sera were diluted . The best results were obtained with 8-10 gl of serum (Fig . 3, panels A and B, lanes 2) . This may reflect presence of inhibitory factors for PCR in the serum (human and animal) .

1

359 .

0

9

8

7

6 5

4

3

2

1

0

0 Fig. 3 . Comparative analysis of slot blot and RTPCR for HDV RNA detection . Tenfold serial dilutions of the serum from a patient with delta chronic active hepatitis . In panels A and B lanes 1, 2, 3, 4, 5, 6, 7 and 8 correspond respectively to 80 µl, 8 pl, 800 nl, 80 nI, 8 nI, 800 pl, 80 pl and 8 pl of the initial serum sample. Amplification was performed with primers 2A, 25 . Panel A : ethidium bromide staining . Panel B :

hybridization with 72 P-labelled oligonucleotide probe 52 ; 131, B2 and B3 : 2 h, 1 day and 3 days exposure . Panel C : slot blot analysis of the same dilutions.

50

A . L. Zignego et al.

Cloning and sequencing of amplification products For one selected sample (infected woodchuck liver) we were able to clone the amplified sequence corresponding to region 2 of HDV genome . The sequence was 359 by and showed a 91 . 8% and 9&4o/o identity with the previously sequenced 2-7 HDV genomes . In summary this study shows that RT PCR can be used for very sensitive and specific HDV RNA detection and rapid generation of HDV probes as well as for HDV RNA cloning and sequencing . This approach should be very useful , when the amount of serum HDV RNA is low as in the first phase of acute infection or late stages of chronic infection . HDV RT PCR will also allow an easy synthesis of HDV probes corresponding to different regions of the viral genome . Finally, HDV RT PCR, with subsequent cloning and sequencing will be an important tool for the study of genetic variability of HDV isolates. To date, only a few sequences are available . In the present investigation we identified a divergence of up to 8 . 2% with one of the available sequences . 7 Further experiments will allow the comparison of a number of different isolates .

ACKNOWLEDGEMENTS We gratefully acknowledge Valérie Thiers for constant assistance in this study, Marie-Annick Buendia, Pierre Sonigo, Hend Farza and Christine Pourcel for helpful discussions . This work was supported by grants from ARC, LNC, CNAM, INSERM .

REFERENCES 1 . Bonino, F ., Hoyer, B ., Shih, J . W .-K ., , Rizzetto, M ., Purcell, R. H . & Gerin, J . L . (1984) . Delta hepatitis agent : structural and antigenic properties of the delta-associated particle . Infectious Immunity 43, 1000-5 . 2 . Wang, K.-S ., Choo, Y .-L ., Weiner, A . J . et al. (1986). Structure, sequence and expression of the hepatitis delta virus genome . Nature 323, 508-13 . 3 . Kos, A ., Dijkema, R., Amberg, A . C., van der Meide, P. H . & Schellekens, M . (1986) . The hepatitis delta virus possesses a circular RNA . Nature 322, 558-60 . 4 . Chen, P .-J ., Kalpana, G ., Goldberg, J . et al. (1986) . The structure and replication of the hepatitis delta virus genome . Proceedings of the National Academy of Sciences U .S.A . 83, 8774-8. 5 . Denniston, K. J ., Hoyer, B. H ., Smedile, A ., Wells, F. V ., Nelson, J . & Gerin, J . L. (1986) . Cloned fragment of the hepatitis delta virus RNA genome ; sequence and diagnostic application . Science 232,873-5 . 6 . Saldanha, J . A ., Tomas, H . C ., Monjiardino, J . P. (1987) . Cloning and characterisation of a delta virus cDNA sequence derived from a human source . Journal of Medical Virology 22, 323-31 . 7 . Makino, S., Chang, M .-F ., Shieh, C .-K. et al. (1987) . Molecular cloning and sequencing of a human hepatitis delta virus RNA . Nature 329, 343-6 . 8 . Kuo, H . Y ., Goldberg, J ., Coates, L., Mason, W ., Gerin, J . & Taylor, J . (1988) . Molecular cloning of hepatitis delta virus RNA from an infected woodchuck liver : sequence, structure and applications . journal of Virology 62, 1855-61 . 9 . Smedile, A ., Rizzetto, M., Demminston, K . et al . (1986). Type D hepatitis : the clinical significance of hepatitis D virus RNA in serum as detected by a hybridisation-based assay . Hepatology 6, 1297302 . 10 . Rasshofer, Buti M ., Esteban, R ., Jardi, R . & Roggendorf, M. (1987) . Demonstration of hepatitis D virus RNA in patients with chronic active hepatitis . journal of Infectious Diseases 157, 191-5 .



Amplification of HDV RNA sequences by PCR

51

11 . Saldanha, J ., di Blasi, F ., Bias, C . et al. (1989) . Detection of hepatitis delta virus RNA in chronic liver disease . Hepatology 9, 23-28 . 12 . Saiki, R. K ., Gelfand, D . H . & Stffel, S . (1988) . Primer directed enzymatic amplification of DNA with a thermostable DNA polymerase . Science 239, 487-91 . 13 . Thiers, V., Nakajima, E ., Kremsdorf, D . et al . (1988) . Transmission of hepatitis B from hepatitis B seronegative subjects . Lancet i3, 1273-6. 14 . Farza, H ., Salmon, A . M., Hadchouel, M . et al . (1987) . Hepatitis B surface antigen gene expression is regulated by sex steroids and glucocorticoids in transgenic mice . Proceedings of the National Academy of Sciences U .S.A . 84, 1187-91 . 15 . Melton, D . A ., Krieg, P . A ., Rebagliati, M. R ., Maniatis, T., Zinn, K . & Green, M . R. (1984) . Efficient in vitro synthesis of biologically active RNA and RNA hybridisation probes from plasmids containing a bacteriophage SP6 promoter . Nucleic Acids Research 12, 7035-56 .