- Email: [email protected]
A novel dicistronic AAV vector using a short IRES segment derived from hepatitis C virus genome
Gene 200 (1997) 157–162
A novel dicistronic AAV vector using a short IRES segment derived from hepatitis C virus genome Masashi Urabe a, Yoko Hasumi a, Yoji Ogasawara a, Takashi Matsushita a,b, Nobuhiko Kamoshita c, Akio Nomoto c, Peter Colosi b, Gary J. Kurtzman b, Kiyotake Tobita d, Keiya Ozawa a,* a Department of Molecular Biology, Institute of Hematology, Jichi Medical School, Tochigi 329-04, Japan b Avigen Inc., Alameda, CA 94502, USA c Department of Microbiology, Institute of Medical Science, University of Tokyo, Tokyo 108, Japan d Department of Virology, Jichi Medical School, Tochigi 329-04, Japan Received 29 January 1997; accepted 19 June 1997; Received by T. Sekiya
Abstract Adeno-associated virus (AAV ) vectors have a limited capacity for packaging DNA. To insert both a therapeutic gene and a selectable marker gene in the same AAV vector efficiently, we developed a novel dicistronic AAV vector containing a 230 base pairs (bp) internal ribosome entry site (IRES ) element derived from hepatitis C virus (HCV ) genome and a 420 bp blasticidin S-resistance gene (bsr) as a small selectable marker in the second cistron. The 650 bp HCV IRES-bsr construct was placed downstream of the 3∞ end of the luciferase gene (Luc) under the control of the human cytomegalovirus (CMV ) promoter. This dicistronic gene conferred blasticidin S-resistance to 293 cells besides luciferase activity, when examined not only by transfection but also by transduction using AAV vectors. The dicistronic AAV vector harbouring HCV IRES-bsr is capable of expressing a therapeutic gene of up to ~3.6 kilobases (kb) (including promoter/enhancer elements) as well as a selectable marker gene. If a selectable marker gene is not necessary, this vector is able to incorporate two different kinds of therapeutic genes more easily than that containing EMCV IRES. The dicistronic AAV vector described here is useful for expressing many kinds of cDNA besides a selectable marker. © 1997 Elsevier Science B.V. Keywords: Blasticidin S-resistance gene; Gene therapy; Gene transfer; Selectable markers
1. Introduction The AAV vector is emerging as highly promising for gene therapy because of several unique properties that the other viral vectors do not share. AAV is a member of the Parvoviridae, of which the genome is only 4.7 kb * Corresponding author. Tel: +81 285 442111, ext. 3481; Fax: +81 285 448675; e-mail: [email protected] Abbreviations: aa, amino acid(s); AAV, adeno-associated virus; bGal, b-galactosidase; BiP, immunoglobulin heavy chain binding protein; bp, base pairs; bsr, blasticidin S-resistance gene; CMV, human cytomegalovirus; EMCV, encephalomyocarditis virus; HCV, hepatitis virus C; IRES, internal ribosome entry site; ITR, inverted terminal repeat(s); kb, kilobase(s); kDa, kilodalton(s); Luc, luciferase; Luc, gene encoding luciferase; NS, non-structural protein; nt, nucleotide(s); pA, early polyadenylation signal of SV40; PA, polyacrylamide; PCR, polymerase chain reaction; PolIk, Klenow ( large) fragment of E. coli DNA polymerase I. 0378-1119/97/$17.00 © 1997 Elsevier Science B.V. All rights reserved. PII S 03 7 8 -1 1 1 9 ( 9 7 ) 0 0 4 12 - 5
in length. Therefore, the size of DNA to be packaged into viral particles is limited, compared to other viral vectors. Our current AAV vector system enables us to insert efficiently a DNA fragment of up to ~4.3 kb including regulatory elements (promoter and polyadenylation signal sequences). In order to insert not only a therapeutic gene but also a selectable marker gene in the same AAV vector, the size of a marker gene should be as small as possible. Among marker genes for selection of transduced cells, bsr encoding blasticidin S deaminase is only 420 bp in length, one of the smallest marker genes (Izumi et al., 1991). An IRES segment that initiates translation by cap-independent manner was first described in the 5∞ non-coding region of poliovirus RNA (Pelletier and Sonenberg, 1988), followed by several studies concerning other IRES segments (Jang et al., 1988; Adam et al., 1991; Molla et al., 1992; Berlioz et al., 1995). Taking
158
M. Urabe et al. / Gene 200 (1997) 157–162
advantage of this unique feature, IRES elements, especially those derived from the EMCV of ~600 nt, have been used to express multiple proteins from a single transcriptional unit (Morgan et al., 1992). Recently, the HCV genome has proved to have an IRES sequence, the size of which is as small as 230 nt ( TsukiyamaKohara et al., 1992), although the minimal HCV IRES segment is still controversial ( Kettinen et al., 1994; Reynolds et al., 1995; Lu and Wimmer, 1996). Moreover, eukaryotic mRNA of BiP has also proved to initiate translation by internal ribosome binding to a 220 nt segment within its 5∞-leader sequence (Macejak and Sarnow, 1991). To our knowledge, these two IRES sequences are the smallest in literature. A shorter IRES segment is particularly valuable for constructing dicistronic AAV vectors to increase the size of space for the insertion of a therapeutic gene. To determine whether these short IRES segments are usable in combination with bsr, we constructed various dicistronic plasmids; the first cistron consisted of a luciferase (Luc) gene and the second was a bsr. Between the two cistrons was inserted an IRES fragment derived from HCV (type 1b, type 2b), BiP, or EMCV. We then analysed the function of these IRES sequences by in vitro translation and in cultured cells. We report here that the dicistronic transgene, harbouring a 230 nt HCV IRES element placed upstream of bsr, efficiently confers blasticidin S-resistance to host cells.
Plasmids constructed are summarized in Fig. 1. A 480 bp HindIII fragment harbouring bsr was separated from pSV2bsr (Funakoshi, Tokyo, Japan), and was then inserted into the HindIII site of pCMV, which harboured the CMV promoter, first intron of human growth hormone, and SV40 early polyadenylation signal (pA) sequences, generating pBSR. HCV IRES (nt 109–341 of HCV genome) was PCR-amplified from type 1b or type 2b HCV cDNA ( Tsukiyama-Kohara et al., 1992) using primers A1 and B, or A2 and B, respectively. The products were digested with XhoI and BglII, and then inserted into the XhoI-BglII site of pBSR, generating p1B and p2B. pT7BiP/Luc (a kind gift from Dr Sarnow, University of Colorado, USA) was digested with NcoI, blunt-ended with T4 DNA polymerase, and further digested with BamHI, cutting out a Luc. p1B and p2B were ClaI digested, blunt-ended with T4 DNA polymerase, and then BamHI digested, which were ligated to the isolated Luc fragment, yielding pL1B and pL2B. These constructs harboured bsr downstream of short HCV IRES (nt 109–341) originating from HCV type 1b or type 2b. Since the first 13 nt sequence of the 5∞ end of the HCV coding region has been suggested to be involved in more efficient translation (9, 10), pL1OB and pL2OB were also constructed containing the first 15 nt sequence of the coding region of the HCV genome, ATG AGC ACA AAT CCT, between the IRES segment
2. Experimental and discussion 2.1. Plasmid construction Primers used in this study were: A1, 5∞–ATATCTCGAGCCTCCAGGACCCCCCCTCCCGGGAGAG–3∞; A2, 5∞–ATATCTCGAGCCTCCAGGCCCCCCCCTCCCGGGAGAG–3∞; B, 5∞–GGCCAGATCTTGTTGAGAAATGTTAAATGTTTTCATGATGCACGGTCT ACGAGACC–3∞; C, 5∞–GGCCAGATCTTGTTGAGAAATGTTAAATGTTTTCATAGGATTTGTGCT CATGATGCACGGTCTACG–3∞; D, 5∞–GCATTCCTAGGGGTCTTTCC–3∞; E, 5∞–GTTAAATGTTTTCATATTATCATCGTGTTT–3∞; F, 5∞–AAACACGATGATAATATGAAAACATTTAAC–3∞; G, 5∞–GTAACTCGAGTTAATTTCGGGTATA–3∞; H, 5∞–ATATCTCGAGAAGCTTCGACGCCG–3∞; I, 5∞–GTTAAATGTTTTCATGGTGCCAGCCAG –3∞; J, 5∞–CTGGCTGGCACCATGAAAACATTTAAC –3∞.
Fig. 1. Schematic representation of the dicistronic plasmids constructed. The first cistron is Luc and the second is bsr. Between the two cistrons is inserted an IRES segment derived from HCV type 1b, 2b, (pL1B, pL2B), BiP (pLBB), or EMCV (pLEB). pL1OB and pL2OB are the same as pL1B and pL2B, respectively, except for an insertion of the first 15 nt sequence of the coding region of HCV between an IRES segment and bsr. T3, T3 promoter for RNA transcription; N, NcoI site.
M. Urabe et al. / Gene 200 (1997) 157–162
and bsr, following the same procedure described above except for the primers used. To generate pL1OB or pL2OB, an HCV fragment (nt 109–356) was amplified from type 1b or type 2b cDNA with primers A1 and C, or A2 and C, respectively. pL1OB and pL2OB were expected to produce chimeric blasticidin S deaminase with 5 aa (MSTNP) added to the N-terminus of the original blasticidin S deaminase. The fusion gene, EMCV IRES-bsr, consisting of an EMCV IRES sequence (nt 259–833 of the EMCV genome) upstream of bsr, was constructed using PCR as follows; two PCR reactions were carried out. One was with primers D and E, using pCMV-I, which harboured EMCV IRES downstream of the CMV promoter, as a template. The other was with primers F and G using pSV2bsr as a template. The amplified fragments were mixed with each other and used for the next PCR with primers D and G. The resulting PCR product was digested with AvrII and XhoI, and inserted into the AvrII-XhoI site of pCMV-I. A Luc gene was further inserted upstream of the 5∞ end of EMCV IRES-bsr (pLEB). pLBB containing the BiP leader placed upstream of bsr was constructed by essentially the same method as was used for constructing pLEB except for the PCR. One PCR was performed with primers H and I, using pT7BiP/Luc as a template, the other was with primers J and G, using pSV2bsr as a template. Two amplified fragments were mixed with each other, used for the second PCR with primers H and G. The resulting amplified fragment was XhoI digested, and inserted into the XhoI site of pCMV. All the PCR-amplified fragments were verified by sequencing. Except for the intron sequence just downstream of the CMV promoter, none of the constructs had a possible splicing donor or acceptor site within dicistronic sequences. 2.2. In vitro transcription and translation Plasmids were linearized by digestion with KpnI, which cuts downstream of the 3∞ end of bsr and the 3∞ protruding ends were removed by a PolIk fragment of Escherichia coli DNA polymerase I. Uncapped RNAs of expected size (2400-2700 nt depending on the length of spacers) were synthesized with T3 RNA polymerase according to a standard protocol and were used as templates for in vitro translation analysis with rabbit reticulocyte lysate (Amersham, Bucks, UK ). Reactions were carried out at 30°C in the presence of [35S]methionine (Amersham), where the final concentration of K+ and Mg2+ was adjusted to 150 and 1.5 mM, respectively. After 1 h of reaction, the translation products were analysed by electrophoresis on a 12.5% SDS-PA gel, showing the translation products of expected sizes (60.7 kDa Luc and 15.6 or 16.1 kDa blasticidin S deaminase) were clearly separated ( Fig. 2). Both of the two cistrons were thus able to undergo
159
Fig. 2. In vitro translation products in a rabbit reticulocyte lysate. RNAs transcribed from linearized dicistronic plasmids were used as templates. Translation products were separated on a 12.5% PA gel. The negative control was carried out without adding transcribed RNA.
efficient translation. Stronger radioactivity of the first cistron might partly reflect the relative abundance of methionine residues in luciferase compared to that of the second cistron (14 to 4 or 5, respectively). The translation efficiency of the second cistron did not differ significantly among the various IRES segments examined. 2.3. Functional analysis of IRES elements in 293 cells by plasmid transfection To determine whether the IRES segments direct translation of the second cistrons in cultured cells, 293 cells were transfected with the dicistronic plasmid vectors described above. pCMVb (Clontech, Palo Alto, CA) encoding bGal was also included in transfection mixtures to monitor the efficiency of transfection. Briefly, 3×105 of 293 cells were plated in six-well plates 24 h before transfection. Transfection of plasmid DNAs were carried out in triplicate by a standard calcium phosphate method ( Wigler et al., 1979). After an additional 24 h-incubation, cells were harvested and replated at appropriate splits, and incubated further for 2 weeks in the presence of 5 mg/ml blasticidin S hydrochloride (Funakoshi). The total number of resistant colonies produced by transfections was estimated. On the other hand, the rest of the cells harvested were used for the assay of bGal activity (MacGregor et al., 1991). The number of resistant colonies was compared among various types of IRES elements after normalization on the basis of bGal activity ( Fig. 3). Dicistronic plasmid vectors harbouring HCV IRES fragments just followed by the coding region of bsr (pL1B, pL2B) were able to develop a similar number of resistant colonies, although pLEB ( EMCV IRES) formed approximately 10 times more colonies. Transfection with pL1OB, pL2OB, or pLBB generated
160
M. Urabe et al. / Gene 200 (1997) 157–162
Fig. 3. Number of blasticidin S-resistant colonies after transfection with dicistronic plasmids (pL1B, pL1OB, pL2B, pL2OB, pLBB, and pLEB). pBSR is a monocistronic vector expressing bsr driven by the CMV promoter. The number of colonies was normalized against bGal activity. The result was obtained from three independent experiments.
a small number of resistant colonies. There is a possibility that the chimeric blasticidin S deaminase was not functional. We transfected 293 cells with the plasmid that expressed the chimeric deaminase directly under the control of the CMV promoter, resulting in efficient resistant colony formation as well as pBSR (data not shown). This evidence indicated that the 5 aa added to the N-terminus of blasticidin S deaminase did not affect its enzymatic activity to catalyse blasticidin S. The in vitro translation study demonstrated that all the dicistronic constructs underwent efficient translation of the second cistron, bsr. However, in 293 cells, the expression level of the second cistron is quite different among the dicistronic transgenes examined, as revealed by comparison of resistant colony formation. The different assays for IRES function might cause different results. Our result, that the translational efficiency of HCV IRES between type 1b and 2b was not different as analysed by in vitro translation and in 293 cells, seems to conflict with a previous report by Tsukiyama-Kohara et al. (1992), showing the more efficient translational activity of type 2b IRES than type 1b when measured using HCV genome from nt 58 to 1781. Moreover, we could not show the positive effect of the first 15 nt of the coding sequence on the translational efficiency of HCV IRES, which is contrasted to the observations by Kettinen et al. (1994) and Lu and Wimmer (1996), showing that nt 40–372 or 18–710 of the HCV genome, respectively, was necessary for more efficient IRES-directed translation. The short HCV IRES element (nt 109–341 or 109–356) was placed upstream of bsr, which is different from other studies with HCV open reading frame ( Tsukiyama-Kohara et al., 1992), truncated influenza NS ( Kettinen et al., 1994; Reynolds et al., 1995) or poliovirus 2A recognition motif followed by VP4 (Lu
and Wimmer, 1996). The different assays for IRES function might also explain those discrepancies. Therefore, our present result may not be directly comparable with the results reported by others. pL1B, pL2B, and pLEB, which showed relatively efficient expression of the second cistron, bsr, in 293 cells were used for further analyses. Five resistant colonies each, obtained by pL1B, pL2B, or pLEB transfection, were isolated, and expanded for the assay of Luc activity. Cell lysates were obtained using a luciferase assay system (Promega, Madison, WI ) according to the manufacturer’s instruction. The Luc activities were analysed on a luminometer (Lumat LB9501; Berthold, Germany). All the colonies examined expressed Luc, demonstrating that both the first (Luc) and the second (bsr) cistrons were translated efficiently in 293 cells ( Fig. 4(A)). 2.4. Transduction with dicistronic AAV vectors The dicistronic genes from pL1B, pL2B and pLEB, were packaged into AAV vectors (vL1B, vL2B and vLEB). AAV vectors were produced using a helper virus-free system (Colosi et al., 1995). Briefly, 293 cells were cotransfected with three plasmids; vector plasmid with a dicistronic gene flanked by AAV-ITR (inverted terminal repeat(s)), helper plasmid (pIM45) containing AAV genome without ITR, and adenovirus-helper plasmid that harboured E2A, VA, and the E4 region of adenovirus, by a calcium phosphate method. Three days after transfection, cells were recovered and AAV vectors produced were collected after three cycles of freeze/thaw and subsequent centrifugation. Vector titre was determined by quantitative DNA dot-blot hybridization of DNase-treated samples.
Fig. 4. (A) Luciferase activity of stable blasticidin S-resistant colonies obtained by transfection with pL1B, pL2B, or pLEB. Five clones each were isolated, and expanded for Luc assay. (B) Luciferase activity of stable blasticidin S-resistant colonies obtained by transduction with AAV vectors (vL1B, vL2B, or vLEB).
M. Urabe et al. / Gene 200 (1997) 157–162
To evaluate the efficiency of Luc expression by the AAV vectors, we transduced 293 cells at different multiplicities of infection, ranging from 102 to 106 particles per cell, and Luc activities were measured 24 h after transduction (Fig. 5). The Luc activities increased in a dose-dependent manner as the amount of AAV vectors infected increased. No significant differences were observed among the AAV vectors examined, suggesting that expression of Luc was not affected by the following sequence, IRES-bsr. The 100-fold decrease in Luc activity compared to that of the stable clones ( Fig. 4) is probably due to the insufficient second-strand synthesis of the AAV vector genome, which is a key step to express transgenes (Fisher et al., 1996). Next, to determine whether these AAV vectors were capable of conferring blasticidin S-resistance, we transduced 293 cells with vL1B, vL2B, or vLEB at a multiplicity of infection of 106. After 2 weeks selection with blasticidin S, resistant colonies were picked up, expanded and analysed for Luc activity. All the clones analysed showed Luc activities ( Fig. 4(B)), although a few clones showed somewhat lower levels of Luc activity.
161
Fig. 6. Southern analysis of representative blasticidin S-resistant clones. Two clones from each group, with high or low level expression of Luc, were analysed. Genomic DNAs were isolated, digested with NcoI, and then separated on a 1% agarose gel. DNA was transferred onto a nylon membrane (Hybond N+, Amersham) and hybridized to a 32P-labelled probe corresponding to the bsr gene. Clones obtained using transgene with HCV-IRES are expected to show a 3.2 kb band. Clones generated by the dicistronic gene with EMCV-IRES should reveal a 2.8 kb fragment. Std, NcoI-digested pL1B showing a 3.2 kb fragment; 293, genomic DNA from 293 cells; arrow, bands corresponding to intact transgene containing HCV-IRES; arrowhead, intact EMCV dicistronic fragment.
2.5. Southern analysis of blasticidin S-resistant clones To investigate whether the stable clones with low Luc activity had DNA rearrangement in the dicistronic transgenes, Southern analysis of representative clones was performed showing high- or low-level expression of Luc ( Fig. 6). Clones obtained by HCV- or EMCVIRES harbouring the dicistronic transgene should show a 3.2 or 2.8 kb band ( Fig. 1), respectively, that hybridizes to a probe derived from the bsr gene. Although clones expressing Luc at a high level harboured intact dicistronic transgenes, clones with lower levels of Luc expression (indicated by an asterisk in Fig. 6) showed
disrupted dicistronic sequences. The DNA rearrangement within the expression cassettes probably resulted in the low level of Luc expression.
3. Conclusions We have developed a novel dicistronic AAV vector that harboured short HCV IRES-bsr, which is capable of expressing a therapeutic gene of up to ~3.6 kb (including promoter/enhancer elements) besides a selectable marker.
Acknowledgement
Fig. 5. Dose responses of Luc activities produced in 293 cells after transduction with AAV vectors. A total of 293 cells were transduced with dicistronic AAV vectors at different multiplicities of infection ranging from 102 to 106. At 24 h post-transduction Luc activities were measured. vL is a monocistronic AAV vector, harbouring Luc alone.
We are deeply grateful to Dr P. Sarnow ( University of Colorado, USA) for providing us with pT7BiP/Luc and to Dr Y. Iwaki ( University of South California, USA) for valuable advice. This work was supported in part by grants from the Ministry of Health and Welfare of Japan, Grants-in-Aid for Scientific Research from the Ministry of Education, Science, Sports and Culture of Japan, a grant from the Japanese Foundation for the Multidisciplinary Treatment of Cancer, CREST (Core Research for Evolutional Science and Technology) of Japan Science and Technology Corporation (JST ), and a grant from Uehara Memorial Foundation.
162
M. Urabe et al. / Gene 200 (1997) 157–162
References Adam, M.A., Ramesh, N., Miller, A.D., Osborne, W.R.A., 1991. Internal initiation of translation in retroviral vectors carrying picornavirus 5∞ non-translated regions. J. Virol. 65, 4985–4990. Berlioz, C., Torrent, C., Darlix, J.-L., 1995. An internal ribosomal entry signal in the rat VL30 region of the Harvey murine sarcoma virus leader and its use in dicistronic retroviral vectors. J. Virol. 69, 6400–6407. Colosi, P., Elliger, S., Elliger, C., Kurtzman, G., 1995. AAV can be efficiently produced without helper virus. Blood 86, 627a Fisher, K.J., Gao, G.-P., Weitzman, M.D., DeMatteo, R., Burda, J.F., Wilson, J.M., 1996. Transduction with recombinant adeno-associated virus for gene therapy is limited by leading-strand synthesis. J. Virol. 70, 520–532. Izumi, M., Miyazawa, H., Kamakura, T., Yamaguchi, I., Endo, T., Hanaoka, F., 1991. Blasticidin S-resistance gene (bsr): a novel selectable marker for mammalian cells. Exp. Cell. Res. 197, 229–233. Jang, S.K., Kra¨usslich, H.-G., Nicklin, M.J.H., Duke, G.M., Palmenberg, A.C., Wimmer, E., 1988. A segment of the 5∞ non-translated region of encephalomyocarditis virus RNA directs internal entry of ribosomes during in vitro translation. J. Virol. 62, 2636–2643. Kettinen, H., Grace, K., Grunert, S., Clarke, B., Rowlands, D. and Jackson, R. (1994). Mapping of the internal ribosome entry site at the 5∞ end of the hepatitis C virus genome. In Viral Hepatitis and Liver Disease, pp. 125–131. Springer–Verlag, Tokyo. Lu, H.-H., Wimmer, E., 1996. Poliovirus chimera replicating under the translational control of genetic elements of hepatitis C virus
reveal unusual properties of the internal ribosomal entry site of hepatitis C virus. Proc. Natl. Acad. Sci. USA 93, 1412–1417. Macejak, D.G., Sarnow, P., 1991. Internal initiation of translation mediated by the 5∞ leader of a cellular mRNA. Nature 353, 90–94. MacGregor, G.R., Nolan G.P., Fiering, S., Roederer, M. and Herzenberg, L.A (1991) Use of E. coli lacZ ( b-galactosidase) as a reporter gene. In Gene Transfer and Expression Protocols, ed E.J. Murray, pp. 21–235, Humana Press, New Jersey. Molla, A., Jang, S.K., Paul, A.V., Reuer, Q., Wimmer, E., 1992. Cardioviral internal ribosomal entry site is functional in a genetically engineered dicistronic poliovirus. Nature 356, 255–257. Morgan, R.A., Couture, L., Elroy-Stein, O., Ragheb, J., Moss, B., Anderson, W.F., 1992. Retroviral vectors containing putative internal ribosome entry site: development of a polycistronic gene transfer system and applications to human gene therapy. Nucl. Acids Res. 20, 1293–1299. Pelletier, J., Sonenberg, N., 1988. Internal initiation of translation of eukaryotic mRNA directed by a sequence derived from poliovirus RNA. Nature 334, 320–325. Reynolds, J.E., Kaminski, A., Kettinen, H.J., Grace, K., Clarke, B.E., Carroll, A.R., Rowlands, D.J., Jackson, R.J., 1995. Unique features of internal initiation of hepatitis C virus RNA translation. EMBO J. 14, 6010–6020. Tsukiyama-Kohara, K., Iizuka, N., Kohara, M., Nomoto, A., 1992. Internal ribosome entry site within hepatitis C virus RNA. J. Virol. 66, 1476–1483. Wigler, M., Pellicer, A., Silverstein, S., Axel, R., Urlaub, G., Chasin, L., 1979. DNA-mediated transfer of the adenine phosphoribosyltransferase locus into mammalian cells. Proc. Natl. Acad. Sci. USA 76, 1373–1376.
A novel dicistronic AAV vector using a short IRES segment derived from hepatitis C virus genome Masashi Urabe a, Yoko Hasumi a, Yoji Ogasawara a, Takashi Matsushita a,b, Nobuhiko Kamoshita c, Akio Nomoto c, Peter Colosi b, Gary J. Kurtzman b, Kiyotake Tobita d, Keiya Ozawa a,* a Department of Molecular Biology, Institute of Hematology, Jichi Medical School, Tochigi 329-04, Japan b Avigen Inc., Alameda, CA 94502, USA c Department of Microbiology, Institute of Medical Science, University of Tokyo, Tokyo 108, Japan d Department of Virology, Jichi Medical School, Tochigi 329-04, Japan Received 29 January 1997; accepted 19 June 1997; Received by T. Sekiya
Abstract Adeno-associated virus (AAV ) vectors have a limited capacity for packaging DNA. To insert both a therapeutic gene and a selectable marker gene in the same AAV vector efficiently, we developed a novel dicistronic AAV vector containing a 230 base pairs (bp) internal ribosome entry site (IRES ) element derived from hepatitis C virus (HCV ) genome and a 420 bp blasticidin S-resistance gene (bsr) as a small selectable marker in the second cistron. The 650 bp HCV IRES-bsr construct was placed downstream of the 3∞ end of the luciferase gene (Luc) under the control of the human cytomegalovirus (CMV ) promoter. This dicistronic gene conferred blasticidin S-resistance to 293 cells besides luciferase activity, when examined not only by transfection but also by transduction using AAV vectors. The dicistronic AAV vector harbouring HCV IRES-bsr is capable of expressing a therapeutic gene of up to ~3.6 kilobases (kb) (including promoter/enhancer elements) as well as a selectable marker gene. If a selectable marker gene is not necessary, this vector is able to incorporate two different kinds of therapeutic genes more easily than that containing EMCV IRES. The dicistronic AAV vector described here is useful for expressing many kinds of cDNA besides a selectable marker. © 1997 Elsevier Science B.V. Keywords: Blasticidin S-resistance gene; Gene therapy; Gene transfer; Selectable markers
1. Introduction The AAV vector is emerging as highly promising for gene therapy because of several unique properties that the other viral vectors do not share. AAV is a member of the Parvoviridae, of which the genome is only 4.7 kb * Corresponding author. Tel: +81 285 442111, ext. 3481; Fax: +81 285 448675; e-mail: [email protected] Abbreviations: aa, amino acid(s); AAV, adeno-associated virus; bGal, b-galactosidase; BiP, immunoglobulin heavy chain binding protein; bp, base pairs; bsr, blasticidin S-resistance gene; CMV, human cytomegalovirus; EMCV, encephalomyocarditis virus; HCV, hepatitis virus C; IRES, internal ribosome entry site; ITR, inverted terminal repeat(s); kb, kilobase(s); kDa, kilodalton(s); Luc, luciferase; Luc, gene encoding luciferase; NS, non-structural protein; nt, nucleotide(s); pA, early polyadenylation signal of SV40; PA, polyacrylamide; PCR, polymerase chain reaction; PolIk, Klenow ( large) fragment of E. coli DNA polymerase I. 0378-1119/97/$17.00 © 1997 Elsevier Science B.V. All rights reserved. PII S 03 7 8 -1 1 1 9 ( 9 7 ) 0 0 4 12 - 5
in length. Therefore, the size of DNA to be packaged into viral particles is limited, compared to other viral vectors. Our current AAV vector system enables us to insert efficiently a DNA fragment of up to ~4.3 kb including regulatory elements (promoter and polyadenylation signal sequences). In order to insert not only a therapeutic gene but also a selectable marker gene in the same AAV vector, the size of a marker gene should be as small as possible. Among marker genes for selection of transduced cells, bsr encoding blasticidin S deaminase is only 420 bp in length, one of the smallest marker genes (Izumi et al., 1991). An IRES segment that initiates translation by cap-independent manner was first described in the 5∞ non-coding region of poliovirus RNA (Pelletier and Sonenberg, 1988), followed by several studies concerning other IRES segments (Jang et al., 1988; Adam et al., 1991; Molla et al., 1992; Berlioz et al., 1995). Taking
158
M. Urabe et al. / Gene 200 (1997) 157–162
advantage of this unique feature, IRES elements, especially those derived from the EMCV of ~600 nt, have been used to express multiple proteins from a single transcriptional unit (Morgan et al., 1992). Recently, the HCV genome has proved to have an IRES sequence, the size of which is as small as 230 nt ( TsukiyamaKohara et al., 1992), although the minimal HCV IRES segment is still controversial ( Kettinen et al., 1994; Reynolds et al., 1995; Lu and Wimmer, 1996). Moreover, eukaryotic mRNA of BiP has also proved to initiate translation by internal ribosome binding to a 220 nt segment within its 5∞-leader sequence (Macejak and Sarnow, 1991). To our knowledge, these two IRES sequences are the smallest in literature. A shorter IRES segment is particularly valuable for constructing dicistronic AAV vectors to increase the size of space for the insertion of a therapeutic gene. To determine whether these short IRES segments are usable in combination with bsr, we constructed various dicistronic plasmids; the first cistron consisted of a luciferase (Luc) gene and the second was a bsr. Between the two cistrons was inserted an IRES fragment derived from HCV (type 1b, type 2b), BiP, or EMCV. We then analysed the function of these IRES sequences by in vitro translation and in cultured cells. We report here that the dicistronic transgene, harbouring a 230 nt HCV IRES element placed upstream of bsr, efficiently confers blasticidin S-resistance to host cells.
Plasmids constructed are summarized in Fig. 1. A 480 bp HindIII fragment harbouring bsr was separated from pSV2bsr (Funakoshi, Tokyo, Japan), and was then inserted into the HindIII site of pCMV, which harboured the CMV promoter, first intron of human growth hormone, and SV40 early polyadenylation signal (pA) sequences, generating pBSR. HCV IRES (nt 109–341 of HCV genome) was PCR-amplified from type 1b or type 2b HCV cDNA ( Tsukiyama-Kohara et al., 1992) using primers A1 and B, or A2 and B, respectively. The products were digested with XhoI and BglII, and then inserted into the XhoI-BglII site of pBSR, generating p1B and p2B. pT7BiP/Luc (a kind gift from Dr Sarnow, University of Colorado, USA) was digested with NcoI, blunt-ended with T4 DNA polymerase, and further digested with BamHI, cutting out a Luc. p1B and p2B were ClaI digested, blunt-ended with T4 DNA polymerase, and then BamHI digested, which were ligated to the isolated Luc fragment, yielding pL1B and pL2B. These constructs harboured bsr downstream of short HCV IRES (nt 109–341) originating from HCV type 1b or type 2b. Since the first 13 nt sequence of the 5∞ end of the HCV coding region has been suggested to be involved in more efficient translation (9, 10), pL1OB and pL2OB were also constructed containing the first 15 nt sequence of the coding region of the HCV genome, ATG AGC ACA AAT CCT, between the IRES segment
2. Experimental and discussion 2.1. Plasmid construction Primers used in this study were: A1, 5∞–ATATCTCGAGCCTCCAGGACCCCCCCTCCCGGGAGAG–3∞; A2, 5∞–ATATCTCGAGCCTCCAGGCCCCCCCCTCCCGGGAGAG–3∞; B, 5∞–GGCCAGATCTTGTTGAGAAATGTTAAATGTTTTCATGATGCACGGTCT ACGAGACC–3∞; C, 5∞–GGCCAGATCTTGTTGAGAAATGTTAAATGTTTTCATAGGATTTGTGCT CATGATGCACGGTCTACG–3∞; D, 5∞–GCATTCCTAGGGGTCTTTCC–3∞; E, 5∞–GTTAAATGTTTTCATATTATCATCGTGTTT–3∞; F, 5∞–AAACACGATGATAATATGAAAACATTTAAC–3∞; G, 5∞–GTAACTCGAGTTAATTTCGGGTATA–3∞; H, 5∞–ATATCTCGAGAAGCTTCGACGCCG–3∞; I, 5∞–GTTAAATGTTTTCATGGTGCCAGCCAG –3∞; J, 5∞–CTGGCTGGCACCATGAAAACATTTAAC –3∞.
Fig. 1. Schematic representation of the dicistronic plasmids constructed. The first cistron is Luc and the second is bsr. Between the two cistrons is inserted an IRES segment derived from HCV type 1b, 2b, (pL1B, pL2B), BiP (pLBB), or EMCV (pLEB). pL1OB and pL2OB are the same as pL1B and pL2B, respectively, except for an insertion of the first 15 nt sequence of the coding region of HCV between an IRES segment and bsr. T3, T3 promoter for RNA transcription; N, NcoI site.
M. Urabe et al. / Gene 200 (1997) 157–162
and bsr, following the same procedure described above except for the primers used. To generate pL1OB or pL2OB, an HCV fragment (nt 109–356) was amplified from type 1b or type 2b cDNA with primers A1 and C, or A2 and C, respectively. pL1OB and pL2OB were expected to produce chimeric blasticidin S deaminase with 5 aa (MSTNP) added to the N-terminus of the original blasticidin S deaminase. The fusion gene, EMCV IRES-bsr, consisting of an EMCV IRES sequence (nt 259–833 of the EMCV genome) upstream of bsr, was constructed using PCR as follows; two PCR reactions were carried out. One was with primers D and E, using pCMV-I, which harboured EMCV IRES downstream of the CMV promoter, as a template. The other was with primers F and G using pSV2bsr as a template. The amplified fragments were mixed with each other and used for the next PCR with primers D and G. The resulting PCR product was digested with AvrII and XhoI, and inserted into the AvrII-XhoI site of pCMV-I. A Luc gene was further inserted upstream of the 5∞ end of EMCV IRES-bsr (pLEB). pLBB containing the BiP leader placed upstream of bsr was constructed by essentially the same method as was used for constructing pLEB except for the PCR. One PCR was performed with primers H and I, using pT7BiP/Luc as a template, the other was with primers J and G, using pSV2bsr as a template. Two amplified fragments were mixed with each other, used for the second PCR with primers H and G. The resulting amplified fragment was XhoI digested, and inserted into the XhoI site of pCMV. All the PCR-amplified fragments were verified by sequencing. Except for the intron sequence just downstream of the CMV promoter, none of the constructs had a possible splicing donor or acceptor site within dicistronic sequences. 2.2. In vitro transcription and translation Plasmids were linearized by digestion with KpnI, which cuts downstream of the 3∞ end of bsr and the 3∞ protruding ends were removed by a PolIk fragment of Escherichia coli DNA polymerase I. Uncapped RNAs of expected size (2400-2700 nt depending on the length of spacers) were synthesized with T3 RNA polymerase according to a standard protocol and were used as templates for in vitro translation analysis with rabbit reticulocyte lysate (Amersham, Bucks, UK ). Reactions were carried out at 30°C in the presence of [35S]methionine (Amersham), where the final concentration of K+ and Mg2+ was adjusted to 150 and 1.5 mM, respectively. After 1 h of reaction, the translation products were analysed by electrophoresis on a 12.5% SDS-PA gel, showing the translation products of expected sizes (60.7 kDa Luc and 15.6 or 16.1 kDa blasticidin S deaminase) were clearly separated ( Fig. 2). Both of the two cistrons were thus able to undergo
159
Fig. 2. In vitro translation products in a rabbit reticulocyte lysate. RNAs transcribed from linearized dicistronic plasmids were used as templates. Translation products were separated on a 12.5% PA gel. The negative control was carried out without adding transcribed RNA.
efficient translation. Stronger radioactivity of the first cistron might partly reflect the relative abundance of methionine residues in luciferase compared to that of the second cistron (14 to 4 or 5, respectively). The translation efficiency of the second cistron did not differ significantly among the various IRES segments examined. 2.3. Functional analysis of IRES elements in 293 cells by plasmid transfection To determine whether the IRES segments direct translation of the second cistrons in cultured cells, 293 cells were transfected with the dicistronic plasmid vectors described above. pCMVb (Clontech, Palo Alto, CA) encoding bGal was also included in transfection mixtures to monitor the efficiency of transfection. Briefly, 3×105 of 293 cells were plated in six-well plates 24 h before transfection. Transfection of plasmid DNAs were carried out in triplicate by a standard calcium phosphate method ( Wigler et al., 1979). After an additional 24 h-incubation, cells were harvested and replated at appropriate splits, and incubated further for 2 weeks in the presence of 5 mg/ml blasticidin S hydrochloride (Funakoshi). The total number of resistant colonies produced by transfections was estimated. On the other hand, the rest of the cells harvested were used for the assay of bGal activity (MacGregor et al., 1991). The number of resistant colonies was compared among various types of IRES elements after normalization on the basis of bGal activity ( Fig. 3). Dicistronic plasmid vectors harbouring HCV IRES fragments just followed by the coding region of bsr (pL1B, pL2B) were able to develop a similar number of resistant colonies, although pLEB ( EMCV IRES) formed approximately 10 times more colonies. Transfection with pL1OB, pL2OB, or pLBB generated
160
M. Urabe et al. / Gene 200 (1997) 157–162
Fig. 3. Number of blasticidin S-resistant colonies after transfection with dicistronic plasmids (pL1B, pL1OB, pL2B, pL2OB, pLBB, and pLEB). pBSR is a monocistronic vector expressing bsr driven by the CMV promoter. The number of colonies was normalized against bGal activity. The result was obtained from three independent experiments.
a small number of resistant colonies. There is a possibility that the chimeric blasticidin S deaminase was not functional. We transfected 293 cells with the plasmid that expressed the chimeric deaminase directly under the control of the CMV promoter, resulting in efficient resistant colony formation as well as pBSR (data not shown). This evidence indicated that the 5 aa added to the N-terminus of blasticidin S deaminase did not affect its enzymatic activity to catalyse blasticidin S. The in vitro translation study demonstrated that all the dicistronic constructs underwent efficient translation of the second cistron, bsr. However, in 293 cells, the expression level of the second cistron is quite different among the dicistronic transgenes examined, as revealed by comparison of resistant colony formation. The different assays for IRES function might cause different results. Our result, that the translational efficiency of HCV IRES between type 1b and 2b was not different as analysed by in vitro translation and in 293 cells, seems to conflict with a previous report by Tsukiyama-Kohara et al. (1992), showing the more efficient translational activity of type 2b IRES than type 1b when measured using HCV genome from nt 58 to 1781. Moreover, we could not show the positive effect of the first 15 nt of the coding sequence on the translational efficiency of HCV IRES, which is contrasted to the observations by Kettinen et al. (1994) and Lu and Wimmer (1996), showing that nt 40–372 or 18–710 of the HCV genome, respectively, was necessary for more efficient IRES-directed translation. The short HCV IRES element (nt 109–341 or 109–356) was placed upstream of bsr, which is different from other studies with HCV open reading frame ( Tsukiyama-Kohara et al., 1992), truncated influenza NS ( Kettinen et al., 1994; Reynolds et al., 1995) or poliovirus 2A recognition motif followed by VP4 (Lu
and Wimmer, 1996). The different assays for IRES function might also explain those discrepancies. Therefore, our present result may not be directly comparable with the results reported by others. pL1B, pL2B, and pLEB, which showed relatively efficient expression of the second cistron, bsr, in 293 cells were used for further analyses. Five resistant colonies each, obtained by pL1B, pL2B, or pLEB transfection, were isolated, and expanded for the assay of Luc activity. Cell lysates were obtained using a luciferase assay system (Promega, Madison, WI ) according to the manufacturer’s instruction. The Luc activities were analysed on a luminometer (Lumat LB9501; Berthold, Germany). All the colonies examined expressed Luc, demonstrating that both the first (Luc) and the second (bsr) cistrons were translated efficiently in 293 cells ( Fig. 4(A)). 2.4. Transduction with dicistronic AAV vectors The dicistronic genes from pL1B, pL2B and pLEB, were packaged into AAV vectors (vL1B, vL2B and vLEB). AAV vectors were produced using a helper virus-free system (Colosi et al., 1995). Briefly, 293 cells were cotransfected with three plasmids; vector plasmid with a dicistronic gene flanked by AAV-ITR (inverted terminal repeat(s)), helper plasmid (pIM45) containing AAV genome without ITR, and adenovirus-helper plasmid that harboured E2A, VA, and the E4 region of adenovirus, by a calcium phosphate method. Three days after transfection, cells were recovered and AAV vectors produced were collected after three cycles of freeze/thaw and subsequent centrifugation. Vector titre was determined by quantitative DNA dot-blot hybridization of DNase-treated samples.
Fig. 4. (A) Luciferase activity of stable blasticidin S-resistant colonies obtained by transfection with pL1B, pL2B, or pLEB. Five clones each were isolated, and expanded for Luc assay. (B) Luciferase activity of stable blasticidin S-resistant colonies obtained by transduction with AAV vectors (vL1B, vL2B, or vLEB).
M. Urabe et al. / Gene 200 (1997) 157–162
To evaluate the efficiency of Luc expression by the AAV vectors, we transduced 293 cells at different multiplicities of infection, ranging from 102 to 106 particles per cell, and Luc activities were measured 24 h after transduction (Fig. 5). The Luc activities increased in a dose-dependent manner as the amount of AAV vectors infected increased. No significant differences were observed among the AAV vectors examined, suggesting that expression of Luc was not affected by the following sequence, IRES-bsr. The 100-fold decrease in Luc activity compared to that of the stable clones ( Fig. 4) is probably due to the insufficient second-strand synthesis of the AAV vector genome, which is a key step to express transgenes (Fisher et al., 1996). Next, to determine whether these AAV vectors were capable of conferring blasticidin S-resistance, we transduced 293 cells with vL1B, vL2B, or vLEB at a multiplicity of infection of 106. After 2 weeks selection with blasticidin S, resistant colonies were picked up, expanded and analysed for Luc activity. All the clones analysed showed Luc activities ( Fig. 4(B)), although a few clones showed somewhat lower levels of Luc activity.
161
Fig. 6. Southern analysis of representative blasticidin S-resistant clones. Two clones from each group, with high or low level expression of Luc, were analysed. Genomic DNAs were isolated, digested with NcoI, and then separated on a 1% agarose gel. DNA was transferred onto a nylon membrane (Hybond N+, Amersham) and hybridized to a 32P-labelled probe corresponding to the bsr gene. Clones obtained using transgene with HCV-IRES are expected to show a 3.2 kb band. Clones generated by the dicistronic gene with EMCV-IRES should reveal a 2.8 kb fragment. Std, NcoI-digested pL1B showing a 3.2 kb fragment; 293, genomic DNA from 293 cells; arrow, bands corresponding to intact transgene containing HCV-IRES; arrowhead, intact EMCV dicistronic fragment.
2.5. Southern analysis of blasticidin S-resistant clones To investigate whether the stable clones with low Luc activity had DNA rearrangement in the dicistronic transgenes, Southern analysis of representative clones was performed showing high- or low-level expression of Luc ( Fig. 6). Clones obtained by HCV- or EMCVIRES harbouring the dicistronic transgene should show a 3.2 or 2.8 kb band ( Fig. 1), respectively, that hybridizes to a probe derived from the bsr gene. Although clones expressing Luc at a high level harboured intact dicistronic transgenes, clones with lower levels of Luc expression (indicated by an asterisk in Fig. 6) showed
disrupted dicistronic sequences. The DNA rearrangement within the expression cassettes probably resulted in the low level of Luc expression.
3. Conclusions We have developed a novel dicistronic AAV vector that harboured short HCV IRES-bsr, which is capable of expressing a therapeutic gene of up to ~3.6 kb (including promoter/enhancer elements) besides a selectable marker.
Acknowledgement
Fig. 5. Dose responses of Luc activities produced in 293 cells after transduction with AAV vectors. A total of 293 cells were transduced with dicistronic AAV vectors at different multiplicities of infection ranging from 102 to 106. At 24 h post-transduction Luc activities were measured. vL is a monocistronic AAV vector, harbouring Luc alone.
We are deeply grateful to Dr P. Sarnow ( University of Colorado, USA) for providing us with pT7BiP/Luc and to Dr Y. Iwaki ( University of South California, USA) for valuable advice. This work was supported in part by grants from the Ministry of Health and Welfare of Japan, Grants-in-Aid for Scientific Research from the Ministry of Education, Science, Sports and Culture of Japan, a grant from the Japanese Foundation for the Multidisciplinary Treatment of Cancer, CREST (Core Research for Evolutional Science and Technology) of Japan Science and Technology Corporation (JST ), and a grant from Uehara Memorial Foundation.
162
M. Urabe et al. / Gene 200 (1997) 157–162
References Adam, M.A., Ramesh, N., Miller, A.D., Osborne, W.R.A., 1991. Internal initiation of translation in retroviral vectors carrying picornavirus 5∞ non-translated regions. J. Virol. 65, 4985–4990. Berlioz, C., Torrent, C., Darlix, J.-L., 1995. An internal ribosomal entry signal in the rat VL30 region of the Harvey murine sarcoma virus leader and its use in dicistronic retroviral vectors. J. Virol. 69, 6400–6407. Colosi, P., Elliger, S., Elliger, C., Kurtzman, G., 1995. AAV can be efficiently produced without helper virus. Blood 86, 627a Fisher, K.J., Gao, G.-P., Weitzman, M.D., DeMatteo, R., Burda, J.F., Wilson, J.M., 1996. Transduction with recombinant adeno-associated virus for gene therapy is limited by leading-strand synthesis. J. Virol. 70, 520–532. Izumi, M., Miyazawa, H., Kamakura, T., Yamaguchi, I., Endo, T., Hanaoka, F., 1991. Blasticidin S-resistance gene (bsr): a novel selectable marker for mammalian cells. Exp. Cell. Res. 197, 229–233. Jang, S.K., Kra¨usslich, H.-G., Nicklin, M.J.H., Duke, G.M., Palmenberg, A.C., Wimmer, E., 1988. A segment of the 5∞ non-translated region of encephalomyocarditis virus RNA directs internal entry of ribosomes during in vitro translation. J. Virol. 62, 2636–2643. Kettinen, H., Grace, K., Grunert, S., Clarke, B., Rowlands, D. and Jackson, R. (1994). Mapping of the internal ribosome entry site at the 5∞ end of the hepatitis C virus genome. In Viral Hepatitis and Liver Disease, pp. 125–131. Springer–Verlag, Tokyo. Lu, H.-H., Wimmer, E., 1996. Poliovirus chimera replicating under the translational control of genetic elements of hepatitis C virus
reveal unusual properties of the internal ribosomal entry site of hepatitis C virus. Proc. Natl. Acad. Sci. USA 93, 1412–1417. Macejak, D.G., Sarnow, P., 1991. Internal initiation of translation mediated by the 5∞ leader of a cellular mRNA. Nature 353, 90–94. MacGregor, G.R., Nolan G.P., Fiering, S., Roederer, M. and Herzenberg, L.A (1991) Use of E. coli lacZ ( b-galactosidase) as a reporter gene. In Gene Transfer and Expression Protocols, ed E.J. Murray, pp. 21–235, Humana Press, New Jersey. Molla, A., Jang, S.K., Paul, A.V., Reuer, Q., Wimmer, E., 1992. Cardioviral internal ribosomal entry site is functional in a genetically engineered dicistronic poliovirus. Nature 356, 255–257. Morgan, R.A., Couture, L., Elroy-Stein, O., Ragheb, J., Moss, B., Anderson, W.F., 1992. Retroviral vectors containing putative internal ribosome entry site: development of a polycistronic gene transfer system and applications to human gene therapy. Nucl. Acids Res. 20, 1293–1299. Pelletier, J., Sonenberg, N., 1988. Internal initiation of translation of eukaryotic mRNA directed by a sequence derived from poliovirus RNA. Nature 334, 320–325. Reynolds, J.E., Kaminski, A., Kettinen, H.J., Grace, K., Clarke, B.E., Carroll, A.R., Rowlands, D.J., Jackson, R.J., 1995. Unique features of internal initiation of hepatitis C virus RNA translation. EMBO J. 14, 6010–6020. Tsukiyama-Kohara, K., Iizuka, N., Kohara, M., Nomoto, A., 1992. Internal ribosome entry site within hepatitis C virus RNA. J. Virol. 66, 1476–1483. Wigler, M., Pellicer, A., Silverstein, S., Axel, R., Urlaub, G., Chasin, L., 1979. DNA-mediated transfer of the adenine phosphoribosyltransferase locus into mammalian cells. Proc. Natl. Acad. Sci. USA 76, 1373–1376.