The primer binding site on the RNA genome of human and simian immunodeficiency viruses is flanked by an upstream hairpin structure
The primer binding site on the RNA genome of human and simian immunodeficiency viruses is flanked by an upstream hairpin structureBenjamin Berkhout
Department of Human Retrovirology, Academic Medical Center, University of Amsterdam, PO Box 22700, 1100 DE Amsterdam, The Netherlands
Received July 16, 1997;Revised and Accepted August 26, 1997
ABSTRACT
Reverse transcription of retroviral genomes is primed by a tRNA molecule that anneals to an 18 nt primer binding site (PBS) on the viral RNA genome. Additional base pair interactions between the tRNA primer and the viral RNA have been proposed. In particular, base pairing was proposed between the anticodon loop of tRNALys3 and the `A-rich' loop of a hairpin located immediately upstream of the PBS site in HIV-1 RNA. In order to judge the importance of this sequence/structure motif, we performed an extensive phylogenetic analysis of this genomic region in a variety of simian and human immunodeficiency viruses (SIV and HIV). Both the phylogeny of natural HIV/SIV sequences and the behaviour of U5-PBS mutant/revertant viruses support the idea that this RNA structure is critical for virus replication. Although this hairpin may play a role in tRNA annealing and/or initiation of reverse transcription, the proposed base pairing interaction between the A-rich loop of the HIV-1 hairpin and the anticodon of the initiator tRNA is not directly supported by this analysis.
INTRODUCTION
Reverse transcription of retroviral genomes is primed by a tRNA annealed to an 18 nt primer binding site (PBS). Different retroviruses use different tRNA species as primer; these natural primers include tRNAPro, tRNATrp, tRNALys1,2 and tRNALys3. Multiple interactions between the tRNA primer and both viral protein and RNA components are involved in this specific primer usage (reviewed in 1 ). First, there is evidence for specific binding of the tRNA primer to reverse transcriptase (RT) enzyme, an interaction that leads to selective encapsidation of the tRNA primer into virion particles. Second, annealing to the viral genome is facilitated by the 18 nt complementarity between the 3'-end of the priming tRNA and the PBS site. Third, annealing of the tRNA may be stimulated by additional base pairing contacts between other parts of the tRNA and viral sequences flanking the PBS site. Such additional contacts were originally proposed for Rous sarcoma virus (2 -4 ), but similar interactions have been reported for human immunodeficiency virus type 1 (HIV-1) (5 ), HIV-2 (6 ) and several retrotransposon elements (7 -9 ). Consistent with the idea that regions of the viral genome outside the PBS sequence participate in selective tRNA primer usage is the observation that retroviruses mutated in the PBS are genetically unstable, as they revert to the wild-type PBS sequence after a few passages in cell culture (10 -13 ). On the other hand, reasonable transduction efficiencies were obtained with murine leukemia virus vectors that use an unnatural or genetically engineered tRNA primer (14 ,15 ).
A structure has been proposed for the HIV-1 RNA-tRNALys3 complex which defined several sequences in the U5 region that interact with different parts of the tRNALys3 primer (5 ). According to this detailed RNA structure model, base pairing occurs between the anticodon loop of tRNALys3 and an `A-rich' loop that is part of a hairpin in the U5 region. Recent studies with mutant HIV-1 genomes have analyzed the role of this additional base pairing interaction (16 -19 ). Specifically, Morrow et al. tested whether PBS variants that use an unnatural tRNA primer could be stabilized by introduction of a compensatory mutation in the U5 loop region that is anticipated to contact the anticodon of the novel tRNA primer. These studies identified viruses that stably maintained the modified PBS site after extended growth in tissue culture. A major problem with this type of analysis is that multiple parameters can influence the genetic stability or reversion capacity of a mutated retrovirus. For instance, it cannot be excluded that genotypes are relatively stable because the corresponding viruses are severely replication impaired, resulting in a low reversion capacity. Alternatively, the U5 mutation may cause an unwanted defect in another replicative function of the HIV-1 leader RNA and revertants may have restored this function. Another obstacle to interpretation of a mutational analysis is that the effect on RNA secondary structure of the U5-PBS region is largely unknown. I therefore performed an extensive phylogenetic analysis of the U5-PBS RNA structures of all types of human and simian immunodeficiency viruses (HIV and SIV).
RESULTS AND DISCUSSION
Phylogenetic conservation of a U5 hairpin, but not of the A-rich loop
RNA secondary structure models for the U5-PBS region of both HIV-1 and HIV-2 have been reported previously. The HIV-1 hairpin contains the A-rich loop (Fig. 1 ) and is based primarily on RNA structure probing experiments (5 ,20 ). Little phylogenetic support is available for this structure because the sequence is relatively well conserved among different virus isolates (21 ). Biochemical RNA structure probing and phylogeny were used to model the HIV-2 RNA structure near the PBS site (6 ). An extended hairpin has been proposed for part of the U5 region immediately upstream of the PBS (Fig. 1 ) and sequence differences in 13 HIV-2 isolates yielded several co-variations that strongly support this base pairing scheme. Interestingly, this hairpin contains two single-stranded A-rich motifs in the internal and external loops that may participate in binding of the U-rich anticodon loop of tRNALys3.
RNA structural effects of U5-PBS mutated viruses and revertants
The HIV-1 mutant His-AC was previously designed (18 ) and contains mutations in both the PBS site and the A-rich sequence of the U5 region to facilitate interaction with the tRNAHis primer (Fig. 2 ). This mutant has a severe replication defect and upon prolonged culture a faster replicating revertant appears. I now propose that both the defect of this mutant and the phenotypic reversion can be explained by RNA structural changes in the flanking U5 hairpin. The U5 mutation is incompatible with formation of the wild-type U5 hairpin, because at least the upper two base pairs are affected by the mutation (Fig. 3 , [Delta]G > 0 kcal/mol). Intriguingly, this mutant is expected to adopt an alternative structure that includes six PBS nucleotides (Fig. 3 , hairpin His-AC*). This rearranged structure not only occludes a significant part of the PBS sequences, but is also more stable than the wild-type U5 hairpin (Fig. 3 , [Delta]G values of -8.0 and -5.4 kcal/mol respectively). Thus the U5 mutation will lead to rearrangement of the U5-PBS structure and I propose that interference with annealing of the tRNA primer contributes to the severe replication defect of this mutant. This idea is supported by the structural changes observed in the revertant virus.
ACKNOWLEDGEMENTS
I thank several members of my laboratory for helpful discussions (Atze Das and Belinda Oude Essink) and Wim van Est for excellent artwork. Research in the Berkhout laboratory is sponsored by the Netherlands Organization for Scientific Research (NWO), the Dutch Cancer Society (KWF), the European Community (grant 950675) and the Dutch AIDS Fund.
REFERENCES
1 Leis,J., Aiyar,A. and Cobrinik,D. (1993) In Skalka,A.M. and Goff,S.P. (eds), Reverse Transcriptase. Cold Spring Harbor Laboratory Press, Cold Spring Harbor, NY, pp. 33-48.
2 Haseltine,W.A., Maxam,A.M. and Gilbert,W. (1977) Proc. Natl. Acad. Sci. USA, 74, 989-993.MEDLINE Abstract
3 Cobrinik,D., Soskey,L. and Leis,J. (1988) J. Virol., 62, 3622-3630.MEDLINE Abstract
4 Aiyar,A., Cobrinik,D., Ge,Z., Kung,H.J. and Leis,J. (1992) J. Virol., 66, 2464-2472.MEDLINE Abstract
5 Isel,C., Ehresmann,C., Keith,G., Ehresmann,B. and Marquet,R. (1995) J. Mol. Biol., 247, 236-250.MEDLINE Abstract
32 Vicenzi,E., Dimitrov,D.S., Engelman,A., Migone,T.-S., Purcell,D.F.J., Leonard,J., Englund,G. and Martin,M.A. (1994) J. Virol., 68, 7879-7890.MEDLINE Abstract
33 Myers,G., Korber,B., Hahn,B.H., Jeang,K.-T., Mellors,J.H., McCutchan,F.E., Henderson,L.E. and Pavlakis,G.N. (1995) Human Retroviruses and AIDS. A Compilation and Analysis of Nucleic Acid and Amino Acid Sequences. Theoretical Biology and Biophysics Group, Los Alamos National Laboratory, Los Alamos, NM.