Kinking of RNA helices by bulged bases, and the structure of the human immunodeficiency virus transactivator response element

J. Mol. Biol. (1992) 226. 305-310

Kinking of RNA Helices by Bulged Bases, and the Structure of the Human Immunodeficiency Virus Transactivator Response Element Fiona A. Riordan,

Anamitra

Bhattacharyyat,

Sean McAteer

and David M. J. Lilleyl

Department of Biochemistry The University Dundee DDl 4HN, U.K. (Received 6 February

1992; accepted 20 March

1992)

We have used gel electrophoresis to show that the pyrimidine bulge of the HIV-l TAR sequence causes a local bending of the helical axis. The TAR bulge caused a retardation in electrophoretic mobility in polyacrylamide gels. When this was placed adjacent to an additional bulged sequence in a linear RNA fragment, the mobility of the molecule varied sinusoidally with the spacing between the two bulges. Electrophoretic mobilities suggested that the TAR sequence context of the pyrimidihe bulge causes a greater degree of axial kinking than in an equivalent randomly chosen sequence. Experiments in which an A, bulge was progressively opposed by adenine bases inserted in the opposite strand showed that even a single opposed adenine markedly reduced electrophoretic mobility, i.e. axial bending, and two adenine bases reduced the mobility virtually to that of a normal duplex. We suggest that the pronounced kinking resulting from an unopposed bulge provides a particularly recognizable feature in RNA, and that this is the basis of the interaction between the HIV Tat protein and the TAR sequence.

Keywords: RNA

structure;

HIV;

AIDS:

structures of RNA three-dimensional The molecules are frequently complex, with extensive secondary and tertiary structure (for a review, see Tinoco et al., 1990). However, this can usually be simplified by division into a series of structural including sections of duplex, hairpin elements, loops, single and multiple-base bulges, internal loops (or “bubbles”), helical junctions and pseudoknots. The collection of such features present in a given RNA species should provide a very distinctive overall structure, and could contribute to the specificity of RNA-protein interaction. The interaction between the human immunodeficiency virus (HIQ) transactivator protein (Tat) and its RNA binding site, the transactivator response element (TAR), provides a simple example. t Present address: Department of Microbiology and Immunology, School of Medicine, University of California. San Francisco. CA 94143 0414, U.S.A. $ Author to whom all correspondence should be addressed $ Abbreviations used: HIV. human immunodeficifnog virus; TAR, t’ransactivator response element. bp. base-pair(s). 0022 -2x3(i;s~/140305-0(i

$03.oo~o

RNA-protein

interaction;

TAR

sequence

Tat is essential for HIV replication (Dayton et al., 1986; Fisher et al., 1986). Upon binding to the TAR RNA sequence located at the 5’ end of each viral mRNA (Berkhout et al., 1989; Feng & Holland, 1988; Hauber & Cullen, 1988; Jakovobits et al., 1988; Rosen et aE., 1985), it causes a substantial increase in transcript levels (Cullen, 1986; Laspia et al., 1989; Muesing et al., 1987; Peterlin et al., 1986; Rice & Mathews, 1988; Sodroski et al., 1985; Wright et al., 1986), possibly by preventing premature termination of the transcriptional elongation complex (Kao et aZ., 1987), or directly at the level of initiation of transcription (Sharp & Marciniak, 1989). The TAR element contains a large degree of self-complementarity, and may be folded to generate a large hairpin loop, the stem of which is interrupted by bulge structures (one or more consecutive, unopposed bases on one strand) (Berkhout & Jeang, 1989; Jakobovits et al., 1988; Muesing et al.. 1987). The sequence of the top of the hairpin loop and a three base bulge is shown in Figure 1. This feature is conserved in both HIV-l and HIV-2 (Emmerman et al., 1987). Studies in a number of laboratories have shown that the critical feature in the specific binding of Tat, and peptides derived

306

F. A. Rio&an

GG U C

G A CG GC AU GC

C u u

: : AU GC AU CG CG 1 II ,

Figure 1. Central part of the base sequence of the HTV-1 TAR element drawn as a hairpin loop (Berkhout & Jeang, 1989; Jakobovits et al., 1988; Muesing et aE., 1987). The full stem contains more than 20 bp, and additional single base bulges. The sequence of the 3 pyrimidine bulge can be either UCU or UUU.

from it, to the TAR element is the three-pyrimidine (Y3) bulge that is located four base-pairs from the terminal hairpin loop (Dingwall et al., 1989, 1990; Feng & Holland, 1988; Roy et al., 1990; Weeks et aE., in 1990; Weeks & Crothers, 1991); trans-activation vivo is dependent on the same feature. We have shown previously that additional adenine or thymine bases introduced into one strand of an RNA duplex (an RNA bulge) cause a retardation of the migration of the species in polyacrylamide gels (Bhattacharyya et aE., 1990). Variation of the position of the bulge, and the periodic variation of mobility with the spacing between two such bulges, indicated that the retardation in electrophoretic mobility was due to a kinking of the helix axis at the position of the bulge. The magnitude of this kinking was a function of the number and type of bases present in the bulge, and similar results have also been obtained in another study (Tang & Draper, 1990). We suggested that the defined kink generated by bulged bases in RNA might provide a very recognizable feature for interaction with proteins (Bhattacharyya et al., 1990). It has been suggested that axial kinking at the pyrimidine bulge might be an important element in the mechanism of the recognition of the TAR sequence by the Tat protein. We therefore examined the gel electrophoretic mobility of double-stranded RNA fragments containing the TAR bulge in its local context, to see if this RNA did contain the proposed kink. We prepared two series of bulgecontaining RNA species (Fig. Z(a)) by transcription

et al.

(Milligan et al., 1987) and compared their electrophoretic mobilities in polyacrylamide gels. The first set of molecules (AB series) comprised a well characterized series of 40 bp duplexes containing central oligouridine bulges of various size, that were used in our earlier studies of RNA bulges (Bhattacharyya et al.: 1990). The second set were molecules of the same length, but differed from the first series in that the central bulge was contained in the TAR flanking sequence (Fig. 1); three such species were prepared, containing UUU or UCU bulges as found in natural viral TAR sequences, or no bulge (i.e. a perfect duplex). These species were electrophoresed in 15% (w/v) is polyacrylamide gels, and the autoradiograph shown in Figure Z(b). As we have shown before. there is a progressive retardation of the doublrstranded AR species as the number of uridines in the bulge is increased, consistent with an increasing axial bending, and this dependence of mobility on bulge size is plotted in Figure 2(c). The G, and UCl: bulges in the TAR context also bring about a significant retardation of the RNA fragments, suggesting that’ these species are also kinked by the TAR bulges. Furthermore, the migration of these species is more retarded in the TAR context than for the equivalent U, bulge in the AB series context. Interpolation on the plot of Figure 2(c) indicates that the U, and UCU bulges generate a retardation in the TAR cont,ext that, is equivalent almost to a U, bulge in the AB context. Thus while the corresponding duplex species have identical electrophoretic mobilities, the Y, bulges have cont.ext,dependent mobility. This suggests that the TAR sequences are indeed kinked by the presence of the Y, bulge, and that the sequence context of the bulge may bring about a larger angle of axial kinking than some other sequences. retardation is Although the electrophoretic strongly suggestive of axial kinking, other explanations are possible, such as a point of flexibility located at the TAR sequence. In view of the importance of the HIV virus and its regulation we sought confirmation of axial kinking by means of a phasing experiment using a second (A,) bulge kink. A series of 60 bp duplexes was prepared (Table l), in which the U, TAR sequence was placed adjacent to an A, variable spacing between the two bulge, with bulges. The overall number of base-pairs was kept constant for all the molecules of the series. If both bulges result in local kinking of the helix axis, the dihedral angle relating the outer sections of the molecule will depend on the spacing between the bulges, and a sinusoidal variation of mobility with spacing should result. We have previously demonstrated this to be the case for two A, bulges et al., 1990). The result is shown in (Bhattacharyya Figure 3(a). It is clear that there is a sinusoidal variation of gel mobility with the spacing between the two bulges, as indicated by the plot shown in Figure 3(b). Fitting to a sine function indicates an approximate helical repeat of 12 bp/turn, consistent with the current state of knowledge about RNA

Communications

307

AB series 5’ 3’

GGGCGUCCUGUGGAUCCAGGUnUCGGAUCCUCUACGCCGCCC 3 CCCGCAGGACACCUAGGUCC..AGCCUAGGAGAUGCGGCGGG 5

TAR sequences 5’ 3’ 5’ 3’

3’ GGGCGUCCUGUGGAUCCAGAUUUGAGCAUCCUCUACGCCGCCC 5’ CCCGCAGGACACCUAGGUCU...CUCGUAGGAGAUGCGGCGGG 3 GGGCGUCCUGUGGAUCCAGAUCUGAGCAUCCUCUACGCCGCCC CCCGcAGGAcACcUAGGucu...CUCGUAGGAGAUGCGGC 5’

(a) AB Un

TAR IIr dupUUU

UCU 0

1

3

5 n

(b)

2

4 Bulge

size

6

8

(bases)

Figure 2. The TAR sequence pyrimidine bulge causes retarded electrophoretic mobility in polyacrylamide gels. (a) Sequences of double-stranded Rh’A species used in these experiments. The AR series comprised 40 bp duplexes interrupted by U, bulges on the top strand. where n was 0. 1, 3. 5 or 7. We have previously characterized bending by these bulged molecules (Bhattacharyya et al.. 1990). The TAR sequences were 40 bp duplexes of closely similar sequence. where the sequence flanking the bulge was that of the natural TAR RNA. The bulge sequences were either UUU or CUT. (b) Comparison of electrophoretic mobility of the TAR-bulge molecules and R8A species were the AB series in 15?~~,polyacrylamide. synthesized by transcription of syrtheticDi%A templates by T7 RNA polymerase (Mllhgan pt ul.. 1987) in 40 miv-Tris. HCI (pH 7.5). 6 mM-MgCI,. 2 mw-spermidine. @5 mM each of &4TP, GTP. CTP and I’TP together with [LX-~‘P]CTP at 37 “C. Synthesized RXA was purified by gel electrophoresis in 7 M-urea. Duplex species were generated by hybridization of equimolar quantities of single strands in @45 M-Ea citrate. PO45 M-SaCl at 6;iY’ followed by slow cooling. Double-stranded RlVA species were electropolyacrylamide (acrylamide : phoresed in 153, bisarrylamide : : 29 : 1) in 90 mM-Tris-borate (pH &3), 50 m,n-KaCl with recirculation of buffer at 1 l/h. One gel plate was constructed to allow serpentine water circulat,ion from a water bath at 15( kO.l)“(‘. and the gel was run for 40 h at 15 V/cm. The gel was then subjected to autoradiography. Tracks L to R: dup. TAR sequence without the 3 base bulge. i.e. perfect duplex species: V.‘uC and UCU, TAR’ sequence with UUU of UC’I’ bulges, respectively; AU C,, AB series of bulged duplexes containing the indicated number of bulged uraril bases. (c) Plot of electrophoretic mobility against uridinr bulge size for the AR series of molecules. The mobility of the U, bulge in the TAR sequence cont,ext is interpolated. Note the relatively slow migration of this species compared with an identical bulge in the ,4B sequence context. The error bars represent the widths of the bands measured on the autoradiograph.

(c)

structure. This is strong evidence that the U, bulge in the TAR context brings about a local kinking of the helical axis. These results suggest that an important feature of

the TAR element may be the local distortion of the helix axis, and that this could be significant in its interaction with the Tat protein. A related feature found in many RNA species is the internal loop or

F. A. Riordan et al.

308 9

II

13

I5

17

20

bp

(a) 780 5

740

;? -5E 700 .g 660 0 h 5 620 580 8

IO

12

14

16

Spacing

(bp)

I8

20

22

(b)

bubble, formed by consecutive base mismatches. This may be symmetrical, or there may be more unpaired bases on one strand than the other. Our studies of DNA lead us to propose that the symmetrical bubble generated much less kinking of the helical axis (Bhattacharyya & Lilley, 1989). Wr therefore examined the effect of opposed but noncomplementary bases on the electrophoretic mobility of a fragment containing a central bulge. Tn this experiment, an A, bulge was progressively opposed by addition of adenine bases on the opposit’e strand, until a bubble of five consecutive A. A mispairs was produced, and the resulting mobility in 20 o/o polyacrylamide was examined (Fig. 4). The result of opposing the A, bulge is striking. Despite the increase in size of the molecule, the mobility increases on opposition of the bulge. Even addition of a single opposing base markedly increases the mobility, and two or more adeninea result in the fragment having virtually the same mobility as t,he unbulged duplex. This suggests that the opposed bulge in RNA is considerably less effective in generating a local kinking of t,he helix axis. Very recently Berkhout & Jeang (1991) have shown that when the UCU bulge of the HIV-l TAR sequence was opposed by a single guanine base, transactivation was reduced by more than fivefold. These experiments show that an unopposed bulge is optimal in generating local axis kinking in doublrstranded RNA, and confirm the supposition that the YS bulges of the TAR elements result in such kinking. Moreover, it appears that the contextual sequence in the TAR RNA is predisposed t.o be more strongly kinked by the bulge? compared t*o a randomly chosen context. Why this should be so is presently unclear. It could be related t’o the strand asymmetry of purines flanking the bulge. Another possibility is that a base-pair on one side of t)he bulge may be broken, as we have observed in n.m.r. studies of a bulged DNA molecule (Aboul-ela. F.. Homans. S. & Lilley, I>. M. J., unpublished result)s); this could be consistent with adenine react’ivity to diethyl pyrocarbonate in the TAR sequence (Weeks & Crothers, 1991). It, is also interesting t’hat t,he adenine

Figure 3. Phasing of base bulges shows that the TAR bulge generates a kinked helix axis. A series of 57 bp duplex species (Table 1) was generated in which the spacing between an A, bulge and the TAR U3 bulge was varied between 9 and 20 bp. The overall size of the molecule was kept constant through the series. (a) Gel electrophoresis of the double-bulge series of RNA duplexes in 20% polyacrylamide in 90 mM-Tris-borate (pH 8.3), 50 mM-NaCl at 15( kO.l)“C. The spacing between bulges (bp) is indicated for each species above the track. (b) Plot showing the variation in electrophoretic mobility wibh the spacing between the A, and TAR bulges in this series of isomeric molecules. The error bars on the data points represent the widths of the bands measured on the autoradiograph. The line was generated by non-linear regression to a sine function. The data show clearly the sinusoidal variation between electrophoretic mobility and bulge spacing, confirming the presence of a defined kink at the TAR bulge. The data do not permit an

immediately

3’ to

the

pyrimidine

bulge

is

selectively converted to inosine when coinjtcted with Tat protein into the nuclei of Xenopwx oorytcs (Sharmeen et nZ.. 1991). The observed kinking ma) be an important determinant of the specificit’y in Tat binding, which is consistent with peptide binding studies in which the TAR sequence has been modified (Weeks & Crothers, 1991). Base bulges constitute a number of RNB-protein interaction sites,

suggesting

able features.

that

they

Exploitation

are particularly

of RNA

recogniz-

secondary

strut-

accurate determination of helical repeat for this RNA species. (c) Schematic diagram showing the dihedral angle (0) generated by 2 kinks when they generate anisotropic kinking in the RKA duplex. This will be a function of the length and helical repeat of the duplex section separating the TAR and A, bulges.

309

Communications Table 1 Sequences of double-stranded RNA species used in electrophoretic experiments in which an A, bulge was phased relative to the TAR bulge 5’ 3’

GGGCGUCCUGUGCGUGGAUCCCUGA5CGUGGCAGAUUUGAGCAUCCUCUGUCUACGCCGCCC CUCGUAGGAGACAGAUGCGGCGGG CCCGCAGGACACGCACCUAGGGAC GCACCGUCU

3’ 5

5’ 3’

GGGCGUCCUGUGCGUGAUCCCUGA5CGUGGCGCAGAUUUGAGCAUCCUCUGUUACGCCGCCC CUCGUAGGAGACAAUGCGGCGGG CCCGCAGGACACGCACUAGGGAC GCACCGCGUCU

3’ 5’

5’ 3’

GGGCGUCCUGUGCGGAUCCCUGA5CGUGGUUC~UUU~AUCCUCUGUACGCCGCCC CUCGUAGGAGACAUGCGGCGGG CCCGCAGGACACGCCUAGGGAC GCACCAAGCGUCU

3’ 5’

5’ 3’

GGGCGUCCUGUGCGAUCCCUGA5CGUGGUGGUCGCAGAUUUGA’XAUCCUCUUACGCCGCCC CUCGUAGGAGAAUGCGGCGGG CCCGCAGGACACGCUAGGGAC GCACCACCAGCGUCU

3’ 5’

5’ 3’

GGGCGUCCUGUGGAUCCCUGA5CGUGGUCUGGUCGCADAUCC CCCGCAGGACACCUAGGGAC GCACCAGACCAGCGUCU

3’ 5

5’ 3’

GGGCGUCCUGGAUCCCUGA5CGUGGUCUUGUGGUCGCAGA-GAGCAUCCUCACGCCGCCC C!UCGUAGGAGUGCGGCGGG CCCGCAGGACCUAGGGAC GCACCAGAACACCAGCGUCXJ

CUCGUAGGAGAUGCGGCGGG

3’ 5

The spacing between the 2 bulges was varied from 9 to 20 bp through the series. The TAR bulge and context are highlighted in bold.

AS/An A5/U5

1

clup

0

I 1

2

3

4

5n

ture by binding proteins may be a consequence of the A-helix of double-stranded RNA, making contacts

with

the

edges

of

the

base-pairs

in

the

deep, narrow major groove rather difficult. Thus, whereas the majority of DNA-binding proteins appear to bind to their targets by direct sequence readout, indirect readout via the secondary structure may be more important in RNA. Weeks & Crothers (1991) have suggested that a function of the TAR bulge may be to widen the major groove locally, in order to permit direct protein-base contact. It has also been suggested that arginine residues could “measure” local groove width (Calnan et al., 1991). Thus selective protein binding may be a complex function of both direct and indirect sequence readout, in which helical distor5' GGGC~~~~G~~~~~AW~~~U~~U~~ACG~XG~CC 3 3' CCCGCAGGAcACCUAGGUCC~GCCuAGGAGAUGcGGcGGG 5

Figure

4. Mismatched

bases opposing a bulge reduce the axial kinking generated by the bulge. A series of double-stranded RNA species based on the AR series containing an A, bulge were generated by transcription and hybridization. The sequence of the lower strand was varied in order to oppose the 4, bulge with 0, 1, 2, 3, 4 or b adenine bases. These species were electrophoresed in a 204& polyacrylamide gel in 90 mM-Tris-borate (pH %3), 50 mM;NaCl at 15( k@l)“C, and the gel subjected to autoradiography. The left’most track contains a perfect duplex species in which the A, bulge is opposed by a U, sequence. thereby generating 5 consecutive A. U basepairs. The remaining tracks contain the A, bulged species, where the bulge is opposed by the indicated number of adenine bases. Note that the effect of opposing the A, bulge is a marked increase in electrophoretic mobility, despite the increased size of the molecules.

tion

of the

kind

we have

described

We thank Alastair Murchie sion, and the MRC and SERC

may

be critical.

and Bob Clegg for discusfor financial support.

References Berkhout. B. & Jeang, K.-T. (1989). Trans-activation of human immunodeficiency virus type 1 is sequence specific for both the single stranded bulge and loop of the trans-acting responsive hairpin: a quantitative analysis. J. Viral. 63, 5501-5504. Berkhout B. & Jeang, K.-T. (1991). Detailed mutational analysis of TAR RNA: critical spacing between the bulge and loop recognition domains. Nucl. Acids Res.

19, 6169-6176. Berkhout. B., Silverman, R. H. & Jeang, K.-T. (1989). Tat trams-activates the human immunodeficiency virus through a nascent RNA target. cell, 59, 273-282.

310

F. A. Riordan

Bhattacharyya, A. & Lilley, D. M. J. (1989). The contrasting structures of mismatched DNA sequences containing looped-out bases (bulges) and multiple mismatches (bubbles). Nucl. Acids Res. 17, 6821-6840. Bhattacharyya, A., Murchie, A. I. H. & Liiley, D. M. J. (1990). RNA bulges and the helical periodicity of double-stranded RNA Nature (London), 343, 484-487. Calnan. B. J., Tidor, B., Biancalana, S.: Hudson. D. & Frankel. A. D. (1991). Arginine-mediated RNA recognition: Science, 252, the arginine fork. 1167-1171. Cullen, B. R. (1986). Trans-activation of human immunodeficiency virus occurs via a bimodal mechanism. Cell, 46, 973-982. Dayton. $. I., Sodroski, J. G., Rosen, C. A., Goh, W. C. & Haseltine, W. A. (1986). The trans-activator gene of human T cell lymphotropic virus type III is required for replication. Cell, 44, 941-947. Dingwall, C., Emberg, I., Gait, M. J., Green, S. M.. Heaphy, S.: Karn, ,J.. Lowe, A. D., Singh, M.. Skinner, M. A. & Valerio, R. (1989). Human immunodeficiency virus 1 tat protein binds trans-activationresponsive region (TAR) RNA in vitro. Proc. Nat. Acad. Sci., C.S.A. 86, 6925-6929. Dingwall, C.. Ernberg, I., Gait, M. J., Green, S. M.; Heaphy, S.. Karn, J., Lowe, A. D., Singh, M. & Skinner, M. $. (1990). HIV-I tat protein stimulates transcription by binding to a U-rich bulge in the stem of TAR RNA structure. EMBO J. 9. 4145-4153. Emmerman, M., Guyader, M., Montagnier. L., Baltimore, D. & Muesing, M. A. (1987). The specificity of the human immunodeficiency virus type 2 trans-activator is different from that of human immunodeficiency virus type 1. EMBO J. 6, 3755-3760. Feng, S. & Holland, E. C. (1988). HIV-l tat trans-activation requires the loop sequence within tar. Nature (London), 334, 165-168. Fisher, A. G., Feinberg, M. B., Josephs, S. F., Harper, M. E., Marselle, L. M., Reyes, G., Gonda, M. A., Aldovini, A., Debouck, C., Gallo, R. C. & Wong-Staal, F. (1986). The trans-activator gene of HTLV-III is essential for virus replication. Nature (London), 320, 367-371. Hauber. J. & Cullen, B. R. (1988). Mutational analysis of the trans-activator responsive region of human immunodeficiency virus type 1 long terminal repeat. J. Virol. 62, 673-679. Jakobovits, A., Smith, D. H., Jakobovits, E. B. & Capon, D. J. (1988). A discrete element 3’ of human immunodeficiency virus 1 (HIV-l) and HIV-2 mRNA initiation sites mediates transcriptional activation by an HIV trans-activator. Mol. Cell Biol. 8, 2555-2561. Kao, S.-V., Calman. A. F., Luciw, P. A. & Peterlin, B. M. (1987). Anitermination of transcription within the long terminal repeat of HIV-l by tat gene product. Xature (London), 330, 489-493. Edited

et al.

Laspia, M. F.. Rice. A. P. & Mathews, M. B. (1989). HIV-l tat protein increases transcriptional initiation and stabilises elongation. Cell, 59, 283-292. Milligan, J. F., Groebe, D. R., Witherall, U. W. & Uhlenbeck, 0. C. (1987). Oligoribonucleotide synthesis using T7 RNA polymerase and synthetic DNA templates. Nucl. Acids Res. 15, 8783-8798. Muesing, M. A.. Smith, D. H. & Capon, D. .J. (1987). Regulation of mRNA accumulation by a human immunodeficiency virus trans-activator protein. Cell, 48, 691-701. Peterlin, B. M., Luciw, P. A.. Barr, P. J. & Walker. M. I). (1986). Elevated levels of mRNA can account for the trans-activation of human immunodeficiency virus. Proc. Nat. Acod. hi., r7.S.A. 83, 9734-9738. Rice. A. P. & Mathews, ,M. B. (1988). Transcriptional but not translational regulation of HIV-l by the tat gene product. Nature (London), 332, 551-553. Rosen. C. A.. Sodroski, J. G. & Haseltine, W. A. (1985). The location of &-acting regulatory sequences in the human T cell lymphotropic virus type III (HTLV-TII/LAV) long terminal repeat. Cell, 41, 8133823. Roy, S.. Parkin. N‘. T., Rosen. C., Itovich. J. & Sonenberg, N. (1990). Structural requirements for tram activation of human immunodeficienry virus type 1 long terminal repeat-directed gene expression by Tat: imporbance of base-pairing, loop sequence and bulges in the Tat-responsive sequence. J. Viral. 64. 1402-1406. Sharmeen, L., Bass, B.. Sonenberg, N., Weintraub. H. & Groudine. M. (1991). Tat-dependent adenosine-toinosine modification of wild-type transactivation response RNA. Proc. Nat. Acad. Sci., U.S.A. 88. 809668100. Sharp; P. A. & Marciniak, R. A. (1989). HIV TAR: an RNA enhancer? Cell, 59, 229-230. Sodroski. -7. G., Rosen, C. A.! Wong-Staal, F., Salahuddin, S. Z.. Popovic, M., Arya, S., Gallo, R. C. & Haseltine. W. A. (1985). Trans-acting transcriptional regulation of human T-cell leukemia virus type ITT long terminal repeat. Science, 227, 171-173. Tang. R. S. & Draper, D. E. (1990). Bulge loops used t,o measure the helical twist of RNA in solut,ion. Riochemistry, 29, 523225237. Tinoco I.. Jr, Puglisi, J. D. & Wyatt, J. R. (1990). RNA folding. Nucl. Acids Mol. Biol. 4, 205-226. Weeks. K. M. bz Crothers, D. M. (1991). RNA recognition by Tat-derived peptides: interaction in the major groove! Cell, 66, 577-588. Weeks, K. M., Ampe, C., Schultz, S. C., Steitz, T. A. & Crothers, D. M. (1990). Fragments of HIV-I Tat protein specifically bind TAR RNA. Science, 249. 1281-1285. Wright, C. M., Felber, B. K., Paskalis, H. & Pavlakis, G. N. (1986). Expression and characterisation of the trans-activator of HTLV-III/LAV virus. fkience, 234, 988-992.

by J. Karn