the major groove of RNA resembles protein interactions in the minor groove ofDNA

A 17-amino acid arginine-rich peptide from the bovine immunodeficiency virus Tat protein has been shown to bind with high affinity and specificity to bovine immunodeficiency virus transactivation response element (TAR) RNA, making contacts in the RNA major groove near a bulge. We show that, as in other peptide-RNA complexes, arginine and threonine side chains make important contributions to binding but, unexpectedly, that one isoleucine and three glycine residues also are critical. The isoleucine side chain may intercalate into a hydrophobic pocket in the RNA. Glycine residues may allow the peptide to bind deeply within the RNA major groove and may help determine the conformation of the peptide. Similar features have been observed in protein-DNA and drug-DNA complexes in the DNA minor groove, including hydrophobic interactions and binding deep within the groove, suggesting that the major groove of RNA and minor groove of DNA may share some common recognition features. A wealth of structural information is now available about DNA-protein recognition, largely from crystallographic and NMR studies of DNA-protein complexes. In comparison, relatively little is known about RNA-protein recognition. The most detailed information is provided by the cocrystal structures of three tRNA synthetase-tRNA complexes (1-5), an R17 coat protein-RNA complex (6), and a UlA ribonucleoprotein domain-RNA complex (7). In the synthetase complexes, sequence-specific contacts occur primarily in the minor groove of tRNAGln (1, 2), in the major groove of tRNAAsP (3, 4), and to the phosphate backbone of tRNASer (5). Important contacts also are made to anticodon loop nucleotides in the glutaminyl and aspartyl complexes. The overall structures of the synthetases are rather different, and aside from RNA conformational changes observed upon binding, few general features of recognition have emerged. In the R17 and UlA complexes, the most important sequence-specific contacts are made to bulge and loop nucleotides of RNA hairpins (6, 7). Recently, several common RNA-binding motifs have been identified (for review, see refs. 8 and 9), suggesting that conserved structural features may be found in some classes of proteins. One of these motifs, the arginine-rich motif, consists of a short region of basic amino acids (8-20 residues long) particularly rich in arginine. This motif has been found in bacterial antiterminators, ribosomal proteins, coat proteins from RNA viruses, the human immunodeficiency virus (HIV) Tat and Rev proteins (10), and the bovine immunodeficiency virus (BIV) Tat protein (11). Studies with model peptides derived from the arginine-rich domains of HIV Rev, HIV Tat, and BIV Tat have emphasized the importance of RNA structure in protein recognition and have provided some details about sequence-specific interactions (11-16). In HIV Rev, a 17-amino acid peptide specifically recognizes the Rev response The publication costs of this article were defrayed in part by page charge payment. This article must therefore be hereby marked "advertisement" in accordance with 18 U.S.C. §1734 solely to indicate this fact. element (RRE) RNA when the peptide is in an a-helical conformation and uses 6 amino acids (4 arginines, 1 threonine, and 1 asparagine) for binding (12). NMR studies have shown that binding induces formation of an internal loop structure in the RRE containing GG and GA base pairs and two loopedout bases (17, 18). In HIV Tat, a single arginine residue within a 9-amino acid peptide is largely responsible for recognition of a bulge region in transactivation response element (TAR) RNA (14). The free amino acid arginine binds to TAR by using a similar set of RNA structural features (19), suggesting that a defined peptide conformation may not be needed for recognition. An NMR model of the arginine-TAR complex suggests that the guanidinium group of arginine hydrogen bonds to a guanine base in the major groove and to two phosphates on the backbone and that the complex is stabilized by a base triple interaction between a uracil in the bulge and an A-U base pair above the bulge (15). In both cases, bulges help widen the major grooves of adjacent A-form RNA helices, thereby increasing accessibility to the proteins (13, 20). BIV Tat is closely related to HIV Tat; however, initial studies with a 17-amino acid peptide suggested that BIV Tat uses a very different set of interactions to recognize BIV TAR (11). The peptide binds to an unusually accessible stem region adjacent to two single-nucleotide bulges in BIV TAR and requires an extensive set of determinants in the major groove (Fig. 1) very different from those of HIV TAR. The sequence of the BIV peptide also is distinct, containing several glycine and proline residues in addition to arginine residues (Fig. 1). In this study, we show that three glycine residues in the peptide are critical for BIV TAR recognition and, most surprisingly, that an isoleucine also is critical. We suggest that the isoleucine intercalates between bases in the major groove of BIV TAR and that glycine residues allow the peptide to bind deeply in the groove, analogous to interactions observed in the minor groove of DNA. MATERIALS AND METHODS Plasmids and Chloramphenicol Acetyltransferase (CAT) Assays. Plasmids encoding BIV Tat peptide mutants were constructed by cloning synthetic oligonucleotide cassettes into a gene encoding a BIV-HIV Tat fusion protein (11). Mutations were confirmed by dideoxynucleotide sequencing. To measure transcriptional activation, mutant BIV-HIV Tat plasmids (50-100 ng) were cotransfected into HeLa cells with an HIV long terminal repeat-CAT reporter plasmid (25-50 ng) in which HIV TAR was replaced with BIV TAR (11), and CAT activity was assayed after 48 hr and quantitated as described (21). Peptides and RNAs. BIV Tat peptides used for in vitro RNA-binding experiments were synthesized with C-terminal amides and acetylated at the N termini on an Applied BioAbbreviations: BIV, bovine immunodeficiency virus; HIV, human immunodeficiency virus; RRE, Rev response element; CAT, chloramphenicol acetyltransferase; Tm, melting temperature; TFE, trifluoroethanol; TAR, transactivation response element.