Is EIAV Tat protein a homeodomain?

Recently two publications (1) reported RNA-binding capacity of the Drosophila ho-meotic bicoid protein bcd (2). We would like to point out that the reverse discovery, namely that a protein turned out to be a homeodomain-like structure after it was discovered as an RNA binding factor, happened in our laboratory. The HIV-1 transactivator protein (hl-tat) is considered a tat response region (TAR) RNA binding protein, and the equine infectious anemia virus (EIAV) tat protein (e-tat) was inferred to play a role similar to the hl-tat protein (3). We recently determined the three-dimensional features of both these proteins by nuclear magnetic resonance (NMR) spectroscopy (4) and found that e-tat contains helical secondary structure as a limit structure. This limit structure was found to be dramatically stabilized by the addition of trifluoroethanol (tfe) (5). The secondary structure of the e-tat protein in tfe showed clearly a helix-loop-helix-turn-helix binding motif similar to the homeodomain protein motif (6, 7). Comparison between helix motif proteins sequences, including bcd, and e-tat showed some resemblance, but no striking homolo-gies, as shown by the alignment of the bcd and e-tat sequences as determined with the ClustalV program (8) and corrected for a continuous helix from Ala 28 to Lys 37 in bcd (7) (Fig. 1). Thus, e-tat and bcd are not closely related on the sequence level (12% identity and 44% similarity). On the level of secondary structure, however , the aligned sequences of bcd and e-tat showed unexpected homology (Fig. 1). The recognition helix of the homeodomain proteins , helix III from the NH2-terminus, corresponds to the basic sequence region of e-tat considered to be responsible for TAR recognition. In addition to the similarity of the helix pattems, both proteins show a strictly conserved turn between helices II and III.