Cell, Vol. 58, 811-812, September 8, 1989, Copyright 0 1989 by Cell Press The Product of the PRP4 Gene of S. cerevisiae Shows Homology to p Subunits of G Proteins The PRW gene of Saccharomyces cerevisiae was cloned by complementation of a conditional lethal mutation that causes a defect in pre-mRNA splicing (Soltyk et al., 1984). The DNA sequence of this gene revealed a 465 amino acid open reading frame that encodes the 52 kd PRP4 protein. Through an immunological approach, the PRP4 protein was shown to be a stable component of the yeast U4/U6 small nuclear ribonucleoprotein particle (snRNP), consistent with its role in RNA splicing (Petersen-Bjorn et al., 1989). We:report here that the deduced PRW amino acid se- quence shows a striking similarity to that of the 8 subunit of bovine transducin (GQ) and other members of an ex- PRP4 bovtrans CDU STE4 PRP4 bovtrans E(vl) coc4 STE4 PRP4 bovtrans E(spl) * CDC4 STE4 PUP4 bovtrans E(SPll CDC4 STE4 PRP4 bovtrans E(spli CDC4 STE4 Letter to the Editor panding group that includes CDC4 (Fong et al., 1986), En- hancer of Split (E(spl); Hartley et al., 1988), and STE4 (Whiteway et al., 1989). An alignment of all five protein se- quences is shown in the figure. The key feature is a 42-64 amino acid repeat structure within which certain amino acids and their spatial relationships are highlly conserved. PRP4 appears to contain five such repeats,. Of the five proteins whose similarities are noted, the functions of transducin and the STE4 gene product are best understood. Transducin is a G protein involved in sig- nal transduction between rhodopsin and a cyclic GMP phosphodiesterase in the mammalian retina (reviewed in Neer and Clapham, 1988). The heterotrimeric protein, composed of Gt,, G,,, and Gt, subunits, is inactive when associated with both membrane-bound rhodopsin and GDP Upon receipt of an extracellular light signal, there is an exchange of GDP for GTP on the CL subunit and a dis- sociation of GtdGTP from both Gtbr and rhodopsin. G,dCJP activates the effector enzyme (cGMP phosphodiesterase) 294- ASTHDELLLQEGHDKGVFSLSFQCD@SL~CSGGMDSLSHLWD 171- IETGfJ~TTTFTGHTGDVHSLSLAPDTRLFVSGACDASAKLWD 42D- IPR"ARQINTL~GEVf/A~TISNP~KY,Yf-jK[]GC~K~ 409- SINKKFLLQ~SGHDGGVYALKYAHGGILV.SGSTDRTVRVWD 121- SASGLKQNAIPLDSQWVLSCAISPSSTLVASAGLNNNCTIYR 37& IRKRDEGQVNQILAHRNIV................... TQVRFSKEDGGKKLVSCGYDNL~NVYS 25% LRADQELHTYSHDNIICGI... ., .., . . . . . . . . . . SD& LASPTPRIKAELTSAAPACYAL TSVSFSKS. jRL~L!#D~~FN~N~ AISP.....................DSKVCFSCCSDGNIAVWD 494- LPKESSVPDHGEEHDYPLVFHT PEENPYFVGV~R~MASVRT~SGH..GNI~VS~SYlZlNTLIVWD 209- IPKAKRVREYSDHLGDVLALAI P EEPN................. LENS~NTFALflC~S~~GYTY~ L GH I I G I SGSD I IWD v L S LTG L L V Y v v A Multiple Alignment of PRP4 with CDC4, E(spl), Bovine Transducin, and STM The alignment was manually constructed from independent pairwise alignments using the University of Wisconsin Genetics Computer Group (UWGCG) LINEUP program. The consensus sequence, which represents the consensus for a single repeat, notes those positions ait which a single amino acid or group of biochemically similar amino acids predominates. The match of PRW to transducin was initially discovered by searching the NBRF data base (v. 19), making use of the UWGCG program WORD- SEARCH (Devereux et al., 1984). However, this match was not the top score and was preceded by alignments that upon further inspection were clearly insignificant. A subsequent search using the FASTA program of Pearson and Lipman (1988) resulted in a list of matches )n which that to transducin was by far the best score. All software was implemented on the SERC SEQNET facility (Daresbury, England),