Evidence for the Absence of the G-T-OP-C Sequence From Two Mammalian Initiator Transfer RNAs (rabbit liver/sheep mammary gland/thin-layer chromatography)

Analyses were performed on the purified initiator tRNA from rabbit liver to test for the presence of V,p and Tp in this molecule. Neither of these nucleotides could be detected after hydrolysis by piperidine, NaOH, or T2 RNase. Similarly* digestion with venom phosphodiesterase plus phospkatase failed to release any pseudouridine or ribothymidine. Identical results were obtained with the initiator tRNA from sheep mammary gland. The absence of these nucleosides was confirmed by pancreatic and Ti RNase digestion of the rabbit-liver initiator tRNA. The classical G-T-VI-C sequence was not detected in this molecule. An A-U-C-G sequence has been identified; it may possibly replace the G-T-#,-C sequence. Initiation of protein biosynthesis in Escherichia coli requires a special formylatable methionine acceptor tRNA, tRNAmt (1). In the cytoplasm of eukaryotic cells, two major tRNAs specific for methionine have been found, one of which functions as an initiator (2-6). This species has been purified from baker's yeast (4), from rabbit liver (7), and from sheep mammary gland (8). The tRNA isolated from baker's yeast initiates the synthesis of proteins in E. coli and the synthesis of hemoglobin in cell-free extracts of rabbit reticulocytes (9). In mammalian cells, the assumption that this species is an authentic initiator is based upon the following properties: (i) the methionyl-tRNA can be formylated by E. coli transformylase; (ii) when formylated, it can form, at low magnesium ion concentrations, an initiation complex with E. coli ribosomes and poly(AUG) in the presence -of the bacterial initiation factors; it can compete with E. coli tRNAMet for the formation of this complex; (iii) it can react with puromycin on bacterial ribosomes to give formylmethionylpuromycin (10). In order to establish the extent of structural homology between initiator tRNAs from prokaryotes and eukaryotes, we have undertaken the primary sequence analysis of the initiator tRNA from rabbit liver. Preliminary results show that, contrary to all the tRNAs sequenced until now, with the exception of tRNA 'G from Salmonella epidermidis (11), this tRNA does not contain the G-T-V/-C sequence. In addition, we have detected neither ribothymidine nor pseudouridine in purified initiator tRNA from sheep mammary gland. Similar results have been obtained by Simsek et al. with the tRNAmet from yeast and wheat germ (23). These authors (23) have also confirmed my results on rabbit-liver and sheep mammary gland initiator tRNAs. MATERIALS AND METHODS AG 50-X12 and AG 1-X8 were purchased from Biorad, DE 23 and DE 81 paper from Whatman, alkaline phosphatase and pancreatic ribonuclease from Worthington, and T1 RNase from Sankyo. The crude T2 RNase preparation was a gift from Dr. G. Keith. Crude and purified rabbit-liver and sheep mammary gland tRNAs were prepared as described (12). Purified E. coli tRNAf e was obtained according to Takeishi et al. (13). Alkaline hydrolysis of tRNAs was performed in 10% piperidine for 2 hr at 1000 or in 0.3 N NaOH for 24 hr at 37°; T2 RNase hydrolysis was performed in 0.05 M ammonium acetate (pH 4.6) for 2 hr at 37°. Hydrolysis to nucleosides was in 0.1 M triethylammonium bicarbonate (pH 8.5) (8 A260 units in 100 ,l) with venom phosphodiesterase (50 ug) and alkaline phosphatase (20 ,g) for 16 hr at 37°. For thin-layer chromatography on cellulose plates, the following solvents were used in the first and in the second dimension, respectively: isobutyric acid-25% NH40H-water 66:1:33 and isopropanol-concentrated HCl-water 70:15:15. The buffer for high-voltage electrophoresis on DEAE paper was 7% HCOOH. Pancreatic and T1 RNase digestion of tRNAs, chromatography on DEAE-cellulose-urea columns, and purification of the oligonucleotides were performed as described by Gangloff et al. (14).