A single base change in the intron of a serine tRNA affects the rate of RNase P cleavage in vitro...

THE JOURNAL OF BIOLOGICAL CHEMISTRY 0 1986 by The American Society of Biological Chemists, Inc.

Vol. 261, No. 13, Issue of May 5, pp. 587&5885,1986 Printed in U.S.A.

A Single Base Change in the Intron of a Serine tRNA Affects the Rate of RNase P Cleavage in Vitro and Suppressor Activity in Vivo in Saccharomyces cerevisiae”

(Received for publication, October 18,1985)

Ian Willis, David FrendeweyS, Mark Nichols, Andrea Hottinger-Werlen, Jerome Schaacks, and Dieter So111 From the Department of Molecular Biophysics and Biochemistry, Yak University, New Haven, Connecticut 0651 1

Differences in the processing of dimeric tRNASe’- tRNAMet precursors derived from the Schizosacchuro- mycespombe sup9 wild-type and opal suppressor genes can be attributed to conformational alterations in the tRNAser anticodonlintron domain. A comparison of the patterns obtained upon transcription of the sup9+ (wild-type) and sup9-e (opal suppressor) genes in a coupled transcriptionlprocessing extract from Succha- romyces cerevisiue reveals that the latter exhibits a greatly reduced efficiency of 5’-end maturation and is susceptible to specific endonucleolytic cleavage(s) within the intron. Free energy calculations indicate that these effects coincide with a destabilization of the wild-type anticodonlintron stem and suggest that the predominant sups-e conformer lacks secondary structure in this region. Evidence in support of this hypothesis was obtained by (i) analyzing the processing of sup9+ and sups-e precursors carrying the intron base substitution, G37:10, which destroys and restores, respectively, the base-pairing potential of the proposed secondary structure and (ii) comparing the strength and temperature sensitivity of sup9-e and sups-e G37:lO suppression in vivo in S. cerevisiae. The data indicate that the anticodon/intron structure of tRNA precursors can influence the rate of RNase P cleavage in vitro and affect tRNA expression in vivo.

In the genomes of eukaryotes, tRNA genes occur almost exclusively as single transcription units (Sharp et al., 1985). Only five examples are known where tandemly arranged tRNA genes give rise to multimeric tRNA precursors. Two of these are in Saccharomyces cerevisiae and involve a tRNAArg- tRNAAsp gene pair (Schmidt et al., 1980). The others occur in Schizosmcharomyces pombe at three genetically defined suppressor loci (sups, sup9, and supl2) and comprise different tRNASer genes, each separated from a tRNA? gene by an identical 7-base pair spacer sequence (Hottinger et al., 1982; Willis et al., 1984; Mao et al., 1980). In all five cases, the synthesis of a dimeric tRNA precursor has been demonstrated using in vitro transcription systems from Xenopus and/or S.

* This work was supported by a grant from the National Institutes of Health. The costs of publication of this article were defrayed in part by the payment of page charges. This article must therefore be hereby marked “advertisement” in accordance with 18 U.S.C. Section 1734 solely to indicate this fact.

$ Present address: Institute of Cell and Tumor Biology, German Cancer Research Center, Heidelberg, Federal Republic of Germany.

§ Present address: Department of Molecular Biology, Princeton University, Princeton, NJ 08540.

ll To whom inquiries should be sent.

cerevisiae (Mao et al., 1980; Schmidt et al., 1980; Willis et al., 1984, Pearson et al., 1985). In addition, Northern analysis of RNA from an S. cerevisiae strain harboring the S. pombe sup9 tRNASe‘-tRNAMet gene pair has shown that dimeric transcripts are also synthesized in vivo (Willis et al., 1984).

In recent years considerable effort has been applied to the investigation of tRNA precursor processing in eukaryotic systems (for a recent review see Deutscher, 1984). These studies have employed both monomeric and dimeric tRNA precursors as substrates to examine (i) the number and nature of tRNA processing nucleases in different eukaryotic cells, (ii) the tRNA structural requirements of these enzymes, and (iii) whether tRNA precursor processing follows an ordered sequence of events. In the most studied systems (yeast, Xen- opus, and Drosophila) the relevant processing enzymes have been identified and in some cases purified (Garber and Alt- man, 1979; Kline et al., 1981; Zasloff et al., 1982; Peebles et al., 1983; Greer et al., 1983; Engelke et al., 1985; Pearson et al., 1985; Castano et al., 1985). The removal of 5’ leader sequences and the separation of tRNAs in a dimeric transcript appears to be carried out by an endonucleolytic activity similar to Escherichia coli RNase P (Koski et al., 1976; Akaboshi et al., 1980; Kline et al., 1981; Engelke et al., 1985; Pearson et al., 1985). Processing of 3’ trailer sequences on the other hand can be accomplished either by a 3 ’ 4 ’ exonuclease as found in yeast (Engelke et al., 1985; Pearson et al., 1985) or by an endonucleolytic 3’ pre-tRNase as in Drosophila (Frendewey et al., 1985) and Xenopus (Castano et al., 1985). Finally, the removal of intervening sequences from those tRNA precursors which contain them requires splicing endonuclease and ligase activities (Peebles et al., 1983; Greer et al., 1983). From these and other studies (Deutscher, 1984 and references therein) it is apparent that the processing of all the tRNA precursors in a given eukaryote is probably achieved by a relatively small number of nucleolytic activities. In accordance with this ex- pectation, studies on the processing of mutant tRNA precursors have shown that the endonucleases involved in tRNA maturation (including the splicing endonuclease) recognize overall tRNA conformation and conserved features of tRNA structure rather than specific sequences (Nishikura et al., 1982; Willis et al., 1984; Pearson et al., 1985). In the case of the intron-containing tRNA precursors from S. cerevisiae, this theme has been extended to include conserved structural features outside the mature tRNA domain (Swerdlow and Guthrie, 1984; Lee and Knapp, 1985). These precursors can be folded into secondary structures where part or all of the anticodon is base paired with complementary sequences in the intron (Ogden et al., 1984). The formation of such S t N C - tures is consistent with calculations of free energy minima and has been confirmed for six of the nine precursors (only

5878

In t ron S t ruc ture of tRNA Precursors Affects RNase P Cleavage 5879

six have been examined so far) by analyzing their conformation in solution with structure-specific probes (Swerdlow and Guthrie, 1984; Lee and Knapp, 1985). These structures and other conserved features of this region have, therefore, been proposed to represent recognition sites for the S. cerevisiae splicing endonuclease (Lee and Knapp, 1985). The view that eukaryotic tRNA precursor processing follows a preferred order of events is consistent with recent data obtained using monomeric tRNA precursors and extracts from human KB cells (Zasloff et al., 1982), Drosophila Kc cells (Frendewey et al., 1985), and Xenopus h u i s ovaries (Castano et al., 1985): 5'-end maturation precedes 3' processing. Equivalent data from yeast is outstanding, but the processing of dimeric tRNA precursors appears to represent an exception (Engelke et al., 1985; Pearson et al., 1985). In both yeast and Xenopus, however, it is known that intron removal follows 5'- and 3'-end maturation (see Deutscher, 1984).

In determining the sequence of tRNA-processing steps and analyzing the processing of mutant tRNA precursors, the use of coupled in vitro transcription and processing systems has been invaluable (see Deutscher, 1984). Using whole cell extracts derived from S. cereuisiae we have recently analyzed the effects of 22 point mutations in the S. pombe sup3 and sup9 genes on the processing of the resulting dimeric tRNA""- tRNA"" precursors (Willis et al., 1984; Pearson et al., 1985). For these studies the identification of tRNA processing de- fects has relied upon the knowledge of the primary structures of the parent sup3-e or sup9-e processing products. In this paper we present these data by way of explanation of the differences in the processing of sup9 wild-type and opal suppressor precursors. We show that a reduction in the efficiency of RNase P cleavage and the production of unusual intron cleavage products result from the inability of sup9-e precursors to form a stable anticodon/intron stem structure.

MATERIALS AND METHODS'

In Vitro Transcription and Processing-The genes encoding sup9' and sup9-e were cloned and sequenced as described by Willis et al. (1984). Whole cell extracts from S. cerevisiae were prepared by the method of Klekamp and Weil (1982) and were used for all transcription and processing experiments. Analytical transcription/processing reactions were carried out under the conditions described by Willis et a/. (1984) for 1 h a t 25 "C. The products were purified and resolved by electrophoresis on 6% polyacrylamide, 8.3 M urea gels.

Prirner-directed Mutagenesis-The methods described by Messing (1983) were used under the conditions of Stewart et ai. (1985) with a synthetic 15-mer to introduce the point mutation G37:10 into the intron of the sup9' and sup9-e genes. MI3 mp8 clones of sup9' and sup.9-e and DNA-sequencing procedures are described by Willis et al. (1984). The 1.5-kilobase HindIII/EcoRI fragment from M13 mp8 clones containing each sup9 allele was recloned into the centromeric plasmid vector YCp50 for in oitro transcription and transformation (Struhl et a/., 1979) into S. cereuisiae strain 3A84 (Schaack and Soll, 1985).

RESULTS

Reduced RNase P Cleavage and Unusual Endonucleolytic Cleavages in the Intron of sup9-e Precursors-Wild-type and opal suppressor alleles of the S. pombe sup9 serine tRNA differ by only a single base pair in the anticodon region (Willis

' Portions of this paper (including part of "Materials and Methods," part of "Results," Figs. 2-6, Tables 1 and 2, and Footnote 2) are presented in miniprint at the end of this paper. Miniprint is easily read with the aid of a standard magnifying glass. Full size photocopies are available from the Journal of Biological Chemistry, 9650 Rockville Pike, Rethesda, MD 20814. Request Document No. 85M-3469, cite the authors, and include a check or money order for $6.00 per set of photocopies. Full size photocopies are also included in the microfilm edition of the Journal that is available from Waverly Press.

9 1 9e

dlmeric tRNA PRE-I precursors [ ngBP 'PRE-2

- "b - a ] PRE-3

tRNASe' 3' half - holf molecules

tRNAse' 5'half -

FIG. 1. Transcription and process ing of sup9+ a n d sup9-e in an S. cereuisiae extract. The RNA products obtained were frac- tionated on a denaturing 6% polyacrylamide gel and visualized by autoradiography. sup9' tarnscription products were identified ini- tially by comparison with RNAs of known primary structure (derived from the supl2+ tRNAF"'-tRNAW'' gene (Mao et al., 1980). These assignments were confirmed by RNase T1 fingerprinting, nearest- neighbor analysis, and 5'-end group analysis (Fig. 5 and data not shown).

et al., 1984), yet upon transcription in an S. cerevisiae extract pronounced differences appear in the type, number, and abundance of the products (Fig. 1). For sup9-e, five major devia- tions from the parent s u p 9 transcription pattern may be distinguished: (i) only two dimeric tRNA precursors (PRE-1 and PRE-2) instead of the usual three, are detectable; (ii) two new products (PRE-3a and PRE-3b) appear; (iii) there is an increase in the amount of PRE-4; (iv) a reduction in the amount of tRNAQr+IVS3 species and mature tRNAQr; and (v) a change in the number, size, and amount of tRNAS"' half- molecules. To understand the nature of these differences we determined the primary structures of the major sup9' and sup9-e produ~ts .~ The results of this work are summarized in Fig. 7. Based on these data and the relative abundance of each species (see Fig. l), the altered processing of sup9-e precursors may be explained by effects on RNase P cleavage and splicing. Three features of the sup9-e processing pattern indicate that the removal of the 5' leader sequence from these precursors occurs less efficiently than for sup9+: (i) the absence of a third dimeric tRNA precursor (Le. one matured at the 5'-end of the tRNAQr as well as at the 3' end of tRNA"', see Fig. 1); (ii) the accumulation of the 5' leader-containing tRNA*r+IVS precursor (PRE-4) and the corresponding reduction in the amount of end-matured tRNA""+IVS product (Fig. 1); and (iii) the presence of tRNAsr 5' half-molecules which contain the 5' leader sequence (half-molecules c and d). In contrast the effect on splicing primarily concerns the appearance of unexpected intron cleavage products. Process-

The abbreviation used is: IVS, intervening sequence.

print at the end of this paper (Figs. 2-6 and Tables 1 and 2). ' The identification of the sup9 transcripts is presented in Mini-

5880 Intron Structure of tRNA Precursors Affects RNase P Cleavage

FIG. 7. Summary of the primary structures of the major sups-e products. The figure shows a secondary structure representation of the primary sup9-e transcript (PRE-1). Nucleotides are numbered from the 5’-end of the transcript. Arrows mark the boundaries between flanking sequences or the intervening sequence and nucleotides of the mature tRNAs. The anticodon in the tRNASe‘ half of the precursor is boxed. Nucleotides of the initial transcript which were deduced from the structural data to be present in the various sup9-e products are indicated below the dimeric precursor. Note that the products PRE- 2, PRE-3b, PRE-4, and tRNAi-a which have been 3’ processed may contain part or a11 of the sequence -CC&H.

Pre-1 2 I - I88

Pre - 2 : I - 181 Pre - 3 a : 55 - 188

ing of PRE-4 and the dimeric tRNA precursors PRE-1 and PRE-2 yields tRNASer 3’ half-molecules a and b and PRE-3a and b, respectively. The 5’ terminus of these four products corresponds to the G residue located three nucleotides 5‘ to the normal,3’ splice site (Fig. 7). The corresponding tRNASe’ 5’ half-molecules (c and d) have heterogeneous 3’ termini. This heterogeneity is presumably due to nibbling by the 3’- 5’ exonuclease in the extract and explains the slow but progressive shortening of these products which is seen during a pulse-chase (data not shown). The largest of these products, half-molecule c, has a 3‘ terminus located 1 nucleotide 3’ to the normal 5’ splice site. Thus the major sites of sup9-e intron cleavage differ both in their location and in the nature of the termini produced (see Table 2, Miniprint) from those made by the S. cerevisiae tRNA-splicing endonuclease (Peebles et al., 1983). Indeed, in vitro processing with partially purified yeast splicing endonuclease (courtesy of Dr. C. Greer) con- firms that other activities in the extract must be responsible for the appearance of these products.‘ Since neither 5’ nor 3’ half-molecules are detectable which could contain any of the remaining 11 nucleotides of intron sequence (Fig. l), it is likely that two endonucleolytic cleavages are required for their removal. Whether one or more activities are involved is not presently known. In addition to these unexpected intron cleavage reactions some normal splicing of sup9-e precursors presumably does occur since a sup9-e product is observed which co-migrates with mature tRNAS“ (Fig. l) , and sup9-e suppressor activity is expressed in vivo in S. cerevisiae.

A Structural Basis for the Altered Processing of sup9-e Precursors-Studies on mutant tRNA precursor processing have shown that reduced efficiencies of RNase P cleavage and splicing can be attributed in almost all cases to alterations in tRNA secondary or tertiary structure (Nishikura et al., 1982; Willis et al., 1984; Pearson et d., 1985). We, therefore, examined the anticodon/intron region of supP and sup9-e for possible conformational differences. sup9+, like all of the

I. Willis, unpublished results.

Pre - 3 b : 55 - 181 Pre - 4 : I - 102

t R N A f -a : 55 - 102

t R N A - c * I - 43 I

t R N A - d $ 1 -42

sup 9 +

\ / U - A U - A

0 - C A -u

C A kc u u

A - U

c u c u 0

sup 9 - e

\ / u “1 A -u U - A

0 ° C .

\ / U - A U - A A - U 0 - c

C U Q

FIG. 8. Potential anticodon/intron conformers of sup9+ and sup9-e. The anticodon stem and loop and intervening sequence of sup9+ and sup9-e precursors are shown schematically in different secondary structure conformations. Solid arrows mark the 5’ and 3’ boundaries of the intron. The open arrow marks the 5’ terminus of PRE-3a and b and tRNA” half-molecules a and b. The anticodons are boxed and the position of the G3T10 base substitution mutation is shown for both precursors.

intron-containing tRNA precursors from S. cerevisiae (Ogden et al., 1984), exhibits anticodon/intron complementarity which enables it to be folded as depicted in Fig. 8, with four potential Watson-Crick base pairs. The G+C change in the anticodon which generates the sup9-e suppressor results in the loss of one of these potential base pairs (Fig. 8). Free energy calculations based on the rules compiled by Ninio (1979) indicate that the stability of the base-paired sup9-e structure (AG = -1.5 kcal) is lower than an unpaired 22- nucleotide loop (AG = -1.7 kcal). For sup9+, however, the greater base pairing potential results in a significantly more stable structure (AG = -4.7 kcal) than either of the sup9-e conformers. Thus these data are consistent with the notion that the conformation in the anticodon/intron region of sup9+ and sup9-e precursors is different and suggest that the equilibrium between alternative sup9-e structures favors an unpaired anticodon/intron loop. The implication, therefore, is that sups-e precursors which are not in the base-paired con-

Intron Structure of tRNA Precursors Affects RNase P Cleavage 5881

figuration are less efficiently processed by RNase P and produce the unusual intron cleavage products (PRE-3a and b and tRNAser half-molecules a-d). To test this hypothesis we have used primer-directed mutagenesis to create the base substitution, G37:10, in the introns of sup9+ and sup9-e and thereby destroy and restore, respectively, the potential for the four base pair anticodon/intron stem (Fig. 8). The products obtained upon transcription of these genes in the S. cereuisiae extract are shown in Fig. 9. sup9-e G37:lO produces a transcription pattern which resembles that for sup9'. Compared to supg-e, the rate of RNase P cleavage of the repair mutant is greatly increased as indicated by the appearance of the third dimeric tRNA precursor, the reduction in the amount of 5' flanked tRNASe'+IVS precursor, and the increase in end-matured tRNASer+IVS precursor. In addition, PRE-3a, PRE-3b, and tRNAser half-molecules a-d, although still apparent, are present in significantly reduced amounts and there is an increase in the amount of mature tRNA"". The base pair disruption mutant, sup9' G37:10, reveals a tendency toward a sup9-e-type processing pattern although the changes are not as dramatic compared to the repair mutant (sup9-e G37:lO) and sup9-e (see "Discussion"). Nevertheless, the unusual intron cleavage products appear as predicted and the efficiency of RNase P cleavage is reduced.

Further evidence supporting the existence of alternative conformers comes from in vivo suppression data. In S. cereuisiae strain 3A84 (Schaack and Soll, 1985) the presence of sup9-e on a centromeric plasmid vector results in almost

9' 9+ 9 e 9e G G 37110 37110

dimeric tRNA precursors yq ' A%4

] P R E - 3 a a b

5' flanked tRNASe'+ IV tRNASe'+ IV

tRNASef -

tRNASCf3' half -

tRNASef5 'ho l f -

FIG. 9. Coupled transcription and processing of sups+ and sups-e containing the intron base substitution mutation G37:lO. Transcription and processing conditions are the same as those used in Fig. 1. The autoradiogram has been overexposed to show the amounts of PRE-3a and h and tRNA%' half-molecules a-d produced.

complete suppression of the adel opal mutation as assayed by the absence of the red pigment which, in the absence of the adel gene product, accumulates when adenine is limiting. When grown a t 30 "C these transformants are light pink. At 37 "C, however, the pink color is more intense implying some temperature sensitivity of sup9-e suppression. For the repair mutant, sup9-e G37:10, colonies are white a t both 30 and 37 "C. Thus, more suppressor tRNA must be made using the G37:lO mutant. This in turn must be due to the improved anticodon/intron stability since the G37:lO mutation is not present in the mature suppressor tRNA. The temperature sensitivity of sup9-e suppression is presumably due to the lower stability of the three base pair anticodon/intron stem compared to the unpaired 22-nucleotide loop (Fig. 8) as expected from free energy calculations (see above).

DISCUSSION

We have identified the primary dimeric transcript of the sup9-e tRNA""-tRNAM" gene and the major nucleolytic products which result from the processing of this precursor in an S. cereuisiae cell-free extract. From these data and a comparison of the processing patterns of the dimeric sup9+ and sup9- e transcripts, it is clear that sup9-e tRNASer precursors (either with or without tRNAM" attached) are less efficiently matured at their 5'-end than their wild-type counterparts and are susceptible to specific endonucleolytic cleavage(s) within the intron by unknown activities in the extract. In order to explain these effects we have proposed that the predominant sup9+ and sup9-e conformers differ ir; regard to their structure in the anticodon/intron region. Several independent lines of evidence are consistent with this conclusion. Calculations of free energy minima for the proposed structures (Fig. 8) as well as comparative data for intron-containing tRNA precursors from S. cereuisiae (Ogden et al., 1984) argue that sup9+ precursors adopt primarily a 4-base pair anticodon/intron stem structure. A comparison between this structure (Fig. 8) and that deduced for S. cereuisiae pre-tRNATyr from structure- probing studies (Swerdlow and Guthrie, 1984) is particularly interesting. Pre-tRNATvr has been shown to form a 3-base pair extended anticodon stem and hairpin comprising one G- C pair, two A-T pairs, and a 9-nucleotide loop. However, the low thermodynamic stability of this structure together with data obtained using different structure-specific probes suggest that a large unpaired anticodon/intron loop is also present as one of the conformers of pre-tRNAT"' (Swerdlow and Guthrie, 1984). Clearly, the 4-base pair stem and loop of sup9' precursors (which includes a terminal G-C base pair) is more stable than the corresponding pre-tRNAT"' structure. However, the base change which generates the sup9-e suppressor eliminates the terminal G-C base pair in this stem (Fig. 8) leaving a potential structure (comprising three A-T pairs) which has a lower stability than that for pre-tRNATvr (Ninio, 1979).

To test directly the hypothesis that an unpaired anticodon/ intron loop was responsible for the altered processing of sup9- e precursors we introduced the base substitution G37:lO to restore the potential for the 4-base pair anticodon/intron stem. The dramatic effect of the G37:lO mutation on sup9-e precursor processing in uitro and also on sup9-e suppressor activity in uiuo confirmed these expectations. Interestingly, the G37:lO mutation did not result in complete reversion to a sup9+-type processing pattern. Furthermore, the effect of the G37:lO mutation on the processing of sup9" precursors, although displaying the correct trend, was far less dramatic than expected (Fig. 9). These data indicate that the thermodynamic stabilities of the mutant structures are not equivalent to the wild-type or opal suppressor structures that they at-


tempt to mimic. Thus, the equilibrium between alternative structures must be different for each of the four precursors. In a comparison between sup9+ and sup9-e (23210 (Figs. 8 and 9) different base-stacking energies resulting from the terminal G-C or C-G base pair may account for the differences observed in in vitro processing. To explain the differences between sup9-e and sup9+ G37:10, we suggest that the latter may involve a nontraditional purine-purine interaction (G35 with G37:lO) or simply, purine base stacking.

These data have obvious implications for RNase P cleavage. We have previously shown that S. cereuisiase RNase P rec- ognizes conserved features of tRNA structure including the anticodon stem region (Pearson et al., 1985). We now add that the conformation in the anticodon/intron domain of tRNA precursors is also important in determining, in vitro and possibly in uiuo, the efficiency of RNase P catalysis. Note that the in vivo data reflects the overall efficiency of suppressor biosynthesis which, in a comparison of sup9-e and sup9-e G37:10, may comprise the individual contributions of RNase P cleavage and splicing. However, recent in vitro analysis of the splicing of sup3-e (which is homologous to sup9-e) and sup3+ (which is structurally analogous to sups-e G37:lO) precursors shows that these substrates are indistinguishable at this level? Thus, in this case, differential RNase P cleavage may be the major factor contributing to the different suppressor activity observed in uiuo. A possible explanation for the reduced efficiency of 5'-end maturation for substrates lacking a stable anticodon/intron stem structure may involve a steric hindrance of RNase P. Thus, we predict that if a stable anticodon/intron interaction cannot be formed, then the rate of RNase P cleavage will decrease above a critical loop size. The unpaired 22-nucleotide loop of sup9-e has clearly already exceeded this limit. Also noteworthy is that the relationship between anticodon/intron structure and RNase P cleavage provides a functional constraint in addition to splicing for the conservation of anticodon/intron complementarity among tRNA precursors in S. cereuisiae.

The identification of the sup9-e-processing products described here also serves to identify the corresponding products of the homologous sup3-e tRNASer-tRNAMet precursor (Pear- son et al., 1985). Surprisingly, a comparison of the processing of these dimeric precursors reveals different rates of 5'-end maturation. sup3-e precursors appear to be more efficiently processed by RNase P than their sup9-e counterparts. This is illustrated by the appearance of three dimeric sup3-e precursors (similar to sup9+) and by the relative amounts of 5' flanked and 5' matured tRNAS"'+IVS precursors (compare Fig. 1 to Pearson et al. (1985), Figs. 2 and 3). These different rates of processing must occur in response to the nature of the 5' leader sequence or to the base at the tip of the extra arm since these are the only differences between sup3-e and sup9-e tRNASe' precursors. Since the analysis of a number of point mutations in the extra arm of both sup3-e and sup9-e has shown that the conformation of this region is unimportant for RNase P cleavage we favor the former explanation. The availability of intergenic convertants in which the 5' flanking sequences of sup3 or sup9 have been transferred to the sup12 tRNA@&-tRNAMet gene or where the base at the tip of the extra arm has been switched will allow us to examine these possibilities (Amstutz et al., 1985; Heyer et al., 1986).

The data described here together with a time course experiments have been used to elaborate a preferred order of nucleolytic processing steps for the sup3-e tRNASer-tRNAMet precursor (Pearson et al., 1985). Similarly, a subcellular ex-

C. Greer and I. Willis, unpublished results.

tract from S. cereuisiae has been used to examine the processing of the tRNAArg-tRNAhp precursor (Engelke et al., 1985). In both studies the removal of the spacer sequence between the two tRNAs was shown to be carried out by a 3'+5' exonuclease activity. In the latter case that activity was partially purified. Evidence is now available which suggests that such a 3'45' exonuclease is responsible for spacer sequence removal in uiuo. This is based on the results of North- ern analyses of RNAs from S. cereuisiae strains transformed with different alleles of sup3-e or sup9-e. High resolution Northern blots reveal the same ladder-like pattern spacer sequence removal as is seen in ~ i t r o . ~ Furthermore, Northern analysis of 29 suppressor-inactive mutant alleles (Willis et al., 1984; Pearson et al., 1985; Heyer et al., 1986) has not produced a single case in which spacer sequence removal is impaired7 (Willis et al., 1984). We can find no evidence, however, that a 3'+5' exonuclease is responsible for the removal of sequences flanking the 3'-end of tRNAMet. In a time course of sup3-e precursor processing no exonucleolytic pattern of bands can be seen between the dimeric tRNA precursors (Pearson et al., 1985). Also, fingerprint analysis of PRE-1 and PRE-2 (Fig. 4) shows that these species do not have heterogeneous 3' termini as might be expected if an exonuclease were involved in their processing. The existence of an endonucleolytic 3' pre-tRNase in yeast has been alluded to previously (Frendewey et al., 1985). We have recently found that such an activity is present in the extracts used in this study? We, therefore, suggest that both an endonuclease and an exonuclease are involved in the 3'-end maturation of dimeric tRNA precursors in yeast.

Acknowledgments-We would like to thank Guido Krupp and Astrid Schon for numerous helpful discussions.

REFERENCES Akaboshi, E., Guerier-Takada, C., and Altman, S. (1980) Bwchem. Biophys.

Amstutz, H., Munz, P., Heyer, W. D., Leupold, U., and Kohli, J. (1985) Cell Res. Commun. 96,831-837

40.879-886 Cas&o, J. G., Tobian, J. A., and Zasloff, M. (1985) J. Biol. Chem. 260,9002-

Deutscher, M. P. (1984) Crit. Reu. Bwchem. 17,45-71 Engelke, D. R., Gegenheimer, P., and Abelson, J. (1985) J. Biol. Chem. 260,

9008

1271-1279 Frendewey, D., Dingermann, T., Cooley, L., and Soll, D. (1985) J. Biol. Chem.

~ ~ ~ ~~~ ~

2m. 449454 Garber, R. L., and Altman, S. (1979) Cell 17,389-397 Greer, C. L., Peebles, C. L., Gegenheimer, P., and Abelson, J. (1983) Cell 3 2 ,

537-546 Heyer, W. D., Munz, P., Amstutz, P., Aebi, R., Gysler, C., Schuchert, P.,

Szankasi, P., Leupold, U., Kohli, J., Gamulin, V., and Soll, D. (1986) J. Mol.

- - - , . " - - -

Hottinger, H., Pearson, D., Yamao, F., Gamulin, V., Cooley, L., Cooper, T., and Biol., in press

SOU. D. (1982) Mol. Gen. Genet. 188.219-224 KlekAp, M . S.; and Weil, P. A. (1982jJ. Bwl Chem. 257,8432-8441 Kline, L., Nishikawa, S., and SOU, D. (1981) J. Biol. Chem. 256,5058-5063 Koski R. A Bothwell A. L. M. and Altman S. (1976) Cell 9,101-116 Lee, M.-C., &d Knapi, G. (1988) J. Biol. Chhm. 260,3108-3115 Mao, J. Schmidt, O., and Soll, D. (1980) Cell 21,509-516 Messing, J. (1983) Methods Enzymol. 101,20-78 Ninio, J. (1979) Biochimie (Paris) 61,1133-1150 Nisbikura, K., Kujan, J., and Hall, B. D. (1982) EMBO J. 1,263-268 Ogden, R. C., Lee, M.-C., and Knapp, G. (1984) Nucleic Acids Res. 12,9367-

Pearson, D., Willis, I., Hottinger, H., Bell, J., Kumar, A., Leupold, U., and Sol],

Peebles C. L. Ge enheimer P and Abelson J. (1983) Cell 32,525-536 Schaaci, J., A d 8511, D. (1~85fNucleicAcids'Res. 13 , 2803-2814 Schmidt 0. Mao J.-I. Ogden R. Beckmann, J., Sakano, H., Abelson, J., and

Sharp, S., Scbaack, J., Cooley, L., Burke, D. J., and So$ D. (1985) Crit. Reu.

Stewart, T. S., Soll, D., andsharp, S. ('1985) Nucleic Acids Res. 13,435-447 Struhl, K., Stinchcomb, D., Scherer, S., and Davis, R. W. (1979) Proc. Nutl.

Willis, I., Hottinger, H., Pearson, D., Chisholm, V., Leupold, U., and Soll, D. Swerdlow, H., and Gutbrie, C. (1984) J. Biol. Chem. 259,5197-5207

Zasloff, M., Santos, T., Romeo, P., and Rosenberg, M. (1982) J. Bid. Chem.

9382

D. (1985) Mol. Cell. Bbl. 5.808-815

Soll, D. (i980) hatu;e 287,'75d-752

Biochem. 19,107-144

Acad. Sci. U. S. A. 76,1035-1039

(1984) EMBO J. 3,1573-1580

257 , 7857-7863

' V. Chisholm, unpublished results.

Intron Structure of tRNA Precursors Affects RNase P Cleavage 5883

I 2 3 4 5 6 7 8 9 1 0 1 1 1 2

P R E - I - P R E - 2 -

- -

P R E - 3 o t

P R E - 4 - tRNA*. I (

tRNAs"-

tRNAY'"

tRNA'" half molecules

PRE -1 f G )


PRE - 2 (U)

7 8

P R E - 4

( G )

Fig. 4: Fingernrint analysis of suo9-e 5 v i t r o transcripts. RNase TI flngerprints are shorn

each panel. The nucleotide used to labrl each R N A specles is shown in brackets. First and for PRE-1 , ?DE-2, PRE-3b and PRE-P-nqerprints are Identified in the upper tight corner of

second dimensions of the flngernrint are indicated in the bottom left comer of the figure. Numbered spots correspond to ollgonucleotider present in [m-32PlGTP-labeled ?RE-1 (Fig. 3 and Tahle 1 ) . NPV oligonucleotides are Identifird alphabetically (see Table 2).

GP -

PUP -

PCP -

origln -

1 2 3 4

rr- 1 2 3 4 5 6 7

PPAP -

u

Intron Structure of tRNA Precursors Affects RNase P Cleavage

tRNASC'

(GI

Ffg. 6: Fingerprint analysis of suP9+ transcripts and sup9-e tRNASer half molecules. The conventions used for the orientation and identification of each fingerprint and the oligonucleotides therein are the same as those used in Figs. 3 and 4.

5885

A single base change in the intron of a serine tRNA affects the rate of RNase P cleavage in vitro...

Documents

Transcript of A single base change in the intron of a serine tRNA affects the rate of RNase P cleavage in vitro...