THE JOURNAL OF BIOLOGICAL CHEMISTRY 0 1987 by The American Society for Biochemistry and Molecular Biology, Inc.

Vol. 262, No. 29, Issue of October 15, pp. 13953-13958,1987 Printed in U.S.A.

Modulation of Yeast 5 S rRNA Synthesis in Vitro by Ribosomal Protein YL3 A POSSIBLE REGULATORY LOOP*

(Received for publication, March 18, 1987)

David A. Brow$ and E. Peter Geiduschek From the Department of Biology, University of California at San Diego, La Jolla, California 92093

Synthesis of yeast 5 S rRNA in a homologous cell- free system has previously been shown to be inhibited by exogenous yeast 5 S rRNA. This inhibition is dimin- ished when 5 S rDNA is first incubated with the cell- free system, implying the presence of a transcription factor IIIA analog in yeast, and is dependent on the nucleotide sequence of the 5 S rRNA added. A two- dimensional gel electrophoresis method was used to demonstrate assembly of newly synthesized 5 S rRNA into a ribonucleoprotein complex containing ribosomal protein YL3. Immunodepletion of the cell-free system with anti-YL3 IgG was found to increase the sensitiv- ity of 5 S rRNA synthesis to inhibition by 5 S rRNA. We propose a model for the regulation of yeast 5 S rRNA synthesis in which YL3 competes with the pre- sumptive transcription factor IIIA analog for binding of 5 S rRNA.

5 S ribosomal RNA (rRNA) is a component of the large subunit of prokaryotic and eukaryotic ribosomes. Its function is unknown, but it is important for the in uitro activity of reconstituted ribosomes (Dohme and Nierhaus, 1976; Zagor- ska et al., 1984) and has a highly conserved secondary struc- ture (Erdmann and Wolters, 1986). Eukaryotic 5 S rRNA and a single ribosomal protein form a complex that can be removed from ribosomes with EDTA (Blobel, 1971). In the yeast Saccharomyces cereuisiae, this ribosomal protein is YL3 (Na- zar et ai., 1979), also known as Lla (Michel et al., 19831, and the ribonucleoprotein complex (RNP)’ is called the YL3 RNP.

Eukaryotic 5 S rRNA genes are transcribed by RNA polym- erase I11 in conjunction with a number of accessory factors (Weinmann and Roeder, 1974). In higher eukaryotes, these include transcription factors (TF) IIIB and IIIC, which are also required for transfer RNA gene transcription, as well as the 5 S rDNA-specific factor, TFIIIA (Segall et al., 1980; Engelke et al., 1980; Shastry et al., 1982). TFIIIA is unique in its ability to bind either the 5 S rRNA gene or 5 S rRNA itself (Pelham and Brown, 1980; Honda and Roeder, 1980).

* This research was supported by a grant from the National Insti- tute for General Medical Sciences. The costs of publication of this article were defrayed in part by the payment of page charges. This article must therefore be hereby marked “aduertisement” in accord- ance with 18 U.S.C. Section 1734 solely to indicate this fact.

3 Predoctoral trainee of a United States Public Health Service training grant in Cell and Molecular Biology. Present address: Dept. of Biochemistry and Biophysics, University of California, San Fran- cisco, CA 94143.

The abbreviations used are: RNP, ribonucleoprotein; TFIIIA, transcription factor IIIA; HEPES, 4-(2-hydroxyethyl)-l-piperazine- ethanesulfonic acid SDS, sodium dodecyl sulfate; PMSF, phenyl- methylsulfonyl fluoride.

Accordingly, 5 S rRNA (but not transfer RNA) gene tran- scription in uitro can be inhibited by addition of 5 S rRNA. Although there is an implication that such “feedback” inhi- bition should be important in 5 S rRNA gene regulation (Pelham, et al., 1981; Gruissem and Seifart, 1982), the details of this presumptive circuit have not yet been elucidated. A TFIIIA analog probably exists in yeast, since homologous in uitro transcription of a yeast 5 S rRNA gene is inhibited by the addition of yeast 5 S rRNA (Klekamp and Weil, 1982) and a yeast cell-free polymerase I11 system has been chro- matographed to yield a fraction required for 5 S rRNA, but not tRNA, gene transcription (Taylor and Segall, 1985).

We have utilized the cell-free system of Klekamp and Weil (1982) to explore the rolle of 5 S rRNA-binding proteins in the biosynthesis of yeast 5 S rRNA in uitro. We provide evidence that the inhibition caused by added 5 S rRNA is due to competition between RNA and DNA for a transcription factor and that the degree of inhibition is dependent on the nucleotide sequence of the 5 S rRNA. YL3 RNP containing newly synthesized 5 S rRNA is assembled in the cell-free system. The ability to form YL3 RNP is correlated with a decreased ability of 5 S rRNA to inhibit 5 S rDNA transcrip- tion, presumably due to competition of YL3 with the yeast TFIIIA analog for 5 S rRNA. This suggests a mechanism for coupling 5 S rRNA gene transcription to a step in ribosome assembly i n uiuo.

EXPERIMENTAL PROCEDURES~

RESULTS

Inhibition of Yeast 5 S rDNA Transcription by 5 S rRNA- It has previously been shown that addition of yeast 5 S rRNA to the crude yeast cell-free system decreases the accumulation of 5 S rDNA transcript (Klekamp and Weil, 1982). The phosphocellulose-purified cell-free system retains this prop- erty. A cloned yeast 5 S rRNA gene was transcribed i n uitro in the presence of varying amounts of yeast 5 S rRNA, added just before the addition of extract (see “Experimental Proce- dures”). The amount of transcript obtained at each RNA concentration was determined and compared to the average of two samples without added RNA. The amount of 5 S rDNA transcript obtained decreased 50% upon addition of 30 ng of yeast 5 S rRNA (see Fig. 2), which corresponds to a 150-fold

The “Experimental Procedures” 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, Bethesda, MD 20814. Request Document No. 87M-0843, cite the authors, and include a check or money order for $1.20 per set of photocopies. Full size photocopies are also included in the microfilm edition of the Journal that is available from Waverly Press.

13953

13954 Participation of YL3 in the Regulation of 5 S rRNA Synthesis

O O I

10 30 100 ng yeast 5s rRNA

FIG. 1. Effect of the order of addition of RNA and template on the inhibition of 5 S rDNA transcription by 5 S rRNA. The in vitro transcription reaction was divided into three 5-min intervals. During the first interval, the cell-free system was incubated with carrier DNA and either 5 fmol pB-1 (“DNA first, ” M) or the indicated amount of yeast 5 S rRNA (“RNA first,” e), or with both (“mixed start,” A). Then yeast 5 S rRNA and pB-1 were added so that all reaction mixtures had both components during the second interval. Ribonucleoside triphosphates were added to allow transcription dur- ing the third interval, and the amount of radioactivity in 5 S rRNA- sized product was determined as described under “Experimental Procedures.” Since the extract loses activity when incubated in the absence of template DNA, the control level of transcription varied DNA first = 5221 cpm, RNA first = 3075 cpm, mixed start = 4847 cpm.

molar excess of RNA over DNA. The effect was completely specific for 5 S rDNA, as 1 pg of 5 S rRNA, which reduced the yield of 5 S rDNA transcript 100-fold, had no effect on the yield of transcript from a cloned yeast tRNALe” gene. To confirm that the added 5 S rRNA was, in fact, inhibiting transcription, rather than influencing the stability of the newly synthesized 5 S rRNA, the order of addition of RNA and template was varied. The 10-min incubation in the ab- sence of nucleotides was split into two 5-min intervals, and the effect of first incubating 5 S rDNA with the cell-free system alone, then with added 5 S rRNA, was determined. The converse order of addition was also tested, and the standard protocol (RNA and template present throughout) was included for comparison. After incubation, nucleoside triphosphates were added and transcription was carried out for only 5 min to minimize the opportunity for factors to reassort between RNA and DNA.

The results of such an experiment are shown in Fig. 1. Preincubation of 5 S rDNA with the cell-free system for 5 min protected 5 S rDNA transcription against inhibition by 5 S rRNA. This is consistent with the known ability of 5 S rDNA to bind TFIIIA, sequestering it in the stable transcrip- tion complex (Bogenhagen et al., 1982; Segall, 1986). The protection was not complete, however, at least partly because 5 min was insufficient time to assemble stable complexes on all the template molecule^.^ In addition, the assembled tran- scription complexes might not be completely stable under these reaction conditions, so that some reassortment of factor could occur in the 5 min before initiation of transcription. Be that as it may, the fact that any protection at all was seen is inconsistent with a post-transcriptional mechanism for the effect of exogenous 5 S rRNA. Adding 5 S rRNA first, instead of together with 5 S rDNA, did not increase inhibition. This suggests that 5 S rRNA did not stably bind TFIIIA under these conditions, so that no advantage was conferred by allowing it access to the factor before addition of 5 S rDNA.

To assess the sequence-specificity of the presumptive inter- action between 5 S rRNA and the TFIIIA analog, 5 S rRNA from three other organisms was also tested for the ability to

G. A. Kassavetis and B. Braun, unpublished results.

inhibit transcription of yeast 5 S rDNA. Human 5 S rRNA (from HeLa cells) and Xenopus laeuis oocyte (Xlo) 5 S rRNA are 60 and 56% identical, respectively, to yeast 5 S rRNA in nucleotide sequence and are 89% identical to each other (Erdmann and Wolters, 1986). Both can form secondary (and, presumably, tertiary) structures that closely resemble the structure proposed for yeast 5 S rRNA (Kjems et al., 1985). Escherichia coli 5 S rRNA is about 50% identical to each of the three eukaryotic RNAs and has a slightly different sec- ondary structure (Stahl et al., 1981). Fig. 2 shows the results obtained when yeast 5 S rDNA transcription was carried out in the presence of these RNAs. The heterologous 5 S rRNAs inhibited transcription to different degrees, each being less effective than yeast 5 S rRNA. HeLa 5 S rRNA was about 6- fold less strong a competitor, whereas Xlo 5 S rRNA was 6 to 7-fold weaker than HeLa RNA. The difference between these two RNAs is striking in view of the similarity of nucleotide sequences. E. coli 5 S rRNA was virtually ineffectual as a competitor of transcription. The effect of 1 pg of each RNA on tDNAL”” transcription was also tested. In no case did the amount of tRNAL”” produced differ by more than 5% from the no-RNA control.

One could argue that the difference in inhibitory activities of the RNAs is due to preferential degradation of the heter- ologous RNAs in the cell-free system, reducing their effective concentration. This possibility was tested by transcribing 5 S rDNA in the presence of 300 ng of each 5 S rRNA, resolving the products on a denaturing gel, and examining the remain- ing RNA by staining the gel with ethidium bromide. A pho- tographic negative of the gel was subjected to densitometry to determine RNA concentration. The added yeast 5 S rRNA was found to be completely stable, while only 100 ng of each heterologous RNA remained at the end of the 40-min incu- bation. Thus, preferential degradation could conceivably con- tribute to the difference between homologous and heterolo- gous RNAs, but not to differences among heterologous RNAs.

Newly Synthesized 5 S rRNA Binds Ribosomal Protein YL3 in Vitro-We examined binding of newly synthesized 5 S rRNA to proteins of the cell-free system. To visualize such an interaction, a two-dimensional polyacrylamide gel electro- phoresis procedure, which utilizes a nondenaturing first dl-

ng 5s rRNA/reactlon

FIG. 2. Effect of homologous and heterologous 5 S rRNA on yeast 6 S rDNA transcription. Transcription of pB-1 was carried out as described under “Experimental Procedures,” but with the addition of the indicated amounts of 5 S rRNA from the following organisms before adding the cell-free system: S. cerevisiae (e), HeLa cell (O), X . laeuis oocyte (B), E. coli (0). After resolving the products on polyacrylamide gels, the percentage yield of 5 S rDNA transcript at each RNA concentration was determined by measuring the radio- activity in the 5 S rRNA-sized band, subtracting the background found in the same region of a lane corresponding to a reaction with only vector DNA as template and normalizing to the average of two duplicate reactions with no added RNA. Each data point represents the average of two independent experiments; the average difference between duplicate determinations was 3.4% of control transcription.

Participation of YL3 in the Regulation of 5 S rRNA Synthesis 13955

mension and a denaturing second dimension, was developed. Stable interaction of a protein with 5 S rRNA slows its migration relative to free 5 S rRNA in the first dimension. The constituents of the ribonucleoprotein complex can then be resolved in the denaturing second dimension. Use of a discontinuous buffer system for the first dimension resulted in sharp bands despite the relatively high salt and protein concentrations of the sample. We refer to these as “RNP gels”.

When transcription in a standard reaction mixture with 5 S rDNA as template was stopped by adding EDTA and cooling to 0 “C, and the sample was then loaded directly onto such a gel, the pattern shown in Fig. 3A was obtained. Approximately 20% of the labeled 5 S rRNA was contained in a band migrating 0.7 times as fast as free 5 S rRNA (a small amount of which was present) in the nondenaturing dimension. The remainder of the 5 S rRNA was unresolved at the top of the gel. Protein co-migrating with the 0.7 RF RNA could not be detected by staining with Coomassie Blue or Silver. However, when the same experiment was done with the whole-cell extract that had not been purified over phosphocellulose, which contained a large amount of endogenous 5 S rRNA, Coomassie staining of the second dimension gel revealed a

m c

Y L W

FIG. 3. Two-dimensional electrophoresis of 5 S rDNA tran- scription products and of purified YL3 RNP. Transcription of pB-1 was stopped with EDTA and the reaction mixture was loaded onto a nondenaturing, discontinuous gel. An aliquot of YL3 RNP purified from ribosomes was loaded into an adjacent lane. After electrophoresis, lanes were cut out of the separating gel and electro- phoresed through a second, denaturing dimension. A, autoradiograph of the entire two-dimensional gel of pB-1 transcription products; the original first dimension gel slice has been replaced by a duplicate not subjected to a second dimension of electrophoresis. B, ethidium bromide stain and C, Coomassie Blue stain of the separating portion of the second dimension gel containing purified YL3 RNP. The positions of YL3 RNP, 5 S rRNA, and ribosomal protein YL3 are indicated.

36-kDa protein aligned with the 0.7 RF band (data not shown). Ethidium bromide staining showed that most of the endoge- nous 5 S rRNA was migrating as this slow form, rather than as free RNA.

These observations suggested that the protein binding newly synthesized 5 S rRNA might be the ribosomal protein YL3 (also known as Lla; Michel et al., 1983), which is abun- dant and has an apparent molecular weight of 36,000. To test this possibility we purified YL3 RNP from ribosomes (see “Experimental Procedures”) and subjected it to electropho- resis in an adjacent lane of the nondenaturing gel. After running the second dimension, the gel was stained for RNA with ethidium bromide (Fig. 3B), then for protein with Coo- massie Blue (Fig. 3C). As can be seen from the positions of RNA and protein, the YL3 RNP was aligned with the 0.7 RF band containing the labeled transcript. Partial dissociation of the RNP generated free 5 S rRNA which served as a marker.

To further explore RNP formation, we raised polyclonal antibody to purified YL3 RNP. Double immunodiffusion in- dicated that the antiserum precipitated YL3 RNP and YL3 protein, but not 5 S rRNA. Western analysis also revealed binding of antibody to YL3 but not to 5 S rRNA (data not shown). Thus it is likely that the antibody recognized epitopes on the protein component of the RNP. Immune IgG was used to immunoprecipitate transcripts from a reaction mixture in which transcription was terminated as for RNP gels. RNA remaining in the supernatant and extracted from the pellet was analyzed by electrophoresis on denaturing gels, and the percentage of 5 S rRNA remaining in the supernatant at different immune IgG concentrations was determined quan- titatively.

In three separate experiments, 53,56, and 60% of the newly transcribed 5 S rRNA was precipitated at saturating IgG concentrations (see Fig. 6 of Brow, 1987). These are probably underestimates of the fraction of 5 S rRNA bound to YL3 because of the opportunity for dissociation of precipitated RNP during washes of the pellet, which were subsequently pooled with the supernatant. In one experiment where the washes were not added in, 73% of the 5 S rRNA was removed from the supernatant. When the 5 S rDNA transcripts were first extracted with phenol, precipitated with ethanol, and resuspended in transcription buffer, no 5 S rRNA was specif- ically precipitated by anti-YL3 RNP IgG. Thus more than half of the newly synthesized repeat 5 S rRNA is bound to YL3.

YL3 Modulates the Inhibition of 5 S rDNA Transcription by 5 S rRNA in Vitro-To further investigate the role of YL3 in 5 S rRNA biosynthesis, the transcription extract was depleted of YL3 by incubation with anti-YL3 RNP IgG and passage through a column of protein A-Sepharose to remove immune complexes. The flow-through was used to transcribe 5 S rDNA and the products were analyzed on a denaturing gel. No difference was seen in the amount of 5 S rRNA produced (see below) or in the efficiency with which it was processed when comparing preimmune IgG-treated and im- mune IgG-treated extracts. However, when the reaction prod- ucts were displayed on nondenaturing gels, the preimmune IgG-treated extract (Fig. 4, lane I ) gave rise to the normal pattern, whereas the immune IgG-treated extract (lane 2 ) showed no YL3 RNP band. (The fact that there is less radioactive material in the whole of lane 2 compared to lane 1 does not mean that less transcript was made; rather, the transcript was most probably in a form that remained in the stacking gel and thus did not appear in the autoradiogram.) The absence of detectable YL3 RNP suggests that the im- mune IgG-treated extract was successfully depleted of YL3.

13956 Participation of YL3 in the Regulation of 5 S rRNA Synthesis

pre imm IgG IgG

I L FIG. 4. Products of in vitro transcription of 5 S rDNA with

IgG-treated extracts displayed on a nondenaturing gel. Cell- free system treated with preimmune ( l a n e 1 ) or anti-YL3 RNP ( l a n e 2) IgG as described under “Experimental Procedures” was used to transcribe pB-1 in uitro. Transcription products were subjected to electrophoresis on a nondenaturing gel equivalent to the first dimen- sion of Fig. 3A. The position of YL3 RNP and free 5 S rRNA are indicated.

The fact that the depleted extract seemed unable to assem- ble newly synthesized 5 S rRNA into YL3 RNP was of interest because of its implications for the regulation of 5 S rDNA transcription. We reasoned that if binding of YL3 and the putative yeast TFIIIA (see above) to 5 S rRNA were mutually exclusive events, then free YL3 present in the extract would tend to relieve the inhibition by titrating out added 5 S rRNA. If this were the case, one would expect the YL3-depleted extract to be more sensitive to inhibition than the preimmune IgG-treated extract. The results of an experiment testing this hypothesis are shown in Fig. 5. Different amounts of yeast 5 S rRNA were added at the beginning of the incubation of 5 S rDNA with preimmune or anti-YL3 RNP IgG-treated extract (the same preparations used in the experiment shown in Fig. 4 , transcription was carried out as usual, and the amount of radioactivity incorporated into 5 S rRNA was determined. As noted above, in the absence of added 5 S rRNA, the amount of transcript produced by the YL3-depleted extract was the same as for the preimmune IgG-treated extract. However, upon adding increasing amounts of 5 S rRNA, 5 S rDNA transcription with the YL3-depleted extract was found to be more readily inhibited than with the preimmune IgG-treated extract. Half-maximal inhibition of 5 S rRNA production required 6.5 times less added RNA in the YL3-depleted sys- tem. This result indicates a competition between YL3 and the putative yeast TFIIIA for binding of 5 S rRNA and demon-

3000 1

“0 ” 3 IO 30 100 300

ng 5s rRNA/reaction

FIG. 5. Effect of depletion of YL3 on the inhibition of 5 S rDNA transcription by 5 S rRNA. IgG-treated extracts (see Fig. 4) were used for transcription of pB-1 in the presence of different amounts of yeast 5 S rRNA. The yield of 5 S rDNA transcript was determined as described under “Experimental Procedures”. +, preim- mune IgG-treated extract; 0, anti-YL3 RNP IgG-treated extract.

strates that transcription of 5 S rDNA in uitro is coupled to a step in ribosome assembly.

DISCUSSION

We have demonstrated the formation of one 5 S rRNA- containing RNP during synthesis of 5 S rRNA in uitro and infer the formation of a second. The presence of one of these RNPs, containing the putative yeast TFIIIA analog, was only assessed indirectly, by the inhibition of 5 S rDNA transcrip- tion upon addition of 5 S rRNA. The fact that this inhibition was diminished by the preincubation of 5 S rDNA with a source of transcription factors (Fig. 1) supports the hypothesis that it is caused by competition between 5 S rRNA and 5 S rDNA for a TFIIIA-like factor (Klekamp and Weil, 1982). We used the transcription inhibition assay to determine the RNA sequence requirements for assembly of this putative RNP. The degree of inhibition observed was found to be dependent on the sequence of the added 5 S rRNA, with S. cereuisiae 5 S rRNA being most inhibitory. HeLa and Xenopus 5 S rRNAs, which differ at only 13 of 120 positions, exhibited a 6 to 7-fold difference in inhibitory activity.

This result contrasts with observations on a Xenopus cell- free transcription system. Pieler et al. (1984) have shown that five heterologous 5 S rRNAs (including those from S. cereuis- iae and E. coli) exhibit inhibitory activity similar to that of the homologous ( X . laeuis) 5 S rRNA. The equivalence of homologous and heterologous eukaryotic 5 S rRNAs in com- peting with 5 S rDNA for binding of Xenopus TFIIIA was also shown by DNase I footprint assays (Hanas et al., 1984), although, in these studies, E. coli 5 S rRNA did not compete well. The greater sequence specificity of the yeast TFIIIA analog may extend to its DNA-binding activity as well. Nei- ther a X . laeuis oocyte 5 S rRNA gene nor a Syrian hamster 5 S rRNA gene are transcribed by the yeast cell-free system (data not shown), yet both Xenopus and HeLa cell-free sys- tems efficiently transcribe the S. cereuisiae 5 S rRNA gene (Drabkin and RajBhandary, 1985): Although it is not known whether TFIIIA has a role in the greater selectivity of the yeast system (which also applies to certain tRNA genes), the possibility exists that the stringent sequence dependence dis- played by the yeast factor in binding RNA also applies to DNA binding.

We were able to detect the formation of another RNP, containing 5 S rRNA and ribosomal protein YL3, in the cell-

‘s. Fuhrman, personal communication.

Participation of YL3 in the Regulation of 5 S rRNA Synthesis 13957

free system by two-dimensional gel electrophoresis. Because most of the newly synthesized 5 S rRNA remained unresolved in the RNP gels, it was not possible to derive quantitative information from this analysis, but immunoprecipitation with anti-YL3 RNP antibody provided an estimate of the fraction of 5 S rRNA associated with this protein (>50%).

The YL3-depleted cell-free system exhibited no YL3 RNP (Fig. 4), but synthesized 5 S rRNA in amount and form indistinguishable from that made by untreated extracts. This suggests that YL3 is not directly involved in transcription of 5 S rDNA and is not required for the short-term stability of newly transcribed 5 S rRNA in uitro. Furthermore, binding of the ribosomal protein must not be required for processing of pre-5 S rRNA, in contrast to such processing in Bacillus subtilis (Stahl et al., 1984). However, the inhibition of 5 S rDNA transcription brought about by adding yeast 5 S rRNA was found to be more severe for the YL3-depleted extract. This observation implies that the binding of YL3 and TFIIIA to 5 S rRNA is mutually exclusive and suggests a model for the regulation of 5 S rRNA gene expression in uiuo. One can imagine that a nucleolar pool of free YL3 is available to bind to newly synthesized 5 S rRNA. The resulting YL3 RNP assembles into preribosomal particles and is transported to the cytoplasm. When free YL3 is limiting, however, 5 S rRNA accumulates and is available for binding to TFIIIA in com- petition with 5 S rDNA and blocking transcription of 5 S rDNA. This hypothetical regulatory circuit is diagrammed in Fig. 6. Simple, and perhaps self-evident as the model is, it nevertheless forces one to confront established ideas about the stability of transcription complexes (reviewed by Lassar et al., 1983; Brown, 1984). The stability of binding of the transcription factor assembly to the internal control region of a 5 S gene sets the time scale for regulation of transcription by 5 S rRNA: a reasonably responsive feedback loop implies a relatively short average lifetime for repetitively initiating transcription complexes. The stability of yeast transcription factor-DNA complexes in vivo is unknown. However, stability in vitro, for tRNA gene transcription by a yeast cell-free system, can be greatly varied by adjusting electrolyte concen- tration (Stillman et al., 1985).

Because purified YL3 protein inhibited transcription non- specifically, i.e. inhibited tDNAL"" as well as 5 S rDNA tran- scription, we were unable to rescue 5 S rRNA synthesis in depleted extracts by its addition. In any case, it has previously been determined that the purified protein is unable to reform an RNP with 5 S rRNA (Nazar and Wildeman, 1983). How- ever, the effect displayed in Fig. 5 is almost certain to be due to lack of YL3 because the antibody used was raised against

7 s RNP YL3 RNP

I ....

5s rDNA Initiation/ Complex

TFIIIB,C FIG. 6. Proposed regulatory role of ribosomal protein YL3

in 5 S rRNA gene transcription. YL3 may facilitate the formation of a stable transcription initiation complex on 5 S rDNA by binding 5 S rRNA to form the YL3 RNP. This would prevent formation of the 7 S RNP, thereby allowing TFIIIA to bind to 5 S rDNA.

apparently homogeneous YL3 RNP. Cross-reactivity of the antibody with TFIIIA seems most unlikely since no effect on the level of 5 S rDNA transcription was seen in the absence of 5 S rRNA.

YL3 is one of the few yeast ribosomal proteins whose synthesis is not coordinately regulated (Gorenstein and War- ner, 1976). In HeLa cells, the YL3 analog L5 is unique among the ribosomal proteins in having a large free pool size (Phillips and McConkey, 1976). Further studies will be required to determine how YL3 synthesis is regulated in yeast.

Our conclusion that YL3 excludes TFIIIA from binding to 5 S rRNA is consistent with the recent finding that rat ribosomal protein L5 and Xenopus TFIIIA protect overlap- ping regions of 5 S rRNA from nuclease (Huber and Wool, 1986a; Huber and Wool, 1986b). We have not yet identified an RNP that might correspond to the yeast TFIIIAo5S rRNA RNP. The possibility that its presence among the products of in uitro transcription is obscured by co-migration with the YL3 RNP is ruled out by the fact that an extract depleted for YL3, which is expected to form more of the TFIIIA RNP, displayed no RNP in the 0.7 RI. region of the native gel (Fig. 4, lane 2 ) . Instead, much of the 5 S rRNA apparently remained in the stacking gel of the native dimension. Further purifica- tion of the yeast TFIIIA analog would allow a more detailed analysis of its interaction with yeast 5 S rRNA. The results displayed in Fig. 2 suggest that yeast 5 S rRNA-affinity chromatography, in the presence of excess E. coli 5 S rRNA, might be a useful step in such a purification.

Acknowledgments-We are indebted to B. Braun, G. Kassavetis, and R. Negri for helpful discussions and comments on our writing. We also thank G. Anders for assistance in raising antisera.

REFERENCES Blobel, G. (1971) Proc. Natl. Acad. Sci. U. S. A. 68, 188-1885 Bogenhagen, D. F., Wormington, W. M., and Brown, D. D. (1982)

Brow, D. A. (1987) J. Biol. Chem. 262, 13959-13965 Brown, D. D. (1984) Cell 37,359-365 Dohme, F., and Nierhaus, K. H. (1976) Proc. Natl. Acad. Sci. U. S.

Drabkin, H. J., and RajBhandary, U. L. (1985) J. Biol. Chern. 260,

Engelke, D. R., Ng, S.-Y., Shastry, B. S., and Roeder, R. G. (1980)

Erdmann, V. A., and Wolters, J. (1986) Nucleic Acids Res. 14, (suppl.)

Gorenstein, C., and Warner, J. R. (1976) Proc. Natl. Acad. Sci. U. S.

Gruissem, W., and Seifart, K. H. (1982) J. Bid. Chem. 267, 1468-

Hanas, J. S., Bogenhagen, D. F., and Wu, C.-W. (1984) Nucleic Acids

Hendrick, J. L., and Smith, A. J. (1968) Arch. Biochem. Biophys.

Honda, B. M., and Roeder, R. G. (1980) Cell 22, 119-126 Huber, P. W., and Wool, I. G. (1986a) J. BiohChem. 261,3002-3005 Huber, P. W., and Wool, 1. G . (1986b) Proc. Natl. Acad. Sci. U. S. A.

Kassavetis, G. A., Elliott, T., Rabussay, D. P., and Geiduschek, E. P.

Kjems, J., Olesen, S. O., and Garrett, R. A. (1985) Biochemistry 24,

Klekamp, M. S., and Weil, P. A. (1982) J. Biol. Chern. 267, 8432-

Lassar, A. B., Martin, P. L., and Roeder, R. G . (1983) Science 222,

Michel, S., Traut, R. R., and Lee, J. C. (1983) Mol. Gen. Genet. 191,

Nazar, R. N., Yaguchi, M., Willick, G. E., Rollin, C. F., and Roy, C.

Cell 28, 413-421

A. 73, 2221-2225

5596-5602

Cell 19, 717-728

rl-r59

A. 73,1547-1551

1472

Res. 12, 2745-2758

126, 155-164

83, 1593-1597

(1983) Cell 33, 887-897

241-250

8441

740-748

251-256

(1979) Eur. J. Biochem. 102, 573-582

13958 Participation of YL3 in the Regulation of 5 S rRNA Synthesis

Nazar, R. N., and Wildeman, A. G. (1983) Nucleic Acids Res. 11 ,

Pelham, H. R. B., and Brown, D. D. (1980) Proc. Natl. Acad. Sci. U.

Pelham, H. R. B., Wormington, W. M., and Brown, D. D. (1981)

Philips, W. F., and McConkey, E. H. (1976) J. Biol. Chem. 251,

Picard, B., and Wegnez, M. (1979) Proc. Natl. Acad. Sci. U. S. A. 76,

Pieler, T., Erdmann, V. A., and Appel, B. (1984) Nucleic Acids Res.

Rubin, G. M. (1975) Methods Cell Bwl. 12,45-64 Segall, J. (1986) J . Biol. Chem 261, 11578-11584 Segall, J., Matsui, T., and Roeder, R. G. (1980) J. Biol. Chem. 2 5 5 ,

3155-3168

S. A. 77, 4170-4174

Proc. Natl. Acad. Sci. U. S. A. 78,4170-4174

2876-2881

241-245

12,8393-8406

11986-11991

Shastry, B. S., Ng, S.-Y., and Roeder, R. G. (1982) J. Bwl. Chem.

Sigel, M. B., Sinha, Y. N., and VanderLaan, W. P. (1983) Methods

Stahl, D. A., Leuhrsen, K. R., Woese, C. R., and Pace, N. R. (1981)

Stahl, D. A., Pace, B., Marsh, T., and Pace, N. R. (1984) J. Biol.

Stillman, D. J., Caspers, P., and Geiduschek, E. P. (1985) Cell 40,

Taylor, M. J., and Segall, J. (1985) J. Biol. Chem 260,4531-4540 Weinmann, R., and Roeder, R. G. (1974) Proc. Natl. Acad. Sci. U. S.

Yaguchi, M., Rollin, C. F., Roy, C., and Nazar, R. N. (1984) Eur. J.

Zagorska, L., Van Duin, J., Noller, H. F., Pace, B., Johnson, K. D.,

257,12979-12986

Enzymol. 93, 3-12

Nucleic Acids Res. 9,6129-6137

Chem. 259, 11448-11453

311-317

A. 7 1 , 1790-1794

Biochem. 139,451-457

and Pace, N. R. (1984) J. Biol. Chem 259,2798-2802