Get 20M+ Full-Text Papers For Less Than $1.50/day. Start a 7-Day Trial for You or Your Team.

Learn More →

Function of the mammalian La protein: evidence for its action in transcription termination by RNA polymerase III.

Function of the mammalian La protein: evidence for its action in transcription termination by RNA... The EMBO Journal vol.8 no.3 pp.851 -861, 1989 of the mammalian La protein: evidence for its Function action in transcription termination by RNA polymerase Ill was recently resolved into two complementing components Ellen Gottlieb1 and Joan A.Steitz (Yoshinaga et al., 1987). An additional 5S-specific factor, Department of Molecular Biophysics and Biochemnistry, Howard TFIIIA (Engelke et al., 1980; Honda and Roeder, 1980; Hughes Medical Institute, Yale University School of Medicine, 333 Pelham and Brown, 1980) has been purified to homogeneity Cedar Street, PO Box 3333, New Haven, CT 06510, USA et al., 1980) and its gene cloned (Ginsberg et al., (Engelke 'Present address: Medical Research Council, Laboratory of Molecular 1984). Prior to initiation, TFIIIA binds to the 5S gene ICR; Biology, Hills Road, Cambridge CB2 2QH, UK TFIIIC component associates with the ICR of both tRNA Communicated by A.A.Travers VA (Engelke et al., 1980; Pelham and Brown, and genes We have tested the hypothesis that the mammalian La 1980; Sakonju et al., 1981; Klemenz et al., 1982; Lassar which appears to be required for accurate and Fuhrman et al., 1984; Stillman and protein, et al., 1983; efficient RNA polymerase III transcription, is a tran- 1984; Stillman et al., 1984; Camier et al., Geiduschek, scription termination factor. Our data suggest that 3' 1985; Carey et al., 1986; Yoshinaga et al., 1987). foreshortened transcripts generated in La's absence are Remaining general factors are believed to associate primarily of a novel transcription intermediate interactions (Lassar et al., 1983; components through protein-protein a These transcripts are Setzer and Brown, 1985; Carey et al., 1986; Klekamp et al., containing paused polymerase. by fractionated transcription complexes, are Wingender et al., 1986). The resulting transcription produced 1986; with kinetics different from full-length remain stable through multiple rounds of tran- synthesized complexes and are chasable to completion from the et al., 1982; Lassar et al., 1983; transcripts, scription (Bogenhagen stalled transcription complexes. Together, these fmdings Schaack et al., 1983). that termination by RNA polymerase HI requires less is known about RNA polymerase III argue Considerably factor(s) and implicate La as such a factor. the process including cessation of elongation, auxilliary termination, La transcript completion and and polymerase dissociation. Four or more Since appears to facilitate transcript release and also binds the RNA product, it may residues embedded in a G-C rich environment release thymidine resulting be a of RNA polymerase III transcription. a termination signal (Bogenhagen and regulator constitute transcription La protein/RNA polymerase Ill/transcription Brown, 1981). Purified Xenopus RNA polymerase Il has Key words: termination shown to be capable of recognizing this simple factor/transcription regulation/transcription been in the absence of auxiliary factors consensus sequence et 1983) and similar properties have been (Cozzarelli al., to calf thymus polymerase 11 (Watson et al., 1984). ascribed Introduction While auxiliary factors facilitating the last two aspects of III responsible for the synthesis of most have not been identified, several observations RNA polymerase is the process yet [see Geiduschek and that exist. First, the simple termination small RNAs in eukaryotic cells suggest they may for a review]. Cellular species its 3' flanking sequences are protected by Tocchini-Valentini (1988) signal and/or 4.5S and Alu transcripts; in transcription factor fractions (Klemenz include 5S rRNA, all tRNAs, 7S-L, components present EBERI and EBER2 are specified by class Ill Camier et 1985; Van Dyke and Roeder, VAI, VAII, et al., 1982; al., carried on viral Immediately following and some 3' deletion mutants compete poorly for the genes genomes. 1987) all RNA III transcripts are packaged (Schaack et al., 1983; Sharp et al., synthesis, polymerase transcription apparatus Wilson et Wu complexes with the 50-kd La protein Allison and Hall, 1985; al., 1985; into RNA-protein 1983; consensus and Mathews, 1982; Rinke and Steitz, 1982; et 1987). Second, seemingly equivalent (Francoeur al., in various environ- Steitz et behave differently sequence al., 1983). sequences et Koski et al., 1982; Initiation of III transcription is directed by ments (DeFranco al., 1982; polymerase and there are hints that terminators sequences within the coding region termed Mazabraud et 1987) multipartite al., et al., et al., 1980; exhibit tissue-specific expression (Mazabraud internal control regions (ICR) (Bogenhagen may while its can be influenced 1987). et 1980), efficiency Sakonju al., and 1989), reviewed in In the accompanying paper (Gottlieb Steitz, by flanking sequences (comprehensively the mammalian La protein may be an RNA and 1988; for apparent we suggested that Geiduschek Tocchini-Valentini, termination factor. This 50-kd see Carbon et 1987; Murphy et al., 1987; III transcription exceptions al., polymerase Steitz et and 1982; al., fractionation, initially phosphoprotein (Francoeur Mathews, Das et al., 1988). Chromatographic Pizer et Francoeur revealed that RNA et al., 1983; HeLa and extracts, 1983; Habets al., 1983; of Xenopus associates with all is of accurate initiation: a Hoch and 1984) III alone incapable et al., 1984; Billings, polymerase III the factor fractions and transcripts, selectively binding minimum of two additional (TFIIIB nascent polymerase stretch encoded the termination for of all class III 3' by signal are transcription genes uridylate TFIIIC) required Its Mathews and et These fractions Francoeur, 1984). presence et al., 1982). (Stefano, 1984; (Segall al., 1980; Shastry and in both in active complexes factors et can be detected transcription contain multiple general (Lassar al., 1983; may TFIIIC and (Gottlieb as evidenced the fact that TFIIIC crude and highly purified preparations Bieker et 1985) by al., ©IRL Press E.Gottlieb and J.A.Steitz ts ........ 1. A model for La action. Schematic representation of a stable transcription complex containing TFIIIB, TFIIIC and a class III gene (top), Fig. can be recognized by RNA polymerase III (middle). (If this were a 5S gene, the additional gene-specific initiation factor, TFIIIA, would which contact the internal control region.) Transcription in the absence of La yields a stalled transcription complex containing the 3' foreshortened directly while the of La produces an active transcription complex capable of multiple rounds of re-use (bottom right). transcript (bottom left), presence of each of the two complexes are summarized below. Class III genes are depicted as stippled boxes with Expected transcriptional properties nascent transcripts are wavy lines; released transcripts are represented as hairpin structures. Other components include: termination signals (TTTT); complex assigned various sizes .300 kd (Lassar et al., 1983; Ruet et al., 1984; Stillman et al., 1985; Yoshinaga et al., TFIIIC, a large globular which over the internal control region and may protect (or associate with something that protects) the terminator; TFIIIB, probably 1987), footprints a molecule of 60 kd or a dimer of two 60-kd subunits (Klekamp and Weil, 1986; P.Martin and R.Roeder, unpublished observations), which single does not to produce a footprint but stabilizes the complex; RNA polymerase III, a multisubunit enzyme of -700 kd (see Roeder, 1976); La, appear a 50-kd which can be found associated. with TFIIIC, with active transcription complexes, and with the 3'-uridylate sequence of full- phosphoprotein, length transcripts. La might induce a conformational change in the polymerase (compare complexes on the bottom left and right), effecting transcript completion. of HeLa-derived tran- 1987). La immunodepletion Steitz, Results extracts results in a dramatic diminution of tran- scription Model and the few synthesized in the scription ability; transcripts 1 model for RNA polymerase III tran- Figure depicts a of La have fewer residues at their 3' ends absence uridylate TFIIIB and TFIIIC associate with a class scription. Initially to those synthesized in the presence of La. These compared III forming a stable transcription complex (top) extents by the addition gene phenomena are reversed to different et al., 1982; Lassar et al., 1983; Schaack of La protein (Gottlieb and Steitz, (Bogenhagen biochemically purified et 1983). RNA polymerase III recognizes this complex 1989). al., We that elongation we a detailed model for La action at the and initiates synthesis (middle). propose Here, present of RNA chain until the polymerase halts termination and employ a variety of in vitro the initial proceeds signal We conclude that auxiliary to the termination signal. In the absence of La, to test this hypothesis. just prior approaches of III transcription a stalled transcription complex (bottom left) and a low level factors facilitate termination polymerase the La protein is a good candidate for of 3' truncated product result. In contrast, when La is present and that mammalian such a termination factor. In addition, we present evidence (the normal situation) the polymerase overcomes the pause right). The for a novel stalled transcription intermediate containing a and transcribes the 3' uridylate stretch (bottom and a 3' foreshortened transcript is released, freeing the polymerase molecule resulting full-length paused for multiple rounds of re-use. This model unreleased transcript. complex 852 La as a termination factor data accommodates all known polymerase III transcription and ascribes both of La's functional properties-its influence on the overall transcription level and the 3' foreshortened transcript termini found in its absence (Gottlieb and Steitz, 1989)-to a single molecular event. To test this model and to assess whether La plays a role in the termination process, we have performed a series of EBEt 2 experiments monitoring the behavior of the EBER2 gene (Rosa et al., 1981; Jat and Arrand, 1982). This prototypical class III gene contains an intragenic control region (Rosa et al., 1981) and requires TFIIIB and TFHIC in conjunction with RNA polymerase III for its transcription (P.Martin, unpublished observations). EBER2 is synthesized at high levels and is quantitatively precipitable by anti-La antibodies in vivo and in vitro (Rosa et al., 1981; E.Gottlieb and Bottor- ' 2 3 J.A.Steitz, unpublished observations). Moreover, the low Fig. 2. Gradient fractionation of EBER2 transcription complexes. level of EBER2 3' foreshortened product synthesized after Panel A: EBER2 DNA was incubated in a HeLa-derived S-l00 for La immunodepletion of a HeLa-derived S-100 extract is 30 min, and the resulting transcription complexes were sedimented from full-length EBER2 (Gottlieb and easily distinguished through a 20-45% glycerol gradient (17 fractions collected). Active We infer that our conclusions with EBER2 Steitz, 1989). transcription complexes were detected by their ability to support the be extended to other class III which behave synthesis of radiolabeled EBER2 following the addition of can genes [ca-32P]GTP. Products are displayed on a 6% polyacrylamide gel. The similarly in La-depleted extracts (Gottlieb and Steitz, 1989). direction of sedimentation is indicated, and the two complex populations are denoted and j3. Panel B: Transcription complexes Transcription complexes differ in the presence and were formed and gradient purified as in (A). Radiolabeled products absence of La synthesized by complexes in peak a (lane 1), peak ,B (lane 2), or by a mixture of used in lanes 1 and 2 (lane 3) after a the fractions 90-min formed on class III genes in vitro Transcription complexes incubation displayed on this 6% gel. Lanes were derived at 30°C are et Lassar et Schaack (Bogenhagen al., 1982; al., 1983; from a single exposure of a single gel. are stable rounds of et al., 1983) through multiple on in La with either of the two complex populations could not be transcription. They can be isolated glycerol gradients active form and contain < of the total detected not shown). Recall that most of the La in cells a synthetically 1% (data Jahn et exists in the form of RNPs (Stefano, 1984). In an alternative extract protein (Wingender et al., 1984; al., 1987), their and we the sedimentation profile of com- enabling us to analyze both catalytic properties approach, compared EBER2 complexes were assembled formed in anti-La versus non-immune depleted molecular constituents. plexes in an untreated HeLa derived extracts. La enhanced levels under transcription conditions Following immunodepletion, et added nucleotides. After of the slower S-100 (Weil al., 1979) containing sedimenting complex population (13) were seen each fraction was while levels of the more rapidly migrating population (a) glycerol gradient fractionation, ribonucleotides were diminished (not shown). This suggests that at least one supplemented with containing [a-32P]GTP for Since additional DNA difference between these two and incubated 90 min. template transcription complex popu- the lations is a deficiency of La protein in 1 relative to c. was not added after ultracentrifugation, resulting products of active reflect the sedimentation position transcription An alternative for the explanation synthesis of shortened to fractionation. complexes formed prior products by the 13 fractions (Figure 2A) is that a nuclease Two distinct populations of complexes can be visualized sediments in this of the To exclude this region gradient. (Figure 2A). One population (denoted a) migrated at 60S possibility, mixing experiments were performed (Figure 2B). and synthesized full-length EBER2. Surprisingly, a second of fractions from regions a (lane 1) and Aliquots gradient slower sedimenting fraction (denoted 13) produced 3' 13 were combined to the 90-min (lane 2) prior transcription of III foreshortened transcripts, characteristic polymerase reaction and the products analyzed as in Figure 2A. Figure in transcription the absence of La (Gottlieb and Steitz, 1989). 2B (lane 3) shows a typical result: about a 20-fold stimulation The levels of RNA synthesized by these two populations are of transcription of full-length product. In other trials, the number of active likely a reflection of the relative complexes lengths of the products reflected the length distribution seen RNA in each population: the ratio of labeled transcripts in the two component fractions. These results argue that the is constant in 13 is not due to a synthesized by these two populations upon appearance of shorter products region from extract in these fractions and further refractionation of a given extract, but varies nuclease specifically present rounds of to extract. suggest that La may recycle between transcription survives a fraction of the total (see Discussion). Only transcription activity observed K.H. fractionation previously by gradient (also loss of one The absence of the La affects the kinetics of protein Seifart, unpublished observations), suggesting To determine whether the transcript accumulation or more essential components. retained in the a To determine whether transcription complexes containing La is protein preferentially by complexes exhibit dissimilar was located in the ELISA or La also transcription region, La protein gradient by lacking might La a extract et kinetics, we subjected transcription (Weil al., (Thoen et al., 1980). Unfortunately, antigenicity peaked with levels in the central with non-immune or anti-La antibodies smeared downwards 1979) to depletion at the and top that of followed to a A of the by exposure protein preparation (Gottlieb portion gradient high enough any repeaking 853 E.Gottlieb and J.A.Steitz absence 3' only foreshortened molecules appear. Full-length products do not seem to be synthesized and subsequently an trimmed, although exceedingly rapid process cannot be excluded. the low Additionally, transcript levels observed in the absence of La cannot be wholly ascribed to transcript instability; there is no initial burst followed by subsequent decay. Instead transcripts build up late in the reaction (cf. lanes f and k, Figure 3B; see Discussion). Visualization of a stalled transcription intermediate The above u4m4 experiments indicate that transcription complexes with different physical and catalytic properties are formed in the or presence absence of the La protein. Our model makes the (Figure 1) further specific prediction that the 3' foreshortened transcripts synthesized in La-depleted extracts are a reflection of stalled transcription complexes in which the nascent transcript remains attached to the DNA template further rounds blocking of synthesis. To look for such a transcription intermediate, S-100 extracts were subjected to immunodepletion with non-immune or anti-La serum as above and then supplemented with labeled and unlabeled in Fig. 3. Time course of EBER2 transcription the presence and nucleotides and the EBER2 After a gene. 45-min incubation, absence of La. Parallel reactions were initiated non-immune by adding (panel A) or anti-La (panel B) depleted extracts (see Materials and reaction mixtures were fractionated on glycerol gradients methods) to pre-mixed transcription reactions the EBER2 including et (Wingender al., 1984; see Figure 2) and the radiolabeled template and [cx-32P]GTP Reactions were incubated at (0 min). 300C, transcripts in each fraction analyzed on denaturing with aliquots removed at time points and treated with immediately polyacrylamide gels (Figure 4). This experiment directly EDTA, SDS and proteinase K to halt synthesis. ethanol Following monitors the precipitation, the radiolabeled products were a sedimentation behavior of transcripts electrophoresed through single 6% polyacrylamide gel. Arrows denote the and 3' full-length synthesized before fractionation, whereas in Figure 2 the foreshortened EBER2 molecules. A small amount of the in sample behavior of preformed transcription complexes was lane was lost e, panel A, upon gel loading. examined. Here, if transcripts migrate into the gradient, they and Steitz, To initiate the are likely to be associated 1989). synthesis, resulting depleted with large complexes since EBER2 were added to assembled RNPs, consisting of the 172 supernatants previously reaction only 1 nucleotide RNA and mixtures salt excess the - containing optimal concentrations, 50-kd La protein, have a sedimentation value of 8S and amounts of and would remain at the ribonucleotides, [a-32P]GTP saturating top of the gradient. template. Radiolabeled at various times Figure 4, panel shows the transcripts appearing B, gradient containing products were resolved on a synthesized in the absence of La. All denaturing polyacrylamide gel (Figure transcripts are short We several 1 that and their level is 3). sampled early points (< h) anticipating low, as expected. While most transcripts these would be most to reveal differences between the remained at the top of the likely gradient (lanes 1-3), some reactions containing and La. Because migrated into the where lacking transcription region competent transcription complexes were not preformed prior to the addition of label, complexes sediment (lanes 6-14; compare with Figure 2A). we could examine both the transcriptional lag and later Agarose gel analysis of of each fraction aliquots confirmed reaction phases previously described for other class IH the presence of the DNA in this genes, template region (data not where similar kinetic were analyses performed (Birkenmeier shown). In in the contrast, gradient containing the labeled et Schaack et Bieker et al., 1978; al., 1983; al., 1985). products synthesized in the presence of La (panel A), all The rate of is transcript appearance strikingly different in of the EBER2 transcripts are full-length and appear at the the presence and absence of of the (Figure 3A) (Figure 3B) La top gradient (lanes 1-4). These are presumably La- protein. In the extract with immunodepleted non-immune containing ribonucleoprotein particles since EBERs serum (Figure 3A), newly synthesized products were synthesized in a La-containing extract are quantitatively La detected after 5 min of incubation (lane b). Product precipitable. Despite the fact that more incorporated counts accumulation lagged over the first 30 min but then increased were loaded onto gradient A than gradient B, no full-length with time. All steadily detectable transcripts, including those EBER2 transcripts are detected cosedimenting with the observed at the earliest time, were full-length. An extract transcription apparatus (panel A, lanes 6-14). Light that was not subjected to the immunodepletion protocol gave exposures of the third fraction from these gradients (panel a similar transcription pattern except that the total C, lanes 1, 2) show that the predominant transcription incorporated counts were slightly greater (not shown). products were full-length EBER2 and 3' foreshortened Therefore, the immunodepletion procedure itself does not EBER2 in the presence and absence of La, respectively. This alter the overall pattern of transcription. In contrast, in the length difference was confirmed by analyzing small samples absence of La (Figure 3B), the first observable transcripts of the total transcription reactions fractionated in A and B appeared 25 min after initiation of the reaction (lane f) and (data not shown). The second most abundant product all detectable EBER transcripts exhibited a rapid gel (denoted by *) has an EBER2 fingerprint; it could represent mobility. a multimer formed during the gradient procedure, since it Consistent with the proposed model (Figure 1) is the fact is absent when these reactions are directly gel fractionated. that even at the earliest time of transcript detection in La's It appears that La-immunodepletion has enabled us to 854 La as a termination factor oc E m Nf,n". Z -a r t 1:- - ; I !e-l- .:kw t.B&K-. .4.r. z& i w .'AAA, VI -.0- "., %-;%.I,.`,..:,!.-- *1.i ... i 4 - .. M c 1 2 To Fig. 4. Visualization of a stalled transcription complex. Immunodepleted extracts (see Materials and methods) were supplemented with [a-32P]GTP and [a-32P]UTP, unlabeled ribonucleotides, salts and the EBER2 template, incubated at 30°C for 45 and fractionated on parallel min, 20-45% glycerol gradients. Each resulting fraction was proteinase K treated, ethanol precipitated and displayed on a 6% polyacrylamide gel. Panels A and B: Migration behavior of radiolabeled transcription products from non-immune and anti-La depleted extracts respectively. Numbers correspond to gradient fractions and arrows denote - 1 the EBER2 transcripts. The direction of sedimentation is indicated (top bottom). Panel C, lanes and 2: Lighter exposures of lanes A3 and B3. The major transcription products include full-length and 3' foreshortened EBERs respectively. Products denoted by * have EBER2 fingerprints, are not observed when aliquots of each reaction are gel fractionated without prior gradient separation, and are likely to be EBER2 multimers formed during gradient fractionation. Most of the additional radioactivity in the top fractions of panels A and B represents overexposure of minor products. An additional radiolabeled RNA with very low gel mobility appears in the middle of both gradients, but we could never retrieve enough of this material to analyze it. visualize unreleased polymerase III transcription products If the latter were the case, the short RNAs presynthesized associated with stalled transcription complexes. The 3' might be susceptible to completion (Maderious and Chen- truncated RNAs sedimenting at the top of the -La gradient Kiang, 1984; Grayhack et al., 1985). Ideally, we would have (Figure 4B) could have dissociated from transcription liked to add La protein to gradient fractions containing these complexes because of stress during gradient fractionation nascent RNAs and analyzed their lengths. However, the low and/or slow but spontaneous release. The latter of in such possibility recovery transcripts retained after complexes would also explain the delayed transcript build-up that we use unfractionated seen at fractionation demanded tran- late reaction times in a La-depleted extract (Figure In for this 3B). scription extracts experiment instead. the of presence La, transcript release appears to be a SA indicates that the 3' foreshortened rapid, Figure transcripts active process coupled to completion; thus, we (Figure 4A) in a extract can be generated La-immunodepleted completed and others et (Ackerman al., 1983; Wingender et al., 1984; addition of a source of La Extracts by subsequent protein. Jahn et have been to al., 1987) unable detect this intermediate with non-immune or anti-La depleted (lanes 1,2) (lanes 3,4) in with complete transcription reactions. serum were the EBER2 supplemented nucleotides, and incubated for 30 and template, [oe-32P]GTP, min, long The stalled transcription can be rescued for labeled 3' foreshortened to complex enough transcripts appear (see The stalled visualized in 4B this short incubation we transcription complexes Figure Figure 3B). By selecting period of could either side of the to maximize the represent non-productive products hoped percentage transcripts remaining or synthetic process bone intermediates. associated with the stalled fide transcription transcription complexes (Figure 855 E.Gottlieb and J.A.Steitz pletion. It is highly unlikely that the full-length molecules detected after the 5-min chase of the stalled transcription complexes result from de novo synthesis since this time for period is too short substantial amounts of new molecules to appear (Figure 3A, lane 2) and the few made would have D Lsi!i 200-fold a specific activity less than the pre-existing 00 3 1~ _7 < these do not 4.. 7 _~ transcripts. [Note that complexes sequester ,_s J ., ,1 r- z < « of et appreciable amounts nucleotides (Wingender al., exhibits the same 1984).] Transcript completion o-amanitin sensitivity as synthesis by RNA polymerase III: inclusion of 300 a-amanitin (Figure 5B) during the 5-min chase itg/ml in period abolishes the shift transcript mobility (lane 3 versus 4). Similar results are seen following chase periods of 15 (lane 5) and 60 min (data not shown). Thus, our results indicate that the discrete 3' foreshortened RNA in La extracts is a bone synthesized depleted fide the of a transcription intermediate, probably product paused polymerase. Transcript completion appears to require active RNA in with the La polymerase Im conjunction protein preparation. Although other data (Gottlieb and Steitz, 1987; this manuscript) indicate that La is normally associated with the the chase data that La can transcription complex, suggest functionally interact even after initiation of This synthesis. conclusion does not contradict the results in 2 since Figure the concentration of La in the tran- gradient-fractionated scription complexes was calculated to be about four orders of magnitude less than in this experiment. Chelating La with RNA also yields stalled transcription complexes termination factor To ask further whether La itself is a that we facilitates transcript completion and release, exploited the protein's documented to bind the U-rich 3' end ability 2 of polymerase III reaction was FBER r# transcripts. transcription EBER2 pre-incubated with excess unlabeled (Figure 6D, lane EBFR labeled EBER) in the hope that incorporation into ribo- nucleoprotein particles would render the La protein to La unavailable the transcription apparatus. is the only protein known to bind the EBERs (J.Stefano, unpublished Glickman et observations; al., 1988). Thus, this depletion is more selective than use of anti-La antibodies since strategy La under that immunodepletion conditions preserve also removes several less abundant transcription activity proteins by virtue of their association with La and/or La- bound RNAs. As a an reaction was control, equivalent pre- Fig. 5. Stalled can be chased to transcription complexes completion. for 30 at Radiolabeled EBER2 transcripts were synthesized min 30°C incubated with an equimolar amount of unlabeled EBER2 with non-immune A in reactions containing extracts depleted (lanes from which the U-rich La binding site had been specifically B or anti-La A B antibodies. 1-2, 1-2) (Lanes 3-4, 3-5) Samples removed. These 3'/EBERs, generated by oligonucleotide- then received buffer Al and Bi and prewarmed transcription (lanes 3, directed RNase H cleavage of EBER2 are shown in Figure or S-100 into buffer A2 3) prewarmed dialyzed transcription (lanes and 4 and either a 200-fold excess of unlabeled 6D. After pre-incubation, the reactions were supplemented 4; B2, 5) containing nucleotides (A) or 300 ig/ml a-amanitin (B). Incubation was for 5 min with the EBER2 gene and both labeled and unlabeled in all cases lane where the chase was extended to a total except B5, nucleotides, incubated for 45 min and fractionated in parallel of 15 min. Resulting radiolabeled products were proteinase K treated on glycerol gradients as in Figure 4. and separated on denaturing 6% polyacrylamide gels. Figure 6B reveals that removal of La by RNA chelation 4B). Then, following the simultaneous addition of a 200-fold again results in visualization of stalled transcription excess of unlabeled ribonucleotides (to dilute the labeled intermediates: a fraction of the radiolabeled transcripts and either 1 precursors) transcription buffer (lanes and 3) synthesized in the EBER2-treated extract migrates into the or a La protein dialyzed into transcription buffer (lanes 2 central region of the glycerol gradient (lanes 4- The 1). and 4), incubation continued for . 5 additional minutes. The of mobility full-length EBER2 transcripts generated in a shift of the presynthesized 3' foreshortened species (lane 3) parallel untreated reaction and loaded onto the far right lane to full-length (lane 4) indicates that at least some of the RNA of the gel demonstrates that these rapidly sedimenting RNAs components of stalled complexes can be chased to com- are short, as expected for paused transcripts. This effect is 856 La as a termination factor RV + EBER' 3' :BER RNA Lii ULJ Fig. 6. A stalled transcription complex can be generated by chelation of the La protein with RNA. Panels A-C: Gradient profiles showing the migration behavior of radiolabeled EBER2 in synthesized extracts preincubated with excess unlabeled EBER2 from which the La binding site had been removed with EBER2 (3'AEBER, A), full-length (B), or no added RNA (C). Equivalent c.p.m. were loaded onto each The direction gradient. of sedimentation in the to glycerol gradient (left right) is indicated (top-bottom); the far right-hand lane of panels A and B contain full-length EBER2 as a marker. In the EBER2 into the transcript panel B, transcript running gradient exhibits the shorter length characteristic of in transcription the absence of La Gottlieb and (see Steitz, 1989), compared to the full-length transcript in panels A and C. Panel D: An ethidium bromide stained of the unlabeled EBER2 of EBER2 gel profile full-length and RNA that has had its site removed RNase H La-binding by oligonucleotide-directed mediated cleavage (3'zEBER). produced by adding EBER2 at an RNA: La protein pretation of these ratio results is that the factor whose activity we estimated to be one-tenth that found in the have been monitoring is the La protein itself. cell and without appreciable diminution of transcription. (In experiments not shown, depression of transcript levels was observed at higher Discussion EBER2 concentrations.) In labeled EBER2 made contrast, Existence of a transcription termination factor for in the extract pre-incubated with 3'AEBERs or (Figure 6A) RNA polymerase I1 in an untreated extract are full and do (Figure 6C) length Two important conclusions emerge from our results: first, not into the significantly migrate gradient. Simply being termination by RNA III does not cause a to move polymerase seems to require incomplete transcription product auxiliary factor(s) and second, the 50-kd mammalian La into the gradient: while RNAs of other are lengths present their 3' terminal Us protein appears to be such a transcription termination factor. in the profiles, those only lacking just Ultimate with confirmation of the latter conclusion will require cosediment transcription complexes (Figure 6B). the addition of in which or 3' biologically active La protein synthesized Moreover, control experiments full-length from a cDNA clone to a reconstituted EBER2 were mixed with unlabeled transcription system foreshortened transcripts from more highly than are now reactions to fractionation revealed that purified components transcription prior available. As discussed in Gottlieb and Steitz neither sediments into the not (1989), detectably gradient (data current RNA must have been protocols for reconstitution of transcription utilize a 3' foreshortened shown). Hence, fractions still is to with containing La. Yet, the fact that stalled in the reaction if it the synthesized migrate can be methods The inter- complexes generated by two independent fractionation. transcription complex upon simplest 857 and J.A.Steitz E.Gottlieb it should involve at least three distinct of La depletion (Figures 4 and 6) implicates La itself as the prokaryotic process, transcript release and termination factor required for their resolution. If another events: polymerase pausing, With and Bear, 1983). molecule were responsible, it would have to meet three polymerase dissociation (Platt with La, allowing complete coli RNA polymerase, pausing is the first step criteria: (i) very tight association Escherichia do not result in anti-La antibodies (Figure 4); (ii) process; while all pauses immunodepletion by in the termination is an for of the U-rich 3' ends of polymerase III pausing obligate prerequisite specific recognition termination, et 1985; 6); and (iii) copurification with La through (von Hippel et al., 1984; Grayhack al., transcripts (Figure termination two is La first fractionates as an RNP and later The order of the remaining processes six steps in which Platt, 1986). not known. as a protein (Gottlieb and Steitz, 1989). of our model (Figure 1) are supported of stalled transcription complexes in the absence Several predictions Detection here. Distinct transcription that La is not required for polymerase pausing. by the analyses presented of La argues to be assembled in the presence and it seems reasonable that RNA polymerase III, like complexes appear Rather, RNA (Grayhack et al., 1985; J.Roberts, absence of La (Figure 2). A fraction of the 3' foreshortened E. coli polymerase the inherent capacity to transcripts generated in La's absence appear to be retained personal communication), possesses complexes (Figure 4-6), providing Parameters inducing pausing by RNA polymerases in stalled transcription pause. that the 3' foreshortened RNA arises via even for the E. coli enzyme. Pausing could the best indication are unknown, of a helical stem between infer that this novel transcription the formation a transcription event. We result from near the 5' and 3' ends minimally contains TFIIIB, TFIIIC (since both present intermediate complementary sequences III factors required for initiation), RNA nascent RNA polymerase transcripts. are non-cycling of most the structure of the DNA the DNA template and the nascent RNA it may be induced by polymerase III, Alternatively, that III termination It would therefore be distinct from previously It is polymerase transcript. template. intriguing residues in the metastable, stable and pre-initiation transcription include 24 adenine coding identified signals (which that are believed to [see Bieker et al. (1985) for distinctions] in at coincide with DNA sequences complexes strand) it would represent a step immediately induce bending (Koo et al., 1986). least two regards: after and it would contain the incomplete mechanisms for La protein action polymerase preceding termination Several our data. Since La does not La to be an integral component of the are compatible with transcript. appears pausing but is in RNA normal transcription complex not only because its presence to bind DNA strongly present appear Gottlieb in gradient purified complexes can be inferred (Figure 2) III transcription complexes (Figure 2; polymerase and in and but also because DNA carrying class III genes can be and 1987) TFIIIC preparations (Gottlieb Steitz, interacts with other protein selectively immunoprecipitated from synthetically active Steitz, 1987), it most likely and 1987). induce conformational extracts with anti-La antibodies (Gottlieb Steitz, components and could thereby be able to interact to Figure 5 suggests that La may also in the transcription complex causing polymerase changes even after This be reflected in the functionally with the transcription complex effect transcript completion. may between initiation of synthesis. substantial S-value difference transcription that short versus full-length transcripts The of the simplest form of the model (as presented complexes synthesize aspect be accounted for solely by the in that may require revision is the prediction that (Figure 2A), which cannot Figure 1) that the mere one round of synthesis should be detected when absence of a 50-kd protein. [Note presence only formation in a HeLa extract transcript completion and release are not La facilitated. In of nucleotides during complex level and length are increases the mean sedimentation rate other words, we observe that transcript significantly we do see a low level of not obligatorily coupled: while et al., 1984) and that E. coli transcription (Wingender of this level does not remain transcript in the absence La, containing nascent RNAs can exist in alternate complexes in 3. This unexpected observation Once the constant over time Figure conformations (Straney and Crothers, 1985).] in at least two ways. Conceivably, has been synthesized, the could be explained transcript's U-rich 3' terminus assembly could be altered by removal be incompatible transcription complex transcription complex configuration may of La-associated transcription initiation factors (e.g. TFIIIC) retention. Alternatively, La may be correctly with transcript upon La immunodepletion, effecting a decrease in transcript complex to bind the newly positioned in the transcription level not directly related to La. This explanation could be RNA 3' terminus when it becomes exposed synthesized viewed as consistent with our reconstitution data (Gottlieb of - heteroduplex instability (Martin and because dA rU and 1989) and with the absence of significantly to the E. coli termination factor rho, Steitz, Tinoco, 1980). Similar is removed by RNA depressed transcription levels when La the nascent transcript (Chen et al., 1986; Bren- upon binding slow transcription complex chelation (Figure 6). However, nan et al., 1987), La might melt the remaining DNA -RNA is entirely responsible since transcript accumu- assembly not using an ATP-dependent helicase activity (Brennan duplex lation in the presence and absence of La (Figure 3) differ et al., 1987). This scenario is supported by recent sequencing in the rate of product accumulation as well as in the dur- of a Xenopus La cDNA clone, which has revealed a potential ation of the transcriptional lag. Alternatively, in a scenario NTP and S.Clarkson, personal binding site (D.Scherly the data in Figures 4 and 6, a single round compatible with Finally, a more indirect role for La in communication). of may simply not be detectable; later transcript synthesis transcript completion and release can also be envisioned. build up in the absence of La may result from slow, spon- pausing at the U tract may result from Here, polymerase taneous release of truncated products. an increased rate of depolymerization versus polymerization by the enzyme (Kassavetis et al., 1986). La could shift the La and transcription termination by sequestering the longer transcripts, driving equilibrium Assuming that eukaryotic transcription termination (like the reaction towards polymerization. initiation) bears some similarity to the comparable additional work is required to determine the Clearly, 858 La as a termination factor details of La's contribution to the termination of RNA syn- substrate RNAs were La be present, would sequestered in thesis by RNA polymerase III. It should be realized that the RNPs, unavailable to further of catalyze rounds transcription. possible mechanisms for La action outlined above are not Such a scenario is reminiscent of the 5S-specific regulatory necessarily mutually exclusive. Further, La's participation for TFIIIA in loop proposed the amphibian oocyte (Pelham in the termination process does not eliminate a subsequent and Brown, 1980). La's relative affinity for various role of the protein in 3' end protection of resulting transcripts and transcription complexes formed on different transcripts; La's unique RNA binding ability may couple class 11 to its state could then genes coupled phosphorylation RNA polymerase transcription III termination to RNA be used to fine-tune this regulation. packaging. Finally, future efforts should also resolve the issue of whether (Bieker al., 1985; et Setzer and Brown, Materials and methods 1985; Carey et al., 1986) or not (Jahn et al., 1987) RNA polymerase III dissociates after each round of synthesis and antibodies and Cells, enzymes HeLa and cells were maintained at in a environment Raji 37°C 5% Co2 (if so) whether dissociation occurs concomitant with or x were cultured in at -7 x 105 or 5 105 cells/ml Cells respectively. following transcript release. in in RPMI 1640 spinners (HeLa) or suspension (Raji) (GIBCO) sup- In a detailed study, Cozzarelli et al. (1983) previously con- with 10% heat-inactivated fetal calf serum 60 plemented (GIBCO), Ag/ml cluded that Xenopus RNA polymerase III alone is capable and 300 penicillin, 100 streptomycin glutamine. Ag/ml Atg/ml Autoimmune sera J.Hardin and Yale of transcription termination. These authors analyzed their patient (kindly provided by J.Craft, University) were selected for their titer and relative high monospecificity 5S transcription products by nuclease mapping, which T, as three determined by independent assays: immunofluorescence, probably was not sensitive enough to discriminate the 1-2 of in vivo RNAs from a immunoprecipitation [32P]orthophosphate-labeled uridylate residue difference between transcripts 5S produced HeLa whole cell Matter et and extract (Hendrick et al., 1981; al., 1982) in the presence and absence of La (Gottlieb and Steitz, Western immunoblots Yen and 1989). (Towbin et al., 1979; Webster, 1981; Mimori Non-immune sera were donated et al., 1984). by healthy laboratory We suggest that the RNAs Cozzarelli et al. analyzed may were obtained ammonium sulfate personnel. IgG preparations by precipitation have been components of stalled transcription complexes. and contained - 17 Anti-La monoclonal and mouse mg/ml protein. IgGs Their data would then strongly argue that RNA polymerase control kind of D.Williams and Smith IgG (the gift P.Venables; et al., 1985) III alone is capable of the first step in termination- were in A - selected grown the ascites of BALB/c mice, protein Sepharose and were - 0.5 recognition of and pausing at the termination signal. Other mg/mi protein. aureus cs-amanitin Lyophilized protein A-bearing Staphylococcus cells, ways of resolving the apparent contradictions between their RNA markers 16S and 23S were from and (MS2, RNAs) purchased and our conclusions are: (i) there might exist species-specific K was obtained from Beckman. Boehringer and Mannheim and proteinase differences between Xenopus and HeLa RNA polymerase RNase H was from Pharmacia. purchased III, although the observations of Watson et al. (1984) on and reactions Extracts, antibody depletions calf thymus RNA polymerase transcription III can likewise be explained extracts were from HeLa Soluble transcription (S-100) prepared log phase by pausing at the termination signal; (ii) polymerase IH alone cells frozen in and stored (Weil et al., 1979), aliquoted, liquid nitrogen might be able to recognize the termination signal and at were at in buffer -70°C. Antibody depletions performed 0°C transcription complete transcription on a naked DNA template (their mM 5 mM 15 mM 0.5 mM as (70 KCI, MgCl2, Hepes pH 7.9, DTT) described and Prior to extracts were experiments) but require additional factors when (Gottlieb Steitz, 1989). confronted transcription, with 0.5 mM each CTP and 0.025 mM supplemented ATP, UTP, GTP, with a protein-covered transcription complex (our experi- and unless otherwise indicated. Each was [a-32P]GTP template, optimized ments); or (iii) Xenopus RNA polymerase III purified to 90% for DNA and concentrations. EBER2 MgCI2 gene containing plasmids, homogeneity might retain some La protein, even though La or R1J et were pJJJ2 (Jat and Arrand, 1982) pEBV (Rosa al., 1981), supplied is not detectable in purified HeLa RNA at 10-40 at polymerase III as supercoiled templates Transcription proceeded 30°C Ag/ml. for indicated. Reactions 25-50 were then the times preparations either by immunochemical (typically Ll) criteria (Gottlieb and with 20 carrier 0.4% SDS and 20 supplemented 4tg yeast RNA, yg proteinase Steitz, 1987) or by activity (W.K.Hoeffler and E.Gottlieb, at for 20-25 made 0.25 M in ammonium acetate K, heated 65°C min, unpublished observations). Radiolabeled RNA were fractionated on and ethanol precipitated. products or M x TBE M 0.5 M 5%, 6% 10% urea/0.5 polyacrylamide/7 (0.5 Tris, sodium 2 mM EDTA which were dried and borate, pH 8.3) gels, at -700C. autoradiographed A regulatory role for La? Figure 1 suggests that La's role in termination may have the a direct affect on initiation by resetting transcription Glycerol gradient analyses of re-use. La's complex for its documented multiple rounds fractionation was a modification of the Gradient performed by procedure et Stable were formed presence might therefore amplify the transcriptional output. of al. Wingender (1984). transcription complexes 60 10 of EBER2 in 150-1l reactions and initiation containing 11 S-100, tsg/ml template, Interaction between transcription termination 1.75 mM sodium GTP and 0.5 mM each UTP, ATP, CTP, phosphate (pH would not be unique to the RNA polymerase III system. of the buffer 15 mM NaCl and the other 6.3), components transcription RNA polymerase I termination affects initiation at the for 30 reactions were chilled listed above. After incubation at min, 30°C adjacent rRNA transcription unit (Grummt et al., and on 20-45% mM 1986; 4.8-ml (v/v) layered glycerol gradients containing 15 mM 0.5 mM 1.5 mM 5 mM DTT, and KCI, MgCl2, Hepes pH 7.9, Henderson and Sollner-Webb, 1986; McStay Reeder, in an 1 rotor at 50K for EDTA. Gradients were SW50. centrifuged r.p.m. formation of U-RNAs 1986) and the 3' end signals the of each 2.2 h at To fraction, 0°C. assay transcription ability resulting the (synthesized by RNA polymerase II) overlap promoter assembled 100 tl of the 0.5 reactions were fraction, mM containing 200-Al et 20 (Hernandez and Weiner, 1986; Neuman de Vegvar al., and 0.025 mM each CTP GTP, ATP, UTP, ACi [a-32P]GTP 70 mM 15 mM 7.9, 5 mM and 1986; Ciliberto et al., 1986). KCI, (410 Ci/mmol), Hepes pH MgCl2 at and K mM DTT. a incubation 0.5 90-min Following 30°C proteinase of is Possible feedback regulation transcription suggested ethanol and resolved on were treatment, transcripts precipitated for by the La is both a factor fact that required transcript wt standards fractionated on a Mol. polyacrylamide gels. parallel gradient completion and binds the resulting products: La could RNA RNA HeLa included E.coti ribosomal (16S, 23S), (28S), MS2 of all RNA III and R17 modulate the ribosomal subunits therefore synthesis polymerase (40S, 60S) (80S). phage were after the collection of If a number of Mixing experiments performed immediately in the mammalian cell. transcripts large 859 E.Gottlieb and J.A.Steitz gradient fractions. A volume of 25 Id from a fraction judged to with the of were synthesize in after 30 exception [a-32P]GTP processed parallel; min full-length RNA and from one judged to of synthesize 3' foreshortened RNA was transcription, added with the other [as-32PJGTP components. were mixed with an equal volume (50 reaction ad) of mixture (1 mM ATP, CTP and 0.05 mM GTP, ,Ci UTP, 10 [a-32P]GTP, 70 mM KCl, 15 mM Hepes pH 7.9, 5 mM MgCI2, 0.5 mM DTT), incubated at 30°C for 2 h Acknowledgements and processed as above. To control for dilution effects, 25 of each fraction Jil We are for advice was also mixed with 25 and from Alan of transcription buffer and the grateful helpful suggestions Jeff reaction treated Weiner, /d Nouria Hernandez, Elisabetta Kim as above; no alterations due to dilution were detected. Roberts, Ullu, Black, La Doug protein Mowry, Daniela concentration Parker, Rhodes, Nelson and David in each fraction was determined by ELISA (Thoen et Kathy Hillary Setzer; David al., 1980) Brow one mechanism we discuss on 100 pl of for La action each gradient fraction. suggested termina- during tion. Materials were John Joe kindly supplied by Hardin, Craft, David Patrick Venables John In vitro kinetics transcript Williams, (antibodies), Arrand and John of accumulation (plasmid) We thank Therese HeLa Yario for technical S-100 extracts were immunodepleted with Flory (oligonucleotides). assistance non-immune or anti-La and Stevens for Research funds were antibodies as described (Gottlieb and Steitz, 1989). Parallel Lynda typing. provided a transcription by grant from the National Institutes of Health reactions (1 ml each) were assembled by adding the (CA depleted supernatant 16038). to pre-compiled a mixture of salts, unlabeled ribonucleotides, [ae-32P]GTP and the EBER2 template. Incubation proceeded at 30°C over a 3 h period. References Aliquots of 50 were removed ul from each reaction at the times indicated, stopped by EDTA and SDS addition, and proteinase K treated. The ethanol and Ackerman,S., Bunick,D., Zandomeni,R. Weinmann,R. Nucleic (1983) precipitated radiolabeled products were fractionated. gel Acids 6041-6064. Res., 11, and Allison,D.S. EMBO Hall,B.D. (1985) J., 2657-2664. 4, Visualization of stalled transcription complexes and Bieker,J.J., Martin,P.L. Roeder,R.G. (1985) Cell, 119-127. 40, Transcription reactions of 100 containing extract ul (immunodepleted with Birkenmeier,E.H., Brown,D.D. and Jordan,E. (1978) Cell, 15, 1077-1086. non-immune or anti-La serum), salts, ribonucleotides, EBER2 template and and Bogenhagen,D.F. Brown,D.D. (1981) Cell, 24, 261-270. 60 each of and were [k-32P]GTP [a-32P]UTP incubated for 45 min and /Ci Bogenhagen,D.F., Sakonju,S. Brown,D.D. (1980) Cell, 19, 27-35. at 30°C and diluted 2-fold with transcription buffer. A volume of 150 pl Bogenhagen,D.F., Wormington,W.M. and Brown,D.D. (1982) Cell, 28, from each reaction was layered on a 4.8-mI, 20-45% glycerol gradient and 413 -421. centrifuged in an SW50. 1 rotor at 50K r.p.m. for 2-2.2 h. Gradients were Brennan,C.A,, Dombroski,A.J. and Platt,T. (1987) Cell, 48, 945-952. fractionated, diluted 2-fold with transcription buffer and proteinase K treated. Camier,S., and Gabrielsen,O., Baker,R. Sentenac,A. (1985) EMBO J., 4, Radiolabeled products were ethanol precipitated and displayed on denaturing 491-500. 6% polyacrylamide gels. To assess the effectiveness of La depletion, the Carbon,P., Murgo,S., Ebel,J.-P., Krol,A., Tebb,G. and Mattaj,I.W. (1987) remaining 50 pl of each transcription reaction, which was not gradient 71-79. Cell, 51, fractionated, was proteinase K treated and gel fractionated. and Carey,M.F., Gerrard,S.P. Cozzarelli,N.R. (1986) J. Biol. Chem., 261, Alternatively, extracts were preincubated on ice for 20 min with 12 ug/ml 4309 -4317. EBER2 or EBER2 devoid of its La protein binding site (3'AEBER). Extracts Chen,C.-Y.A., and Galluppi,G.R. Richardson,J.P. (1986) Cell, 46, were then supplemented for transcription and processed as above. EBER2 1023-1028. for this experiment was prepared by anti-La immunoprecipitation from Raji Ciliberto,G., Dathan,N., Frank,R., Philipson,L. and Mattaj,I.W. (1986) cells (Matter et al., 1982) followed by purification on a 5% poly- EMBO 2931-2937. J., 5, acrylamide/7 M urea/1 x TBE gel, visualized with ethidium bromide, ex- Cozzarelli,N.R., Gerrard,S.P., Schlissel,M., and Brown,D.D. Bogenhagen, cised and eluted in 0.5 M sodium acetate, 10 mM EDTA, 10 mM Tris-HCl D.F. 829-835. (1983) Cell, 34, pH 7.5, 0.1% SDS, PCA extracted, ether washed and ethanol precipitated. Das,G., Henning,D., Wright,D. and Reddy,R. (1988) EMBO J., 7, Although the vast majority of EBER2 in the cell is complexed with the La 503 -512. protein, the use of anti-La selected RNA ensured that the isolated RNA DeFranco,D., Burke,K.B., Hayashi,S., Tener,G.M., Miller,R.C. and was intact and capable of binding La; non-immunoselected RNA was, Nucleic Soll,D. (1982) Acids Res., 10, 5779-5808. however, successfully employed as well. To generate EBER2 lacking the Engelke,D.R., Ng,S.-Y., Shastry,B.S. and Roeder,R.G. (1980) Cell, 19, La binding site, gel-purified EBER2 was mixed in water with a 220-fold 717-728. molar excess of a 20 mer oligonucleotide (5'-AAAAATAGCGGACAAGC- Francoeur,A.M. and Mathews,M.B. (1982) Proc. Natl. Acad. Sci. USA, CGA-3') complementary to its 3' end, heated to 80°C for 5 min to facilitate 6772-6776. 79, hybridization, equilibrated to 30°C for 15 min, made 40 mM Tris-HCI Francoeur,A.M., Chan,E.K.L., Garrels,J.I. and Mathews,M.B. (1984) Mol. pH 8.0, 5 mM MgCI2, 1 mM DTT, and incubated with RNase H Cell. 586-590. Biol., 5, (20 U/pg RNA) at 300C. After 60 min, additional RNase H (10 U/pg RNA) Fuhrman,S.A., and Engelke,D.R. Geiduschek,E.P. (1984) J. Biol. Chem., was added and incubation continued for 30 to min facilitate quantitative 259, 1934-1943. cleavage. The resulting 3'-deleted EBER2 was gel purified, eluted, PCA Geiduschek,E.P. and Tocchini-Valentini,G.P. (1988) Annu. Rev. Biochem. extracted, ether washed and ethanol precipitated. A shorter oligonucleotide 873-914. 57, (5'-AAAAATAGC-3'), complementary to the terminal 9 residues of EBER2, Ginsberg,A.M., and King,B.O. Roeder,R.G. (1984) Cell, 479 -489. 39, was also tried, but only 20% cleavage was obtained at a 500-fold molar Glickman,J.N., Howe,J.G. and Steitz,J.A. (1988) J. Virol., 62, 902-911. excess; presumably the high proportion of A-U base pairs resulted in and Gottlieb,E. Steitz,J.A. (1987) In Reznikoff,W.S., Burgess,R.B., instability of the hybrid. The concentrations of all RNAs were determined Dahlberg,J.E., Gross,C.A., Record,M.T. and Wickens,M.P. (eds), RNA by both UV absorption and intensity of ethidium bromide staining compared and the Polymerase Regulation of Transcription. Elsevier, New York, to RNA standards. The addition of synthetic polymers and had (U5 A5) 465 -468. pp. no effect (either transcriptionally or post-transcriptionally) in our system. and Gottlieb,E. Steitz,J.A. (1989) EMBO J., 8, 841-850. Grayhack,E.J., Yang,X., Lau,L.F. and Roberts,J.W. (1985) Cell, 42, Chase 259-269. Non-immune and anti-La immunodepleted extracts were supplemented with Grummt,I., Kuhn,A., Bartsch,I. and Rosenbauer,H. (1986) Cell, 47, the EBER2 template, salts, nucleotides pM (600 each ATP, CTP, UTP, 901 -911. 25 GTP) and ptM Transcription proceeded at 400,uCi/ml [ca-32P]GTP. Habets,W.J., den Brok,J.H., Boerbooms,A.M.T., van de Putte,L.B.A. and 30°C for 30 in 25 min reactions, which were then supplemented with van Al Venrooij,W.J. (1983) EMBO J., 2, 1625-1632. either 12.5 transcription ul buffer (70 mM KCl, 5 mM MgCl2, 15 mM and Henderson,S. Sollner-Webb,B. (1986) Cell, 47, 891-900. Hepes pH 7.9, 0.5 mM DTT) or S-100 dialyzed into transcription buffer. Hendrick,J.P., Wolin,S.L., Rinke,J., Lerner,M.R. and Steitz,J.A. (1981) Each minxture (prewarmed to 300C) contained unlabeled ribonucleotides such Mol. Cell. Biol., 1, 1138-1149. that the final 37.5 p,l reaction was 600 pM each in ATP, CTP, and UTP Hernandez,N. and Weiner,A.M. (1986) Cell, 47, 249-258. and 5 mM in GTP. Some reactions were supplemented with 300 pg/ml Hoch,S.O. and Billings,P.B. (1984) J. Immunol., 133, 1397-1403. x-amanitin instead of excess GTP. unlaoeled Incubation proceeded at 30°C Honda,B.M. and Roeder,R.G. (1980) Cell, 22, 119-126. for an additional chase period and was then halted by the addition of SDS, Jahn,D., and Seifart,K.H. Wingender,E. (1987) J. Mol. Biol., 193, EDTA and K. To assess the level of proteinase radiolabeled 303 -313. products the chase synthesized reactions during period, all containing P. and components Jat, Arrand,J.R. (1982) Nucleic Acids Res., 10, 3407-3425. 860 La as a termination factor Kassavetis,G.. Zentner,P.G. and Geiduschek,E.P. (1986) J. Biol. Chem., Wingender,E., Jahn,D. and Seifart,K.H. (1986) J. Biol. Chem., 261, 261, 14256-14265. 1409-1413. Klekamp,M.S. and Weil,P.A. (1986) J. Biol. Chem., 261, 2819-2827. Wu,G.-J., Railey,J.F. and Cannon,R.E. (1987) J. Mol. Biol., 194, and Geiduschek,E.P. (1982) Proc. Acad. Klemenz,R., Stillman,D.J. Natl. 423-442. Sci. USA, 79, and 6191-6195. Yen,T.S. Webster,R.E. (1981) J. Biol. Chem., 256, 11259-11265. Koo,H.-S., Wu,H.-M. and Crothers,D.M. (1986) Nature, 320, 501-506. Yoshinaga,S.K., Boulanger,P.A. and Berk,A.J. (1987) Proc. Natl. Acad. Wormington,M. and Hall,B.D. (1982) Nucleic Sci. Koski,R.A., Allison,D.S., USA, 84, 3585-3589. Acids Res., 10, 8127-8143. Martin,P.L. and Roeder,R.G. (1983) Science, 222, 740-748. Received on on Lassar,A.B., September 27, 1988; revised December 7, 1988 McStay,B. and Reeder,R.H. (1986) Cell, 47, 913-920. Maderious,A. and Chen-Kiang,S. (1984) Proc. Natl. Acad. Sci. USA, 81, 5931 -5935. Martin,F.H. and Tinoco,l.,Jr (1980) Nucleic Acids Res., 8, 2295-2299. Mathews,M.B. and (1984) Mol. Cell. Biol., 4, 1134-1140. Francoeur,A.M. Matter,L., Schopfer,K., Wilhelm,J.A., Nyffenegger,T., Parisot,R.F. and DeRobertis,E.M. (1982) Arth. Rheum., 25, 1278-1283. Mazabraud,A., Scherly,D., Muller,F., Rungger,D. and Clarkson,S.G. 197, 1-11. (1987)J. Mol. Biol., Mimori,T., Hinterberger,M., Pettersson,I. and Steitz,J.A. (1984) J. Biol. 259, 560-565. Chemii., Murphy,S., Di Liegro,C. and Melli,M. (1987) Cell, 51, 81-87. Neuman de Vegvar,H.E., Lund,E. and Dahlberg,J.E. (1986) Cell, 47, 259-266. Pelham,H. and Brown,D.D. (1980) Proc. Natl. Acad. Sci. USA, 77, 4170-4174. Pizer,L.I., Deng,J.-S., Stenberg,R.M. and Tan,E.M. (1983) Mol. Cell. Biol., 3, 1235-1245. Platt,T. (1986) Annu. Rev. Biochem., 55, 339-372. and Bear,D.G. (1983) In Beckwith,J., Davies,J. and Platt,T. Gallant,J.A. (eds), Gene Function in Cold Spring Procarvotes. Harbor Laboratory Press, Cold Spring Harbor, NY, pp. 123-161. Rinke,J. and Steitz,J.A. (1982) Cell, 29, 149-159. Roeder,R.G. (1976) In Losick,R. and Chamberlin,M. (eds), RNA Cold Spring Harbor Laboratory Press, Cold Spring Harbor, Polymerase. NY, pp. 285-329. Rosa,M.D., Gottlieb,E., Lerner,M.R. and Steitz,J.A. (1981) Mol. Cell. Biol., 1, 785-796. Ruet,A., Camier,S., Smagowicz,W., Sentenac,A. and Fromageot,P. (1984) EMBO J., 3, 343-350. Sakonju,S., Bogenhagen,D.F. and Brown,D.D. (1980) Cell, 19, 13-25. Sakonju,S., Brown,D.D., Engelke,D., Ng,S.-Y., Shastry,B.S. and Roeder,R.G. (1981) Cell, 23, 665-669. Schaack,J., Sharp,S., Dingermann,T. and So1l,D. (1983) J. Biol. Chem., 258, 2447-2453. Segall,J., Matsui,T. and Roeder,R.G. (1980) J. Biol. Chem., 255, 11986-11991. Setzer,D.R. and Brown,D.D. (1985) J. Biol. Chem., 260, 2483-2492. Dingermann,T., Schaack,J., DeFranco,D. and Soll,D. (1983) Sharp,S., J. Biol. Chem., 258, 2440-2446. Shastry,B.S., Ng,S.-Y. and Roeder,R.G. (1982) J. Biol. Chem., 257, 12979- 12986. Smith,R.P., Williams,D.G., Venables,P.J.W. and Maini,R.N. (1985) J. Immunol. Methods, 77, 63-76. Stefano,J.E. (1984) Cell, 36, 145-154. Steitz,J.A., Wolin,S.L., Rinke,J., Pettersson,I., Mount,S.M., Lerner,E.A., Hinterberger,M. and Gottlieb,E. (1983) Cold Spring Harbor Symp. Quant. Biol., 47, 893-900. Stillman,D.J., Better,M. and Geiduschek,E.P. (1985) Mol. J. Biol., 185, 451 -455. Stillman,D.J. and Geiduschek,E.P. (1984) EMBO J., 3, 847-853. Stillman,D.J., Silversten,A.L., Zenter,P.G. and Geiduschek,E.P. (1984) J. Biol. Chem., 259, 7995-7962. Straney,D.C. and Crothers,D.M. (1985) Cell, 43, 449-459. Thoen,C.O., Mills,K. and Hopkins,M. (1980) Am. J. Vet. Res., 41, 833 -837. Towbin,H., Staehelin,T. and Gordon,J. (1979) Proc. Natl. Acad. Sci. USA, 76, 4350-4354. Van Mol. Cell. 1021-1031. Dyke,M.W. and Roeder,R.G. (1987) Biol., 7, von Hippel,P.H., and Bear,D.G., Morgan,W.D. McSwiggen,J.A. (1984) Annu. 389-446. Rev. Biochem., 53, Watson,J.B., Gralla,J.D. Nucleic Actids Chandler,D.W. and (1984) Res., 12, 5369-5384. Weil,P.A., Segall,J., Harris,B., Ng, S-Y. and Roeder,R.G. (1979) J. Biol. Chem., 254, 6163-6173. Larson,D., Young,L.S. and Sprague,K.U. J. Mol. Wilson,E.T., (1985) Biol., 183, 153-163. Wingender,E., Shi,X.P., Houpert,A. and Seifart,K.H. (1984) EMBO J., 3, 1761-1768. http://www.deepdyve.com/assets/images/DeepDyve-Logo-lg.png The EMBO Journal Springer Journals

Function of the mammalian La protein: evidence for its action in transcription termination by RNA polymerase III.

Gottlieb, E.; Steitz, J. A.
The EMBO Journal , Volume 8 (3) – Mar 1, 1989
Download PDF
11 pages

Loading next page...
 
/lp/springer-journals/function-of-the-mammalian-la-protein-evidence-for-its-action-in-Qrp4xAOqlf

References (1)

Publisher
Springer Journals
Copyright
Copyright © European Molecular Biology Organization 1989
ISSN
0261-4189
eISSN
1460-2075
DOI
10.1002/j.1460-2075.1989.tb03446.x
Publisher site
See Article on Publisher Site

Abstract

The EMBO Journal vol.8 no.3 pp.851 -861, 1989 of the mammalian La protein: evidence for its Function action in transcription termination by RNA polymerase Ill was recently resolved into two complementing components Ellen Gottlieb1 and Joan A.Steitz (Yoshinaga et al., 1987). An additional 5S-specific factor, Department of Molecular Biophysics and Biochemnistry, Howard TFIIIA (Engelke et al., 1980; Honda and Roeder, 1980; Hughes Medical Institute, Yale University School of Medicine, 333 Pelham and Brown, 1980) has been purified to homogeneity Cedar Street, PO Box 3333, New Haven, CT 06510, USA et al., 1980) and its gene cloned (Ginsberg et al., (Engelke 'Present address: Medical Research Council, Laboratory of Molecular 1984). Prior to initiation, TFIIIA binds to the 5S gene ICR; Biology, Hills Road, Cambridge CB2 2QH, UK TFIIIC component associates with the ICR of both tRNA Communicated by A.A.Travers VA (Engelke et al., 1980; Pelham and Brown, and genes We have tested the hypothesis that the mammalian La 1980; Sakonju et al., 1981; Klemenz et al., 1982; Lassar which appears to be required for accurate and Fuhrman et al., 1984; Stillman and protein, et al., 1983; efficient RNA polymerase III transcription, is a tran- 1984; Stillman et al., 1984; Camier et al., Geiduschek, scription termination factor. Our data suggest that 3' 1985; Carey et al., 1986; Yoshinaga et al., 1987). foreshortened transcripts generated in La's absence are Remaining general factors are believed to associate primarily of a novel transcription intermediate interactions (Lassar et al., 1983; components through protein-protein a These transcripts are Setzer and Brown, 1985; Carey et al., 1986; Klekamp et al., containing paused polymerase. by fractionated transcription complexes, are Wingender et al., 1986). The resulting transcription produced 1986; with kinetics different from full-length remain stable through multiple rounds of tran- synthesized complexes and are chasable to completion from the et al., 1982; Lassar et al., 1983; transcripts, scription (Bogenhagen stalled transcription complexes. Together, these fmdings Schaack et al., 1983). that termination by RNA polymerase HI requires less is known about RNA polymerase III argue Considerably factor(s) and implicate La as such a factor. the process including cessation of elongation, auxilliary termination, La transcript completion and and polymerase dissociation. Four or more Since appears to facilitate transcript release and also binds the RNA product, it may residues embedded in a G-C rich environment release thymidine resulting be a of RNA polymerase III transcription. a termination signal (Bogenhagen and regulator constitute transcription La protein/RNA polymerase Ill/transcription Brown, 1981). Purified Xenopus RNA polymerase Il has Key words: termination shown to be capable of recognizing this simple factor/transcription regulation/transcription been in the absence of auxiliary factors consensus sequence et 1983) and similar properties have been (Cozzarelli al., to calf thymus polymerase 11 (Watson et al., 1984). ascribed Introduction While auxiliary factors facilitating the last two aspects of III responsible for the synthesis of most have not been identified, several observations RNA polymerase is the process yet [see Geiduschek and that exist. First, the simple termination small RNAs in eukaryotic cells suggest they may for a review]. Cellular species its 3' flanking sequences are protected by Tocchini-Valentini (1988) signal and/or 4.5S and Alu transcripts; in transcription factor fractions (Klemenz include 5S rRNA, all tRNAs, 7S-L, components present EBERI and EBER2 are specified by class Ill Camier et 1985; Van Dyke and Roeder, VAI, VAII, et al., 1982; al., carried on viral Immediately following and some 3' deletion mutants compete poorly for the genes genomes. 1987) all RNA III transcripts are packaged (Schaack et al., 1983; Sharp et al., synthesis, polymerase transcription apparatus Wilson et Wu complexes with the 50-kd La protein Allison and Hall, 1985; al., 1985; into RNA-protein 1983; consensus and Mathews, 1982; Rinke and Steitz, 1982; et 1987). Second, seemingly equivalent (Francoeur al., in various environ- Steitz et behave differently sequence al., 1983). sequences et Koski et al., 1982; Initiation of III transcription is directed by ments (DeFranco al., 1982; polymerase and there are hints that terminators sequences within the coding region termed Mazabraud et 1987) multipartite al., et al., et al., 1980; exhibit tissue-specific expression (Mazabraud internal control regions (ICR) (Bogenhagen may while its can be influenced 1987). et 1980), efficiency Sakonju al., and 1989), reviewed in In the accompanying paper (Gottlieb Steitz, by flanking sequences (comprehensively the mammalian La protein may be an RNA and 1988; for apparent we suggested that Geiduschek Tocchini-Valentini, termination factor. This 50-kd see Carbon et 1987; Murphy et al., 1987; III transcription exceptions al., polymerase Steitz et and 1982; al., fractionation, initially phosphoprotein (Francoeur Mathews, Das et al., 1988). Chromatographic Pizer et Francoeur revealed that RNA et al., 1983; HeLa and extracts, 1983; Habets al., 1983; of Xenopus associates with all is of accurate initiation: a Hoch and 1984) III alone incapable et al., 1984; Billings, polymerase III the factor fractions and transcripts, selectively binding minimum of two additional (TFIIIB nascent polymerase stretch encoded the termination for of all class III 3' by signal are transcription genes uridylate TFIIIC) required Its Mathews and et These fractions Francoeur, 1984). presence et al., 1982). (Stefano, 1984; (Segall al., 1980; Shastry and in both in active complexes factors et can be detected transcription contain multiple general (Lassar al., 1983; may TFIIIC and (Gottlieb as evidenced the fact that TFIIIC crude and highly purified preparations Bieker et 1985) by al., ©IRL Press E.Gottlieb and J.A.Steitz ts ........ 1. A model for La action. Schematic representation of a stable transcription complex containing TFIIIB, TFIIIC and a class III gene (top), Fig. can be recognized by RNA polymerase III (middle). (If this were a 5S gene, the additional gene-specific initiation factor, TFIIIA, would which contact the internal control region.) Transcription in the absence of La yields a stalled transcription complex containing the 3' foreshortened directly while the of La produces an active transcription complex capable of multiple rounds of re-use (bottom right). transcript (bottom left), presence of each of the two complexes are summarized below. Class III genes are depicted as stippled boxes with Expected transcriptional properties nascent transcripts are wavy lines; released transcripts are represented as hairpin structures. Other components include: termination signals (TTTT); complex assigned various sizes .300 kd (Lassar et al., 1983; Ruet et al., 1984; Stillman et al., 1985; Yoshinaga et al., TFIIIC, a large globular which over the internal control region and may protect (or associate with something that protects) the terminator; TFIIIB, probably 1987), footprints a molecule of 60 kd or a dimer of two 60-kd subunits (Klekamp and Weil, 1986; P.Martin and R.Roeder, unpublished observations), which single does not to produce a footprint but stabilizes the complex; RNA polymerase III, a multisubunit enzyme of -700 kd (see Roeder, 1976); La, appear a 50-kd which can be found associated. with TFIIIC, with active transcription complexes, and with the 3'-uridylate sequence of full- phosphoprotein, length transcripts. La might induce a conformational change in the polymerase (compare complexes on the bottom left and right), effecting transcript completion. of HeLa-derived tran- 1987). La immunodepletion Steitz, Results extracts results in a dramatic diminution of tran- scription Model and the few synthesized in the scription ability; transcripts 1 model for RNA polymerase III tran- Figure depicts a of La have fewer residues at their 3' ends absence uridylate TFIIIB and TFIIIC associate with a class scription. Initially to those synthesized in the presence of La. These compared III forming a stable transcription complex (top) extents by the addition gene phenomena are reversed to different et al., 1982; Lassar et al., 1983; Schaack of La protein (Gottlieb and Steitz, (Bogenhagen biochemically purified et 1983). RNA polymerase III recognizes this complex 1989). al., We that elongation we a detailed model for La action at the and initiates synthesis (middle). propose Here, present of RNA chain until the polymerase halts termination and employ a variety of in vitro the initial proceeds signal We conclude that auxiliary to the termination signal. In the absence of La, to test this hypothesis. just prior approaches of III transcription a stalled transcription complex (bottom left) and a low level factors facilitate termination polymerase the La protein is a good candidate for of 3' truncated product result. In contrast, when La is present and that mammalian such a termination factor. In addition, we present evidence (the normal situation) the polymerase overcomes the pause right). The for a novel stalled transcription intermediate containing a and transcribes the 3' uridylate stretch (bottom and a 3' foreshortened transcript is released, freeing the polymerase molecule resulting full-length paused for multiple rounds of re-use. This model unreleased transcript. complex 852 La as a termination factor data accommodates all known polymerase III transcription and ascribes both of La's functional properties-its influence on the overall transcription level and the 3' foreshortened transcript termini found in its absence (Gottlieb and Steitz, 1989)-to a single molecular event. To test this model and to assess whether La plays a role in the termination process, we have performed a series of EBEt 2 experiments monitoring the behavior of the EBER2 gene (Rosa et al., 1981; Jat and Arrand, 1982). This prototypical class III gene contains an intragenic control region (Rosa et al., 1981) and requires TFIIIB and TFHIC in conjunction with RNA polymerase III for its transcription (P.Martin, unpublished observations). EBER2 is synthesized at high levels and is quantitatively precipitable by anti-La antibodies in vivo and in vitro (Rosa et al., 1981; E.Gottlieb and Bottor- ' 2 3 J.A.Steitz, unpublished observations). Moreover, the low Fig. 2. Gradient fractionation of EBER2 transcription complexes. level of EBER2 3' foreshortened product synthesized after Panel A: EBER2 DNA was incubated in a HeLa-derived S-l00 for La immunodepletion of a HeLa-derived S-100 extract is 30 min, and the resulting transcription complexes were sedimented from full-length EBER2 (Gottlieb and easily distinguished through a 20-45% glycerol gradient (17 fractions collected). Active We infer that our conclusions with EBER2 Steitz, 1989). transcription complexes were detected by their ability to support the be extended to other class III which behave synthesis of radiolabeled EBER2 following the addition of can genes [ca-32P]GTP. Products are displayed on a 6% polyacrylamide gel. The similarly in La-depleted extracts (Gottlieb and Steitz, 1989). direction of sedimentation is indicated, and the two complex populations are denoted and j3. Panel B: Transcription complexes Transcription complexes differ in the presence and were formed and gradient purified as in (A). Radiolabeled products absence of La synthesized by complexes in peak a (lane 1), peak ,B (lane 2), or by a mixture of used in lanes 1 and 2 (lane 3) after a the fractions 90-min formed on class III genes in vitro Transcription complexes incubation displayed on this 6% gel. Lanes were derived at 30°C are et Lassar et Schaack (Bogenhagen al., 1982; al., 1983; from a single exposure of a single gel. are stable rounds of et al., 1983) through multiple on in La with either of the two complex populations could not be transcription. They can be isolated glycerol gradients active form and contain < of the total detected not shown). Recall that most of the La in cells a synthetically 1% (data Jahn et exists in the form of RNPs (Stefano, 1984). In an alternative extract protein (Wingender et al., 1984; al., 1987), their and we the sedimentation profile of com- enabling us to analyze both catalytic properties approach, compared EBER2 complexes were assembled formed in anti-La versus non-immune depleted molecular constituents. plexes in an untreated HeLa derived extracts. La enhanced levels under transcription conditions Following immunodepletion, et added nucleotides. After of the slower S-100 (Weil al., 1979) containing sedimenting complex population (13) were seen each fraction was while levels of the more rapidly migrating population (a) glycerol gradient fractionation, ribonucleotides were diminished (not shown). This suggests that at least one supplemented with containing [a-32P]GTP for Since additional DNA difference between these two and incubated 90 min. template transcription complex popu- the lations is a deficiency of La protein in 1 relative to c. was not added after ultracentrifugation, resulting products of active reflect the sedimentation position transcription An alternative for the explanation synthesis of shortened to fractionation. complexes formed prior products by the 13 fractions (Figure 2A) is that a nuclease Two distinct populations of complexes can be visualized sediments in this of the To exclude this region gradient. (Figure 2A). One population (denoted a) migrated at 60S possibility, mixing experiments were performed (Figure 2B). and synthesized full-length EBER2. Surprisingly, a second of fractions from regions a (lane 1) and Aliquots gradient slower sedimenting fraction (denoted 13) produced 3' 13 were combined to the 90-min (lane 2) prior transcription of III foreshortened transcripts, characteristic polymerase reaction and the products analyzed as in Figure 2A. Figure in transcription the absence of La (Gottlieb and Steitz, 1989). 2B (lane 3) shows a typical result: about a 20-fold stimulation The levels of RNA synthesized by these two populations are of transcription of full-length product. In other trials, the number of active likely a reflection of the relative complexes lengths of the products reflected the length distribution seen RNA in each population: the ratio of labeled transcripts in the two component fractions. These results argue that the is constant in 13 is not due to a synthesized by these two populations upon appearance of shorter products region from extract in these fractions and further refractionation of a given extract, but varies nuclease specifically present rounds of to extract. suggest that La may recycle between transcription survives a fraction of the total (see Discussion). Only transcription activity observed K.H. fractionation previously by gradient (also loss of one The absence of the La affects the kinetics of protein Seifart, unpublished observations), suggesting To determine whether the transcript accumulation or more essential components. retained in the a To determine whether transcription complexes containing La is protein preferentially by complexes exhibit dissimilar was located in the ELISA or La also transcription region, La protein gradient by lacking might La a extract et kinetics, we subjected transcription (Weil al., (Thoen et al., 1980). Unfortunately, antigenicity peaked with levels in the central with non-immune or anti-La antibodies smeared downwards 1979) to depletion at the and top that of followed to a A of the by exposure protein preparation (Gottlieb portion gradient high enough any repeaking 853 E.Gottlieb and J.A.Steitz absence 3' only foreshortened molecules appear. Full-length products do not seem to be synthesized and subsequently an trimmed, although exceedingly rapid process cannot be excluded. the low Additionally, transcript levels observed in the absence of La cannot be wholly ascribed to transcript instability; there is no initial burst followed by subsequent decay. Instead transcripts build up late in the reaction (cf. lanes f and k, Figure 3B; see Discussion). Visualization of a stalled transcription intermediate The above u4m4 experiments indicate that transcription complexes with different physical and catalytic properties are formed in the or presence absence of the La protein. Our model makes the (Figure 1) further specific prediction that the 3' foreshortened transcripts synthesized in La-depleted extracts are a reflection of stalled transcription complexes in which the nascent transcript remains attached to the DNA template further rounds blocking of synthesis. To look for such a transcription intermediate, S-100 extracts were subjected to immunodepletion with non-immune or anti-La serum as above and then supplemented with labeled and unlabeled in Fig. 3. Time course of EBER2 transcription the presence and nucleotides and the EBER2 After a gene. 45-min incubation, absence of La. Parallel reactions were initiated non-immune by adding (panel A) or anti-La (panel B) depleted extracts (see Materials and reaction mixtures were fractionated on glycerol gradients methods) to pre-mixed transcription reactions the EBER2 including et (Wingender al., 1984; see Figure 2) and the radiolabeled template and [cx-32P]GTP Reactions were incubated at (0 min). 300C, transcripts in each fraction analyzed on denaturing with aliquots removed at time points and treated with immediately polyacrylamide gels (Figure 4). This experiment directly EDTA, SDS and proteinase K to halt synthesis. ethanol Following monitors the precipitation, the radiolabeled products were a sedimentation behavior of transcripts electrophoresed through single 6% polyacrylamide gel. Arrows denote the and 3' full-length synthesized before fractionation, whereas in Figure 2 the foreshortened EBER2 molecules. A small amount of the in sample behavior of preformed transcription complexes was lane was lost e, panel A, upon gel loading. examined. Here, if transcripts migrate into the gradient, they and Steitz, To initiate the are likely to be associated 1989). synthesis, resulting depleted with large complexes since EBER2 were added to assembled RNPs, consisting of the 172 supernatants previously reaction only 1 nucleotide RNA and mixtures salt excess the - containing optimal concentrations, 50-kd La protein, have a sedimentation value of 8S and amounts of and would remain at the ribonucleotides, [a-32P]GTP saturating top of the gradient. template. Radiolabeled at various times Figure 4, panel shows the transcripts appearing B, gradient containing products were resolved on a synthesized in the absence of La. All denaturing polyacrylamide gel (Figure transcripts are short We several 1 that and their level is 3). sampled early points (< h) anticipating low, as expected. While most transcripts these would be most to reveal differences between the remained at the top of the likely gradient (lanes 1-3), some reactions containing and La. Because migrated into the where lacking transcription region competent transcription complexes were not preformed prior to the addition of label, complexes sediment (lanes 6-14; compare with Figure 2A). we could examine both the transcriptional lag and later Agarose gel analysis of of each fraction aliquots confirmed reaction phases previously described for other class IH the presence of the DNA in this genes, template region (data not where similar kinetic were analyses performed (Birkenmeier shown). In in the contrast, gradient containing the labeled et Schaack et Bieker et al., 1978; al., 1983; al., 1985). products synthesized in the presence of La (panel A), all The rate of is transcript appearance strikingly different in of the EBER2 transcripts are full-length and appear at the the presence and absence of of the (Figure 3A) (Figure 3B) La top gradient (lanes 1-4). These are presumably La- protein. In the extract with immunodepleted non-immune containing ribonucleoprotein particles since EBERs serum (Figure 3A), newly synthesized products were synthesized in a La-containing extract are quantitatively La detected after 5 min of incubation (lane b). Product precipitable. Despite the fact that more incorporated counts accumulation lagged over the first 30 min but then increased were loaded onto gradient A than gradient B, no full-length with time. All steadily detectable transcripts, including those EBER2 transcripts are detected cosedimenting with the observed at the earliest time, were full-length. An extract transcription apparatus (panel A, lanes 6-14). Light that was not subjected to the immunodepletion protocol gave exposures of the third fraction from these gradients (panel a similar transcription pattern except that the total C, lanes 1, 2) show that the predominant transcription incorporated counts were slightly greater (not shown). products were full-length EBER2 and 3' foreshortened Therefore, the immunodepletion procedure itself does not EBER2 in the presence and absence of La, respectively. This alter the overall pattern of transcription. In contrast, in the length difference was confirmed by analyzing small samples absence of La (Figure 3B), the first observable transcripts of the total transcription reactions fractionated in A and B appeared 25 min after initiation of the reaction (lane f) and (data not shown). The second most abundant product all detectable EBER transcripts exhibited a rapid gel (denoted by *) has an EBER2 fingerprint; it could represent mobility. a multimer formed during the gradient procedure, since it Consistent with the proposed model (Figure 1) is the fact is absent when these reactions are directly gel fractionated. that even at the earliest time of transcript detection in La's It appears that La-immunodepletion has enabled us to 854 La as a termination factor oc E m Nf,n". Z -a r t 1:- - ; I !e-l- .:kw t.B&K-. .4.r. z& i w .'AAA, VI -.0- "., %-;%.I,.`,..:,!.-- *1.i ... i 4 - .. M c 1 2 To Fig. 4. Visualization of a stalled transcription complex. Immunodepleted extracts (see Materials and methods) were supplemented with [a-32P]GTP and [a-32P]UTP, unlabeled ribonucleotides, salts and the EBER2 template, incubated at 30°C for 45 and fractionated on parallel min, 20-45% glycerol gradients. Each resulting fraction was proteinase K treated, ethanol precipitated and displayed on a 6% polyacrylamide gel. Panels A and B: Migration behavior of radiolabeled transcription products from non-immune and anti-La depleted extracts respectively. Numbers correspond to gradient fractions and arrows denote - 1 the EBER2 transcripts. The direction of sedimentation is indicated (top bottom). Panel C, lanes and 2: Lighter exposures of lanes A3 and B3. The major transcription products include full-length and 3' foreshortened EBERs respectively. Products denoted by * have EBER2 fingerprints, are not observed when aliquots of each reaction are gel fractionated without prior gradient separation, and are likely to be EBER2 multimers formed during gradient fractionation. Most of the additional radioactivity in the top fractions of panels A and B represents overexposure of minor products. An additional radiolabeled RNA with very low gel mobility appears in the middle of both gradients, but we could never retrieve enough of this material to analyze it. visualize unreleased polymerase III transcription products If the latter were the case, the short RNAs presynthesized associated with stalled transcription complexes. The 3' might be susceptible to completion (Maderious and Chen- truncated RNAs sedimenting at the top of the -La gradient Kiang, 1984; Grayhack et al., 1985). Ideally, we would have (Figure 4B) could have dissociated from transcription liked to add La protein to gradient fractions containing these complexes because of stress during gradient fractionation nascent RNAs and analyzed their lengths. However, the low and/or slow but spontaneous release. The latter of in such possibility recovery transcripts retained after complexes would also explain the delayed transcript build-up that we use unfractionated seen at fractionation demanded tran- late reaction times in a La-depleted extract (Figure In for this 3B). scription extracts experiment instead. the of presence La, transcript release appears to be a SA indicates that the 3' foreshortened rapid, Figure transcripts active process coupled to completion; thus, we (Figure 4A) in a extract can be generated La-immunodepleted completed and others et (Ackerman al., 1983; Wingender et al., 1984; addition of a source of La Extracts by subsequent protein. Jahn et have been to al., 1987) unable detect this intermediate with non-immune or anti-La depleted (lanes 1,2) (lanes 3,4) in with complete transcription reactions. serum were the EBER2 supplemented nucleotides, and incubated for 30 and template, [oe-32P]GTP, min, long The stalled transcription can be rescued for labeled 3' foreshortened to complex enough transcripts appear (see The stalled visualized in 4B this short incubation we transcription complexes Figure Figure 3B). By selecting period of could either side of the to maximize the represent non-productive products hoped percentage transcripts remaining or synthetic process bone intermediates. associated with the stalled fide transcription transcription complexes (Figure 855 E.Gottlieb and J.A.Steitz pletion. It is highly unlikely that the full-length molecules detected after the 5-min chase of the stalled transcription complexes result from de novo synthesis since this time for period is too short substantial amounts of new molecules to appear (Figure 3A, lane 2) and the few made would have D Lsi!i 200-fold a specific activity less than the pre-existing 00 3 1~ _7 < these do not 4.. 7 _~ transcripts. [Note that complexes sequester ,_s J ., ,1 r- z < « of et appreciable amounts nucleotides (Wingender al., exhibits the same 1984).] Transcript completion o-amanitin sensitivity as synthesis by RNA polymerase III: inclusion of 300 a-amanitin (Figure 5B) during the 5-min chase itg/ml in period abolishes the shift transcript mobility (lane 3 versus 4). Similar results are seen following chase periods of 15 (lane 5) and 60 min (data not shown). Thus, our results indicate that the discrete 3' foreshortened RNA in La extracts is a bone synthesized depleted fide the of a transcription intermediate, probably product paused polymerase. Transcript completion appears to require active RNA in with the La polymerase Im conjunction protein preparation. Although other data (Gottlieb and Steitz, 1987; this manuscript) indicate that La is normally associated with the the chase data that La can transcription complex, suggest functionally interact even after initiation of This synthesis. conclusion does not contradict the results in 2 since Figure the concentration of La in the tran- gradient-fractionated scription complexes was calculated to be about four orders of magnitude less than in this experiment. Chelating La with RNA also yields stalled transcription complexes termination factor To ask further whether La itself is a that we facilitates transcript completion and release, exploited the protein's documented to bind the U-rich 3' end ability 2 of polymerase III reaction was FBER r# transcripts. transcription EBER2 pre-incubated with excess unlabeled (Figure 6D, lane EBFR labeled EBER) in the hope that incorporation into ribo- nucleoprotein particles would render the La protein to La unavailable the transcription apparatus. is the only protein known to bind the EBERs (J.Stefano, unpublished Glickman et observations; al., 1988). Thus, this depletion is more selective than use of anti-La antibodies since strategy La under that immunodepletion conditions preserve also removes several less abundant transcription activity proteins by virtue of their association with La and/or La- bound RNAs. As a an reaction was control, equivalent pre- Fig. 5. Stalled can be chased to transcription complexes completion. for 30 at Radiolabeled EBER2 transcripts were synthesized min 30°C incubated with an equimolar amount of unlabeled EBER2 with non-immune A in reactions containing extracts depleted (lanes from which the U-rich La binding site had been specifically B or anti-La A B antibodies. 1-2, 1-2) (Lanes 3-4, 3-5) Samples removed. These 3'/EBERs, generated by oligonucleotide- then received buffer Al and Bi and prewarmed transcription (lanes 3, directed RNase H cleavage of EBER2 are shown in Figure or S-100 into buffer A2 3) prewarmed dialyzed transcription (lanes and 4 and either a 200-fold excess of unlabeled 6D. After pre-incubation, the reactions were supplemented 4; B2, 5) containing nucleotides (A) or 300 ig/ml a-amanitin (B). Incubation was for 5 min with the EBER2 gene and both labeled and unlabeled in all cases lane where the chase was extended to a total except B5, nucleotides, incubated for 45 min and fractionated in parallel of 15 min. Resulting radiolabeled products were proteinase K treated on glycerol gradients as in Figure 4. and separated on denaturing 6% polyacrylamide gels. Figure 6B reveals that removal of La by RNA chelation 4B). Then, following the simultaneous addition of a 200-fold again results in visualization of stalled transcription excess of unlabeled ribonucleotides (to dilute the labeled intermediates: a fraction of the radiolabeled transcripts and either 1 precursors) transcription buffer (lanes and 3) synthesized in the EBER2-treated extract migrates into the or a La protein dialyzed into transcription buffer (lanes 2 central region of the glycerol gradient (lanes 4- The 1). and 4), incubation continued for . 5 additional minutes. The of mobility full-length EBER2 transcripts generated in a shift of the presynthesized 3' foreshortened species (lane 3) parallel untreated reaction and loaded onto the far right lane to full-length (lane 4) indicates that at least some of the RNA of the gel demonstrates that these rapidly sedimenting RNAs components of stalled complexes can be chased to com- are short, as expected for paused transcripts. This effect is 856 La as a termination factor RV + EBER' 3' :BER RNA Lii ULJ Fig. 6. A stalled transcription complex can be generated by chelation of the La protein with RNA. Panels A-C: Gradient profiles showing the migration behavior of radiolabeled EBER2 in synthesized extracts preincubated with excess unlabeled EBER2 from which the La binding site had been removed with EBER2 (3'AEBER, A), full-length (B), or no added RNA (C). Equivalent c.p.m. were loaded onto each The direction gradient. of sedimentation in the to glycerol gradient (left right) is indicated (top-bottom); the far right-hand lane of panels A and B contain full-length EBER2 as a marker. In the EBER2 into the transcript panel B, transcript running gradient exhibits the shorter length characteristic of in transcription the absence of La Gottlieb and (see Steitz, 1989), compared to the full-length transcript in panels A and C. Panel D: An ethidium bromide stained of the unlabeled EBER2 of EBER2 gel profile full-length and RNA that has had its site removed RNase H La-binding by oligonucleotide-directed mediated cleavage (3'zEBER). produced by adding EBER2 at an RNA: La protein pretation of these ratio results is that the factor whose activity we estimated to be one-tenth that found in the have been monitoring is the La protein itself. cell and without appreciable diminution of transcription. (In experiments not shown, depression of transcript levels was observed at higher Discussion EBER2 concentrations.) In labeled EBER2 made contrast, Existence of a transcription termination factor for in the extract pre-incubated with 3'AEBERs or (Figure 6A) RNA polymerase I1 in an untreated extract are full and do (Figure 6C) length Two important conclusions emerge from our results: first, not into the significantly migrate gradient. Simply being termination by RNA III does not cause a to move polymerase seems to require incomplete transcription product auxiliary factor(s) and second, the 50-kd mammalian La into the gradient: while RNAs of other are lengths present their 3' terminal Us protein appears to be such a transcription termination factor. in the profiles, those only lacking just Ultimate with confirmation of the latter conclusion will require cosediment transcription complexes (Figure 6B). the addition of in which or 3' biologically active La protein synthesized Moreover, control experiments full-length from a cDNA clone to a reconstituted EBER2 were mixed with unlabeled transcription system foreshortened transcripts from more highly than are now reactions to fractionation revealed that purified components transcription prior available. As discussed in Gottlieb and Steitz neither sediments into the not (1989), detectably gradient (data current RNA must have been protocols for reconstitution of transcription utilize a 3' foreshortened shown). Hence, fractions still is to with containing La. Yet, the fact that stalled in the reaction if it the synthesized migrate can be methods The inter- complexes generated by two independent fractionation. transcription complex upon simplest 857 and J.A.Steitz E.Gottlieb it should involve at least three distinct of La depletion (Figures 4 and 6) implicates La itself as the prokaryotic process, transcript release and termination factor required for their resolution. If another events: polymerase pausing, With and Bear, 1983). molecule were responsible, it would have to meet three polymerase dissociation (Platt with La, allowing complete coli RNA polymerase, pausing is the first step criteria: (i) very tight association Escherichia do not result in anti-La antibodies (Figure 4); (ii) process; while all pauses immunodepletion by in the termination is an for of the U-rich 3' ends of polymerase III pausing obligate prerequisite specific recognition termination, et 1985; 6); and (iii) copurification with La through (von Hippel et al., 1984; Grayhack al., transcripts (Figure termination two is La first fractionates as an RNP and later The order of the remaining processes six steps in which Platt, 1986). not known. as a protein (Gottlieb and Steitz, 1989). of our model (Figure 1) are supported of stalled transcription complexes in the absence Several predictions Detection here. Distinct transcription that La is not required for polymerase pausing. by the analyses presented of La argues to be assembled in the presence and it seems reasonable that RNA polymerase III, like complexes appear Rather, RNA (Grayhack et al., 1985; J.Roberts, absence of La (Figure 2). A fraction of the 3' foreshortened E. coli polymerase the inherent capacity to transcripts generated in La's absence appear to be retained personal communication), possesses complexes (Figure 4-6), providing Parameters inducing pausing by RNA polymerases in stalled transcription pause. that the 3' foreshortened RNA arises via even for the E. coli enzyme. Pausing could the best indication are unknown, of a helical stem between infer that this novel transcription the formation a transcription event. We result from near the 5' and 3' ends minimally contains TFIIIB, TFIIIC (since both present intermediate complementary sequences III factors required for initiation), RNA nascent RNA polymerase transcripts. are non-cycling of most the structure of the DNA the DNA template and the nascent RNA it may be induced by polymerase III, Alternatively, that III termination It would therefore be distinct from previously It is polymerase transcript. template. intriguing residues in the metastable, stable and pre-initiation transcription include 24 adenine coding identified signals (which that are believed to [see Bieker et al. (1985) for distinctions] in at coincide with DNA sequences complexes strand) it would represent a step immediately induce bending (Koo et al., 1986). least two regards: after and it would contain the incomplete mechanisms for La protein action polymerase preceding termination Several our data. Since La does not La to be an integral component of the are compatible with transcript. appears pausing but is in RNA normal transcription complex not only because its presence to bind DNA strongly present appear Gottlieb in gradient purified complexes can be inferred (Figure 2) III transcription complexes (Figure 2; polymerase and in and but also because DNA carrying class III genes can be and 1987) TFIIIC preparations (Gottlieb Steitz, interacts with other protein selectively immunoprecipitated from synthetically active Steitz, 1987), it most likely and 1987). induce conformational extracts with anti-La antibodies (Gottlieb Steitz, components and could thereby be able to interact to Figure 5 suggests that La may also in the transcription complex causing polymerase changes even after This be reflected in the functionally with the transcription complex effect transcript completion. may between initiation of synthesis. substantial S-value difference transcription that short versus full-length transcripts The of the simplest form of the model (as presented complexes synthesize aspect be accounted for solely by the in that may require revision is the prediction that (Figure 2A), which cannot Figure 1) that the mere one round of synthesis should be detected when absence of a 50-kd protein. [Note presence only formation in a HeLa extract transcript completion and release are not La facilitated. In of nucleotides during complex level and length are increases the mean sedimentation rate other words, we observe that transcript significantly we do see a low level of not obligatorily coupled: while et al., 1984) and that E. coli transcription (Wingender of this level does not remain transcript in the absence La, containing nascent RNAs can exist in alternate complexes in 3. This unexpected observation Once the constant over time Figure conformations (Straney and Crothers, 1985).] in at least two ways. Conceivably, has been synthesized, the could be explained transcript's U-rich 3' terminus assembly could be altered by removal be incompatible transcription complex transcription complex configuration may of La-associated transcription initiation factors (e.g. TFIIIC) retention. Alternatively, La may be correctly with transcript upon La immunodepletion, effecting a decrease in transcript complex to bind the newly positioned in the transcription level not directly related to La. This explanation could be RNA 3' terminus when it becomes exposed synthesized viewed as consistent with our reconstitution data (Gottlieb of - heteroduplex instability (Martin and because dA rU and 1989) and with the absence of significantly to the E. coli termination factor rho, Steitz, Tinoco, 1980). Similar is removed by RNA depressed transcription levels when La the nascent transcript (Chen et al., 1986; Bren- upon binding slow transcription complex chelation (Figure 6). However, nan et al., 1987), La might melt the remaining DNA -RNA is entirely responsible since transcript accumu- assembly not using an ATP-dependent helicase activity (Brennan duplex lation in the presence and absence of La (Figure 3) differ et al., 1987). This scenario is supported by recent sequencing in the rate of product accumulation as well as in the dur- of a Xenopus La cDNA clone, which has revealed a potential ation of the transcriptional lag. Alternatively, in a scenario NTP and S.Clarkson, personal binding site (D.Scherly the data in Figures 4 and 6, a single round compatible with Finally, a more indirect role for La in communication). of may simply not be detectable; later transcript synthesis transcript completion and release can also be envisioned. build up in the absence of La may result from slow, spon- pausing at the U tract may result from Here, polymerase taneous release of truncated products. an increased rate of depolymerization versus polymerization by the enzyme (Kassavetis et al., 1986). La could shift the La and transcription termination by sequestering the longer transcripts, driving equilibrium Assuming that eukaryotic transcription termination (like the reaction towards polymerization. initiation) bears some similarity to the comparable additional work is required to determine the Clearly, 858 La as a termination factor details of La's contribution to the termination of RNA syn- substrate RNAs were La be present, would sequestered in thesis by RNA polymerase III. It should be realized that the RNPs, unavailable to further of catalyze rounds transcription. possible mechanisms for La action outlined above are not Such a scenario is reminiscent of the 5S-specific regulatory necessarily mutually exclusive. Further, La's participation for TFIIIA in loop proposed the amphibian oocyte (Pelham in the termination process does not eliminate a subsequent and Brown, 1980). La's relative affinity for various role of the protein in 3' end protection of resulting transcripts and transcription complexes formed on different transcripts; La's unique RNA binding ability may couple class 11 to its state could then genes coupled phosphorylation RNA polymerase transcription III termination to RNA be used to fine-tune this regulation. packaging. Finally, future efforts should also resolve the issue of whether (Bieker al., 1985; et Setzer and Brown, Materials and methods 1985; Carey et al., 1986) or not (Jahn et al., 1987) RNA polymerase III dissociates after each round of synthesis and antibodies and Cells, enzymes HeLa and cells were maintained at in a environment Raji 37°C 5% Co2 (if so) whether dissociation occurs concomitant with or x were cultured in at -7 x 105 or 5 105 cells/ml Cells respectively. following transcript release. in in RPMI 1640 spinners (HeLa) or suspension (Raji) (GIBCO) sup- In a detailed study, Cozzarelli et al. (1983) previously con- with 10% heat-inactivated fetal calf serum 60 plemented (GIBCO), Ag/ml cluded that Xenopus RNA polymerase III alone is capable and 300 penicillin, 100 streptomycin glutamine. Ag/ml Atg/ml Autoimmune sera J.Hardin and Yale of transcription termination. These authors analyzed their patient (kindly provided by J.Craft, University) were selected for their titer and relative high monospecificity 5S transcription products by nuclease mapping, which T, as three determined by independent assays: immunofluorescence, probably was not sensitive enough to discriminate the 1-2 of in vivo RNAs from a immunoprecipitation [32P]orthophosphate-labeled uridylate residue difference between transcripts 5S produced HeLa whole cell Matter et and extract (Hendrick et al., 1981; al., 1982) in the presence and absence of La (Gottlieb and Steitz, Western immunoblots Yen and 1989). (Towbin et al., 1979; Webster, 1981; Mimori Non-immune sera were donated et al., 1984). by healthy laboratory We suggest that the RNAs Cozzarelli et al. analyzed may were obtained ammonium sulfate personnel. IgG preparations by precipitation have been components of stalled transcription complexes. and contained - 17 Anti-La monoclonal and mouse mg/ml protein. IgGs Their data would then strongly argue that RNA polymerase control kind of D.Williams and Smith IgG (the gift P.Venables; et al., 1985) III alone is capable of the first step in termination- were in A - selected grown the ascites of BALB/c mice, protein Sepharose and were - 0.5 recognition of and pausing at the termination signal. Other mg/mi protein. aureus cs-amanitin Lyophilized protein A-bearing Staphylococcus cells, ways of resolving the apparent contradictions between their RNA markers 16S and 23S were from and (MS2, RNAs) purchased and our conclusions are: (i) there might exist species-specific K was obtained from Beckman. Boehringer and Mannheim and proteinase differences between Xenopus and HeLa RNA polymerase RNase H was from Pharmacia. purchased III, although the observations of Watson et al. (1984) on and reactions Extracts, antibody depletions calf thymus RNA polymerase transcription III can likewise be explained extracts were from HeLa Soluble transcription (S-100) prepared log phase by pausing at the termination signal; (ii) polymerase IH alone cells frozen in and stored (Weil et al., 1979), aliquoted, liquid nitrogen might be able to recognize the termination signal and at were at in buffer -70°C. Antibody depletions performed 0°C transcription complete transcription on a naked DNA template (their mM 5 mM 15 mM 0.5 mM as (70 KCI, MgCl2, Hepes pH 7.9, DTT) described and Prior to extracts were experiments) but require additional factors when (Gottlieb Steitz, 1989). confronted transcription, with 0.5 mM each CTP and 0.025 mM supplemented ATP, UTP, GTP, with a protein-covered transcription complex (our experi- and unless otherwise indicated. Each was [a-32P]GTP template, optimized ments); or (iii) Xenopus RNA polymerase III purified to 90% for DNA and concentrations. EBER2 MgCI2 gene containing plasmids, homogeneity might retain some La protein, even though La or R1J et were pJJJ2 (Jat and Arrand, 1982) pEBV (Rosa al., 1981), supplied is not detectable in purified HeLa RNA at 10-40 at polymerase III as supercoiled templates Transcription proceeded 30°C Ag/ml. for indicated. Reactions 25-50 were then the times preparations either by immunochemical (typically Ll) criteria (Gottlieb and with 20 carrier 0.4% SDS and 20 supplemented 4tg yeast RNA, yg proteinase Steitz, 1987) or by activity (W.K.Hoeffler and E.Gottlieb, at for 20-25 made 0.25 M in ammonium acetate K, heated 65°C min, unpublished observations). Radiolabeled RNA were fractionated on and ethanol precipitated. products or M x TBE M 0.5 M 5%, 6% 10% urea/0.5 polyacrylamide/7 (0.5 Tris, sodium 2 mM EDTA which were dried and borate, pH 8.3) gels, at -700C. autoradiographed A regulatory role for La? Figure 1 suggests that La's role in termination may have the a direct affect on initiation by resetting transcription Glycerol gradient analyses of re-use. La's complex for its documented multiple rounds fractionation was a modification of the Gradient performed by procedure et Stable were formed presence might therefore amplify the transcriptional output. of al. Wingender (1984). transcription complexes 60 10 of EBER2 in 150-1l reactions and initiation containing 11 S-100, tsg/ml template, Interaction between transcription termination 1.75 mM sodium GTP and 0.5 mM each UTP, ATP, CTP, phosphate (pH would not be unique to the RNA polymerase III system. of the buffer 15 mM NaCl and the other 6.3), components transcription RNA polymerase I termination affects initiation at the for 30 reactions were chilled listed above. After incubation at min, 30°C adjacent rRNA transcription unit (Grummt et al., and on 20-45% mM 1986; 4.8-ml (v/v) layered glycerol gradients containing 15 mM 0.5 mM 1.5 mM 5 mM DTT, and KCI, MgCl2, Hepes pH 7.9, Henderson and Sollner-Webb, 1986; McStay Reeder, in an 1 rotor at 50K for EDTA. Gradients were SW50. centrifuged r.p.m. formation of U-RNAs 1986) and the 3' end signals the of each 2.2 h at To fraction, 0°C. assay transcription ability resulting the (synthesized by RNA polymerase II) overlap promoter assembled 100 tl of the 0.5 reactions were fraction, mM containing 200-Al et 20 (Hernandez and Weiner, 1986; Neuman de Vegvar al., and 0.025 mM each CTP GTP, ATP, UTP, ACi [a-32P]GTP 70 mM 15 mM 7.9, 5 mM and 1986; Ciliberto et al., 1986). KCI, (410 Ci/mmol), Hepes pH MgCl2 at and K mM DTT. a incubation 0.5 90-min Following 30°C proteinase of is Possible feedback regulation transcription suggested ethanol and resolved on were treatment, transcripts precipitated for by the La is both a factor fact that required transcript wt standards fractionated on a Mol. polyacrylamide gels. parallel gradient completion and binds the resulting products: La could RNA RNA HeLa included E.coti ribosomal (16S, 23S), (28S), MS2 of all RNA III and R17 modulate the ribosomal subunits therefore synthesis polymerase (40S, 60S) (80S). phage were after the collection of If a number of Mixing experiments performed immediately in the mammalian cell. transcripts large 859 E.Gottlieb and J.A.Steitz gradient fractions. A volume of 25 Id from a fraction judged to with the of were synthesize in after 30 exception [a-32P]GTP processed parallel; min full-length RNA and from one judged to of synthesize 3' foreshortened RNA was transcription, added with the other [as-32PJGTP components. were mixed with an equal volume (50 reaction ad) of mixture (1 mM ATP, CTP and 0.05 mM GTP, ,Ci UTP, 10 [a-32P]GTP, 70 mM KCl, 15 mM Hepes pH 7.9, 5 mM MgCI2, 0.5 mM DTT), incubated at 30°C for 2 h Acknowledgements and processed as above. To control for dilution effects, 25 of each fraction Jil We are for advice was also mixed with 25 and from Alan of transcription buffer and the grateful helpful suggestions Jeff reaction treated Weiner, /d Nouria Hernandez, Elisabetta Kim as above; no alterations due to dilution were detected. Roberts, Ullu, Black, La Doug protein Mowry, Daniela concentration Parker, Rhodes, Nelson and David in each fraction was determined by ELISA (Thoen et Kathy Hillary Setzer; David al., 1980) Brow one mechanism we discuss on 100 pl of for La action each gradient fraction. suggested termina- during tion. Materials were John Joe kindly supplied by Hardin, Craft, David Patrick Venables John In vitro kinetics transcript Williams, (antibodies), Arrand and John of accumulation (plasmid) We thank Therese HeLa Yario for technical S-100 extracts were immunodepleted with Flory (oligonucleotides). assistance non-immune or anti-La and Stevens for Research funds were antibodies as described (Gottlieb and Steitz, 1989). Parallel Lynda typing. provided a transcription by grant from the National Institutes of Health reactions (1 ml each) were assembled by adding the (CA depleted supernatant 16038). to pre-compiled a mixture of salts, unlabeled ribonucleotides, [ae-32P]GTP and the EBER2 template. Incubation proceeded at 30°C over a 3 h period. References Aliquots of 50 were removed ul from each reaction at the times indicated, stopped by EDTA and SDS addition, and proteinase K treated. The ethanol and Ackerman,S., Bunick,D., Zandomeni,R. Weinmann,R. Nucleic (1983) precipitated radiolabeled products were fractionated. gel Acids 6041-6064. Res., 11, and Allison,D.S. EMBO Hall,B.D. (1985) J., 2657-2664. 4, Visualization of stalled transcription complexes and Bieker,J.J., Martin,P.L. Roeder,R.G. (1985) Cell, 119-127. 40, Transcription reactions of 100 containing extract ul (immunodepleted with Birkenmeier,E.H., Brown,D.D. and Jordan,E. (1978) Cell, 15, 1077-1086. non-immune or anti-La serum), salts, ribonucleotides, EBER2 template and and Bogenhagen,D.F. Brown,D.D. (1981) Cell, 24, 261-270. 60 each of and were [k-32P]GTP [a-32P]UTP incubated for 45 min and /Ci Bogenhagen,D.F., Sakonju,S. Brown,D.D. (1980) Cell, 19, 27-35. at 30°C and diluted 2-fold with transcription buffer. A volume of 150 pl Bogenhagen,D.F., Wormington,W.M. and Brown,D.D. (1982) Cell, 28, from each reaction was layered on a 4.8-mI, 20-45% glycerol gradient and 413 -421. centrifuged in an SW50. 1 rotor at 50K r.p.m. for 2-2.2 h. Gradients were Brennan,C.A,, Dombroski,A.J. and Platt,T. (1987) Cell, 48, 945-952. fractionated, diluted 2-fold with transcription buffer and proteinase K treated. Camier,S., and Gabrielsen,O., Baker,R. Sentenac,A. (1985) EMBO J., 4, Radiolabeled products were ethanol precipitated and displayed on denaturing 491-500. 6% polyacrylamide gels. To assess the effectiveness of La depletion, the Carbon,P., Murgo,S., Ebel,J.-P., Krol,A., Tebb,G. and Mattaj,I.W. (1987) remaining 50 pl of each transcription reaction, which was not gradient 71-79. Cell, 51, fractionated, was proteinase K treated and gel fractionated. and Carey,M.F., Gerrard,S.P. Cozzarelli,N.R. (1986) J. Biol. Chem., 261, Alternatively, extracts were preincubated on ice for 20 min with 12 ug/ml 4309 -4317. EBER2 or EBER2 devoid of its La protein binding site (3'AEBER). Extracts Chen,C.-Y.A., and Galluppi,G.R. Richardson,J.P. (1986) Cell, 46, were then supplemented for transcription and processed as above. EBER2 1023-1028. for this experiment was prepared by anti-La immunoprecipitation from Raji Ciliberto,G., Dathan,N., Frank,R., Philipson,L. and Mattaj,I.W. (1986) cells (Matter et al., 1982) followed by purification on a 5% poly- EMBO 2931-2937. J., 5, acrylamide/7 M urea/1 x TBE gel, visualized with ethidium bromide, ex- Cozzarelli,N.R., Gerrard,S.P., Schlissel,M., and Brown,D.D. Bogenhagen, cised and eluted in 0.5 M sodium acetate, 10 mM EDTA, 10 mM Tris-HCl D.F. 829-835. (1983) Cell, 34, pH 7.5, 0.1% SDS, PCA extracted, ether washed and ethanol precipitated. Das,G., Henning,D., Wright,D. and Reddy,R. (1988) EMBO J., 7, Although the vast majority of EBER2 in the cell is complexed with the La 503 -512. protein, the use of anti-La selected RNA ensured that the isolated RNA DeFranco,D., Burke,K.B., Hayashi,S., Tener,G.M., Miller,R.C. and was intact and capable of binding La; non-immunoselected RNA was, Nucleic Soll,D. (1982) Acids Res., 10, 5779-5808. however, successfully employed as well. To generate EBER2 lacking the Engelke,D.R., Ng,S.-Y., Shastry,B.S. and Roeder,R.G. (1980) Cell, 19, La binding site, gel-purified EBER2 was mixed in water with a 220-fold 717-728. molar excess of a 20 mer oligonucleotide (5'-AAAAATAGCGGACAAGC- Francoeur,A.M. and Mathews,M.B. (1982) Proc. Natl. Acad. Sci. USA, CGA-3') complementary to its 3' end, heated to 80°C for 5 min to facilitate 6772-6776. 79, hybridization, equilibrated to 30°C for 15 min, made 40 mM Tris-HCI Francoeur,A.M., Chan,E.K.L., Garrels,J.I. and Mathews,M.B. (1984) Mol. pH 8.0, 5 mM MgCI2, 1 mM DTT, and incubated with RNase H Cell. 586-590. Biol., 5, (20 U/pg RNA) at 300C. After 60 min, additional RNase H (10 U/pg RNA) Fuhrman,S.A., and Engelke,D.R. Geiduschek,E.P. (1984) J. Biol. Chem., was added and incubation continued for 30 to min facilitate quantitative 259, 1934-1943. cleavage. The resulting 3'-deleted EBER2 was gel purified, eluted, PCA Geiduschek,E.P. and Tocchini-Valentini,G.P. (1988) Annu. Rev. Biochem. extracted, ether washed and ethanol precipitated. A shorter oligonucleotide 873-914. 57, (5'-AAAAATAGC-3'), complementary to the terminal 9 residues of EBER2, Ginsberg,A.M., and King,B.O. Roeder,R.G. (1984) Cell, 479 -489. 39, was also tried, but only 20% cleavage was obtained at a 500-fold molar Glickman,J.N., Howe,J.G. and Steitz,J.A. (1988) J. Virol., 62, 902-911. excess; presumably the high proportion of A-U base pairs resulted in and Gottlieb,E. Steitz,J.A. (1987) In Reznikoff,W.S., Burgess,R.B., instability of the hybrid. The concentrations of all RNAs were determined Dahlberg,J.E., Gross,C.A., Record,M.T. and Wickens,M.P. (eds), RNA by both UV absorption and intensity of ethidium bromide staining compared and the Polymerase Regulation of Transcription. Elsevier, New York, to RNA standards. The addition of synthetic polymers and had (U5 A5) 465 -468. pp. no effect (either transcriptionally or post-transcriptionally) in our system. and Gottlieb,E. Steitz,J.A. (1989) EMBO J., 8, 841-850. Grayhack,E.J., Yang,X., Lau,L.F. and Roberts,J.W. (1985) Cell, 42, Chase 259-269. Non-immune and anti-La immunodepleted extracts were supplemented with Grummt,I., Kuhn,A., Bartsch,I. and Rosenbauer,H. (1986) Cell, 47, the EBER2 template, salts, nucleotides pM (600 each ATP, CTP, UTP, 901 -911. 25 GTP) and ptM Transcription proceeded at 400,uCi/ml [ca-32P]GTP. Habets,W.J., den Brok,J.H., Boerbooms,A.M.T., van de Putte,L.B.A. and 30°C for 30 in 25 min reactions, which were then supplemented with van Al Venrooij,W.J. (1983) EMBO J., 2, 1625-1632. either 12.5 transcription ul buffer (70 mM KCl, 5 mM MgCl2, 15 mM and Henderson,S. Sollner-Webb,B. (1986) Cell, 47, 891-900. Hepes pH 7.9, 0.5 mM DTT) or S-100 dialyzed into transcription buffer. Hendrick,J.P., Wolin,S.L., Rinke,J., Lerner,M.R. and Steitz,J.A. (1981) Each minxture (prewarmed to 300C) contained unlabeled ribonucleotides such Mol. Cell. Biol., 1, 1138-1149. that the final 37.5 p,l reaction was 600 pM each in ATP, CTP, and UTP Hernandez,N. and Weiner,A.M. (1986) Cell, 47, 249-258. and 5 mM in GTP. Some reactions were supplemented with 300 pg/ml Hoch,S.O. and Billings,P.B. (1984) J. Immunol., 133, 1397-1403. x-amanitin instead of excess GTP. unlaoeled Incubation proceeded at 30°C Honda,B.M. and Roeder,R.G. (1980) Cell, 22, 119-126. for an additional chase period and was then halted by the addition of SDS, Jahn,D., and Seifart,K.H. Wingender,E. (1987) J. Mol. Biol., 193, EDTA and K. To assess the level of proteinase radiolabeled 303 -313. products the chase synthesized reactions during period, all containing P. and components Jat, Arrand,J.R. (1982) Nucleic Acids Res., 10, 3407-3425. 860 La as a termination factor Kassavetis,G.. Zentner,P.G. and Geiduschek,E.P. (1986) J. Biol. Chem., Wingender,E., Jahn,D. and Seifart,K.H. (1986) J. Biol. Chem., 261, 261, 14256-14265. 1409-1413. Klekamp,M.S. and Weil,P.A. (1986) J. Biol. Chem., 261, 2819-2827. Wu,G.-J., Railey,J.F. and Cannon,R.E. (1987) J. Mol. Biol., 194, and Geiduschek,E.P. (1982) Proc. Acad. Klemenz,R., Stillman,D.J. Natl. 423-442. Sci. USA, 79, and 6191-6195. Yen,T.S. Webster,R.E. (1981) J. Biol. Chem., 256, 11259-11265. Koo,H.-S., Wu,H.-M. and Crothers,D.M. (1986) Nature, 320, 501-506. Yoshinaga,S.K., Boulanger,P.A. and Berk,A.J. (1987) Proc. Natl. Acad. Wormington,M. and Hall,B.D. (1982) Nucleic Sci. Koski,R.A., Allison,D.S., USA, 84, 3585-3589. Acids Res., 10, 8127-8143. Martin,P.L. and Roeder,R.G. (1983) Science, 222, 740-748. Received on on Lassar,A.B., September 27, 1988; revised December 7, 1988 McStay,B. and Reeder,R.H. (1986) Cell, 47, 913-920. Maderious,A. and Chen-Kiang,S. (1984) Proc. Natl. Acad. Sci. USA, 81, 5931 -5935. Martin,F.H. and Tinoco,l.,Jr (1980) Nucleic Acids Res., 8, 2295-2299. Mathews,M.B. and (1984) Mol. Cell. Biol., 4, 1134-1140. Francoeur,A.M. Matter,L., Schopfer,K., Wilhelm,J.A., Nyffenegger,T., Parisot,R.F. and DeRobertis,E.M. (1982) Arth. Rheum., 25, 1278-1283. Mazabraud,A., Scherly,D., Muller,F., Rungger,D. and Clarkson,S.G. 197, 1-11. (1987)J. Mol. Biol., Mimori,T., Hinterberger,M., Pettersson,I. and Steitz,J.A. (1984) J. Biol. 259, 560-565. Chemii., Murphy,S., Di Liegro,C. and Melli,M. (1987) Cell, 51, 81-87. Neuman de Vegvar,H.E., Lund,E. and Dahlberg,J.E. (1986) Cell, 47, 259-266. Pelham,H. and Brown,D.D. (1980) Proc. Natl. Acad. Sci. USA, 77, 4170-4174. Pizer,L.I., Deng,J.-S., Stenberg,R.M. and Tan,E.M. (1983) Mol. Cell. Biol., 3, 1235-1245. Platt,T. (1986) Annu. Rev. Biochem., 55, 339-372. and Bear,D.G. (1983) In Beckwith,J., Davies,J. and Platt,T. Gallant,J.A. (eds), Gene Function in Cold Spring Procarvotes. Harbor Laboratory Press, Cold Spring Harbor, NY, pp. 123-161. Rinke,J. and Steitz,J.A. (1982) Cell, 29, 149-159. Roeder,R.G. (1976) In Losick,R. and Chamberlin,M. (eds), RNA Cold Spring Harbor Laboratory Press, Cold Spring Harbor, Polymerase. NY, pp. 285-329. Rosa,M.D., Gottlieb,E., Lerner,M.R. and Steitz,J.A. (1981) Mol. Cell. Biol., 1, 785-796. Ruet,A., Camier,S., Smagowicz,W., Sentenac,A. and Fromageot,P. (1984) EMBO J., 3, 343-350. Sakonju,S., Bogenhagen,D.F. and Brown,D.D. (1980) Cell, 19, 13-25. Sakonju,S., Brown,D.D., Engelke,D., Ng,S.-Y., Shastry,B.S. and Roeder,R.G. (1981) Cell, 23, 665-669. Schaack,J., Sharp,S., Dingermann,T. and So1l,D. (1983) J. Biol. Chem., 258, 2447-2453. Segall,J., Matsui,T. and Roeder,R.G. (1980) J. Biol. Chem., 255, 11986-11991. Setzer,D.R. and Brown,D.D. (1985) J. Biol. Chem., 260, 2483-2492. Dingermann,T., Schaack,J., DeFranco,D. and Soll,D. (1983) Sharp,S., J. Biol. Chem., 258, 2440-2446. Shastry,B.S., Ng,S.-Y. and Roeder,R.G. (1982) J. Biol. Chem., 257, 12979- 12986. Smith,R.P., Williams,D.G., Venables,P.J.W. and Maini,R.N. (1985) J. Immunol. Methods, 77, 63-76. Stefano,J.E. (1984) Cell, 36, 145-154. Steitz,J.A., Wolin,S.L., Rinke,J., Pettersson,I., Mount,S.M., Lerner,E.A., Hinterberger,M. and Gottlieb,E. (1983) Cold Spring Harbor Symp. Quant. Biol., 47, 893-900. Stillman,D.J., Better,M. and Geiduschek,E.P. (1985) Mol. J. Biol., 185, 451 -455. Stillman,D.J. and Geiduschek,E.P. (1984) EMBO J., 3, 847-853. Stillman,D.J., Silversten,A.L., Zenter,P.G. and Geiduschek,E.P. (1984) J. Biol. Chem., 259, 7995-7962. Straney,D.C. and Crothers,D.M. (1985) Cell, 43, 449-459. Thoen,C.O., Mills,K. and Hopkins,M. (1980) Am. J. Vet. Res., 41, 833 -837. Towbin,H., Staehelin,T. and Gordon,J. (1979) Proc. Natl. Acad. Sci. USA, 76, 4350-4354. Van Mol. Cell. 1021-1031. Dyke,M.W. and Roeder,R.G. (1987) Biol., 7, von Hippel,P.H., and Bear,D.G., Morgan,W.D. McSwiggen,J.A. (1984) Annu. 389-446. Rev. Biochem., 53, Watson,J.B., Gralla,J.D. Nucleic Actids Chandler,D.W. and (1984) Res., 12, 5369-5384. Weil,P.A., Segall,J., Harris,B., Ng, S-Y. and Roeder,R.G. (1979) J. Biol. Chem., 254, 6163-6173. Larson,D., Young,L.S. and Sprague,K.U. J. Mol. Wilson,E.T., (1985) Biol., 183, 153-163. Wingender,E., Shi,X.P., Houpert,A. and Seifart,K.H. (1984) EMBO J., 3, 1761-1768.

Journal

The EMBO Journal – Springer Journals

Published: Mar 1, 1989

There are no references for this article.