Proc. Natl. Acad. Sci. USA Vol. 90, pp. 9006-9010, October 1993 Biochemustry RNase E activity is conferred by a single polypeptide: Overexpression, purification, and properties of the ams/rne/hmpl gene product (/RNA processing) R. S. CORMACK, J. L. GENEREAUX, AND G. A. MACKIE* Department of Biochemistry, University of Western Ontario, London, Ontario, Canada N6A 5C1 Communicated by Myron K. Brakke, June 29, 1993 (receivedfor review May 10, 1993)

ABSTRACT E, an that processes encoding a high molecular mass protein hypothesized to be pre-SS rRNA from Its precursor, is now believed to be the involved in nucleoid partitioning or cell wall invagination major endoribonuclease participating in mRNA turnover in during division (16). The revised DNA sequence, coterminal . The product of the ams/rne/hmpl gene, with the 5' end of the two previous sequences, encodes a which is required for RNase E activity, was overexpressed, protein of 1025 amino acid residues whose predicted relative purified to near homogeneity by electroelution from an molecular mass would be 114,000. This polypeptide displays SDS/polyacrylamide gel, and renatured. The purified poly- anomalous electrophoretic mobility in SDS/polyacrylamide peptide possesses nucleolytic activity in vitro with a specificity gels, however, and migrates with an apparent size of 180 kDa identical to that observed for crude RNase E preparations. In (16). It has not been proven that the polypeptide encoded by addition, both UV crosslinking and RNA-protein blotting the ams/rne/hmpl gene is, in fact, a ribonuclease or is unambiguously showed that the Ams/Rne/Hmpl polypeptide intimately associated with the cleavage ofRNA either in vitro has a high affnity for RNA. Our results demonstrate that or in vivo. Indeed, it has been suggested that the gene product RNase E activity is directly attributable to, and is an inherent is not a ribonuclease but rather a regulatory protein or an property of, an RNA-binding protein, the ams/rne/hmpl gene auxiliary factor (8). It has also been proposed that RNase E product. activity requires the association of an unidentified catalytic subunit with other proteins, which could include the heat Regulating the balance between mRNA synthesis and decay shock protein GroEL (17). is a significant means of controlling gene expression. In the In an effort to understand the role ofRNase E with respect bacterium Escherichia coli, mRNA decay is usually initiated to substrate recognition and RNA processing, we have over- by endonucleolytic cleavage and followed by exonucleolytic expressed and purified the ams/rne/hmpl gene product to degradation at the new 3' ends by polynucleotide phospho- near homogeneity and have assayed its activity on various rylase and ribonuclease II (1). The endoribonuclease RNase substrate . In addition, we have utilized UV crosslink- E, which processes pre-5S rRNA from a larger precursor, 9S ing and RNA-protein blots to assess the binding affinity of RNA (2), has been demonstrated both genetically and bio- the full-length or truncated protein. Our results demonstrate chemically to be necessary for the decay of many specific that the ams/rne/hmpl gene product, to which we will refer mRNAs (for reviews, see refs. 3 and 4). The discovery that as Ams/Rne/Hmpl, has a high affinity for RNA and is an essential gene required for RNase E activity, rne, is allelic independently capable ofcorrectly processing several known to ams (altered mRNA stability) (5-8), manifested by ams-1, substrates of RNase E, including the 9S precursor to pre-5S a temperature-sensitive mutation responsible for a 6-fold rRNA. increase in the chemical half-life of pulse-labeled mRNA under nonpermissive conditions (9), constitutes strong evi- MATERIALS AND METHODS dence that RNase E is the rate-limiting factor in the turnover of many mRNAs. The site of cleavage by RNase E occurs Bacterial Strains and Plasmids. The E. coli strain 18-11BP A A (18) was obtained from M. P. Deutscher (University of within the consensus sequence GAUUU (3), which has been Connecticut Health Center, Farmington). The vector pET-11 found to be single stranded and flanked by stem-loop struc- and its host strain for expression, E. coli BL21(DE3) (19), tures (3, 10, 11). In vivo experiments have demonstrated that were obtained from Novagen. The recombinant bacterio- a substrate for RNase E, RNA I of the ColEl plasmid, phage A234 (20), originally AE3G11 (21) and the plasmid requires several unpaired nucleotides at its 5' terminus for pFMK33 (20), both containing all or part of the ams/rne/ efficient cleavage and suggest that RNase E may be sensitive hmpl gene, were obtained from D. Steege (Duke University, to 5'-terminal base pairing (12). Durham, NC) and S. R. Kushner (University of Georgia, Studies of the biochemical properties of the enzyme have Athens), respectively. pFMK33 was digested with Xmn I and generated conflicting data. RNase E was originally thought to BamHI. The 2.7-kbp fragment spanning residues 329-3068 of be a 70-kDa protein that could be rendered temperature the ams/rne/hmpl sequence (coordinates of ref. 16) was sensitive by the rne-3071 mutation (13). Genes capable of ligated into the unique BamHI site of pET-li. Plasmid complementing the ams-) or rne-3071 mutations have been pGM101 contained the 2.7-kbp fragment correctly oriented cloned, but small errors in sequence determination led to the relative to the vector, with the blunt end at the Xmn I site prediction of the ams/rne gene product being a polypeptide ligated directly to the BamHI site of the vector and the ofeither91 kDaor62 kDa, respectively (14, 15). The ams/rne four-residue gap on one strand repaired. To reconstruct the gene was recently recloned and sequenced as hmpl, a gene complete ams/rne/hmpl gene, a fragment of DNA spanning

The publication costs ofthis article were defrayed in part by page charge Abbreviations: IPTG, isopropyl ,B-D-thiogalactopyranoside; DTT, payment. This article must therefore be hereby marked "advertisement" dithiothreitol. in accordance with 18 U.S.C. §1734 solely to indicate this fact. *To whom reprint requests should be addressed. 9006 Downloaded by guest on September 24, 2021 Biochemistry: Cormack et al. Proc. Natl. Acad. Sci. USA 90 (1993) 9007

residues 3008-3824 was prepared by PCR amplification ofthe followed by the addition of an equal volume of SDS sample corresponding region of A234 with oligonucleotide primers buffer. The samples were separated in an SDS/7% polyacryl- spanning residues 3008-3031 (5'-AGTCGCCAATGCCGT- amide gel, stained with Coomassie brilliant blue R-250, and TGACCGTAG) and complementary to residues 3801-3824, exposed to x-ray film. with a C -- G mismatch at position 3811 to form a new BamHI RNA-Protein Blots. SDS/7% PAGE gels were used to site (5'-GGTTAGCAAGGATCCCATTCGATG). The prod- separate proteins prepared by boiling the equivalent of 50 p1 uct of amplification was digested with BamHI, purified, and of midlogarithmic phase cultures in SDS sample buffer. The ligated into pGM101 cleaved with BamHI to form pGM102, proteins were electroblotted to a nitrocellulose membrane which contained the complete ams/rne/hmpl gene (see Fig. using a carbonate transfer buffer (24) for 1-3 hr at a constant 1). The orientation and complete nucleotide sequence of the current of 250 mA. After transfer, the blot was placed in 750-bp BamHI fragment in pGM102 were checked by restric- TEN50 buffer (10 mM Tris-HCl, pH 8.0/1 mM EDTA/50 mM tion mapping and by DNA sequencing. Strain BL21(DE3) NaCI) and stored at 4°C for a minimum of 16 hr. Marker lanes was transformed with pGM101, pGM102, and pET-li to form were stained with 0.2% Ponceau S in 3% trichloroacetic acid. strains GM400, GM402, and GM403, respectively. Individual strips of the blot were incubated in binding buffer Preparation of Ams/Rne/Hmpl. Appropriate cultures [TEN50/0.02% Ficoll 400/0.02% polyvinylpyrrolidone/ were grown in a rich medium (19) at 28°C to early logarithmic 0.02% bovine serum albumin (fraction V)/yeast RNA (250 phase and induced with 0.5 mM isopropyl f-D-thiogalacto- pg/ml)] for 90 min at 44°C. 32P-labeled probe RNA generated pyranoside (IPTG). Cultures were harvested 5 hr later, and from in vitro transcription (specific activity 1.5 x 108 the supernatant from a 30,000 x g centrifugation was (S-30) was added x 105 and were precipitated with 26% (wt/vol) ammonium sulfate to yield the cpm/pug) (5 cpm/ml), the blots incubated for 90 min at 44°C. The blots were washed at AS-26 fraction as described (22) and dialyzed for 5 hr. ambient temperature for 10 min in TEN50 buffer, 10 min in Samples containing 2.5-3.0 mg of protein were heated for 5 min at 100°C in SDS sample buffer [60 mM Tris-HCl, pH TEN50 with 200 mM NaCl, and then for 10 min in TENS0 6.8/1.5% (wt/vol) SDS/50 mM dithiothreitol (DTT)/5% (wt/ with 500 mM NaCl. The blots were dried and exposed to x-ray vol) glycerol], applied to a 100 mm x 1.5 mm x 4 mm well film. in an SDS/5.5% polyacrylamide (49:1, acrylamide to N,N'- methylenebisacrylamide) gel, and separated by electropho- RESULTS resis. The outer edges of the gel were stained and used as a guide to excise the band containing the overexpressed poly- Overexpression and Purification of Ams/Rne/Hmpl. The peptide, which was subsequently eluted from the gel slice structure of a plasmid containing most ofthe ams/rne/hmpl using the Bio-Rad model 422 electroeluter. The eluted protein gene is illustrated in Fig. 1. Plasmid pGM102 contains 213 was precipitated with 4-5 vol of acetone containing 1 mM residues ofthe 5'-noncoding region, the entire predicted open DTT and recovered by centrifugation. The pellet of protein reading frame (16), and 199 residues of the 3'-noncoding and SDS was washed several times with 80%o acetone in 1 mM region, including the putative p-independent terminator, un- DTT, air dried, dissolved in 300-500 Al of denaturing buffer der control ofthe T7 lac promoter-operator region in pET-11 (6 M guanidine hydrochloride/50 mM Hepes-NaOH, pH (19). The predicted ams/rne/hmpl gene product encoded by 7.5/150 mM NaCl/1 mM DTT/0.1 mM EDTA/20% glycer- pGM102 is identical to that of ref. 16 except for two changes: ol), and renatured by dilution into the same buffer lacking Arg -* Pro at codon 905 and Ala -* Gly at codon 1020. These guanidine hydrochloride as described (23). After dialysis, the alterations may reflect errors produced during amplification. dilute material was concentrated to =300 A1 using an Amicon Plasmid pGM101 is similar to pGM102 except that deletion of Centricon-30 centrifugal microconcentrator. a BamHI fragment removes the C-terminal 20%6 of the In Vitro Transcription and Assay of RNase E Activity. ams/rne/hmpl open reading frame. Synthesis of RNA substrates was carried out as described Induction of cultures of GM402, which contains pGM102, (11). For the synthesis of RNA probes, 40 ,Ci (1 Ci = 37 resulted in the overexpression of a protein whose apparent GBq) of[a-32P]CTP (3000 Ci/mmol) and 5 ,uM unlabeled CTP size in a crude extract (S-30) was -180 kDa (Fig. 2, lane 2), were used in the incubation in place of 100 pM CTP. Assays consistent with the aberrant mobility of Ams/Rne/Hmpl for RNase E activity were performed at 30°C as described noted by others (14, 16). RNase E activity is difficult to with 0.02 nM substrate (22). The products were denatured quantify accurately in crude fractions such as the S-30. Based and resolved in a sequencing gel. AS-26 fractions from strain on our previous experience that RNase E activity is quanti- GM402 or from nonoverproducing strains were assayed at 5 tatively precipitated by 26% (wt/vol) ammonium sulfate, we ,ug/ml or at 100 pg/ml, respectively; renatured polypeptide fractionated the S-30 from GM402 with ammonium sulfate was assayed at -0.5 pg/ml on S20 mRNAs and at 2-3 pg/ml (22) to yield an AS-26 fraction. The specific activity ofRNase on 9S RNA. Assays with renatured polypeptide also con- E in the AS-26 fraction from the overproducing strain GM402 was 50- than in an AS-26 fraction tained 0.1% Triton X-100 and 10% (wt/vol) polyethylene to 60-fold greater prepared glycol 8000. B B UV Crosslinking. Approximately 4-5 ng of 32P-labeled 9S 329 541 3068 3618 3816 RNA generated from in vitro transcription (-1.5 x 108 P-T7 0-lac .M cpm/pg) was incubated in 25 mM Tris-HCl, pH 8.0/5 mM I CaCl2/60 mM KCl/100 mM NH4CI/0.1 mM DTT/5% glyc- ATG TGA erol containing a 1000-fold weight excess ofyeast RNA (5 pg) N C and AS-26 protein (0.18 mg/ml) in a final volume of 20 pd for 10 min at 300C and then for 10 min on ice. The samples were FiG. 1. Schematic maps of plasmids containing the ams/rne/ irradiated on ice with a 254-nm light source (Stratagene UV hmpl gene. The construction of pGM102, which contains the com- Stratalinker 1800) for 5 min (0.78 J/cm2) at a distance of5 cm. plete coding sequence, is described in Materials and Methods. The 1 2 M 1 of 10mM 2 units solid box represents the cloned sequence, and the line represents Subsequently, ,ul of NaCl, Id MgCl2, sequences derived from the vector pET-1l (19). Coordinates of the of RNase T1, and 0.45 unit of RNase V, were added to the ams/rne/hmpl sequence (16) are shown above the solid box. The irradiated samples and incubated for 30 min at 370C. Half of bent arrow shows the direction of transcription, and B denotes the the sample was added to an equal volume of SDS sample positions of BamHI cleavage sites. The stippled box indicates the buffer, and the remaining sample was treated with S jug of extent of the polypeptide products. The 750-bp BamHI frgment proteinase K in the presence of 1% SDS for 10 min at 65°C (residues 3068-3816) is deleted in pGM101. Downloaded by guest on September 24, 2021 9008 Biochemistry: Cormack et al. Proc. Natl. Acad Sci. USA 90 (1993) described in Materials andMethods. A sample ofthe purified (0 polypeptide is shown in Fig. 2, lane 4. The prominent band CD G\DQG Co 0< of 180 kDa is about 95% pure, as only traces of other kDa polypeptides are visible when the sample is overloaded (data not shown).

200-- The renatured Ams/Rne/Hmpl polypeptide was assayed for RNase E activity on several substrates including 9S RNA and S20 mRNA (Fig. 3D), and the products were separated 97 - ___; on a 40-cm sequencing gel, which could distinguish single- 66- - residue differences in size. The data in Fig. 3A show that 9S ...... RNA is converted via several intermediates to a product whose electrophoretic mobility is identical to that of pre-5S 45- rRNA (126 residues). Both the 207 ("8S")- and 165 ("7S")- residue intermediates corresponding to single cleavages at 1 2 3 4 the "b" and "a" sites, respectively, have been characterized previously as the products of RNase E digestion of 9S RNA FIG. 2. Purity of Ams/Rne/Hmpl. Cultures of GM402 were induced with IPTG and grown for 5 hr at 28°C prior to harvest, lysis, by primer extension (11) and by RNA fingerprinting (25). In and purification as described in Materials and Methods. The follow- addition, the 81-residue 5' product and the 3' terminator ing protein samples were denatured and separated in an SDS/7% region also accumulated as expected, although the latter polyacrylamide gel: lane 1, molecular mass standards (400 ng each; product has run off the gel shown in Fig. 3A. Bio-Rad); lane 2, S-30 extract (12 ytg); lane 3, AS-26 fraction (10 g); The mRNA for ribosomal protein S20 is known to be a lane 4, renatured Ams/Rne/Hmpl polypeptide (-250 ng) prepared substrate for an ams/rne-dependent in vitro by preparative electrophoresis. (10, 22). The data in Fig. 3B show that renatured Ams/Rne/ Hmpl cleaves the full-length S20 mRNA substrate (t87D) from strain GM403 (vector without insert; data not shown). efficiently. The observed 147-residue product also accumu- The putative Ams/Rne/Hmpl polypeptide was also recov- lates in vivo when activity is suppressed ered in the AS-26 fraction (Fig. 2, lane 3). The reduced (22). Other products generated by the renatured polypeptide mobility of Ams/Rne/Hmpl and its lability prompted us to comigrate precisely with those produced by crude RNase E attempt its purification by preparative electrophoresis. Ac- (compare bands in Fig. 3B, lanes 2 and 3 with lanes 5 and 6). cordingly, proteins in the AS-26 fraction were boiled in SDS There are, however, differences in relative intensities, par- sample buffer and resolved by electrophoresis. The putative ticularly among the products of 147-173 residues. These Ams/Rne/Hmpl polypeptide was eluted from the prepara- products correspond to cleavages in the region between tive gel, renatured by dilution (23), and concentrated as residues 270 and 300 (coordinates given in ref. 10). A B C D 9S a b - - 365 _40 aDw *4- 246 a .... - 331 _~~~ 332 - I- 246 ~' ... *...... --- 165 __M- - 207 126 f- 4)". fs n ff, :iii. :.:.l-.:.: .: 207 257 .. n... :,::.::.:::::. _ -- 165 - 81 * = 224 _0 -, O t87D

I i t.s - 365 *- 126 *__ m- 331 ,173 -* 257 - - 173 4. .. .156 ------_.156 * *:*I 4147 147 t95D - - 81 I-I 332 v 224 -p- 114 as i,,,_ 0-- 114

1 2 3 4 5 6 1 2 3 4 5 6 1 2 3 4 5 6 FIG. 3. RNase activity of renatured Ams/Rne/Hmpl on three RNA substrates. RNA substrates were assayed at 30°C as described in Materials and Methods with either an AS-26 fraction or with the renatured Ams/Rne/Hmpl polypeptide as indicated. The products were separated in an 8% (A) or 6% (B and C) polyacrylamide sequencing gel. (A) 9S RNA digested with an AS-26 fraction from IPTG-induced strain GM402 for 0 (lane 1), 15 (lane 2), and 45 (lane 3) min or with purified Ams/Rne/Hmpl for 0 (lane 4), 60 (lane 5), and 240 (ane 6) min. (B) t87D digested with an AS-26 fraction from strain 18-11BP for 0 Oane 1), 20 (lane 2), and 60 (lane 3) min or with purified Ams/Rne/Hmpl for 0 (lane 4), 40 (lane 5), and 120 (lane 6) min. (C) t95D digested with an AS-26 fraction from strain 18-11BP for 0 (lane 1), 20 Oane 2), and 60 Oane 3) min or with purified Ams/Rne/Hmpl for 0 (lane 4), 40 (lane 5), and 90 (lane 6) min. Arrows to the right of the gels denote the sizes (nucleotide residues) of prominent products. (D) Illustration of the extent of each RNA substrate, the relative positions of cleavage sites, and the sizes of prominent products. The template RNAs shown are 9S RNA (Top) (11); t87D, a 365-residue S20 mRNA (Middle); and t95D, which is similar to t87D except for a 33-residue in-frame deletion, which removes a single stem-loop (stem V) in the S20 mRNA (Bottom) (10). The vertical arrows indicate the cleavage sites. Downloaded by guest on September 24, 2021 Biochemistry: Cormack et al. Proc. Natl. Acad. Sci. USA 90 (1993) 9009 A derivative oft87D lacking residues 304-336, t95D, is also RNA-binding capacity of the polypeptide by probing protein cleaved efficiently and accurately by renatured Ams/Rne/ blots with 32P-labeled RNA. To avoid complications due to Hmpl (Fig. 3C). The spectrum ofproducts obtained with the partial proteolysis during fractionation, whole cell lysates of crude activity (Fig. 3C, lanes 1-3) is identical in mobility and the strains overproducing Ams/Rne/Hmpl were separated in relative intensities to the products produced by the purified by SDS/PAGE, blotted onto nitrocellulose, and then probed protein (Fig. 3C, lanes 4-6). Taken together, these data with 32P-labeled 9S RNA (Fig. 5). The probe specifically and demonstrate that the purified, renatured polypeptide product strongly bound to both full-length and truncated Ams/Rne/ ofthe ams/rne/hmpl gene is capable ofproducing authentic Hmpl polypeptides but only in lysates of IPTG-induced RNase E cleavages on three substrates. The relative rates of cultures of GM400 or GM402, respectively (Fig. 5, compare cleavage ofthese substrates differed considerably, however, lanes 4 and 6 with lanes 3 and 5). The time ofinduction in this from the rates measured when AS-26 fractions were the experiment was relatively short so that the overproduced source ofRNase E activity. In particular, complete cleavage truncated or full-length proteins migrating as 110-kDa and of the 9S RNA required longer times and 2- to 3-fold more 180-kDa polypeptides, respectively, constitute only a small renatured Ams/Rne/Hmpl polypeptide than the S20 portion (<2%) ofthe total protein in each lane. The probe also mRNAs. bound to the endogenous 180-kDa protein found in each UV Crossilnking. As a further test of whether the amsl extract, albeit at a reduced level, which is visible with longer rne/hmpl gene product recognizes and associates with its exposures (Fig. 5, lanes 1-5). The signals were resistant to substrates, crosslinking experiments were performed to iden- competitor yeast RNA up to concentrations of 1 mg/ml. Two tify proteins involved in the cleavage of 9S RNA in vitro. other unknown proteins (1420 kDa and =72 kDa) were also AS-26 fractions containing RNase E activity were incubated capable of binding the 9S RNA probe. Both of these bands with 32P-labeled 9S RNA, irradiated with UV, digested with appear at nearly equivalent levels in each lane and are not , and then separated by SDS/PAGE (Fig. 4). affected by IPTG induction. Both the 120-kDa and 72-kDa The strain containing only the vector (GM403) gave rise to bands were not detectable when blots of AS-26 fractions bands migrating at 180 kDa and at 90 kDa (Fig. 4, lane 1). Use enriched for RNase E activity prepared from wild-type or of extracts from the two overproducing strains, GM400 and overproducing strains were probed (data not shown), which GM402, led to the efficient labeling of a 110-kDa band (lane implies that these polypeptides are unrelated to RNase E 3) or a 180-kDa band (lane 5), respectively, which comigrated activity. RNA-protein blots were also performed on AS-26 with the truncated and full-length forms ofAms/Rne/Hmpl. fractions of the two overproducing strains, GM400 and All labeled bands were proteinase K-sensitive (Fig. 4, lanes GM402, using other 32P-labeled nucleic acid probes (data not 2, 4, and 6). Several smaller bands of weaker intensity were shown). Both the 180-kDa and 110-kDa polypeptides were also observed in lanes 1, 3, and 5. The intensity oflabeling of selectively bound by RNAs of various sequences including some of the minor bands as well as the bands in the 90-kDa S20 mRNA, 9S RNA deletion mutants (11), vector-derived region increased upon IPTG induction (compare lanes 3 and transcripts, and poly(A) RNA. In contrast, neither single- 5 with lane 1 in Fig. 4), suggesting that they may arise from stranded nor double-stranded DNA bound to either polypep- degradation of the induced polypeptides. Similar patterns of tide (data not shown). labeling were obtained when S20 mRNA was used as the labeled substrate RNA (data not shown). About 6-fold greater signals were obtained when CaCl2 was used in the incubation DISCUSSION in lieu of MgCl2 or in the absence of either salt (magnesium A variety of circumstantial evidence has accumulated to ions are required for enzymatic cleavage). suggest that the ams/rne/hmpl gene is required for the RNA Binding by Ams/Rne/Hmpl. Analysis of the pre- activity associated with RNase E (3-8). This work now dicted amino acid sequence ofAms/Rne/Hmpl revealed the provides direct evidence that the polypeptide product ofthe presence of a putative RNA-, the so-called RNA ams/rne/hmpl gene possesses two discrete activities: it is a recognition motif(14, 26). This prompted us to investigate the ribonuclease that catalyzes authentic cleavages on 9S RNA and S20 mRNA previously attributed to RNase E in vivo and GM403 GM400 GM402 . if in vitro, and it binds RNA with a high affinity. Both properties ± - + ± kDa GM403 GM400 GM402 I IF I I 200- 4- I- - + 180 kDa 116- 110 97 - 200 - 90 El 4180 66 - 116- _I *4- 1 10 97 -

45- 66-

2 3 4 5 6 45 - 1 2 3 4 5 6 FIG. 4. UV crosslinking of 9S RNA and proteins in AS-26 fractions. 32P-labeled 9S RNA was incubated with AS-26 fractions FIG. 5. RNA-protein blots probed with 9S RNA. Whole cell prepared from strains GM403, GM400, and GM402, irradiated with lysates of exponentially growing cultures without (-) or with (+) UV, digested with ribonucleases, and then separated by SDS/PAGE induction by IPTG were separated by SDS/PAGE, electroblotted as described in Materials andMethods. Prior to electrophoresis, half onto nitrocellulose, and probed with 32P-labeled 9S RNA as de- of each sample was treated (+) with proteinase K. Lanes 1 and 2, scribed in Materials andMethods. Lanes 1 and 2, lysates from strain AS-26 from strain GM403; lanes 3 and 4, AS-26 from strain GM400; GM403; lanes 3 and 4, lysates from strain GM400; lanes 5 and 6, lanes 5 and 6, AS-26 from strain GM402. Arrowheads on the rigot lysates from strain GM402. Arrowheads on the right indicate the indicate the 180-kDa, 110-kDa, and 90-kDa protein bands discussed 180-kDa and 110-kDa protein bands. Molecular mass standards are in the text. Molecular mass standards are indicated on the left. indicated on the left. Downloaded by guest on September 24, 2021 9010 Biochemistry: Cormack et al. Proc. Natl. Acad Sci. USA 90 (1993) can be manifested clearly in the absence of other proteins a recent report has claimed that GroEL functionally interacts and/or factors. Surprisingly, the Ams/Rne/Hmpl polypep- with RNase E (17). The slow rate of processing of 9S RNA tide is renaturable after boiling in SDS with up to 15% (relative to the S20 mRNA substrates t95D and t87D) by the recovery of activity. purified polypeptide implies that other protein(s) could en- The renatured Ams/Rne/Hmpl polypeptide faithfully re- hance the rate of cleavage or alter the structure of the produced known RNase E cleavage products on the three substrate into a more favorable conformation. If this is the RNA substrates tested. The only differences compared to the case, then the S20 mRNA substrates are either more acces- activity in crude extracts are reflected in the relative rate of sible to RNase E or do not require such putative accessory cleavage and the relative abundances of the cleavage inter- proteins. mediates. The 81-residue (5') and 126-residue (p5) cleavage products of 9S RNA catalyzed by the renatured Ams/Rne/ Note Added in Proof. Direct sequencing has shown that the changes Hmpl polypeptide were recovered in approximately equi- affecting codons 905 and 1020 in pGM102 relative to the sequence molar yields, consistent with an endonucleolytic mechanism. reported in ref. 16 are also present in the DNA of A234 and thus do In contrast, the anticipated 5' products of S20 mRNA cleav- not reflect errors in amplification. age are faint or invisible with both t87D and t95D. A number of ams/rne-dependent cleavage sites have previously been We thank our colleagues for providing strains and comments. mapped to the 5' half of the S20 mRNA, where they occur R.S.C. acknowledges support provided by an Ontario Graduate 20-30 residues apart (22). The RNAs released by such Scholarship. This research was funded by a grant from the Medical successive endonucleolytic cleavages would be small, Research Council of Canada to G.A.M. weakly labeled, and would have run off the gels in Fig. 3 B 1. Belasco, J. G. & Higgins, C. F. (1988) Gene 72, 15-23. and C. 2. Ghora, B. K. & Apirion, D. (1978) Cell 15, 1055-1066. The RNA-binding properties of Ams/Rne/Hmpl are best 3. Ehretsmann, C. P., Carpousis, A. J. & Krisch, H. M. (1992) illustrated in the UV-crosslinking and RNA-protein blotting Genes Dev. 6, 149-159. experiments. There is a good correlation between the inten- 4. Higgins, C. F., Peltz, S. W. & Jacobson, A. (1992) Curr. Opin. sity of the signal obtained from UV crosslinking and the Genet. Dev. 2, 739-747. amount of Ams/Rne/Hmpl and RNase E activity found in 5. Mudd, E. A., Krisch, H. M. & Higgins, C. F. (1990) Mol. the irradiated extract. Overexpression of the gene product Microbiol. 4, 2127-2135. resulted in proportionally greater signals than in the wild-type 6. Babitzke, P. & Kushner, S. R. (1991) Proc. Natl. Acad. Sci. extract. The exclusion of MgCl2 (and replacement with USA 88, 1-5. CaCl2) to inhibit cleavage by RNase E shows that the binding 7. Taraseviciene, L., Miczak, A. & Apirion, D. (1991) Mol. Microbiol. 5, 851-855. of RNA by the ams/rne/hmpl gene product is independent 8. Melefors, 0. & von Gabain, A. (1991) Mol. Microbiol. 5, of cleavage. RNA-protein blots showed that the RNA- 857-864. binding domain of Ams/Rne/Hmpl is contained within the 9. Ono, M. & Kuwano, M. (1979) J. Mol. Biol. 129, 343-357. first 844 amino acids of its sequence as the polypeptide 10. Mackie, G. A. (1992) J. Biol. Chem. 267, 1054-1061. encoded by pGM100 displays the same apparent affinity for 11. Cormack, R. S. & Mackie, G. A. (1992) J. Mol. Biol. 228, RNA as the complete polypeptide of 1025 residues. This 1078-1090. affinity for RNA must be relatively high since the RNA- 12. Bouvet, P. & Belasco, J. G. (1992) Nature (London) 360, protein complex is stable to washes with 500 mM NaCl. 488-491. Our results demonstrate that no other subunits 13. Misra, T. K. & Apirion, D. (1979) J. Biol. Chem. 254, 11154- obligatory 11159. are involved in the binding ofRNA substrates by Ams/Rne/ 14. Claverie-Martin, F., Diaz-Torres, M. R., Yancey, S. D. & Hmpl or in RNase E activity. First, RNase E activity can be Kushner, S. R. (1991) J. Biol. Chem. 266, 2843-2851. overexpressed significantly, as the AS-26 fraction from 15. Chauhan, A. K., Miczak, A., Taraseviciene, L. & Apirion, D. GM402 is 50- to 60-fold more active than that from GM403, (1991) Nucleic Acids Res. 19, 125-129. and there was a commensurate increase in the intensity ofthe 16. Casar6gola, S., Jacq, A., Laoudj, D., McGurk, G., Margarson, 180-kDa band as judged by Coomassie blue staining of S., Tempte, M., Norris, V. & Holland, I. B. (1992) J. Mol. extracts. Second, RNA-protein blots show that the full- Biol. 227, 30-40. length or C-terminally truncated Ams/Rne/Hmpl polypep- 17. Sohlberg, B., Lundberg, U., Hartl, F.-U. & von Gabain, A. tides are to bind RNA. No other (1993) Proc. Natl. Acad. Sci. USA 90, 277-281. individually competent 18. Deutscher, M. P., Marshall, G. T. & Cudny, H. (1988) Proc. proteins contained in the AS-26 fraction in which virtually all Natl. Acad. Sci. USA 85, 4710-4714. the RNase E activity resides can significantly bind RNA (data 19. Studier, F. W., Rosenberg, A. H., Dunn, J. J. & Dubendorff, not shown). Third, although UV-crosslinking experiments J. W. (1990) Methods Enzymol. 185, 60-89. demonstrate that one or more proteins of about 90 kDa 20. Claverie-Martin, F., Diaz-Torres, M. R., Yancey, S. D. & present in the AS-26fractions can bind RNA in addition to the Kushner, S. R. (1989) J. Bacteriol. 171, 5479-5486. Ams/Rne/Hmpl polypeptide, there is no evidence that the 21. Kohara, Y., Akiyama, K. & Isono, K. (1987) Cell 50, 495-508. two proteins are interacting directly. The identity of the 22. Mackie, G. A. (1991) J. Bacteriol. 173, 2488-2497. 90-kDa protein(s) remains unknown. We believe that they, 23. Hager, D. A. & Burgess, R. R. (1980) Anal. Biochem. 109, and other minor crosslinked are most 76-86. proteins, likely degra- 24. Dunn, S. D. (1986) Anal. Biochem. 157, 144-153. dation products of Ams/Rne/Hmpl. Alternatively, one or 25. Roy, M. K., Singh, B., Ray, B. K. & Apirion, D. (1983) Eur. more of these proteins could comprise component(s) of a J. Biochem. 131, 119-127. much larger nucleolytic complex (27). The renaturable 26. Kenan, D. J., Query, C. C. & Keene, J. D. (1991) Trends RNase E activity displayed by the purified Ams/Rne/Hmpl Biochem. Sci. 16, 214-220. polypeptide, however, does not exclude the possibility that 27. Miczak, A., Srivastava, R. A. K. & Apirion, D. (1991) Mol. other proteins or subunits may be associated with it. Indeed, Microbiol. 5, 1801-1810. Downloaded by guest on September 24, 2021