Nucleotide sequence and in vitro processing of a precursor molecule to Escherichia coli 4.5 S RNA.

A precursor molecule to the stable 4.5 S RNA species of Escherichia coli has been found to accumulate at 42 degrees in a strain thermosensitive for the function of ribonuclease P. The precursor molecule is 130 nucleotides long. Twenty-two extra nucleotides, starting with pppGp, precede the mature sequence at its 5' terminus. At least 1 extra uridine residue can be found at the 3' terminus. The precursor to 4.5 S RNA is cleaved in vitro by RNase P to generate a 5' end identical to that of the mature 4.5 S RNA.

A precursor molecule to the stable 4.5 S RNA species of Escherichia coli has been found to accumulate at 42" in a strain thermosensitive for the function of ribonuclease P. The precursor molecule is 130 nucleotides long. Twenty-two extra nucleotides, starting with pppGp, precede the mature sequence at its 5' terminus. At least 1 extra uridine residue can be found at the 3' terminus. The precursor to 4.5 S RNA is cleaved in uitro by RNase P to generate a 5' end identical to that of the mature 4.5 S RNA.
Several low molecular weight RNAs which have no known function have been identified in Escherichia coli. One of these is designated 4.5 S RNA because of its electrophoretic mobility in polyacrylamide gels. The primary structure of this RNA has been determined by Griffin (1). The 4.5 S RNA is very stable in uiuo and the quantity in a bacterial cell is comparable to that of a single tRNA species. It contains no modified nucleotides and does not appear to be associated with ribosomes.
The accumulation of 4.5 S RNA in uiuo, as for tRNA and 5 S I-RNA, is under stringent control (2). A precursor to 4.5 S RNA has been tentatively identified by other investigators at restrictive temperature in strains of E. coli thermosensitive for RNase P function (3,4). Since both the transcription and the processing of transcripts from genes coding for 4.5 S RNA are apparently under the same control as the genes and gene transcripts for tRNA, it is possible that 4.5 S RNA plays some role in protein synthesis. In order to study in detail one aspect of this control, endonucleolytic processing by RNase P, we have determined the nucleotide sequence of a 4.5 S RNA precursor molecule which accumulates in E. coli strain A49 at restrictive temperatures (4). Even though the structure of the mature portion of this molecule is different from that of the tRNA portion of tRNA precursor RNAs, it is in leed a substrate for RNase P. In fact, the availability of the products generated both by RNase P and human KB cell RNase NU cleavage of the 4.5 S RNA precursor molecule has greatly facilitated the nucleotide sequence studies. Furthermore, determination of the cleavage sites in 4.5 S RNA precursor of these enzymes has provided interesting new information concerning the factors governing specific ribonuclease-sub- Polyribonucleotides-SzP-labeled M, RNA (7), 4 S RNA, and 5 S rRNA were prepared as described by Bothwell and Altman (6,8). The ST-labeled precursor to 4.5 S RNA was prepared as follows. A culture of strain A49 was grown at 30" to a density of about ' Abbreviation used is: CMCT RNA to E. coli 4.5 S RNA 4 x lo8 cells/ml in TPGA medium. The cells were shifted to 42" for 45 min and then labeled with 5 to 10 mCi of carrier-free 32P0,3-for 20 min. The RNA was extracted by the addition of an equal volume of 88% phenol at room temperature, followed by vigorous shaking for 10 min. The RNA was precipitated with ethanol and then separated in and extracted from 10% polyacrylamide slab gels as described previously (9).

Nucleotide
Sequence Analysis--Sequence analysis was performed according to the methods developed by Sanger and collaborators (10,11). CMCT blocking was done according to Barrel1 (12). When alkaline phosphatase and CMCT blocking were used together, the substrate was first incubated in 5 ~1 of 0.1 mg/ml of alkaline phosphatase in 0.05 M Tris/HCl (pH 8.9) and 0.01 M MgCl, and then heated for 3 min at 90" to inactivate the alkaline phosphatase. The sample was dried, resuspended in the CMCT blocking buffer, and processed according to Barre11 (12).

Identification
of Precursor to 4.5 S RNA-A labeled RNA species having the same mobility in 10% polyacrylamide gels as the $80 psu3+ phage-encoded tRNA?" precursor (9) can be found in WP-labeled RNA preparations made at restrictive temperatures in uninfected Escherichia coli A49 cells (Fig. 1). Fingerprint analysis shows that this species (Fig. 3, panel A) contains all of the RNase T, oligonucleotides characteristic of mature 4.5 S RNA (l), except it lacks the mature 5'-pGp terminus. Furthermore, this RNA can be cleaved in vitro by purified RNase'P to yield two products: a molecule which is the same size as 4.5 S RNA and a smaller fragment 22 nucleotides in length (Fig. 2, lane 6). Finally, no modified nucleotides were found in the RNA after digestion with RNase Tz and two-dimensional thin layer chromatography.
Griffin found this to be the case for mature 4.5 S RNA (1). These observations identified the molecule as a precursor to 4.5 S RNA.
Preparations of the 4.5 S RNA precursor also contain two tRNA precursors amounting to about 10% of the total radioactive RNA. This was judged by the relative amounts of 4.5 S and tRNA-size products resulting from RNase P cleavage of the precursor band. Therefore, the intact precursor is at least 90% radiochemically pure as eluted from a gel. RNase NU, an endoribonuclease isolated from human KB cells, can also cleave the 4.5 S RNA precursor to yield fragments 111 and 19 nucleotides long (Fig. 2, lane 7). Because their size and electrophoretic mobilities differ from the cleavage products of the contaminating tRNA precursors, the RNase P and RNase NU-derived fragments of the 4.5 S precursor are radiochemitally pure. This has greatly facilited the determination of the primary sequence of the "extra" part of the precursor molecule.
Nucleotide Sequence Analysis of Precursor to 4.5 S RNA -Fingerprints of the RNase T, and RNase A digests of the 4.5 S RNA precursor and of the products of RNase P and RNase NU cleavage of the precursor are shown in Fig. 3. Panel A shows a RNase T, fingerprint of the intact precursor, while panel B gives the fingerprint of the 4.5 S RNA generated by RNase P digestion of the precursor for comparison.
The fingerprints look identical because several of the precursorspecific oligonucleotides coincidentally co-migrate with oligonucleotides found in the mature molecule. Characterization of the precursor-specific sequences required separation of the 5' extra piece from the mature segment by digestion of the precursor with RNase P or RNase NU.
Some information concerning the extra sequence in the 4.5 S RNA precursor can be obtained by comparing the molar yields of the RNase T, oligonucleotides from the intact precursor with RNA from strain A49 was by polyacrylamide gel electrophoresis. The labeling conditions have been described under "Experimental Procedure." 6 S tRNA precursor is a mixture of precursors to at least two tRNA species. These precursors, which have been identified as such through fingerprint analysis (13), co-migrate in this gel system with 6 S RNA, which is found as a small proportion of the total "6 S" mixture from pulse-labeled A49 cells.
the yields from'the 4,5 S fragment resulting from RNase P cleavage of the precursor. Increased molar yields of products 1, 2, 15 and 16 (see Table I) were observed in the intact precursor compared with those found in the part of the molecule destined to become mature 4.5 S RNA. The product CpApApGp, found in 4.5 S RNA from strain MREGOO (l), has never been observed in the 4.5 S RNA precursor or the mature 4.5 S RNA from strain A49. However, an extra mole of CpApGp was found. This product was probably the result of a guanosine occurring at position 98 instead of an adenosine (see legend to Table I).
Griffin has observed some variation in the molar yield of certain RNase T, and RNase A products in separate preparations of mature 4.5 S RNA from a single E. coli strain. This was especially true for RNase T, and RNase A products containing the sequence ApApGp. Similar variation has been detected in the mature segment of RNA derived from RNase P cleavage of the 4.5 S precursor. The values in Table I are, therefore, the average molar yields from several preparations. The reactions were performed as described previously (6). The RNase P used in these experiments was the DEAE-Sephadex fraction described by Robertson et al. (5) and the RNase NU was the ammonium sulfate fraction (6). The cleavage products obtained by RNase NU digestion of the 4.5 S RNA precursor or the M, RNA were the same using either the ammonium sulfate fraction or the more pure DEAE-Sephadex fraction. The substrate used in the experiments shown in lanes I to 3 was the 6 S tRNA precursor indicated in Fig. 1 and is shown here only as a mobility marker which can also be cleaved by RNase P; to lane 4, the mobility markers 4 S RNA, 5 S RNA, and 6 S tRNA precursor were added; in lanes 5 to 7, the substrate used was 4.5 S RNA precursor and in lanes 8 to 10 it was M, RNA which is shown for comparison (14). The three-nucleotide size difference between the 4.5 S RNA precursor cleavage products produced by RNase P and RNase NU appears as a mobility difference in lanes 6 and 7. The two contaminating tRNAs produced by RNase P cleavage of the 4.5 S RNA precursor preparation are visible in lane 6.

RNase
T, fingerprints of the extra 5' RNA fragments produced by RNase P and RNase NU are shown in Fig. 3, panels D and E, respectively.
The products are numbered corresponding to those listed in Table II The products are numbered to correspond to the numbers in Table II. Panel E, the RNase NU-generated 5' fragment of the 4.5 S RNA precursor.
The products are numbered to correspond to those listed in Table II. Panel F, the RNase P-generated 5' fragment of 4.5 S RNA precursor.
The identity of the products has been indicated. The RNase NU-generated 5' fragment has an identical RNase A fingerprint except it lacks the product labeled ApAp&n. In each fingerprint, separation was by electrophoresis on cellulose acetate in pyridine acetate, 7 M urea, 0.001 M EDTA, pH 3.5, from right to left, and on DEAE-paper in 7% formic acid (v/v), from top to bottom.  The numbering of the T, products corresponded to the products labeled in Fig. 3, panels D and E. The molar yields were expressed relative to CpGp or GpUp taken as 1.0. CMCT blocking, followed by RNase A digestion, was used to locate the positions of uridine in oligonucleotides 5, 6, and 7. Products of digestion were separated by one-dimensional electrophoresis on Whatman No. 3MM paper, at pH 3.5. CMCT-modified products migrated toward the cathode and are given a positive sign relative to the origin; unmodified products migrated toward the anode and are given a negative sign. Product identity (after removal of blocking group) is given with migration distances: UpUpCp = +15.4 cm; Gp = +13.3 cm; UpCp = +10.4 cm; ApApU,,, = +8.6 cm; Cp = -2.8 cm; ApApCp = -15.3 cm; unblocked Gp = -26.6 cm. The final composition of these products was determined by alkali digestion and analysis of the products of this treatment. "The sequence of these products was determined by Griffin (1). RNase A and RNase U, digests of these products gave the products expected from the indicated sequence. No further verification of their sequence was done.
b The radioactivity found in this spot in the precursor contained 4.4 mol of octanucleotides.
The separation of the large oligonucleotides found in the extra 5' sequence in the 4.5 S RNA precursor was accomplished only by the use of RNase P and RNase NU to produce fragments containing the 5' sequence alone (see Table II). c These products were not separable in this electrophoretic system. d The products ApCpGp and CpApGp could not be separated in this system. Further analysis of this product revealed the presence of ApGp and ApCp in the ratio of 3:l. This meant that there were 3 mol of CpApGp to 1 mol of ApCpGp in the 4.5 S RNA sequence. Strain MREGOO had only 2 mol of CpApGp (1). These facts combined with the lack of the product CpApApGp found in strain MREGOO suggested that the 4.5 S RNA produced in strain A49 has a guanosine in position 98 instead of adenosine which is found in strain MREGOO. 4.5 S RNA from strain CA265 also lacks CpApApGp hut has an extra mole of CpApGp which was interpreted as being due to the presence of a guanosine at position 98 (1).
'The 3' end of this precursor resolved into two streaks instead of the one observed in the mature 4.5 S RNA in strain MREGOO (1). Evidence described in the text showed that the second 3'.oligonucleotide contained an extra uridine residue at the 3' end. cleotide 6 is CpUpCpCpApApUoH. (A control experiment was done to ensure the alkaline phosphatase treatment combined with CMCT and RNase A treatments would give the expected results under these conditions on a known dinucleotide, CpGp.
As expected, the products without phosphatase treatment were Rn'ase T,-produced RNase T, digestion of the RNase P-produced 5' fragment.
The products of the partial digestion (designated t, to t,) were separated by homochromatography (Fig. 5) and further characterized in the standard fashion. The composition of product t, shows that the RNase A product GpUp (Table IV)  must connect the RNase T, products 2 and 4 ( Table  II). The composition of products tz and t, indicate that the product GpGpUp (Table II) must connect  complete  RNase T, products 3, 4 and 5. Similarly, t3 and t, show that GpGp (Table  II) must connect RNase T, products 5 and 6 (Table  II). These data were then used to construct the complete sequence of the 5 fragment shown in Fig. 6. As shown by the determination of the sequence of oligonucleotide 7, the RNase NU cleavage site is located between nucleotides 19 and 20 of the precursor (Fig. 7)  Spleen phosphodiesterase partial digestion was carried out as described by Brownlee (16) and the products separated in 7% (v/v) formic acid on DEAE-paper.
Two oligonucleotides were isolated and then treated with CMCT and RNase A (see "Experimental Procedure") and the mobilities of these further products correlated with known oligo-or mononucleotides.
Other oligonucleotides could not be recovered in sufficiently large amounts for further analysis. Lane I shows the products of alkaline hydrolysis of M, RNA, which has pppGp at its 5' terminus. The position where pppAp would migrate has been indicated; lane 2 contained the product labeled 1 in Fig. 3, panel D and has the same mobility as pppGp. Lane 3 contained ppGp which was eluted from the fingerprint shown in Fig. 3 Analysis of 3' Terminus of Precursor to 4.5 S RNA-The two RNase T, products from the precursor containing the 3'  Table IV) and of RNase T, (top line above the sequence) and RNase A (second line above the sequence) complete digestion (see Fig.  3 and Table II). terminus (Fig. 3, panel C) were subjected to RNase A digestion and found to contain ApCp and Cp residues as well as variable amounts of Up. In a separate experiment, when a RNase T, d;gest of the large fragment of RNase P digestion of 4.5 S RNA precursor molecule was examined using homochromatography, two distinct products were seen which contained 3'-terminal sequences. Neither of these products yielded Up after RNase A digestion, but the product moving faster in the pH 3.5 electrophoresis cellulose acetate step yielded pU after snake venom phosphodiesterase digestion. Thus, it is apparent that at least one additional uridine residue may be present at the 3' terminus of the precursor molecule and the occasional presence of more than one 3'-terminal uridine residue in uiuo cannot be excluded. sequence plus three extra nucleotides at the 5' terminus (and thus the RNase P site) was not further cleaved by RNase P. After this substrate was exposed to RNase P, it was repurified by gel electrophoresis and analyzed by digestion with RNase T,. The RNA yielded a RNase T, fingerprint identical to the one shown in Fig. 3, panel C. The intact 5'.terminal of this RNase T,-generated product, ApApUpGp, was still observed and no pGp was apparent. Thus, no nucleotides were removed from either end of the molecule.

DISCUSSION
The precursor to the 4.5 S RNA has several interesting features. The 5' terminus begins with pppGp, showing that this molecule is a gene transcript unaltered at its 5' terminus. The The mature sequence starts with the guanosine residue on the 3' side of the RNase P cleavage site. The long hairpin in the mature sequence was presented as a hypothetical stable hydrogen-bonded structure by Griffin (1). The identity of the residue in position 98 of the mature 4.5 S RNA is not unique (see Table I, Footnote d). An extra uridine residue is shown at the 3' terminus (see text). The feathered arrows in the 5' hairpin loop indicate sites of cleavage during RNase T, partial digestion. sequence proximal to the 3' terminus of the molecule indicates it may be a remnant of transcription termination (7, 17, 18) since it can terminate with at least one and possibly more uridine residues. If the 3' terminus in the 4.5 S RNA precursor