Characterization of ribonuclease NU cleavage sites in a bacteriophage phi80-induced ribonucleic acid.

Ribonuclease NU, an endoribonuclease isolated from human KB tissue culture cells, can cleave a bacteriophage phi80-induced RNA at four distinct sites. Nucleotide sequence analysis of the eight cleavage products has shown that the enzyme produces oligonucleotides terminating in 3'-phosphate groups, and that the four cleavage sites are in the only nonhydrogen-bonded region of the substrate. Various aspects of the cleavage reaction with this RNA and with other substrates are discussed.


Ribonuclease
NU, an endoribonuclease isolated from human KB tissue culture cells, can cleave a bacteriophage 480-induced RNA at four distinct sites. Nucleotide sequence analysis of the eight cleavage products has shown that the enzyme produces oligonucleotides terminating in 3'-phosphate groups, and that the four cleavage sites are in the only nonhydrogen-bonded region of the substrate. Various aspects of the cleavage reaction with this RNA and with other substrates are discussed.
RNase NU,1 an endoribonuclease isolated from human KB tissue culture cells, was detected by virtue of its ability to cleave an Escherichia coli tRNATyr precursor molecule (1). During the course of the purification and characterization of this enzyme, it was found that substrates normally stable in tivo, such as tRNA, 5 S RNA, and rRNA, were resistant to attack by RNase NU, but substrates unstable in Go, such as the tRNATyr precursor and various bacterial mRNAs, were susceptible to attack. It is of obvious interest, if the above generalization concerning the action of RNase NU is correct, to understand what governs the specificity of this enzyme. To this end, it is useful to examine the cleavage sites in radiochemically pure substrates of known nucleotide sequence. Such an analysis has been carried out with the tRNATyr precursor molecule; two cleavage sites were found but no obvious similarities between these sites were identified. We have now extended this kind of analysis to include another substrate susceptible to RNase NU cleavage and of known nucleotide sequence. This substrate, a bacteriophage @O-induced RNA (MS) 62 nucleotides long (2), is unstable in tivo. A comparison of the cleavage sites in Ma with those in the tRNA precursor suggests that cleavage of RNA by RNase NU will occur only in regions devoid of secondary or tertiary structure, or both. 1 The abbreviations used are: RNase NU, endoribonuclease specific for unstable RNA (1); precursor, the RNA precursor (128 nucleotides) to E. coli tRNATyrsu3+ mutant A25; MS, bacteriophage @O-induced RNA 62 nucleotides long called I% by Pieczenik et al. (2) : 5 S RNA. ribosome-associated RNA with sedimentation coefficie& of 5 S.

EXPERIMENTAL PROCEDURE
All biological materials, chemicals, chromatographic media, enzymes, and biochemical methods were as described in the accompanying paper (1). Since MI is coded for by bacteriophage &30 SUX+ mutant A25. and it can be labeled with 32POF in Escheiichia coli infected w&h this phage, it was extracted from the same acrylamide gels used to prepare the tRNATyr precursor (1). of an acrylamide gel separation of the products of the reaction of RNase NU with Ma. The reaction was carried out and the products analyzed in a 10% acrylamide gel as described in the accompanying paper (1). 32P-labeled bulk Escherichia coli 4 S tRNA, which was eluted as a single band from another gel and which had been subjected to electrophoresis as such again in 10% polyacrylamide gels, was added to the completed reaction before applying the sample to the gel. The various RNA species were eluted from the gel and identified by nucleotide sequence analysis. Numbers in parentheses refer to distance, in centimeters, of each band from the origin.

RESULTS
The products of the reaction of RNase NU with Mt, as visualized by autoradiography after acrylamide gel electrophoresis of the reaction mixture which had been incubated under standard conditions (l), are shown in Fig. 1. The bands numbered 1 to 8 have been identified as cleavage products of MO by st.andard nucleotide sequencing techniques. The production of eight bands suggests four discrete cleavage sites and this notion has been verified through the sequencing studies (see below).
The electrophoretic mobilities of the various bands are also in agreement with the hypothesis that there are four cleavage sites which yield four pairs of bands, i.e., Site A (see Fig. 2, Structure A) gives Bands 1 and 8; Site B gives Bands 2 and 7; Site C gives Bands S and 6' and Site D gives Bands 4 and 5. Data from several preparative cleavage experiments indicate that the rates of cleavage at the different sites are in the following approximate ratios, A :B : C : D as 2 : 3 : 3 : 1. With our purest enzyme preparations, we never observed small fragments of M3 which would result from cleavage at one of these sites (e.g. A, see Fig. 2) followed by cleavage in the same molecule at another (e.g. D). Therefore, it appears that RNase NU cleavage at any one site is not a prerequisite for cleavage at another site.
The nucleotide sequence analysis of the RNase NU cleavage products was made simple because the complgte nucleotide sequence of Mt has been determined by Pieczenik et al. (2). The complete sequence, drawn in a hypothetical structure, is shown in Fig. 2. There are 12 different products resulting from the action of Ti ribonuclease on Ms. These are listed in Table I in the order (aside from the mononucleotide Gp) they appear in _ the sequence as it is read in the 5' to 3' direction.
Table-1 sum-1461 marizes those products of T1 digestion found in each of the eight RNase NU cleavage products and lists also any new products found. The four RNase NU cleavage sites of Ma were unequivocally identified as follows.
Site A-Band 8 contains Ti-produced oligonucleotides 1 to 3 as well as a new product which was further identified by ribonuclease A digestion as being UpApApCpApCp. Band 1 contains all the T1 products expected for MB except for those found in Band 8 and the Product 4; it has instead an extra mole of ApGp.
Thus Site A is identified as being between nucleotides 17 and 18 from the 5' end of the molecule.
Site B-Band 7 contains the first four (from the 5' end) Ti products in the Ma sequence as well as an extra spot in the position of Up migration.
Band 2 contains all the Ti products missing in Band 7 as well as an extra mole of Gp. The location of Site B is defined by the products of ribonuclease A digestion of Bands 2 and 7 (Table II).
That is, if the cleavage site is located between nucleotides 20 and 21 from the 5' end of the molecule, a ribonuclease A digestion of Band 2 should contain 1 mol only of GpUp and not, for example, ApGpUp, and this is seen to be the case.
Xite C-Band 6 contains the same four T, products as found in    (2) in their determination of the nucleotide sequence of 14 If a product occurs more than once, it is not listed twice.

$-A-G-G-C-C-A-G-U-A-A-A-A-G-C-A-U-U-
A number 1 indicates th 2' t the normal molar yield of the product (a~ listed in reference 2) was found and 0 indicates absence of the product. +'fhese products were found in one mole excess of the normal molar yield compared to the standard fingerprint.
The'quantitation is seldom exact: for di-and tri-nucleotides the excess yield was 0.6-1.0 mole; for Gp the value varied from 0.3-0.5 excess moles.
-TWO moles Of this product are normally found in one mole of i4 mole of UpGp is found.
These products are found in one mole ess than the normal molar ?' In bands 2 and 6, only one yield.
Yendof iV13aswellasanextraproduct, ApUp. Band4contains two-dimensional sketches. The difficulty in defining the preferthe Ti products missing in Band 5, except for Product 6; 2 mol ence of a ribonuclease for one kind of structure or another can of ApApGp were, however, found in the Ti analysis of Band 4. also be illustrated by the reaction of E. coli RNase III with an The RNase A digest of Band 5 indicated that all the nucleotides rRNA precursor molecule. This 30 8 molecule is a precursor to up to and including nucleotide 25 from the 5' end (the product 16 S and 23 S rRNA and is cleaved only once by RNase III (3, GpApUp was found) were present.
Therefore, Site D must be 4); the remaining rRNA is untouched. However, previous studbetween nucleotides 25 and 26.
ies have shown that RNase III will digest double-stranded re-In all analyses no evidence was found for the generation of new gions of RNA (5, 6) and sequences do exist in the mature rRNA 5'-phosphate-terminated oligo-or mononucleotides. Only 3'-which can be drawn in potential hydrogen-bonded structures (7). phosphate-terminated products of cleavage were found as was So RNase III seems to select, in a manner not well understood, also the case for the Escherichia coli tRNATy' precursor molecule particular hydrogen-bonded structures for attack. after reaction with RNase NU.
We have also found that in a preparation of 32P-labeled Q/3 RNA (which contained primarily one-half to one-quarter sized DISCUSSION intact molecules) RNase NU was capable of producing several The four cleavage sites of R4, by RNase NU are depicted in discrete cleavage products.
It is not clear, however, that this Fig. 2, StmLcture A. Pieczenik et al. (2) have provided evidence violates our general rule concerning in z&o instability of subthat the hydrogen-bonded structure of MS, as shown, probably strates susceptible to RNase NU attack since it is known, for does exist in z&o.
The four Rh'ase NU cleavage sites are located example, that there are more ribosomal binding sites exposed in in the only single-stranded region in this molecule. Similarly, the cleavage sites of RI%ase NU in the E. coli tRNATy' precursor the fragmented, compared to intact, RNA of the related bacterioare also located in the single-stranded region of a hypothetical phage RI7 (8). Extensive nucleotide sequence studies of a rehydrogen-bonded structure for that molecule (Fig. 2, Structure lated bacteriophage RNA reveal that at least one long segment B). The similarity in RNase NU specificity with the two sub-of this molecule is almost totally involved in secondary structure strates cannot be further detailed since the two regions of single (9). Shorter segments may be devoid of such structure but strandedness have different lengths and different primary se-could still be protected from nuclease action by tertiary folding of the molecule. quence. In both cases, however, it can be said that cleavage In addition to the result cited above, we have occurs no more than seven nucleotides removed from the stem of found that one particular fragment of Qp RNA (which is 26 a region of secondary structure. It may be that the property of nucleotides long and cannot be represented in a hydrogen-bonded RNase NU, that enables it to distinguish between RNAs (or parts structure*) is not cleaved at all by RNase NU, suggesting again thereof) that are quickly degraded in viva from those that are not, that an appropriate substrate for this enzyme must be more is the ability to recognize single-stranded regions adjacent to regions of higher order structure.
The specification of the details 2 A. Senear and J. A. Steitz, personal communication. The of such "higher order" structure cannot be made from studying experiments with Qp RNAs were done in collaboration with A. Senear. These are the critical di-or trinucleotides in determining the exact RN&se NU cleavage sites. All other RNase A products were found in the molar yields expected for each band assuming the four cleavage sites shown in Figure 2, structure A. No RNsse A digestion was carried out on Bands 7 or 6. complex than single-stranded RNA lacking any secondary or