ribonuclease digestion followed by paper electrophoresis of released hydra- zones, all as previously described.2,8. RESULTS AND DISCUSSION. Fig. 1 shows ...
Volume 3 no.5 May 1976
Nucleic Acids Research
The 3'-terminal sequence of mitochondrial 13S ribosomal RNA
Donald T.Dubin and John Shine*'
Department of Microbiology, CMDNJ-Rutgers Medical School, Piscataway, NJ 08854, USA Received 18 February 1976 ABSTRACT We have examined the 3'-terminal sequence of the "small" structural ribosomal RNA ("13S") of hamster cell mitochondria, using a procedure involving [3H]isoniazide labeling of samples subjected to sequential periodate oxidation and s-elimination. The terminus was found to be PyUAUJAOH, which is similar, but not identical, to the corresponding terminus of eukaryotic cytoplasmic 18S rRNA.
INTRODUCTION The 3'-termini of 16S rRNA from E.coli2'3 and several other bacteria4 contain polypyrimidine sequences (in some cases proximal to a terminal A residue) that are complementary to purine-rich sequences in initiator regions of bacterial and bacteriophage mRNA. It has been proposed that this complementarity plays a role in mRNA-ribosome binding, 4 and a recent study5 provides direct experimental support for this idea. On the other hand, 18S rRNA from several eukaryotes terminates in .._AUUAOH (refs 6-10), to which a complementary sequence has been reported near the 5'-terminus of the mRNA of a plant virus, brome mosaic virus ("BMV") (ref. 11; see also discussions in refs. 5, 10 and 12.) Mammalian mitochondrial ribosomes occupy a peculiar evolutionary niche. They resemble bacterial ribosomes in antibiotic a fact that supports the idea that mitochondria may have evolved from primitive prokaryotic endosymbionts (see, e.g. ref. 14). On the other hand, mammalian mitochondrial ribosomal RNA has certain unique properties: low molecular weight [approx. 550,000 and 350,000 (ref. 15)]; absence of a con16; and few methylated The present ventional 5S studies address the question of whether the "small" 18 rRNA of hamster mitochondria (nominally 13S) nevertheless contains a 3'-terminus with homology to other "small" rRNA's; and if so whether it resembles 16S or 18S RNA in this respect.
sensitivity,13
species13,
residues.17
C Information Retrieval Umited 1 Falconberg Court London WI V 5FG England
1225
Nucleic Acids Research METHODS For isolation of 13S RNA, hamster (BHK-21) cells were grown and processed as previously described, 16, 17 with minor modifications to facilitate scaling up. Approx. 10 1 of culture at 8 x 105 cells/ml in exponential growth were handled at a time. Cell densities for mechanical disruption were increased to 4 to 7 x 107/ml. Crude mitochondrial pellets were suspended at a density of 5 to 7 x 108 cell equivalents per ml, and 2-ml portions were layered onto 30-ml sucrose gradients in Spinco SW-27 tubes for isopycnic centrifugation (25,000 rev/min, lh, O°). The mitochondrial bands were removed, combined and diluted with an equal volume of "hypotonic buffer"16 for subsequent pelleting. The final pellet was treated with SDS-containing extraction mixture (approx. one ml per 108 cell equivalents) and then subjected to 3 phenol extractions followed by ethanol precipitation.17 13S RNA was purified by density gradient sedimentation, alternately in sucrose containing low, and isotonic, levels of salt. 17 This procedure effectively separates 13S RNA from contaminating DNA, cytoplasmic 18S rRNA, and mitochondrial 17S rRNA (ref. 17). Yields of total mitochondrion-associated RNA were approx. 25 pg per 108 cells, 8-9% of which was 13S RNA (see Fig. 1 below).
1.0
1S0
4S
II.\o,o I
0.5 0 0
~
~
0
10 20 Fraction Number
30P
Fig. 1. Density Gradient Pattern of Mitochondrion-Associated RNA. The mitochondrion-associated RNA from 16 1 of cells, obtained as described in Methods, was divided into six aliquots, each of which was layered onto "low salt" gradients in SW-41 tubes and sedimented for 18h at 36,000 rev/min. (ref. 17). 30 0.4-ml fractions were collected from each gradient and the pellets were taken up in 0.4 ml. Absorbancy at 260 mp was read on appropriate dilutions of each fraction and is expressed relative to the undiluted sample (1 cm path length); a typical pattern is presented. Contaminating cytoplasmic rRNA, which constitutes approx. 60% of the total RNA of such preparations, appears in the pellet fraction ("P"). Sedimentation was from left to right.
1226
Nucleic Acids Research BHK 18S rRNA was prepared from similar cultures, essentially as previously described. 19 Sequencing was performed using a cyclic procedure especially adapted for accurate analysis of pyrimidine-rich 3'-termini. Samples were subjected to sequential periodate oxidation and s-elimination; at each step 3'-termini were labeled with [3H]isoniazid and characterized by pancreatic ribonuclease digestion followed by paper electrophoresis of released hydrazones, all as previously described. 2,8 RESULTS AND DISCUSSION Fig. 1 shows a representative mitochondrial RNA sedimentation pattern, from a "low salt" gradient. "Cuts" from the 13S regions of such gradients were combined, and after two serial recentrifugations in "standard salt" (cf ref. 17) yielded the pattern of Fig. 2. The final preparation (bracket) was estimated to contain less than 5% contaminating 17S or 18S RNA.
3.0 -a
2.0
S.
-135 1.0~~~~~~~
0ooy 10
7S 18S
20 Fraction Number
I
\'0 2.0 at 0.5 cm path length). 500 pg of purified 13S RNA was recovered (bracketed fractions). The positions of mitochondrial 17S rRNA and cytoplasmic 18S rRNA in parallel runs are indicated by arrows [18S RNA runs much closer to mitochondrial rRNA in standard salt than in low salt (17)].
1227
Nucleic Acids Research Table 1.
3'-Terminal Sequence Determination of 13S RNA
Number of Stepwise Degradations before Labelling with [3H]isoniazid
Labelled Hydrazone Released by Pancreatic Ribonuclease
0 1 2 3
4 *iNicHz:
A-iNicHz* U-iNicHz AU-iNicHz A-iNicHz U-iNicHz
Sequence
PYAOH
PyUAOH PyAUUAOH PyAUJAOH PyUAUUAOH
isonicotinoylhydrazone
The results of a 3'-terminal analysis on this preparation are shown in Fig. 2 and are summarized in Table 1. A second preparation (not illustrated) yielded equivalent findings. The patterns of Fig. 3 indicate that 13S RNA is essentially homogeneous with regard to its 3'-terminus. The sequence obtained, PyUAUUAOH, differs from the 3'-terminal hexanucleotide sequences of all 16S and 18S ribosomal RNA's thus far examined in this regard (Table 2). The 3'-terminus of the homologous (18S) rRNA, as determined by the same method, was identical to that found for other 18S RNA's, . . GAUCAUUAOH (Table 2). Interestingly, the 3'-terminus reported for another organelle rRNA, Euglene gtacilis chloroplast 16S RNA, is even more distinctive than that of mitochondrial 13S rRNA (ref. 20 and Table 2). The 13S 3'-terminus does bear a significant resemblance to the terminus of 18S RNA in that both contain the . . AUUAOH tetranucleotide that has been postulated to pair with the ..UAAU.. sequence of the proposed BMV mRNA ribosome-binding site (cf Introduction). In fact, the 13S sequence carries this complementarity one or perhaps two nucleotides further, since the mRNA sequence in question continues as (5') ... UAAUAA.... (ref. 11). The present results contribute to evidence16' 17 that mammnalian mitochondrial ribosomes did not evolve, at least not in straightforward fashion, from ribosomes of the prokaryotic type. The homology between the 13S and the 18S 3'-termini, and the complementarity of both to the putative ribosome-binding sequence of BMV RNA, suggest parallel evolution under constraint(s) common to the mitochondrial and extra-mitochondrial compartments of eukaryotic cells. An obvious example of such a constraint might
1228
Nucleic Acids Research
0
GU C N HA I
41
i G
GUJ AC N Ii 'I AU
ACN
2
a2
2
0 x
a-
C-4
4
GU~~~~~~~~~~~~~~~~~~~~~~~~ A_ 4 GU C N G ACN
3
II II
0
4
tUIUII
n.k JLIL
2
15 30 0 15 30 Distance from Origin(Cm) Fig. 3. Paper Electrophoresis of Pancreatic Ribonuclease Digests of Stepwise-degraded, Terminally-labelled Mitochondrial 13S FAA. Purified 13S RNA (Fig. 2, bracket) was subjected to four stepwise degradations. After the removal of successive 3'-nucleotides, a sample was removed, labelled with [3H]isoniazid and digested with pancreatic ribonuclease. UIder the conditions used (8) the reaction between isoniazid and periodate-oxidised 13S RNA was quantitative, as shown previously for cytoplasmic rDNA (19). The removal of 3'-terminal nucleotides was greater than 951 up to 4 cycles of stepwise degradation with little change in the sedimentation profile of the RNA (see ref. 6). A portion of each digest was mied with unlabelled marker nucleoside hydrazones and subjected to paper electrophoresis (8). The arrows labelled G, U, A, C, AU and N refer to the positions of the corresponding sarker hydrazones and free isoniazid. The hydrazone of AC (not added in these runs) migrates between those of U and A (ref. 8). Differences in the absolute c.p.a. present on each electropherogram reflect differing sample sizes used for electrophoresis. 0-4 refer to the number of stepwise degradations before labelling with isoniazid.
Table 2.
3'-Termini of "Small" Ribosomal RNA's
Sequence Source PyUAUUAOH Mitochondria (hamster) Eukaryotic Cytoplasm* GAUCAUUAOH Bacteria PyCUCCUJUAOH Eschericha coli, Pseudomonas aeruginosa Bacillus sterothermophilus CCUUUCUAOH Bacillus subtilis, Caulobacter crescentis PyCUUUCUOH Chloroplast (Euglena gracilis) ACAACUCNOH
Reference This work
5-9 & this work
2-4 4 4 20
*yeast, drosophila, mouse, rabbit, moth, hamster
1229
Nucleic Acids Research be a requirement that mitochondrial ribosomes translate certain mRNA's that are of nuclear origin, and that (perhaps consequently) have a "cytoplasmic" type of 5'-terminus. There is in fact evidence that RNA of rat nuclei can enter mitochondria, and that it associates with the mitochondrial ribosomes; 21 and that tetrahymena mitochondrial ribosomes can translate rabbit hemoglobin mRNA.22 On the other hand, we have obtained evidence from hamster cells23 that mitochondrial messenger RNA lacks the "m7G-capped" 5'-terminus that appears to be a rather general requirement for the proper binding of mRNA to cytoplasmic ribosomes (e.g., ref. 12, 24). These several observations are compatible with the hypothesis that eukaryotic cytoplasmic ribosomes utilize two kinds of mRNA interaction for formation of initiation complexes: one involving m 7G and the other involving the messenger . .UAAU.. sequence; whereas animal mitochondrial ribosomes use only the latter type. Further studies on mitochondrial mRNA, including 5'-terminal sequence analysis, should provide a test for this hypothesis, and experiments in this direction are planned.
ACKNOWLEDGMENTS This work was supported by NIH grant No. GM14957 and by a grant to Dr. L. Dalgarno from the Australian Research Grants Committee. We thank Mrs. K. Timko and Mr. R. Baer for expert technical assistance; Dr. G. Cleaves for help in setting up procedures for large scale isolation of mitochondrial RNA; and Dr. L. Dalgarno for advice on 3'-sequence analysis.
*Department of Biochemistry, School of General Studies, Australian National University, Canberra, A.C.T. 2600, Australia REFERENCES 1. 2.
Present Address: Department of Biochemistry and Biophysics, University of California School of Medicine, San Francisco, California 94143, U.S.A. Shine, J. and Dalgarno, L. (1974) Proc. Natl. Acad. Sci. U.S.A., 71,
1342-1346. 3. 4. 5.
Sprague, K.U. and Steitz, J.A. (1975) Nucleic Acids Res. 2, 787-798. Shine, J. and Dalgarno, L. (1975) Nature 254, 34-38. Steitz, J.A. and Jakes, K. (1975) Proc. Natl. Acad. Sci. U.S.A., 72,
6. 7. 8. 9.
4734-4738. Hunt, J.A. (1970) Biochem. J. 120, 353-363. Dalgarno, L. and Shine, J. (1973) Nature New Biol. 245, 261-262. Shine, J. and Dalgarno, L. (1974) Biochem. J. 141, 609-615. Eladari, M.E. and Galibert, F. (1975) Eur. J. Biochem. 55, 247-255.
1230
Nucleic Acids Research 10.
11. 12. 13. 14. 15. 16. 17. 18.
19. 20. 21. 22.
23. 24.
Sprague, K.U., Kramer, R.A. and Jackson, M.B. (1975) Nucleic Acids Res. 2, 2111-2118. Dasgupta, R., Shih, D.S., Saris, C. and Kaesberg, P. (1975) Nature
256, 624-628. Both, G.W., Furuichi, Y., Muthukrishnan, S. and Shatkin, A.J. (1975) Cell 6, 185-195. Borst, P. and Grivell, L.A. (1971) FEBS Letters 13, 73-88. Raven, P.H. (1970) Science 169, 641-646. Borst, P. (1972) Ann. Rev. Biochem. 41, 333-376. Dubin, D.T. and Montenecourt, B.S. (1970) J. Mol. Biol. 48, 279-295. Dubin, D.T. (1974) J. Mol. Biol. 84, 257-273. Terminology: "Small" rRNA refers to the smaller of the two structural RNA's of ribosomes (18S, 16S, or 13S, depending on the source) as opposed to the "large" (28S, 23S, or 17S) rRNA's. Shine, J. and Dalgarno, L. (1973) J. Mol. Biol. 75, 57-72. Zalben, L.B., Kissil, M.S., Woese, C.R., and Buetow, D.E. (1975) Proc. Natl. Acad. Sci., U.S.A. 72, 2418-2422. Gaitskhoki, V.S., Kisselev, O.I., and Neifakh, S.A. (1973) FEBS Letters, 31, 93-96. Dimitriadis, G.J. and Georgatsos, J.G. (1974) FEBS Letters 46, 96100. Taylor, R.H. and Dubin, D.T. (1975) J. Cell Biol. 67, 428a. Muthukrishnan, S., Both, G.W., Furuichi, Y., and Shatkin, A.J. (1975) Nature 255, 33-37.
1231
