The 3′ end of large ribosomal subunit RNA from mosquito mitochondria: Homogeneity of transcribed moieties

PLASMID

13, 139-144 (1985)

The 3’ End of Large Ribosomal Subunit RNA from Mosquito Mitochondria: Homogeneity of Transcribed Moieties DONALD Department of Microbiology,

T. DUBIN AND CHUEN-CHIN

UMDNJ-Rutgers

HSUCHEN

Medical School, P.O. Box 101, Piscataway, New Jersey 08854

Received December 26, 1984 3’ Terminal sequences of mosquito mitochondrial 16 S ribosomal RNA, which is posttranscriptionally polyadenylated, have been examined with the aim of determining the degree of homogeneity of transcribed moieties. 3’ End-labeled sampleswere subjected to oligonucleotide fingerprint analysis and to ladder gel analysis after primary and secondary nuclease digestion; and products of reverse transcriptase reactions were characterized using 16 S RNA as template and selected oligodeoxynucleotides as primers. The results indicated a remarkable degree of homogeneity compared to homologous mammalian mitochondrial systems, and suggested differences in modes of expression of insect, versus mammalian, mitochondriaf genomes. 0 1985 Academic

Ress Inc.

The large ribosomal subunit (16 S) RNAs of mammalian mitochondria have unusually heterogeneous 3’ termini. A portion of this heterogeneity is the result of post-transcriptional oligoadenylation, but a portion is also ascribable to ragged transcribed moieties (Dubin et al., 1981, 1982a, 1982b; Van Etten et al., 1983). This latter phenomenon has contributed substantially to the idea that most mammalian mitochondrial (“mit”) 16 S RNA molecules arise via termination of transcription, rather than via processing (see Attardi, 1984, for review). Mosquito mit 16 S rRNA is also posttranscriptionally adenylated (in this case largely polyadenylated; see Dubin et al., 1982a; HsuChen and Dubin, 1984). In the present work we examine the 3’-terminal sequencesof mosquito mit 16 S RNA with the aim of elucidating the degree of homogeneity of the transcribed moieties. The results indicate little or no transcript heterogeneity, and suggestthat there are significant differences between insects and mammals with regard to expression of mit rRNA and of the mit genomes in general. MATERIALS

AND

METHODS

Mosquito (Cedes albopictus) cells were grown, and the mitochondrial fraction pre-

pared, as described in Dubin and HsuChen (1983). 16 S RNA was purified using multiple phenol extractions followed by serial sucrose density gradient sedimentation through “low” and then “standard” salt (Dubin, 1974). Samples were 3’ end-labeled with [32P]pCp and repurified by acrylamide gel electrophoresis as described by Baer and Dubin (198 1). No oligo-T treatment was employed. For “complete” digestion with RNAses Tr or A, end-labeled samples containing 10 pg of RNA (mainly unlabeled carrier) were treated for 30 min at 37’C with 0.08 unit of RNAse T, or 0.001 pg of RNAse A in 5 ~1 of 0.0 1 M Tris . HCl, pH 8. These low enzyme levels [lo% of those used in similar studies on HeLa RNA species(Dubin et al., 1982b)] sufficed for essentially complete releaseof 3’terminal oligonucleotides with minimal artifactual attack of A-A bonds; even so, RNase A digests generally had levels of small oligonucleotides (especially AzCp and ACp) that, although low, were more abundant than expected on the basis of results of RNAse T1 results (cf. Dubin and HsuChen, 1983). Thus, for quantitative examination of 3’ adenylation in the oligo-A range, we relied on RNAse T, (v.i.). Fingerprint and ladder analyses were as described earlier (Baer and Dubin, 1980, 1981).

139

0147-619X/85 $3.00 Copyright 0 1983 by Academic Press. Inc. All rights of reproduction in any form reserved.

140

DUBIN AND HSU CHEN

Reversetranscriptase reactions were essentially as described by Ojala et al. (198 l), except that, presumably due to the high A + T content of Aedes mit 16 S RNA, we found it necessaryto incubate at 20°C rather than 39°C; the time of incubation was extended to 45 min. For primers, synthetic oligodeoxynucleotides with free 5’- and 3’hydroxyl groups (New England Biolabs) were 5’ end-labeled using [T-~~P]ATP and T4 polynucleotide kinase, and were purified by gel electrophoresis (Ojala et al., 1981). Seikagaku American, Inc., was the source of reverse transcriptase (avian myeloblastosis virus) and Boehringer-Mannheim Biochemicals was the source of deoxynucleosidetriphosphates. RESULTS

We have previously presented data localizing the end of the Aedes mit 16 S RNA gene to a T residue (T1332 of Fig. 1) preceding the next gene (that for tRNAle”) or to one of the immediately downstream A residues (HsuChen and Dubin, 1984). These studies, employing nuclease S, protection analysis, and ladder gel comparisons of products of RNase T,, vs. RNAse A digestion of 3’ end-labeled RNA, left open the question of homogeneity of the transcribed moieties of the gene. For example, although ladders generated from end-labeled RNA by RNases T, or A showed discrete and regularly spaced bands (HsuChen and Dubin, 1984) similarly generated ladders from HeIa mit 16 S RNA appeared similarly “clean” (our unpublished observations). Yet, when RNase T, digests of the HeLa RNA were subjected to fingerprint analysis, it became clear that they contained the products of two main classes of transcribed moiety, with either two or three terminal U residues (Dubin et al., 1982b); and that each apparently discrete band seen

1260

1290

1300

FIG. 2. Fingerprint analysis of RNase Ti-digested, 3’ end-labeled Aedes mitochondrial 16 S RNA. The digest was subjected to electrophoresis at pH 3.5 (right to left), and then homochromatography on a PEI plate (downward), as described in the text.

in ladder gels was a mixture of U2A,Cp and U3A,-,Cp. Fingerprint analysis (Fig. 2) of the Aedes RNA yielded a pattern compatible with the presence of a single adenylated Tireleased family corresponding to the end of the gene as presented in Fig. 1. However, such a family would have the theoretical composition (Ai2, U,,JA,Cp, and the occurrence of minor famihes arising from transcribed moieties whose ends correspond to residues upstream from T1332 or downstream from Al of Fig. 1 would have been difficult to detect in this system (seeVolckaert and Fiers, 1976). More revealing were experiments such as that illustrated in Fig. 3. RNase Ti-released, 3’ end-labeled bands were recovered from ladder gel separations, subjected to secondary digestion with RNase A, and rerun together with RNAse A digests of intact RNA to

1310

AAGGACCTAATATAAAAAATATAAATTTTAAATTTATGAATTATATTAATATTTAATAAA

1320

1330

1 ACTAT

FIG. 1. The 3’-terminal region of the Aedes mitochondrial 16 S RNA gene. We number from the 5’ end of the 16 S RNA gene and then, after a space, from the 5’ end of the tRNApAo gene, which follows.

3’ END OF INSECT MITOCHONDRIAL

A

I2

A

T,,AT

TEA -T

-C

141

rRNA

-G -T

-C

-G

FIG. 3. Secondary analysis of T,-released oligonucleotides. The left panel shows a ladder gel separation of Tidigested, 3’ end-labeled 16 S RNA. The bands corresponding to N&p and NSsCp(cf HsuChen and Dubin, 1984) were recovered, rerun to eliminate contamination with neighboring bands, and then digested with RNAse A. These secondary digestswere then subjected to ladder gel electrophoresis in parallel with primary RNase A digests of the original RNA sample (lanes “A” of right panel) to provide markers. Lanes 1 and 2 of this panel show, respectively, RNase A digestsof N&p and N&p.

provide size markers. As shown in Fig. 3, T1 bands corresponding to N&p and N&p yielded, after secondary digestion, bands running with A&p and A&p, respectively, and no (~2%) other bands. Similar results were obtained by secondary digestion of putative N&p, N&p, N&p, and N&p bands, which yielded A$Zp, A&p, A&p, and A&p, respectively. We conclude that each oligonucleotide size class releasedby T, digestion of 3’ end-labeled Aedes mit 16 S RNA is homogeneous with regard to the length of 3’-terminal poly- or oligoadenylate. Experiments employing reverse transcriptase were confirmatory and, in addition, supported our assignment for the sequence of the transcribed moiety. As shown in Fig. 4, T12A and TlzAT were efficient primers when highly purified 16 S RNA was used as template. Furthermore, when dCTP or dGTP were omitted from reaction mixes, ladder gels contained prominent bands that were one nucleotide shorter than oligonucleotides corresponding to each of the first several G’s or C’s (respectively) of the predicted template sequence (Fig. 4). This is presumed to result from “pausing” of the enzyme prior to incorporation of the omitted triphosphates (apparently present as trace contaminants; see Fig. 5, below). Most relevant to the present

FIG. 4. Reverse transcriptase analysis of the 3’ end of 16 S RNA. The left half of the gel displays reactions primed with Ti2A and the right half with Ti2AT, in all cases lacking the dNTP indicated at the tops of each channel. We indicate the inferred sequence of the complementary DNA using Aedes 16 S RNA gene numbering (Fig. 1).

work is the fact that the same band patterns were generated by T,zA as by TrZAT, as if both were priming at, and only at, the same

142

DUBIN AND HSU CHEN

(Maxam and Gilbert, 1980), a clear sequence corresponding precisely to the complement of the putative gene-end was obtained, as shown in Fig. 5 for products of a T12Aprimed reaction. Thus, within the limits of uncertainty resulting from the occurrence of post-transcriptional adenylation, transcribed moieties of all Aedes mit 16 S RNA molecules correspond to the end of the gene. Although the adenylation-derived uncertainty cannot be eliminated in the absence of a system-not currently available-in which post-transcriptional adenylation is selectively prevented, some suggestiveevidence on this point was obtained by detailed analysis of the smaller oligonucleotides released by RNase T, from 3’ end-labeled samples. The distribution of label among the various sized oligonucleotides (and hence among chainlength classesof 3’ poly- and oligoadenylate) was surprisingly complex, and in a manner that was consistent from preparation to prep__^__ aration (Fig. 6). The bulk (80-90%) of the A I332 end label appeared as a cluster of five main bands that correspond to RNase T1-released T oligonucleotides N5&Zp (or RNase A-reT leased A34-38Cp). There were also minor modes corresponding to T,-released N35-39Cp, T N4,&p, and N22-27Cp.Most pertinent, a sharp cutoff occurred below N&p (which T would correspond to RNase A-released A3Cp); this latter band contained about 0.5% FIG. 5. Sequence analysis of a T,*A-primed DNA of total label, whereas the regions corresegment. A reaction mixture was constructed as for Fig. sponding to N&p or N&p contained 4, except that both dCTP and dGTP were omitted. After ~0.1%. Considerations of parsimony lead us gel electrophoresis as for Fig. 4, bands corresponding to positions 1279 and 1280 were recovered and sequenced to propose that N&p arises from molecules, (Maxam and Gilbert, 1980). Numbering is as for Fig. 4. none of which have been post-transcription(The radioopacities in the region of residue 1305 arose ally adenylated, and that the end of the gene from contaminated intensifying screens inadvertently and the ends of all transcribed moieties may used for this series of experiments.) well correspond to Al 335 of the sequence of Fig. 1. unique UAIZ stretch. Finally, we directly DISCUSSION determined sequences for products both of Three setsof observations support the idea T12A- and TlzAT-primed reactions, run in the absence of dCTP and dGTP. Such sam- that most mammalian mit 16 S rRNA tranples yielded prominent bands corresponding scripts are generated by termination of tranto both the “C” and “G” sets of Fig. 4. scription (followed, or accompanied, by adWhen the bands corresponding to presumed enylation) rather than by processing of a pausing before C 1278/C 1279 were recovered larger transcript. (1) In hamster (Dubin et and subjected to partial chemical degradation al., 1981; Kotin and Dubin, 1984) and mouse G MG CM

C

3’ END OF INSECT MITOCHONDRIAL

FIG. 6. Distribution of end label among chain-length classes of 16 S RNA. Samples of 3’ end-labeled 16 S RNA were digested with RNase T, , followed by electrophoresis through a 20% acrylamide ladder gel; a region of 15 cm from about the center of the gel is shown. The left two lanes represent differing amounts of T, digest. The right lane shows an RNase A digest run to facilitate counting; we counted back from A&p, A&p (regularly the most prominent RNase A products) to locate A&p on a darker exposure; this oligonucleotide runs marginally slower than N&p.

(Van Etten et al., 1983) cells about 30% of the transcribed moieties of 16 S rRNA terminate at residues two to four nucleotides within what would normally be the next gene, and in human cells the transcript termini are split about evenly between the last and the next-to-last residues of the 16 S RNA gene (Dubin et al., 1982b). This contrasts sharply with the apparent precision with which precursor transcripts are processed to yield the mitochondrial messengerRNAs (see Attardi, 1984). (2) There exist near the 3’ ends of mammalian mit 16 S RNA genes highly conserved potential “hairpins,” with G * C-rich stems, that are followed by runs of two or more T’s. These constructs resemble abbreviated versions of bacterial transcription attenuators (Rosenberg and Court, 1979).

rRNA

143

(3) In mammalian mitochondria, the gene following the 16 S RNA gene encodes one of only three mammalian mit tRNAs (tRNA$J that retain relatively normal D and T arms and hence can fold into a conventional tight three-dimensional configuration; it has been proposed that this feature, too, is related to special constraints imposed by a crucial regulatory role of this region of the genome (Van Etten et al., 1983). The mosquito mit 16 S RNA gene lacks a 3’-terminal structure with features of a transcription attenuator (HsuChen and Dubin, 1984) the following gene (for tRNApAG) lacks conventional D and T arms (HsuChen and Dubin, 1984) and we have shown here that the transcribed moieties of 16 S RNA are strikingly homogeneous compared to mammalian counterparts. The “transcription attenuation” hypothesis is attractive as applied to mammalian mitochondria in that it explains the much higher rate of transcription of ribosomal RNA vis a vis the mRNAs, most of which are encoded downstream from the rRNA genes on the same strand (the H strand) (Gelfand and Attardi, 1981). This differential transcription is difficult to explain on the basis of a strong promoter for the rRNA genes vs. weak downstream promoter(s) for mRNA genes. No evidence for downstream initiation has been obtained (Montoya et al., 1982) nor has sequence analysis provided candidates for downstream promoters (Anderson et al., 1981, 1982; Bibb et al., 1981). Aedes mitochondria resemble mammalian in that rRNA occurs at a great molar excessrelative to putative mRNA (our unpublished observations; also Eaton and Randlett, 1978). No direct information is available on mechanisms of expression of insect mit genomes, and these genomes (at least the Drosophila) differ from the mammalian in that mRNA genesare more equally distributed between the two strands (seeClary et al., 1984). Nevertheless, as in mammals, the rRNA and many of the mRNA genesare on the same strand, and by far the largest noncoding intergenic stretch (which would be the most likely candidate to contain promoters) occurs upstream from the rRNA

144

DUBIN AND HSU CHEN

genes, which in turn are upstream of the mRNAs (Clary et al., 1984). Thus, transcription attenuation would seem a potentially useful strategy for regulation of expression of insect mit genomes as well. We propose that either this strategy is dispensed with, or that the molecular mechanism of attenuation in insect mitochondria differs significantly from that in mammals. Studies on kinetics of transcription, and the nature of primary transcripts, in insect systemsare clearly indicated.

DUBIN, D. T., HSUCHEN,C.-C., TIMKO, K. D., AzzoLINA, T. M., PRINCE, D. L., AND RANZINI, J. L. (1982a). 3’-termini of mammalian and insect cell mitochondrial ribosomal RNA. In “Mitochondrial Genes” (G. Attardi, P. Borst, and P. Slonimski, eds.), pp. 89-98. Cold Spring Harbor Press, Cold Spring Harbor, New York. DUBIN, D. T., MONTOYA, J., TIMKO, K. D., AND ATTARDI, G. (1982b). Sequence analysis and precise mapping of the 3’-ends of HeLa cell mitochondrial ribosomal RNAs. J. Mol. Biol. 157, 1-19. DUBIN, D. T., TIMKO, K. D., AND BAER, R. J. (1981). The 3’-terminus of the large ribosomal subunit (“ 17s”) RNA from hamster mitochondria is ragged and oligoadenylated. Cell 23, 27 l-278. ACKNOWLEDGMENTS This work was supported by Grant GM 14957 of the EATON, B. T., AND RANDLETT, D. J. (1978). Origin of the Actinomycin D insensitive RNA speciesin Aedes National Institutes of Health, USPHS. We thank Ms. K. albopictus cells. Nucleic Acids Res. 5, 1301- 1313. Timko for her expert technical assistance. GAINES,G., AND ATTARDI, G. (1984). Highly efficient RNA-synthesizing system that uses isolated human REFERENCES mitochondria: New initiation events and in viva-like ANDERSON,S., BANKIER, A. T., BARRELL, B. G., DE processing patterns. Mol. Cell Biol. 4, 1605-16 17. BRWIJN,M. H. L., COULSON,A. R., DROUIN, J., GELFAND, R., AND A~ARDI, G. (1981). Synthesis and EPERON,I. C., NEIRLICH,D. P., ROE, B. A., SANGER, turnover of mitochondrial ribonucleic acid in HeLa F., SCHREIER,P. H., SMITH, A. J. H., STADEN,R., cells: The mature ribosomal and messengerribonucleic AND YOUNG, I. G. ( 1981). Sequenceand organization acid speciesare metabolically unstable. Mol. Cell Biol. of the human mitochondrial genome. Nature (London) 1,497-51 I. 290,457-465. HSUCHEN,C.-C., AND DUBIN, D. T. (1984). Sequences ANDERSON,S., DE BRUIJN,M. H. L., COUISON,A. R., of the cooing and flanking regions of the huge ribosomnl EPERON,I. C., SANGER,F., AND YOUNG, I. G. (1982). subunit RNA gene of mosquito mitochondria. Nucleic Complete sequence of bovine mitochondrial DNA. Acids Res. 12, 7771-7785. Conserved features of the mammalian mitochondrial KOTIN, R. M., AND DUBIN, D. T. (1984). Sequences genome. J. Mol. Biol. 156, 683-717. around the 3’-end of a ribosomal RNA gene of hamster ATTARDI, G. (1984). RNA synthesis and processing in mitochondria: Further support for the “transcription mitochondria. In “Processing of RNA” (D. Apirion, attenuation” model. Biochim. Biophys. Acta 782, 106ed.), pp. 227-290. CRC Press,Boca Raton, Florida. 108. BAER,R. J., AND DUBIN, D. T. (1980). The 3’-Terminal MAXAM, A. M., AND GILBERT, W. (1980). Sequencing sequence of the small subunit ribosomal RNA from end-labeled DNA with base-specilicchemical cleavages. hamster mitochondria. Nucleic Acids Res. 8, 4927In “Methods in Enzymology” (L. Grossman and K. 4941. Moldave, eds.), Vol. 65, pp. 449-560. Academic Press, BAER, R. J., AND DUBIN, D. T. (1981). Methylated New York. regions of hamster mitochondrial ribosomal RNA: MONTOYA, J., CHRISTIANSON,T., LEVENS, D., RABIStructural and functional correlates. Nucleic Acids NOWITZ,M., AND A?‘TARDI,G. (1982). Identification Rex 9, 323-337. of initiation sites for heavy-strand and light-strand BIBB, M. J., VAN ET~EN, R. A., WRIGHT, C. T., WALtranscription in human mitochondrial DNA. Proc. BERG,M. W., AND CLAYTON,D. A. (1981). Sequence Natl. Acad. Sci. USA 79, 7195-7199. and gene organization of mouse mitochondrial DNA. OJALA, D., MONTOYA, J., AND ATTARDI, G. (1981). Cell 26, 167-180. tRNA punctuation model of RNA processingin human CLARY, D. O., WAHLEITHNER,J. A., AND WOLSTENmitochondria. Nature (London) 290,470-474. HOLME, D. R. (1984). Sequence and arrangement of ROSENBERG,M., AND COURT, D. (1979). Regulatory the genes for cytochrome B, URFl, URF4L, URF4, sequencesinvolved in the promotion and termination URFS, URF6 and five tRNAs in Drosophila mitoof RNA transcription. Annu. Rev. Genet. 13,319-353. chondrial DNA. Nucleic Acids Res. 12, 3747-3762. VAN ETTEN, R. A., BIRD, J. W., AND CLAYTON, D. A. DUBIN, D. T. (1974). Methylated nucleotide content of (1983). Identification of the 3’-ends of the two mouse mitochondrial ribosomal RNA from hamster cells. J. mitochondrial rihosomnl RNAs. J. Biol. Chem. 258, Mol. Biol. 84, 257-273. 10104-10110. DUBIN, D. T., AND HSUCHEN, C.-C. (1983). The 3’- VOLCKAERT,G., MIN Jou, W., AND FIERS,W. (1976). Analysis of 3ZP-laheledbacteriophage MS2 RNA by a terminal region of mosquito mitochondrial small riminifingerprinting procedure. Anal. Riochem. 72,433bosomal subunit RNA: Sequence and localization of 446. methylated residues. Plasmid 9, 307-320.