THE JOURNAL OF BIOL~CICAL CHEMISTRY 0 1994 by T h e American Society for Biochemistry and Molecular Biology, Inc.

Vol. 269,No. 46,Issue of November 18,pp. 29112-29120, 1994 Printed in U.S.A.

Human Mitochondrial Transcription Termination Exhibits RNA Polymerase Independence and Biased Bipolarityin Vitro* (Received for publication, July 19, 1994, and in revised form, August 25, 1994)

Jin Shang and David A. Clayton From the Department of Developmental Biology, Stanford University School of Medicine, Stanford, California 94305-5427

29112

Downloaded from www.jbc.org by on April 14, 2008

Human mitochondrial 16 S rRNA 3'-end formationre- complex, release of the nascent RNAchain, and dissociation the quires a tridecamer template sequence and a trans-actof RNA polymerase from the DNA template. Three general its mechanisms of transcriptionterminationhavebeen ing proteinof -34 kDa. This protein binds tightly to elabotarget sequence and further analysis of the protein-DNA rated. The first employs an intrinsic terminator within the complex revealed that the DNA is bent. Either T3, T7, RNA molecule itself which prevents RNA polymerase from Escherichia coli, or yeast mitochondrial RNA polymer- maintaining its usual stable interaction with the nascent RNA ase produced transcripts mapping at this termination chain and results in dissociation the of the polymerase from the site. With these heterologousRNA polymerases, RNA 3'- RNA. The second mechanism involves an RNA-binding protein end formation was detected only in the transcription (exemplified by rho factor in bacteria) that directs dissociation of mitochondrial rRNA synthesis; of the nascent RNA. The third mechanism requires the action polarity opposite that the efficiencyof termination in the homologous human of a site-specific DNA-binding protein (like the termination RNA polymerase systemis approximately 2-fold greater factor for RNA polymerase I) which binds to a DNA sequence in this same opposite polarity. These results suggested downstream of the 3'-end of a nascent transcript. It appears the possible importance of biased bipolar transcription that termination of transcription initiated with human mitotermination in vivo. For wild-type mtDNA, the apparent chondrial RNA polymerase (h-mtRNA polymerase) uses the relative efficiency of termination in vivo reflected the third mechanism. a pathogenic values determined in vitro. Examination of The development of a n in vitro transcriptiodtermination human mtDNA mutation known to result in impaired termination in vitro showed no significant differences in assay that faithfully reflects the in vivo termination process relative transcript abundances in vivo, despite a lossof facilitated the initial characterization of mitochondrial tranin vitro termination efficiency in both directions. Re- scription termination (2). A tridecamer sequence (human cently, six additional mitochondrial disease-associated mtDNA nucleotides 3237-3249) located within thetRNAL'u(wR' point mutations have been reported that cluster at the gene was identified by mutagenesis to be an essential compohumanmitochondrialtranscriptiontermination site. nent of the termination process (3). Evidence for a titratable None of these resulted in significantly impaired tran- factor responsiblefor termination furthersuggested that a sitespecific DNA-binding protein plays a crucial role in terminascription termination in vitro. tion (3). Chromatographic separation of human mitochondrial lysates resulted in theidentification of a termination activity The two strands of mammalian mtDNA are each transcribed that footprinted a 28-base pair region centered atthe in the form of polycistronic RNA molecules which are subse- tridecamer sequence (4, 5). This termination activity, termed quently processed to yield the maturemRNA, tRNA, and rRNA mitochondrial transcription termination factor (mtTERM, also species. Transcription initiates from two distinct major sites mTERF), has been found t o reside within a group of proteins within the mitochondrial genome, either the heavy (H) strand ranging inmolecular mass from 33 to 36 kDa, among which the promoter (HSP)l or the light(L) strand promoter (LSP). Tran- most abundant species is a 34-kDa protein (5). Using DNA scripts initiated from the HSP fall into two classes reflecting affinity chromatography, three polypeptides, two of molecular were found t o differing degrees of expression. The promoter-proximal tran- mass 34 kDa andone of molecular mass 31 kDa, be associated with the footprinting capacity. The terminationscripts, which encompass the region of the genome from the tRNAPhegene t o the 16 S rRNA gene, are present at a level promoting activity appears toreside in the34-kDa components 50-100-fold higher than thepromoter-distal transcripts (1).It has (6). Knowledge of the precise mechanism by which is Termibeen suggested that transcription termination at the boundary of mtTERM mediates transcription termination limited. the 16 S rRNA gene and the downstream tRNAbum) gene is at nation is independent of the normal mitochondrial promoter, sequence was found t o function bidirectionleast partially responsible for the relative abundance of the pro- and the tridecamer ally in vitro(2). In addition,it has been reported thatmtTERM moter-proximal transcripts. a half-life of 40-50 min onits cognate In general, transcript elongation is very processive and re- binds tightly to DNAwith ( 7 ) . DNA binding site quires special signals for termination. Transcription terminaThe DNA sequence footprinted by mtTERM (human mtDNA tion usually includes three steps: pausing of the elongation nucleotides 3229-3256) is referred to as themtTERM binding site. A naturally occurring point mutation in themiddle of the * This investigation was supported by Grant GM33088-24from the mtTERM binding site(nucleotide position 3243) is found in the National Institute of General Medical Sciences. The costs of publication of this article were defrayed in part by the payment of page charges. majority of patients with a mitochondrial disorder known as This article must therefore be hereby marked "advertisement"in ac- mitochondrial myopathy, encephalopathy, lactic acidosis, and cordance with 18 U.S.C. Section 1734 solely t o indicate this fact. stroke-like episodes (MELAS) (8). The 3243 mutation hasbeen The abbreviations used are: HSP, heavy strand promoter; LSP, light shown to result in a deficiency in protein synthesis ( 7 , Q), alstrand promoter; bp, base paifis); MELAS, a mitochondrial disorder described as mitochondrial myopathy, encephalopathy, lactic acidosis, though the precise mechanism by which this point mutation leads tocellular dysfunction and ultimatelyclinical syndromes and stroke-like episodes.

Mitochondrial TFanscription Termination

29113

Downloaded from www.jbc.org by on April 14, 2008

Fortheconstruction of plasmidscontainingdisease-associated is poorly understood. This mutation results in impairment of point mutations, site-directed mutagenesis was carried out proper transcription termination in vitro due to a reduced af- tRNAL”u(wR’ 0 , 3 2 5 1 (A to G), finity of mtTERM for the mutatedmtDNA template (5).In the in pTERM(F)at the positions 3243(A to G), 3250 (T to 3252 (A to G), 3256 (C to T), 3260(A to G), and 3271 (T C) to to generate a rapid increase in the num- sevenmutatedversions:p3243(F),p3250(F), past several years, there has been p3251(F), p3252(F), ber of point mutations identified in human mtDNA that are p3256(F),p3260(F), and p3271(F), respectively. Thecorresponding believed to be disease-associated. Remarkably, seven of these seven mutated versionsof pTERM(R) were derived from these plasmids mutations (including the 3243 mutation) are clustered in the by placing the 163-bp EcoRV-HincII fragments in the reverse orientatRNAL”u(UUR’ gene, within or very close to themtTERM binding tion. The nucleotide sequences of these constructs were confirmed by site a t nucleotides 3250 (lo), 3251 (161, 3252 (121, 3256 (111, DNA sequencing. The plasmid pCY4-TM containing the mtTERM binding site in a 3260 (131, and 3271 (14). circularly permuted nucleotide sequence was constructed by inserting In this paper, we show that mtTERM bound to its mtDNA the 55-bp EcoRV-HzncII DNAfragment from pTM(F) intothe SmaI site binding site can promote the termination of transcription ini- of permutation vector pCY4 (18). The nucleotide sequence of the insert was by DNAsequencing. tiated with heterologous RNA polymerases in an orientation- in the circular permutation plasmid confirmed DNA Templates-The templates used for run-off transcription reaca physical blockdependent manner. This finding implies that tions with h-mtRNA polymerase and T3RNA polymerase were generage mechanism is involved intranscriptiontermination ated by restriction digestionof appropriate plasmids as indicated in the mediated by mtTERM. Our results also suggest that the ter- figurelegends.Forsc-mtRNApolymerase,HindIII-digested pSCmination mechanism involves a conformational change in the TERM(F) and EcoRV-digested pSCTERM(R) were used. For T7 RNA DNA. Interestingly, termination of complementary L-strand polymerase,XhoI-digestedpTM(F) and XhoI-digested pTM(R) were RNA synthesis (the strand of opposite sense compared with used. Termination of transcription by E. coli RNA polymerase was tested rRNA) is more efficient than 16 S rRNA 3’-end formation; this EcoRI, was the caseboth in vitro and in vivo.Although the previously using 3’-end dC-tailed templates. pTERM(F) was digested with dC-tailed, and then digested with HincII to generate a template with characterized 3243 MELAS mutation was defective in bipolar the mtTERM binding site in the forward orientation relative t o the dCtermination in vitro,there wasno obvious change in RNAabun- tail. For the construction of the template with the mtTERM binding site dance in vivo to implicate faulty termination duet o this muta- in the reverse orientation relative to dC tail, the pTERM(F) was digested EcoRV The 178-bpdCtion. All other additional reported mutations in this location of with HindIII, dC-tailed, and then digested with the mitochondrial genome did not result significantly in altered tailedEcoRI-HincII and 165-bpdC-tailedHindIII-EcoRVfragments were isolated and used as templates for transcription reactions. efficiencies of transcription termination in vitro.

Preparation of the Termination Factor and h-mtRNA PolymerasemtTERM and h-mtRNA polymerase were prepared as described previously (5). Mitochondria were isolated from 12 liters of human KB cells Enzymes-Restriction endonucleases, T3 and T7 RNA polymerases, using a sucrose gradient. The 5-130 mitochondrial lysate was chromatoand Oligonucleotide Tailing Kit werepurchased from Boehringer graphed on DEAE-Sephacel and P11 phosphocellulose. h-mtRNA poMannheim. RNase T2 was purchased fromLife Technologies, Inc. SP6 lymerase was assayed by using poly(dA-dT) templates and measuring RNA polymerase and Escherichiacoli RNA polymerase were purchased the incorporation of trichloroaceticacid-precipitablecountsusing from Pharmacia Biotech Inc. Purified h-mtTFA protein was provided by [a-32PlUTP(19). The termination factor was assayed by gel mobility Daniel J. Dairaghi of this laboratory.2 Purified yeast mtRNApolymershift analysis, DNaseI footprinting, and the transcriptiodtermination ase (sc-mtRNApolymerase) and sc-mtTFB were provided by Baoji Xu of reaction as described previously ( 5 ) . The peak fractions were pooled, this 1aborato1-y.~ and bovine serum albumin was added to 100 pg/ml. Samples were Cells a n d Cell Culture-The wild-type cybrid lineML3 mo6 (contain- concentrated using Centricon10 (Amicon) by centrifugation at 5000 x g ing 100% wild-type mtDNA) and the 3243 MELAS mutation cybrid line for 40 minfollowed by dialysis against a buffer containing 10 mM TrisML3 mo2-3-f (containing 90% mutated mtDNA) were kindlyprovided C1, pH 8.0, 50 mM KCl, 0.1 mM EDTA, 1mM dithiothreitol, 1 mM phenby Dr. Jun-ichi Hayashi (Universityof Tsukuba, Tsukuba, Japan). Cells ylmethylsulfonyl fluoride, and 50%glycerol. were grown in glucose-rich medium RPMI 1640 supplemented with To obtain mtTFA-free termination factor for circular permutation pyruvate (0.1 mg/ml) and fetal bovine serum (10%) (17). assays (since mtTFA binds and bends the vector DNA (ZO)), denaturConstruction of Rasmids-Plasmid pTERM(F) was constructed by ation-renaturation chromatography was performed as described (21). inserting the 308-bp BamHI-Hind111 fragment from plasmidpLSPIsolated mitochondria were lysed in boiling sodium dodecyl sulfateTERM (5) intothe BamHI-Hind111 polylinkersites of pBluescript containing buffer. The mitochondrial extracts were loaded onto a hyKS(+). The inserted fragment contains the human mitochondrial LSP droxylapatite column, and the proteins were eluted with a linear graregion(nucleotides 445-3231, a n EcoRI linker, and the termination dient from 0.1 to0.5 M NaCl (mtTERM and mtTFA were separated a t region (nucleotides 3160-3314). Plasmid pTERM(R) was derived from this step of purification). The mtTERM-containing fractions, identified pTERM(F) by placing the 163-bp EcoRV-HincII fragment in the inby gel mobility shift assays, were bulk-renatured by addition of a n verted orientation such that the termination site was in the reverse excess amount of Triton X-100 and chromatographed on P11 phosphoorientation relative to the LSP (see Fig. 1, schematic diagram). cellulose with a linear 0.1-1.0 M NaCl gradient. The fractions containThe plasmid pTM(F) and pTM(R) were constructedby cloning a 34-bp ing mtTERM activity were dialyzed in a buffer containing 10 mM Trissyntheticdouble-strandedfragment (5’-AGC”TGTTAAGATGGCAC1, pH 8.0, 50 mM NaCl, 0.1 mM EDTA, 1 m~ dithiothreitol, 1 mM GAGCCCGGTAATCGCA-3‘) into the HindIII site of plasmid pBluephenylmethylsulfonyl fluoride, and 50% glycerol. script I1 KS(-). The sequence and orientation of the insert were conIn Vitro Danscription Reactions-Run-off transcription reactionsusfirmed byDNA sequencing. The plasmids were named either pTM(F) ing or h-mtRNA polymerase were carried out as described ( 5 ) . Standard pTM(R) to reflect the orientation of the mtTERM binding site in the reactions were carried out in 25 p1 containing 10 mM Tris-C1, pH 8.0, 10 forward or reverse orientation, respectively, relative to the directionof mM MgCl,, 1 mM dithiothreitol, 100 pg/ml bovine serum albumin, 400 transcription from the T3 promoter. PM ATP, 150 p~ CTP, 150 p~ UTP, 4 p~ GTP, 4 pCi of [(Y-”P]GTP(800 To construct a plasmid with the termination region downstream of a CUmmol), and 1 pg/ml DNA template, 10 ng of h-mtTFA, 1 pl of hyeast mitochondrial promoter, pTERM(F) was digested with EcoRI, mtRNA polymerase fraction in thepresence or absence of I-pl mtTERM blunt-ended withKlenow fragment and then digested with HindIII. The fraction. After a 30-min incubation a t 30 “C,the reactions were stopped 312-bp EcoRI-Hind111 fragment was inserted into the HpaI-Hind111 by adding an equal volume of a solution containing 20 nm EDTA, 1% sites of plasmid pO5-36g3 such that the orientationof the termination SDS, and 0.5 mg/ml proteinase K and incubated at 37 “C for 30 min. site was forward relative to the yeast promoter. This plasmid was Ribonucleic acids were phenol/chloroform-extracted, precipitated with named pSCTERM(F1. Plasmid pSCTERM(R) was constructedby insert- ethanol, denatured in 80%formamide at 95 “C for 5 min, and separated ing the146-bp EcoRV-Hind111 fragment of pTERM(F) intothe HindIII- on 6% polyacrylamide,7 M urea gels. Transcription reactions using EcoRV sites of plasmid pO5-36g3. other RNA polymerases were carried out underthe samereaction conditions exceptthat radiolabeled UTP was used insteadof GTP, and the appropriate amount of KC1 was added: 2 0 4 0 mM for T3 and T7 RNA D. J . Dairaghi, G. S. Shadel,and D.A. Clayton,submitted for publication. polymerases, 100 mM for E. coli RNA polymerase, and 30 mM for scB. Xu and D. A. Clayton, submitted for publication. mtRNA polymerase. Either 0.06 unitof T3 RNA polymerase, or 3 units EXPERIMENTAL PROCEDURES

29114

Termination lFanscription Mitochondrial

of T7 RNA polymerase, or 0.1 unitof E. coli RNA polymerase, or 20ng of sc-mtRNA polymerase with 20 ng of sc-mtTFB was used in each of these reactions. Quantitationof autoradiographs was accomplished using a n AMBIS Radioanalytic Imaging System (AMBIS System, Co.) or nn a PhosphorImager (Molecular Dynamics). - + + n + rntTERM For time course experiments, transcription reactions were initiated 1 in 100-pl volumes with 800 p~ ATP, 800 p~ CTP, 800 p~ GTP, 80 pCi of [a-”P]UTP (800 CVmmol), 0.24 unit of T3 RNA polymerase, and8 pl of the mtTERM peak fraction. Aftera 1-min 45-s incubation a t 30 “C, 2 1.11 of 100 mx unlabeled UTP was added toa final concentration of 2 mM, and thereaction was continueda t 30 “C.Aliquots were withdrawnfrom transcriptionreactions a t t h e indicatedtimesandwereassayed as RO described above. Gel Mobility Shift Assays-Mobility shift assays were performed a s described (5). A typical reaction contained 1.0 ng of an a-”P-5’-end labeled EcoRI-Hind111 fragment of pTERM(F). The competitor DNA, C T unlabeled EcoRI-Hind111 fragments of pTERM(F), or of the plasmids carrying each of the disease-associated mutations were added in the amounts as listedin the figure legends. Protein-DNA complexes were T + analyzed by electrophoresis through 4 8 nondenaturing polyacrylamide gels. Circular Permutation Assays-Circularly permuted DNA plasmid pCY4-TM was digested with one of the enzymes indicated in Fig. 3, 5‘-end-labeled, and gel-isolated. These DNA fragments were used gel in mobility shift assays with mtTERM purifiedby denaturation-renaturation chromatography. The bending angle was estimated by calculating LSP the ratio of the mobilities of the fastest and the slowest migrating d +b e+complexes in the mobility shift assay and by linear interpolation bepTERM(F) I M I tween points obtained with A-tractDNA standards (22). RO (262 nt) RNA Analysis-The acid guanidinium thiocyanate-phenol-chloroT (170 nt) form extraction method (23) was used for the isolation of total RNAfrom ML3 mo6 and ML3 mo2-3-f cells. RNase protection assays were carried out as described by Saccomanno et al. (24). Specific complementary RNA probes were synthesized in vitro using either T7 or SP6 RNA LSP ,.A .AG polymerases. Plasmid pBS-16s was constructed by cloning the DraI&O +Jb+b HincII fragment of human mtDNA nucleotides 2050-2424 into the I pTERM(R) EcoRV site of pBluescript I1 KS(-). Polymerasechainreactionwas I I t 1 I carried out using primer 1 (3960-3979, forward) and primer 2 (4152pRO (284 nt) 4171, reverse) to amplify partof the ND1 gene (human mtDNA nucleT (210 nt) otides 3960-4171). The TA Cloning Kit (Invitrogen) was used to subFIG.1. Run-off transcriptiodtermination assays wereperclone the polymerasechainreactionproductsintoplasmid pCR to generate plasmidpCR-ND1. The nucleotide sequences andthe orienta- formed using h-mtRNApolymerase andh-mtTFAin the absence tion of the insert were confirmed by DNA sequencing. The 16 S rRNA (-) or presence (+) of mtTERM on linearized templates. The four pTERM(F1, HindIIIprobe was synthesizedby T7 RNA polymerase from the templatepBS- DNA templatestestedare:HindIII-digested p3243(F), XhoI-digested pTERM(R), and XhoI-digested 1 6 s linearized at theAvaII site. The ND1 probe was synthesized by SP6 digested RNApolymerase from pCR-ND1 linearized at theNot1 site. Probes were p3243(R). RO, run-off transcripts; T, terminated transcripts.In the purified on denaturing polyacrylamide gels. 20 of pgtotal RNA isolated schematic diagram, the open box indicates DNA template; bent arrow from each cell line was hybridized with 5 x lo5cpm of each probe for represents transcription promoter; filledbox, termination region; wavy 12-16 h at 50 “C. After treatment with50 unitdm1 RNase T2, RNA was lines, transcripts; Eu, EcoRV; H d , HindIII; Hc, HincII;Xo, XhoI. precipitated with isopropyl alcohol and analyzed by denaturing polyacrylamide gel electrophoresis followed by autoradiography. a supercoiled circular template. The results were consistent

-

Human mtTERM Functions Bidirectionally with h-mtRNA Polymerase with a Higher Termination Efficiency in theReverse Orientation-The human mtDNA transcription terminationsequence functions bidirectionally in vitro (2). We designate the orientation as “forward” (F) when the mtTERM binding site as it is with relative to thepromoter is in the same orientation respect to transcription across the human mtDNA rRNA gene region and as“reverse” (R) when the mtTERM binding site was placed in theopposite polarity. In order to compare termination efficiency with the mtTERM DNA binding site in each of the two opposite orientations, we performed transcription reactions using templates with the mtTERM binding site cloned downstream of the standard test promoter in either the forward or reverse orientation.When the mtTERM binding site was in the forward orientation, 20-30% of the transcriptswere terminated in the presence of mtTERM (Fig. 1).With the mtTERM binding site in the reverse orientation, but under the same reaction conditions and in thepresence of the same amountof mtTERM, the termination efficiency was 2-3-fold higher (5070%) (Fig. 1).Since mtDNA is a closed circular molecule, we tested if mtTERM directs transcription terminationin vitro on

with thoseperformed using thelinearized templates in that the termination efficiency was higher when the mtTERM binding site was in the reverse orientation than in the forward orientation (data not shown). Theseresults indicated that mtTERMmediated termination is more efficient in the reverse rather than theforward orientation in vitro and raised thepossibility that this unexpected bias was operative i n vivo as well. The 3243 MELAS mutation results in reduced affinity of mtTERM for its target site and, in turn, decreased in transcription termination in theforward orientation in vitro (5).To test whether this mutationalso causes defective termination for the reverse orientation, transcription reactions werecarried outon DNA templates carrying the3243 MELAS mutation. Thismutation affected transcription termination in the reverse orientation; however, the effect was not as dramatic as seen for the forward orientation. The 3243 MELAS mutation resulted in a >85% decrease in termination efficiency for the forward orientation, whereas for the reverse orientation therewas an -50% decrease in the terminationefficiency (Fig. 1).Similar results were observed when closed circular DNA templates were used (data not shown). Human mtTERM Mediates Termination of Danscription by Heterologous R N A Polymerases in an Orientation-dependent

Downloaded from www.jbc.org by on April 14, 2008

RESULTS

5

29115

Mitochondrial Danscription Termination

* RO

o-

RO

-4-r

T3 Promoter

T3 Promtcr A

pTERM(R)

Not1

RO (227 nt) T(130nt)

Manner-In order to determine whether mtTERM is capable of polymerase in a manner similar to thatwith its cognate RNA mediating terminationof transcription by any RNA polymerase polymerase. As with h-mtRNA polymerase, transcription termination efficiency with these heterologous RNA polymerases or whether it is specific for h-mtRNA polymerase, plasmids is also dependent on the orientation of the mtTERM binding wereconstructed such that the mtTERM bindingsitewas placed downstream of various promoters. The standard pro- site, with an even stronger bias for the reverse orientation moters for T3 RNA polymerase, T7 RNA polymerase, and sc- (Table I). The shorter transcripts could result from transcription termtRNA polymerase wereused, butfor E. coli RNA polymerase, 3“end-tailed templates were employed, since the addition of mination ortranscription pausing. To distinguish between the mtTERM protein fraction significantlyinhibited transcrip- these two possibilities, a time course experiment was carried out. The transcription reaction was initiated for a very short tion initiation at the bacterial promoter. Each template was transcribed with its corresponding RNA polymerase in the ab- time in the presence of radiolabeled UTP followed by a chase sence or presence of mtTERM and using eitherof two orienta- with a 2000-fold excess of nonradiolabeled UTP (Fig. 2B ). The tions of the mtTERM binding site. When the termination site shorter transcripts persisted for up t o 60 min of incubation, was inthe forward orientation,transcription across the longer than the half-life of mtTERM on its DNA binding site, were not formed by tranmtTERM binding site on linearized templates generated only implying that the shorter transcripts full-length run-off transcripts regardlessof the presence of mt- scription pausing. We conclude that the shorterRNAs are the TERM (data not shown). The length of the transcripts was predicted terminated transcripts. consistent with transcription startingat the promoter and exSince the templatesused for heterologous RNA polymerases tending to the endof the template. When the termination site contained additional human mtDNA sequences flanking the was placed in the reverse orientation, there were two classes of mtTERM binding site, we synthesized an oligonucleotide fragtranscripts generated in thepresence of mtTERM: longer run- ment containing only the mtTERM binding site and inserted it off transcripts and shorter truncated transcripts whose length into heterologous sequence contexts in order to test for the corresponded to termination occurring immediately upstream ability to support mtTERM-directed termination of transcripof the mtTERM binding site (Fig. 2, A and C). To confirm that tion by T3 RNA polymerase. When the polarity of the the formation of the shorter transcripts was due tomtTERM mtTERM binding sequence was placed in the reverse orientabinding at its DNA binding site, a template with the 3243 tion relative to the T3promoter, -20-30% of the transcription terminatedtranscripts MELAS mutation in the mtTERM binding site was transcribed productsweremtTERM-dependent with T3 RNA polymerase inthe absence or presence of (Fig. 2C). Alow abundance RNA species shorter than the runmtTERM. As shown in Fig. 2 4 , the formation of shorter tran- off transcripts was present regardless of the presence of mtscripts wassignificantly impaired using the mutated template TERM, and it wasprobably the result of either degradation of transcribed by T3 RNA polymerase, as compared with the tran- the run-off transcripts ormtTERM-independent pausing or terscription products from the wild-type template. The mtTERM- mination. Since the termination efficiency using this template dependent formation of the shorter RNAs indicated that mt- was not significantly different from that using the template TERM likely terminates transcription by heterologous RNA containing additional 5’ and 3’ sequences of human mtDNA

Downloaded from www.jbc.org by on April 14, 2008

FIG.2. mtTERM is capable of mediating termination of transcription by T3 RNA polymerase. A, termination is directed by the interaction of mtTERM withits DNA binding site. The in vitro run-off transcription assay wasperformed with T3RNA polymerase inthe absence (-) or presence (+) of mtTERM. Two linearized plasmids containing the mtTERM binding site in the reverse orientation relative to the T3 promoter were used: EcoRI-digested pTERM(R) and EcoRI-digested p3243(R). B, time course of transcription elongation by T3 RNA polymerase. The of T3 RNA polymerase and mtTERM followed by chasing with an EcoRI-digested pTERM(R1 was pulse-labeled with [a-:”PIUTP in the presence excess amount of unlabeled UTP. Aliquots were removed from the transcription reaction at thetimes indicated above the lanes. C , the mtTERM binding site is sufficient for supporting termination. NotI-digested pTM(R) was transcribed by T3 RNA polymerase in the absence (-) or presence (+) of mtTERM. The filled circle indicates a mtTERM-independent RNA species shorter than therun-off transcript. In each figure, the positions of mtTERM-dependent terminated transcripts(2’) and run-off transcripts (RO are indicated by arrows. In theschematic diagrams:open box, DNA template; bent arrow, transcription promoter; filled box, mtTERM binding sequence, withan arrow indicating the orientationof the termination site relative to the promoter;urauy lines, transcripts.

29116

Termination Dunscription Mitochondrial TABLE I

Termination of transcription by selected RNA polymerases RNA polymerase Orientation" Eficiency

T3 RNAP

T7 RNAP

E. coli RNAP sc-mtRNAP h-mtRNAP

of terminationb

+

-

t

20-30%

t

10-20% -

t