American Journal of Medical Genetics 71:443–452 (1997) High Proportions of mtDNA Duplications in Patients With Kearns-Sayre Syndrome Occur in the Heart Bernard Fromenty,1 Rosalba Carrozzo,1 Sara Shanske,1 and Eric A. Schon1,2* 1 2 Department of Neurology, College of Physicians and Surgeons, Columbia University, New York, New York Department of Genetics and Development, College of Physicians and Surgeons, Columbia University, New York, New York Kearns-Sayre syndrome (KSS) is a sporadic multisystem mitochondrial disorder characterized by progressive external ophthalmoplegia, pigmentary retinopathy, onset before age 20, and severe cardiac conduction defects that can lead to death. KSS patients harbor partial deletions of mitochondrial DNA (D-mtDNA), sometimes associated with the corresponding mtDNA duplication (dup-mtDNA). As reports on the distribution of dup-mtDNAs among KSS tissues are scarce, we searched for the presence of dupmtDNAs in different autopsy tissues of two such patients, one of whom carried the socalled ‘‘common deletion.’’ Using a newly developed long polymerase chain reaction (PCR) protocol in conjunction with Southern blot analyses, we found dup-mtDNAs in most of the examined tissues from both patients. The proportion of dup-mtDNA in these tissues was much lower than the proportion of D-mtDNA, with one notable exception: in both patients, we found an unusually high level of dup-mtDNA in the heart. These data suggest that dup-mtDNAs may be more stable in heart tissue of KSS patients than in other long-lived postmitotic tissues. Am. J. Med. Genet. 71:443–452, 1997. © 1997 Wiley-Liss, Inc. Contract grant sponsor: National Institutes of Health; Contract grant numbers: NS28828, NS11766, AG12131, HD32062; Contract grant sponsor: The Muscular Dystrophy Association; Contract grant sponsor: The Dana Foundation; Contract grant sponsor: The Procter and Gamble Company; Contract grant sponsor: Ministère Français des Affaires Etrangères. Bernard Fromenty’s current address is INSERM U24, Hôpital Beaujon, 92118 Clichy Cedex, France. *Correspondence to: Eric A. Schon, Department of Neurology, Room 4-431, Columbia University, 630 West 168th Street, New York, NY 10032. E-mail: [email protected] Received 17 October 1996; Accepted 6 March 1997 © 1997 Wiley-Liss, Inc. KEY WORDS: DNA; KSS; long PCR; mitochondria; mitochondrial disease; progressive external ophthalmoplegia; polymerase chain reaction INTRODUCTION Kearns-Sayre syndrome (KSS) is a sporadic multisystem mitochondrial disorder characterized by the invariant triad of progressive external ophthalmolplegia (PEO), pigmentary retinopathy, and onset before age 20, with at least one of the following additional manifestations: complete heart block, cerebellar dysfunction, or cerebrospinal fluid protein above 100 mg/dl [Rowland et al., 1991]. While normal individuals contain a single population of mtDNAs, KSS patients harbor two populations of mtDNAs: wild-type mtDNA (wt-mtDNA; 16.6 kb) and mtDNAs with large-scale rearrangements. The rearrangement is usually a partial deletion of mtDNA (DmtDNA) [Holt et al., 1988; Zeviani et al., 1988; Moraes et al., 1989]; its size, location, and proportion differs among patients, and does not appear to be correlated with the presentation or the severity of the disease phenotype. One particular species of D-mtDNA, removing 4,977 bp, has been found in about one-third of all patients studied, and has therefore been called the ‘‘common deletion’’ [Schon et al., 1989; Mita et al., 1990]. Although it is well known that some KSS patients can harbor both a D-mtDNA and the corresponding partial tandem duplication of mtDNA (dupmtDNA) [Poulton et al., 1989a,b, 1993, 1994; Brockington et al., 1995], there is little information concerning the tissue distribution of both rearranged species in such patients. We recently developed a novel long polymerase chain reaction (PCR)-based approach that allows for the specific and efficient detection of dup-mtDNAs, even in tiny amounts of tissue [Fromenty et al., 1996]. We used this method as a screening tool to search rapidly for the presence of dup-mtDNAs in different autopsy tissues of two KSS patients. One of these patients was known to contain a 26.6-kb species of dup-mtDNA in skeletal 444 Fromenty et al. muscle [Fromenty et al., 1996], whereas the second patient was previously thought to harbor only D-mtDNA, specifically, the common deletion, in the autopsy tissues that were analyzed [Nakase et al., 1990; Shanske et al., 1990]. We report here a detailed analysis of the distribution of the various rearranged mtDNA species in autopsy tissues from both patients, using both long PCR and Southern blot approaches. MATERIALS AND METHODS Patients Patient KB had typical KSS and was described in detail previously [Shanske et al., 1990]. Briefly, this girl developed pigmentary retinopathy at age 8 and external ophthalmoplegia and neurosensory hearing loss at age 13, when a pacemaker was implanted because of a worsening cardiac conduction defect. CSF protein was 132 mg/dl. At age 19 she developed diabetes mellitus and was confined to a wheelchair. She died at age 22. Patient KK also had typical KSS, but in addition had cerebral lesions consistent with Leigh disease. The boy’s problems began at age 3, when right ptosis was noted. At age 6 he developed pigmentary degeneration of the retinas and mild hearing loss, and external ophthalmoplegia at age 8. At this age the electrocardiogram showed bifascicular block, which rapidly evolved into complete third-degree block. When hospitalized for biventricular pacemaker placement he was also diagnosed with juvenile diabetes mellitus. At age 10 he began to require a hearing aid. At age 14 he reported worsening hearing, some muscle weakness, fatigue, and some tremors, and his speech was noted to be slurred. He required a wheelchair and could no longer walk at age 17. He died at age 18. Isolation of DNA Total DNA from tissues was isolated from 10–20 mg of frozen tissues taken at autopsy (3 hours postmortem), as described [Shanske et al., 1990]. Total DNA was also extracted from the blood of normal subjects. Long PCR All long PCR reactions were performed with the Expand™ Long Template PCR system (Boehringer Mannheim, Indianapolis, IN) using the manufacturer’s recommendations, except that ‘‘Buffer 3’’ (10× concentrated buffer containing 22.5 mM MgCl2, 20% DMSO, 1% Tween 20) was always used; enzyme concentration was reduced to 1.2 units (unless otherwise indicated), and 200 mg/ml of purified bovine serum albumin (BSA; New England Biolabs, Beverly, MA) was added to all reactions. The primer concentration was 300 nM (equivalent to 15 pmol in a 50-ml reaction volume) unless specified otherwise. Unless otherwise indicated, long PCR reactions were performed with 10–100 ng of total DNA in a final volume of 50 ml and in GeneAmp™ thin-walled reaction tubes (Perkin Elmer, Oak Brook, IL) using a thermal cycler (Perkin Elmer). The tem- plate used for the long PCR reaction was either uncut total DNA or total DNA previously digested with the restriction enzyme PmeI (one site in mtDNA, at nt10414 [Anderson et al., 1981]), in which case 200 ng of total DNA was digested with 1.6 units of PmeI in a final volume of 20 ml for 3 hr, followed by heat denaturation at 65°C for 20 min. Long PCR reactions were electrophoresed through 0.8% agarose gels containing ethidium bromide. Some long PCR products were digested with the restriction enzymes HindIII or PvuII (Boehringer Mannheim), and the products were visualized in agarose gels. Long PCR for the Amplification of Wild-Type and Dup-mtDNAs Primer pairs FP1 (nt 10364–10377)/BP1 (nt 8783– 8764) and FP2 (nt 13949–13972)/BP2 (nt 8020–7994) were used for patients KB and KK, respectively, in order to amplify simultaneously both wt- and dupmtDNAs (Fig. 1). The PCR conditions were: initial denaturation at 93°C for 2 min, followed by 30 cycles of 30 s at 93°C, 30 s at 62°C, and 30 min at 68°C, with a final extension at 68°C for 10 min; the extension time was increased by 6 s/cycle. Primer pair FP3 (nt 13330– 13350)/BP3 (nt 8582–8561) was also used for both patients KB and KK when the DNA samples were previously digested with PmeI (Fig. 1). The PCR conditions were as described above, except that the annealing temperature was 60°C. Long PCR for the Specific Amplification of Dup-mtDNA To amplify the dup-mtDNAs specifically, we used a novel approach employing a forward ‘‘breakpoint’’ or ‘‘breakpoint loop-out’’ primer annealing across the abnormal breakpoint junction, and a backward primer annealing in the non-duplicated region of the dupmtDNA [Fromenty et al., 1996]. For patient KB, the forward ‘‘breakpoint loop-out’’ primer FP4 was a 22-mer spanning the breakpoint junction, and which omitted an 18-nt region containing a 13-bp perfect repeat (see Fig. 1): 58-TTAAACACAAACTACTTGGCAG-38. The first 15 nt of this primer are specific for the ATPase 8 gene (nt 8450– 8464), whereas the last 7 nt (bold characters) are specific for the ND5 gene (nt 13460–13466). Long PCR with the forward ‘‘breakpoint loop-out’’ primer FP4 used a ‘‘hemi-nested’’ PCR approach. For the first PCR, the initial denaturation at 93°C for 2 min was followed by five cycles performed with 8 pmol of primer FP4 and backward primer BP4 (nt 9950–9931) and 0.6 units of enzyme mixture, with denaturation at 93°C for 30 s and annealing at 38°C for 20 s, followed by a slow increase in temperature (10°C/min) up to 68°C (the extension step), which was maintained for 9 min. At the end of this first PCR, tubes were kept in ice (or stored at −20°C) and 5 ml of the PCR reaction were removed and used as template for the second PCR, which was performed in a second reaction tube with 8 pmol of primer FP4 and backward primer BP1 and the following PCR conditions: initial denaturation at 93°C for 2 min, followed by 35 cycles of 30 s at 93°C, 30 s at 58°C, mtDNA Duplications in KSS 445 Fig. 1. Description of the rearranged mtDNAs. Shown are circular maps [notation of Anderson et al., 1981 with counterclockwise gene order] of the dup-mtDNAs and their corresponding D-mtDNAs, as well as the wt-mtDNAs (genome sizes in boxes, in bp). Only the genes (names [ND5 and ND6, subunits 5 and 6 of NADH dehydrogenase-CoQ oxidoreductase; COX II, subunit II of cytochrome c oxidase; A8/6, subunits 8 and 6 of ATP synthase] in italics; solid and open boxes) involved in each rearrangement are shown. The origins of replication are indicated by right-angled arrows located outside (OH) and inside (OL) the circles. The duplicated portion of the mtDNA (D1 and D2 ‘‘pie sections’’) and its size (in bp), and the non-duplicated region (large D) and its size (16,569 minus the size of the duplication, which is equal to the size of the region deleted in D-mtDNA [small D]), are indicated. The map positions at the insertion breakpoint boundaries straddle the pie section boundaries. For each D-mtDNA, the deleted region (triangle), its size, and its breakpoint coordinates, are indicated. Primers (arrows) are located at approximate map positions (not to scale). The restriction sites used in the RFLP analyses (PvuII at nt-2650; PmeI at nt-10414) are shown. The breakpoint loop-out primer FP4 is denoted by the ‘‘interrupted’’ and ‘‘broken’’ arrows; breakpoint primer FP5 is denoted by a straight or ‘‘broken’’ arrow, as appropriate. The deletion-specific regional probe (nt 10354–12413) used in the Southern blot analyses (Fig. 2) is indicated by the thick arc. The bottom panels show the sequences flanking the abnormal breakpoint junctions (brackets) and their map positions. The sequences of the upstream gene are in plain text; those of the downstream gene are in bold. The perfect 13-bp direct repeats in KB are in italics. Sequences of the breakpoint primers are shown below the junction sequences and are derived from the underlined areas. and 11 min at 68°C, with a final extension at 68°C for 10 min. The extension time was increased by 6 s/cycle. In preliminary experiments, primer FP4, in conjunction with appropriate backward primers, yielded a specific amplification product from a patient [patient GP in Baerlocher et al., 1992] known to harbor the common duplication [Poulton et al., 1995a] (data not shown). For patient KK, the forward ‘‘breakpoint’’ primer FP5 was a 23-mer spanning the breakpoint junction entirely, as described previously [Fromenty et al., 1996]: 58-CCATCCCTACGCATCCCAACCAC-38 (Fig. 1). The first 16 nt of this primer are specific for the COX II gene (nt 7820–7835), whereas the last 7 nt (bold characters) are specific for the ND6 gene (nt 14331– 14337). Forward ‘‘breakpoint’’ primer FP5 was used with backward primer BP2 and the following PCR conditions: initial denaturation at 93°C for 2 min, followed by 30 cycles of 30 s at 93°C, 30 s at 64°C, and 15 min at 68°C, with a final extension at 68°C for 10 min; the extension time was increased by 6 s/cycle. 446 Fromenty et al. Southern Blot Hybridization Analyses One to 3 mg of total DNA were digested with the restriction enzyme PvuII (Boehringer Mannheim) or PmeI (New England Biolabs) according to the manufacturer’s recommendations. Gel electrophoresis and blotting were performed as described [Shanske et al., 1990]. The blots were hybridized sequentially with two random-primed 32P-labelled human mtDNA probes derived either from full-length placental mtDNA or from a 2,060-bp PCR product spanning nt 10354–12413. The quantitation of the different mtDNA species was performed on the fragments digested with PmeI and hybridized with the total mtDNA probe (see Fig. 2) using a GS-363 Molecular Imager System (Bio-Rad, Richmond, CA). The densitometric data were corrected for the sizes of the relevant mtDNA species [Zeviani et al., 1988], including presumed deletion dimers. RESULTS The deletion in KSS patient KK is located between nt-7836 in the COX II gene and nt-14331 in the ND6 gene [Fromenty et al., 1996], and removes 6,495 bp of mtDNA (i.e., a partially deleted molecule of 10,074 bp); the deletion is not flanked by direct repeats (i.e., a ‘‘class II’’ deletion [Mita et al., 1990]). Patient KK also harbored the corresponding dup-mtDNA, which is 26.6-kb-long (i.e., 16,569 bp + 10,074 bp; see Fig. 1), in skeletal muscle [Fromenty et al., 1996]. The deletion in KSS patient KB is the ‘‘common deletion’’ [Nakase et al., 1990; Shanske et al., 1990]. It is located between nt-8483 in the ATPase 8 gene and nt13460 in the ND5 gene, and removes 4,977 bp of mtDNA (i.e., a partially deleted molecule of 11,592 bp); the deletion is flanked by a perfect direct repeat of 13 bp (i.e., a ‘‘class I’’ deletion [Mita et al., 1990]). The corresponding species of dup-mtDNA (henceforth referred to as the ‘‘common duplication’’ for convenience) is 28.2-kb-long (i.e., 16,569 kb + 11,592 bp; see Fig. 1). Use of Long PCR to Detect Dup-mtDNAs In an attempt to amplify the dup-mtDNAs in the various samples, we used two different long PCR-based approaches. In our first approach, we used two primers annealing in the non-duplicated region and pointing towards the putative duplicated region (FP1/BP1 for patient KB and FP2/BP2 for patient KK; see Fig. 1), theoretically allowing for the amplification of both dupmtDNA and wt-mtDNA, but not D-mtDNA [Fromenty et al., 1996]. Using this method, we were unable to amplify any PCR product corresponding to a dupmtDNA (i.e., a 26.6-kb or a 20.7-kb fragment in patients KB and KK, respectively) from any of the DNA samples tested (Figs. 3A and 3B, upper gels). On the other hand, the wt-mtDNA (15.0 kb and 10.6 kb for patients KB and KK, respectively) was amplified efficiently (Figs. 3A and 3B, upper gels). Efforts to amplify the dup-mtDNA species, either by varying the time of extension and/or the template concentrations [Fromenty et al., 1996] or by linearizing the mtDNA prior to amplification [Linz et al., 1990; Stewart et al., 1995] were unsuccessful (data not shown). In order to overcome these difficulties, we employed a second long PCR-based approach recently developed by us [Fromenty et al., 1996]. This novel methodology uses a forward primer designed to anneal specifically across the abnormal breakpoint junction (‘‘breakpoint primer’’) and a backward primer annealing in the nonduplicated region of dup-mtDNA (see Fig. 1), thereby allowing for the specific amplification of the dupmtDNA without interference from wt-mtDNA templates. The design of the forward breakpoint primers is dictated by the nature of the sequence at the abnormal breakpoint junction [Fromenty et al., 1996]. For patient KK, who harbored a class II rearrangement, we used forward breakpoint primer FP5 and backward primer BP2 (located in the deleted region, so D-mtDNAs cannot be amplified) to amplify specifically a 10.3-kb PCR product diagnostic for the 26.6-kb dupmtDNA present in the psoas skeletal muscle [Fromenty et al., 1996 and Fig. 3B, lower gel]; note that wt-mtDNA was not amplified. We were also able to amplify this dup-mtDNA from heart, liver, cerebellum, and frontal lobe (Fig. 3B, lower gel). The yield of this fragment was higher from skeletal muscle, heart, and TABLE I. Proportion of mtDNA Species in Autopsy Tissues From Two KSS Patients Present work Shanske et al.  WT Dela Dup Rearranged WT Rearranged Patient KB Skeletal muscleb Smooth muscleb Liver Kidney Heartb Frontal lobe 64 99 89 87 77 59 31 1 11 8 2 35 5 0 0 5 21 6 36 1 11 13 23 41 50 96 86 60 60 56 50 4 14 40 40 44 Patient KK Skeletal muscleb Liver Heart Frontal lobe Cerebellum 69 18 70 74 82 25 77 16 24 11 6 5 14 2 7 31 82 30 26 18 a Sum of deletion monomer plus dimer. Average of two Southern blot experiments. b Fig. 2. Detection of rearranged mtDNAs in KSS tissues by Southern blot hybridization analysis. Total DNA from the indicated tissues was digested with the indicated enzymes (U, uncut; Pm, PmeI; Pv, PvuII), electrophoresed through a 0.8% agarose gel, transferred to a membrane, and hybridized with a regional probe from nt 10354–12413 (upper panels) or with total mtDNA probe (lower panels). Markers (M) are HindIII-digested phage l (sizes at left, in kb). Assignment of the various hybridizing bands (WT, normal length 16.6-kb molecule; Dup, duplicated mtDNA; D, deleted mtDNA monomeric circle; DD, presumed deletion dimeric circle; D*, linearized deleted mtDNA species derived from the various rearranged mtDNAs) and their sizes (in kb) is based on the hybridization patterns and the relative positions of migration in the gel. A: Patient KB. B: Patient KK. Sm, smooth muscle; L, liver; K, kidney; H, heart; FL, frontal lobe; Sk, skeletal muscle; C, cerebellum. 448 Fromenty et al. PCR product appeared to be higher in the heart than in skeletal muscle or kidney (Fig. 3A, lower gel). The 11.9-kb PCR product was deemed to be derived from the putative 28.2-kb dup-mtDNA species by three criteria. First, the length of the long PCR product corresponded to the distance between FP4 and BP1 (11,893 bp), the primer pair used in the second PCR of our hemi-nested long PCR protocol. Second, the identity of the PCR product was confirmed by RFLP analysis (data not shown). Third, none of the DNA controls gave an 11.9-kb PCR product (Fig. 3A lower gel and data not shown), whereas they gave the expected long PCR products (15.0 kb) under other long PCR conditions (Fig. 3A, upper gel and data not shown). In addition, we verified that the 11.9-kb PCR fragment produced by our hemi-nested protocol did not derive from a recombination event [Pääbo et al., 1990] which could theoretically have occurred between the D-mtDNA and the wt-mtDNA templates during the PCR reaction. For this purpose we used as a template a mixture of two DNA samples, one containing nearly 100% of the common deletion (assessed by Southern blot) and the other containing only wt-mtDNA (3 different PmeI-digested DNA samples were used in three independent experiments). We did not observe any long PCR product corresponding to the 11.9-kb fragment, indicating that ‘‘jumping PCR’’ did not occur with our long PCR protocol (data not shown). Quantitation of Rearranged mtDNA Species by Southern Blot Hybridization Fig. 3. Detection of rearranged mtDNAs in KSS tissues by long PCR. A: Patient KB. Upper gel: Long PCR with two primers annealing in the non-duplicated region in dup-mtDNA (FP1/BP1) and whose 38 ends point away from each other (i.e., towards the duplicated area; see Fig. 1). Shown are PCR products from blood DNA from two controls (C1, C2), and from DNA from smooth muscle (Sm), liver (L), kidney (K), heart (H), frontal lobe (FL), and skeletal muscle (Sk). The size of the PCR product, in kb, is at right, and is consistent with that of wt-mtDNA (WT) only. The 0.8% agarose gel was loaded with 6 ml of the 50-ml PCR reaction. Markers (M) are HindIII-digested phage l (sizes at left, in kb). Lower gel: Specific detection of dup-mtDNA (Dup) using hemi-nested long PCR with ‘‘breakpoint loop-out’’ forward primer FP4 (i.e., FP4/BP4 followed by FP4/BP1) on DNAs digested with PmeI prior to PCR. The 0.8% agarose gel was loaded with 25 ml of the second 50-ml PCR reaction. B: Patient KK. Upper gel: Long PCR with primers FP2/BP2. Samples include cerebellum (C). Lower gel: Long PCR with ‘‘breakpoint’’ forward primer FP5 (i.e., FP5/BP2). All other notation and amounts loaded as in A. liver than from cerebellum and frontal lobe (Fig. 3B, lower gel). For patient KB, who harbored a class I rearrangement, we used a hemi-nested PCR protocol (forward primer FP4 and backward primers BP4/BP1, annealing in the non-duplicated region of the putative dupmtDNA) to amplify specifically an 11.9-kb PCR product, diagnostic for the 28.2-kb dup-mtDNA, from heart, skeletal muscle (psoas), and kidney (Fig. 3A, lower gel). Amplification of this fragment was found to be more reproducible on mtDNA templates previously linearized with PmeI (Fig. 3A, lower gel). The amount of the The presence of dup-mtDNAs in some tissues of our KSS patients was confirmed by Southern blot hybridization analysis of PvuII- and PmeI-digested total DNA probed with total or regional mtDNA probes. PmeI cuts at nt-10414, which is located inside the region deleted in both patients KB and KK, thus allowing one to distinguish between dup-mtDNA and D-mtDNA species. In patient KB, we found that 4 out of 6 tissues, skeletal muscle, kidney, frontal lobe, and heart, contained, in addition to the common deletion, significant amounts of the corresponding dup-mtDNA, whose proportion of total mtDNA species ranged from 5% in skeletal muscle and kidney to 21% in heart (Fig. 2A; Tables I and II). In patient KK, dup-mtDNA was found in all tissues studied, ranging from 2% in frontal lobe to 14% in heart (Fig. 2B, Tables I and II). DISCUSSION It had been thought that dup-mtDNAs are present infrequently in patients with sporadic KSS [Poulton et al., 1989a,b], but it has become increasingly clear that this is not the case, as a significant number of KSS patients harbor both a D-mtDNA and the corresponding dup-mtDNA [Poulton et al., 1993, 1994, 1995a,b; Brockington et al., 1995; our unpublished data]. However, few reports exist on the tissue distribution of dupmtDNAs in KSS patients, present either alone or together with the corresponding D-mtDNAs. We therefore examined tissues from two KSS autopsies for the presence and relative amounts of these two rearranged species. We note that both of these patients had diabe- mtDNA Duplications in KSS 449 TABLE II. Distribution of Rearranged mtDNA Species in Autopsy Tissues From Four KSS Patients Patient KB Skeletal muscle Smooth muscle Liver Kidney Heart Cerebellum Cerebrumc Patient KK Poulton et al. [1995b] Brockington et al.  Dup Dela %Dupb Dup Del %Dup Dup Del %Dup Dup Del %Dup 5 0 0 5 21 — 6 31 1 11 8 2 — 35 14 0 0 38 91 — 15 6 — 5 — 14 7 2 25 — 77 — 16 11 24 19 — 6 — 47 39 8 1 — 1 16 6 10 3 47 — 64 54 8 23 37 2 — 2 23 43 30 8 2 — — 19 7 — — 22 — — 0 10 — — 8 — — 100 41 — — a Sum of deletion monomer plus dimer. Percentage of dup-mtDNAs among all rearranged species (i.e., Dup/[Dup + Del] × 100). Frontal lobe in patients KB and KK. b c tes mellitus, which has also been associated specifically with the presence of maternally inherited dupmtDNAs [Rötig et al., 1992; Dunbar et al., 1993; Ballinger et al., 1994; Poulton et al., 1995b]. One patient (KK) had recently been found by us to harbor dup-mtDNAs in skeletal muscle [Fromenty et al., 1996], but the other tissues had not been evaluated. We now show that dup-mtDNAs were present in all examined tissues from this subject (Table I). The other patient (KB) had been analyzed by us previously [Shanske et al., 1990], but the analysis had been done ambiguously: although all KB tissues had been assayed by Southern blot analysis of total DNA digested with a battery of enzymes (EcoRI, HindIII, PstI, PvuII, XbaI), none of them were diagnostic for the detection of a duplicated species. The reassessment of tissues from subject KB performed here has shown that patient KB also harbored dup-mtDNAs in significant quantities, in 4 of 6 tissues examined (Table I). High Levels of Dup-mtDNAs Are Present in Heart We made two surprising observations regarding dupmtDNAs in the heart. First, there were high absolute amounts of dup-mtDNA in this tissue in both patients (14% in KK, 21% in KB), higher than in any other tissue examined (Table I). Second, dup-mtDNA represented a significant fraction (47% in KK, 91% in KB) of all rearranged species in heart, whereas in the other tissues the dup-mtDNAs were in the minority (ranging from 0% in smooth muscle and liver in KB to 39% in cerebellum in KK [see Table II]). Only two other papers report the proportion of dup-mtDNA in heart tissue from KSS patients harboring duplications [Brockington et al., 1995; Poulton et al., 1995b]. In those patients, the absolute amount of total rearranged mtDNAs (i.e., dup- plus D-mtDNAs) was lower (17% [Brockington et al., 1995] and 14% [Poulton et al., 1995b]) than the levels found in KB and KK (23% and 30%, respectively), but the proportion of rearranged molecules that were dup-mtDNAs was still higher in the heart (41% [Brockington et al., 1995] and 43% [Poulton et al., 1995b]; see Table II) than in all other tissues examined, save one. The pathogenic significance of dup-mtDNAs remains uncertain. The deleterious effects of D-mtDNAs are probably due to the lack of the indispensible tRNAs which are required for translation [Nakase et al., 1990; Hayashi et al., 1991], but dup-mtDNAs contain all 37 mtDNA-encoded genes, albeit with some in excess. Thus, dup-mtDNAs might be pathogenic in an indirect way. For example, they might generate the corresponding D-mtDNA species via a recombination event, at least in some tissues [Poulton et al., 1994, 1995a]. Alternatively, transcription and/or translation of the fusion gene formed at the duplication breakpoint might interfere with normal mitochondrial function (note that the possibility of formation of aberrant translation products also applies to the corresponding D-mtDNA, but only in those situations where other mtDNAs within the organelle, either wt- or dup-mtDNAs can provide the requisite tRNAs). It may even be that duplications are not pathogenic at all, or are somehow ‘‘less pathogenic’’ than deleted mtDNAs, on a permolecule basis. Patients KB and KK presented with severe cardiac conduction defects, necessitating implantation of a pacemaker. The patients described by Brockington et al.  and Poulton et al. [1995b] suffered from milder heart block and had lower proportions of rearranged mtDNA species in their hearts than did KB or KK (Table II). These data suggest that the overall mutant load correlates with the severity of the cardiac conduction defect in KSS patients. Alternatively, it is possible that the conduction defects are due to the focal presence of D-mtDNAs in the conduction system, as hypothesized previously [Remes et al., 1992]. It is noteworthy in this regard that in skeletal muscle from a patient with a late-onset myopathy who harbored dupmtDNAs as the major rearranged mtDNA species in this tissue, the few respiratory-deficient fibers that were present contained almost exclusively D-mtDNAs [Manfredi et al., 1997]. Taken together, our data extend previous reports [Brockington et al., 1995; Poulton et al., 1995b] that the proportion of dup-mtDNAs can vary greatly among tissues. Furthermore, it seems that the fraction of total rearranged mtDNAs that are dup-mtDNAs is consistently high in heart, whereas this fraction is low in other examined tissues (Table II). Factors Controlling the Steady-State Levels of Rearranged mtDNAs The above observation suggests that dup-mtDNAs may be more stable in the heart than in other longlived postmitotic tissues, such as skeletal muscle and 450 Fromenty et al. brain. For example, it had been shown that the proportion of dup-mtDNAs in skeletal muscle in KSS patients fell over time [Poulton et al., 1993, 1995a], resulting in low proportions of dup-mtDNAs and high levels of D-mtDNAs in this tissue at the time of death (Table II). Unfortunately, the factors controlling the generation, stability, and turnover of dup-mtDNAs are not known. At least four possibilities come to mind. First, the size of the molecule itself may affect stability, with large dup-mtDNAs being less stable than smaller ones [Poulton et al., 1994, 1995a]. However, the 28.2-kb common duplication is one of the largest dup-mtDNAs described to date, but was present at high levels in the heart of patient KB (Table I). Second, the cell type may affect stability and turnover of dup- and D-mtDNAs. Cardiac cells, unlike skeletal muscle fibers, are not syncytia, and they also have greater oxidative energy requirements. For these two reasons, it is reasonable to assume that the time period necessary for D-mtDNAs to accumulate and reach levels causing cellular dysfunction in cardiac cells is shorter than in skeletal muscle, and that the threshold level of D-mtDNAs required to cause such dysfunction is lower in heart than in skeletal muscle. Cells appear to sense the total number of mitochondrial genomes they contain and to regulate their numbers based on their particular energetic requirements [e.g., see King and Attardi, 1988; Tritschler et al., 1992], although the mechanisms by which this occurs are obscure. Thus, as the level of dup-mtDNAs falls, the level of D-mtDNAs plus wt-mtDNAs should rise, and vice versa. If cardiac tissue can indeed tolerate fewer D-mtDNAs than can skeletal muscle, this ‘‘lever principle’’ could then explain why dup-mtDNAs are elevated in this tissue. Third, it is possible that dup-mtDNAs are generated relatively more frequently in heart than are DmtDNAs, perhaps because cardiomyocytes contain extremely high numbers of mitochondria and mtDNAs [Shoffner and Wallace, 1995]; this environment might favor organellar fusion and/or intermolecular recombination events in which duplicated molecules are generated. There are no data in support of such a scenario, and no such advantage of duplicated molecules over deleted molecules was observed in a prokaryotic model of the generation of deletions and duplications between direct repeats [Trinh and Sinden, 1993, and references therein]. Finally, the rate of cell division, together with the respective rates of replication of the different rearranged species present in the cell, may also affect stability of the various mtDNA species [but see Moraes and Schon, 1995]. Replication of wild-type mtDNA requires two origins of replication [Clayton, 1992], one, near nt-200, priming heavy-strand synthesis (OH) in one direction and the other, near nt-5750, priming light-strand synthesis (OL) in the opposite direction. A dup-mtDNA usually contains not two, but four, replication origins (2 OHs and 2 OLs); note that the dupmtDNAs in patients KB and KK fall into this category (see Fig. 1). The proportion of such a ‘‘4-origin’’ dupmtDNA fell rapidly in both uncloned and cloned fibroblasts [Poulton et al., 1993], implying that a dupmtDNA containing two pairs of replication origins might have difficulty in replicating. On the other hand, the proportion of a dup-mtDNA containing two OHs but only one OL (i.e., the corresponding D-mtDNA contained OH but lacked OL) increased during passage in transformed lymphoblasts [Ballinger et al., 1994]. While the absence of OL in a D-mtDNA should impair or even prevent the ability of such a molecule to replicate, the presence of a single functional OL in a ‘‘3origin’’ dup-mtDNA might actually compensate for any potential impairment of this molecule’s ability to replicate. In support of this idea, another patient harboring a different ‘‘3-origin’’ dup-mtDNA with only one OL [patient CH in Fromenty et al., 1996] contained an unusually high proportion of dup-mtDNA and a very low amount of the corresponding D-mtDNA (with no OL) in skeletal muscle [Manfredi et al., 1997]. Long PCR Can Amplify Dup-mtDNAs Specifically The long PCR results reported here were confirmed by Southern blot analysis. However, we consistently had problems trying to amplify dup-mtDNA in the frontal lobe of patient KB, even though Southern blot analysis demonstrated the presence of dup-mtDNA in this tissue. While we cannot omit the possibility that contaminating elements inhibited PCR amplifcation, it may also be that brain tissue is particularly difficult to analyze. Indeed, long PCR amplified dup-mtDNAs efficiently from other tissues containing similar proportions of this rearrangement (e.g., 5% in kidney and skeletal muscle in patient KB; Fig. 3A, lower gel; Table I). Similarly, although dup-mtDNA-derived long PCR products were obtained from the cerebellum and the frontal lobe DNA of patient KK, the yields were lower in those tissues than in other tissues harboring essentially similar proportions of dup-mtDNAs, such as liver and skeletal muscle (Fig. 3B, lower gel; Table I). We note that dup-mtDNAs are very long molecules which may be more prone to damage (e.g., oxidation, depurination, chain scission) than are shorter wt- or D-mtDNA templates, either in vivo or in vitro (e.g., following preservation of the brain tissue or during the isolation of the DNA). Such damage could lead in turn to a dramatic decrease in the efficiency of long PCR, which requires high template integrity [Barnes, 1994; Cheng et al., 1994]. It is therefore conceivable that brain tissues or brain-derived DNA contained a high proportion of ‘‘Southern blot-observable’’ dup-mtDNA templates, and yet contained so few intact full-length molecules that long PCR could not amplify them. Although we do not know the lower limit of detection of dup-mtDNAs by long PCR, we note that we were able to use the ‘‘breakpoint loop-out primer’’ method to amplify a short 5.5-kb fragment derived from a 21.2-kb class I mtDNA duplication in a sample of blood DNA containing less than 1% dup-mtDNA [Fromenty et al., 1996]. Since about one-third of all KSS/PEO patients harbor the common deletion [Schon et al., 1994], a long PCR method for the specific amplification of the corresponding 28.2-kb common duplication should be useful in detecting this prevalent rearrangement, especially mtDNA Duplications in KSS when the amount of DNA available is a limiting factor (e.g., analysis on endomyocardial biopsy specimens). The common deletion is also found at low levels in tissues (mainly brain, skeletal muscle, and heart) of aged individuals [Cortopassi et al., 1992; Schon et al., 1996; Pallotti et al., 1996]. However, the real nature of this mtDNA rearrangement in aged tissues is not known, because the standard PCR methodologies previously used for their detection cannot differentiate a D-mtDNA from its corresponding dup-mtDNA [Fromenty et al., 1996]. Our novel approach to amplify dupmtDNAs in general, and the 28.2-kb common duplication in particular, should help clarify this important issue. In conclusion, we have used long PCR and classic Southern blotting to detect dup-mtDNA species in different tissues of two KSS patients. Our data confirm that the relative amounts of D- and dup-mtDNAs vary widely among tissues, making the relative contribution of each species of rearranged mtDNA to the overall clinical picture even more difficult to sort out than had been appreciated previously, as heteroplasmy for both rearranged mtDNA species raises obvious questions regarding the relative pathogenicity of each molecule. Furthermore, it seems that dup-mtDNAs are more stable in the heart of KSS patients than in other longlived postmitotic tissues. Our long PCR methodology for the specific amplification of dup-mtDNAs could be a convenient and powerful tool to help attack these problems. ACKNOWLEDGMENTS We thank Drs. E. Bonilla, S. DiMauro, M. Hirano, and A. Rötig for providing tissues, and Drs. M. P. King and J. Poulton for helpful comments. B.F. was supported by the Lavoisier Fellowship from the Ministère Français des Affaires Etrangères. REFERENCES Anderson S, Bankier AT, Barrell BG, de Bruijn MHL, Coulson AR, Drouin J, Eperon IC, Nierlich DP, Roe BA, Sanger F, Schreier PH, Smith AJH, Staden R, Young IG (1981): Sequence and organization of the human mitochondrial genome. Nature 290:457–465. Baerlocher KE, Feldges A, Weissert M, Simonsz HJ, Rötig A (1992): Mitochondrial DNA deletion in an 8-year-old boy with Pearson syndrome. J Inher Metab Dis 15:327–330. 451 mitochondrial DNA associated with diabetes mellitus. Hum Mol Genet 2:1619–1624. Fromenty B, Manfredi G, Sadlock J, Zhang L, King MP, Schon EA (1996): Efficient and specific amplification of identified partial duplications of human mitochondrial DNA by long PCR. Biochim Biophys Acta 1308: 222–230. Hayashi J-I, Ohta S, Kikuchi A, Takemitsu M, Goto Y-I, Nonaka I (1991): Introduction of disease-related mitochondrial DNA deletions into HeLa cells lacking mitochondrial DNA results in mitochondrial dysfunction. Proc Natl Acad Sci USA 88:10614–10618. Holt IJ, Harding AE, Morgan-Hughes JA (1988): Deletions of mitochondrial DNA in patients with mitochondrial myopathies. Nature 331: 717–719. King MP, Attardi G (1988): Injection of mitochondria into human cells leads to a rapid replacement of the endogenous mitochondrial DNA. Cell 52:811–819. Linz U, Delling U, Rübsamen-Waigmann H (1990): Systematic studies on parameters influencing the performance of the polymerse chain reaction. J Clin Chem Clin Biochem 28:5–13. Manfredi G, Vu T, Bonilla E, Schon EA, DiMauro S, Arnaudo E, Zhang L, Rowland LP, Hirano M (1997): Association of myopathy with largescale mitochondrial DNA duplications and deletions: Which is pathogenic? Ann Neurol, in press. Mita S, Rizzuto R, Moraes CT, Shanske S, Arnaudo E, Fabrizi G, Koga Y, DiMauro S, Schon EA (1990): Recombination via flanking direct repeats is a major cause of large-scale deletions of human mitochondrial DNA. Nucl Acids Res 18:561–567. Moraes CT, Schon EA (1995): Replication of a heteroplasmic population of normal and partially-deleted human mitochondrial genomes. In Palmieri F, Papa S, Saccone C, Gadaleta MN (eds): ‘‘Progress in Cell Research, Vol. 5. Amsterdam: Elsevier, pp 209–215. Moraes CT, DiMauro S, Zeviani M, Lombes A, Shanske S, Miranda AF, Nakase H, Bonilla E, Wernec LC, Servidei S, Nonaka I, Koga Y, Spiro A, Brownell KW, Schmidt B, Schotland DL, Zupanc MD, DeVivo DC, Schon EA, Rowland LP (1989): Mitochondrial DNA deletions in Progressive External Ophthalmoplegia and Kearns-Sayre syndrome. N Engl J Med 320:1293–1299. Nakase H, Moraes CT, Rizzuto R, Lombes A, DiMauro S, Schon EA (1990): Transcription and translation of deleted mitochondrial genomes in Kearns-Sayre syndrome: Implications for pathogenesis. Am J Hum Genet 46:418–427. Pääbo S, Irwin DM, Wilson AC (1990): DNA damage promotes jumping between templates during enzymatic amplification. J Biol Chem 265: 4718–4721. Pallotti F, Chen X, Bonilla E, Schon EA (1996): Evidence that specific mtDNA point mutations may not accumulate in skeletal muscle during normal human aging. Am J Hum Genet 59:591–602. Poulton J, Deadman ME, Gardiner RM (1989a): Duplications of mitochondrial DNA in mitochondrial myopathy. Lancet 1:236–240. Poulton J, Deadman ME, Gardiner RM (1989b): Tandem direct duplications of mitochondrial DNA in mitochondrial myopathy: Analysis of nucleotide sequence and tissue distribution. Nucl Acids Res 17:10223– 10229. Ballinger SW, Shoffner JM, Gebhart S, Koontz DA, Wallace DC (1994): Mitochondrial diabetes revisited. Nat Genet 7:458–459 (letter). Poulton J, Deadman ME, Bindoff L, Morten K, Land J, Brown G (1993): Families of mtDNA rearrangements can be detected in patients with mtDNA deletions: Duplications may be a transient intermediate form. Hum Mol Genet 2:23–30. Barnes WM (1994): PCR amplification of up to 35-kb DNA with high fidelity and high yield from l bacteriophage templates. Proc Natl Acad Sci USA 91:2216–2220. Poulton J, Deadman ME, Bindoff L, Morten K, Land J, Brown G (1994): Are duplications of mitochondrial DNA characteristic of Kearns-Sayre syndrome? Hum Mol Genet 3:947–951. Brockington M, Alsanjari N, Sweeney MG, Morgan-Hughes JA, Scaravilli F, Harding AE (1995): Kearns-Sayre syndrome associated with mitochondrial DNA deletion or duplication: A molecular genetic and pathological study. J Neurol Sci 131:78–87. Poulton J, Morten KJ, Marchington D, Weber K, Brown KG, Rötig A, Bindoff L (1995a): Duplications of mitochondrial DNA in Kearns-Sayre syndrome. Muscle Nerve Suppl. 3:S154–S158. Cheng S, Fockler C, Barnes WM, Higuchi R (1994): Effective amplification of long targets from cloned inserts and human genomic DNA. Proc Natl Acad Sci USA 91:5695–5699. Clayton DA (1992): Transcription and replication of animal mitochondrial DNAs. Int Rev Cytol 141:217–232. Cortopassi GA, Shibata D, Soong N-W, Arnheim N (1992): A pattern of accumulation of a somatic deletion of mitochondrial DNA in aging human tissues. Proc Natl Acad Sci USA 89:7370–7374. Dunbar DR, Moonie PA, Swingler RJ, Davidson D, Roberts R, Holt IJ (1993): Maternally transmitted partial direct tandem duplication of Poulton J, O’Rahilly S, Morten KJ, Clark A (1995b): Mitochondrial DNA, diabetes and pancreatic pathology in Kearns-Sayre syndrome. Diabetologia 38:868–871. Remes AM, Hassinen IE, Majamaa K, Peuhkurinen KJ (1992): Mitochondrial DNA deletion diagnosed by analysis of an endomyocardial biopsy specimen from a patient with Kearns-Sayre syndrome and complete heart block. Br Heart J 68:408–411. Rötig A, Bessis J-L, Romero N, Cormier V, Saudubray J-M, Narcy P, Lenoir G, Rustin P, Munnich A (1992): Maternally inherited duplication of the mitochondrial genome in a syndrome of proximal tubulopathy, diabetes mellitus, and cerebellar ataxia. Am J Hum Genet 50:364–370. 452 Fromenty et al. Rowland LP, Blake DM, Hirano M, DiMauro S, Schon EA, Hays AP, DeVivo DC (1991): Clinical syndromes associated with ragged red fibers. Rev Neurol (Paris) 147:467–473. Schon EA, Hirano M, DiMauro S (1994): Mitochondrial encephalomyopathies: Clinical and molecular analysis. J Bioenerg Biomemb 26:291– 299. Schon EA, Rizzuto R, Moraes CT, Nakase H, Zeviani M, DiMauro S (1989): A direct repeat is a hotspot for large-scale deletions of human mitochondrial DNA. Science 244:346–349. Schon EA, Sciacco M, Pallotti F, Chen X, Bonilla E (1996): Mitochondrial DNA mutations and aging. In Holbrook NJ, Martin GR, Lockshin RA (eds): ‘‘Cellular Aging and Cell Death.’’ New York: Wiley-Liss, Inc., pp. 19–34. Shanske S, Moraes CT, Lombes A, Miranda AF, Bonilla E, Lewis P, Whelan MA, Ellsworth CA, DiMauro S (1990): Widespread tissue distribution of mitochondrial DNA deletions in Kearns-Sayre syndrome. Neurology 40:24–28. Shoffner JM, Wallace DC (1995): Oxidative phosphorylation diseases. In Scriver CR, Beaudet AL, Sly WS, Valle D (eds): ‘‘The Metabolic and Molecular Basis of Inherited Diseases, 7th Edition.’’ New York: McGraw-Hill, pp 1535–1609. Stewart ACM, Gravitt PE, Cheng S, Wheeler CM (1995): Generation of entire human papillomavirus genomes by long PCR: Frequency of errors produced during amplifiation. Genome Res 5:79–88. Trinh TQ, Sinden RR (1993): The influence of primary and secondary DNA structure in deletion and duplication between direct repeats in Escherichia coli. Genetics 134:409–422. Tritschler H-J, Andreetta F, Moraes CT, Bonilla E, Arnaudo E, Danon MJ, Glass S, Zelaya BM, Vamos E, Telerman-Toppet N, Kadenbach B, DiMauro S, Schon EA (1992): Mitochondrial myopathy of childhood associated with depletion of mitochondrial DNA. Neurology 42:209–217. Zeviani M, Moraes CT, DiMauro S, Nakase H, Bonilla E, Schon EA, Rowland LP (1988): Deletions of mitochondrial DNA in Kearns-Sayre syndrome. Neurology 38:1339–1346.