Mitochondrial DNA (mtDNA) deletion mutations co-localize with electron transport system (ETS) abnormalities in rhesus monkey skeletal muscle fibers. Using laser capture microdissection in conjunction with PCR and DNA sequence analysis, mitochondrial genomes from single sections of ETS abnormal fibers were characterized. All ETS abnormal fibers contained mtDNA deletion mutations. Deletions were large, removing 20–78% of the genome, with some to nearly all of the functional genes lost. In one-third of the deleted genomes, the light strand origin was deleted, whereas the heavy strand origin of replication was conserved in all fibers. A majority (27/39) of the deletion mutations had direct repeat sequences at their breakpoints and most (36/39) had one breakpoint within or in close proximity to the cytochrome b gene. Several pieces of evidence support the clonality of the mtDNA deletion mutation within an ETS abnormal region of a fiber: (a) only single, smaller than wild-type, PCR products were obtained from each ETS abnormal region; (b) the amplification of mtDNA from two regions of the same ETS abnormal fiber identified identical deletion mutations, and (c) a polymorphism was observed at nucleotide position 16103 (A and G) in the wild-type mtDNA of one animal (sequence analysis of an ETS abnormal region revealed that mtDNA deletion mutations contained only A or G at this position). Species-specific differences in the regions of the genomes lost as well as the presence of direct repeat sequences at the breakpoints suggest mechanistic differences in deletion mutation formation between rodents and primates.
Mitochondrial DNA (mtDNA) deletion mutations accumulate with age in skeletal muscle triggering the concomitant generation of mitochondrial enzymatic abnormalities, intrafibre atrophy and, ultimately, fiber loss. These mutations therefore contribute to age-related fiber loss, an important component of sarcopenia. Initial studies characterizing age-associated mtDNA abnormalities were based on tissue homogenate studies. Although these studies identified, in numerous species, an age-associated accumulation of mtDNA deletion mutations, an association between genotype and phenotype was not possible due to the heterogeneous distribution of the abnormalities in aged tissue. In situ hybridization studies (Muller-Hoecker, 1990; Muller-Hoecker et al., 1992, 1993; Lee et al., 1998; Lopez et al., 2000) determined that mtDNA deletion mutations and associated enzymatic abnormalities were distributed in a mosaic pattern in aged muscle. In addition, the abnormalities are segmental, contained within a portion of the affected muscle fiber. Consistent with these findings, the calculated abundance of a given mtDNA deletion mutation (compared with wild-type genome) varies with the number of fibers included in the analysis (Schwarze et al., 1995). This study also indicated the accumulation of mtDNA deletion mutations to high levels in muscle fibers. The molecular analysis of these abnormalities requires technologies that isolate discrete sections of individual fibers.
Laser capture microdissection provides a tool with which single fibers from a histological slide can be sampled and assayed for mitochondrial genotype. Through the use of microdissection, we previously demonstrated, in aged rat skeletal muscle, that mtDNA deletion mutations accumulate in electron transport system (ETS) abnormal regions of muscle fibers. These deletion mutations were large, with 4.4–9.7 kb of mtDNA lost, and were clonal within a given affected fiber region (Cao et al., 2001). All of the deletion mutations analysed from aged rat muscle were unique (although clonal within a fiber, the size of the deletion and breakpoints varied between fibers) and were contained within the mitochondrial major arc region (i.e. maintaining both light- and heavy-strand origins of replication). Direct repeat sequences, often found to flank mtDNA deletion mutations in human myopathies and aging tissues, were rare in rats. In contrast to the rodent studies, age-associated mtDNA deletion mutations in humans have emphasized a few ‘common’ deletion mutations that have direct repeat sequences at their breakpoints (Cortopassi & Arnheim, 1990; Linnane et al., 1990; Corral-Debrinski et al., 1992). Our studies in aged rhesus monkey skeletal muscle identified an age-induced accumulation of mtDNA deletion mutations, some of which were ‘common deletions’ with direct repeat sequences flanking the breakpoints (Lee et al., 1993, 1994). This PCR/DNA sequencing approach was, however, based on tissue homogenate studies. Few conclusions could be drawn regarding the phenotypic impact of the mutant genomes. Subsequent in situ hybridization experiments, although linking deletion mutations with phenotypic change (alterations in mitochondrial enzymatic function), did not allow precise delineation of the breakpoint. In this study, we demonstrate that mtDNA deletion mutations accumulate clonally in rhesus monkey muscle fibers exhibiting abnormalities in the ETS. By defining the mtDNA deletion breakpoints from different fibers exhibiting mitochondrial abnormalities, rhesus ‘common’ deletion mutations were identified with direct repeat sequences present at the breakpoints. These studies suggest fundamental differences in the mechanisms that initiate, and/or the subsequent selection of, mtDNA deletion mutations that accumulate with age between rodents and primates.
Sequence of rhesus mitochondrial genome
The first rhesus mtDNA primer pairs used in this study were based on the published human mtDNA sequence (GenBank accession no. J01415). Segments of the rhesus mitochondrial genome were amplified, cloned and sequenced as denoted in Experimental procedures below. The insert sequences were aligned, and rhesus-specific primers created based on the monkey mitochondrial genome. The remainder of the sequence was determined using rhesus-specific primers and direct sequencing methods. During the course of the study, we completed the sequencing of the rhesus monkey mtDNA genome (GenBank accession no. AY612638). The Macaca mulatta mitochondrial genome is 16 564 bp and has 80.4% nucleotide sequence similarity with the human mitochondrial genome (Brudno et al., 2003).
ETS abnormal regions of fibers are concomitant with mtDNA deletion mutations
Muscle fibers exhibiting ETS abnormalities (COX−/SDH++, Fig. 1) accumulate with age in rhesus monkey skeletal muscle (Lee et al., 1993; Lopez et al., 2000). These abnormalities occur within discrete segments of the affected fiber, are flanked by enzymatically normal regions and are distributed mosaically within the tissue. We microdissected and examined a total of 39 ETS abnormal muscle fibers, 23 from the 39-year-old monkey and 16 from the 29-year-old monkey. Following laser capture microdissection of the ETS abnormal fibers, the mtDNA was selectively amplified and sequenced. Single, smaller than full-length, PCR amplification products were observed in all abnormal fibers (Fig. 2). The deleted genomes ranged in size from 3372 to 12 884 bp. The deletions were large, eliminating 20–78% of the mtDNA (Table 1 & 2), resulting in the loss of some to nearly all of the functional genes. There was no significant difference between the 29- and 39-year-old monkeys with respect to the size of the deletion products (P = 0.38). One ETS abnormal region (fiber 16) was microdissected from two slide sections located 40 µm apart. PCR amplification and direct DNA sequencing of the two regions identified identical amplification products (4.5-kb product, indicating a 11.9-kb deletion). The use of extension PCR on microdissected ETS normal fibers (n = 11) from aged monkeys amplified only wild-type (full-length) mtDNA.
Table 1. Sequence analysis of mtDNA deletion breakpoints
Breakpoint sequences are based upon the published rhesus monkey mitochondrial genome (GenBank accession no. AY612638). The left and right breakpoints are separated by periods, direct repeat nucleotides are in parentheses, and deleted nucleotides are underlined.
Breakpoint numbers are based upon the published AY612638 rhesus mitochondrial genome. The left and right breakpoints are separated by a dash; any numbers between the dashes indicate direct repeat sequences.
Characteristics of mtDNA deletion mutations
The truncated mtDNA genomes were categorized based upon the presence/absence of direct repeat sequences at the deletion breakpoints (Table 1). Direct repeat sequences, defined in this study as being ≥ 4 bp, ranged in size from 4 to 32 bp and were present in 27/39 deletion mutations sequenced (Table 1). The majority (16/27) had repeats with complete sequence identity, whereas 11 had imperfect repeats. In all cases, one repeat was deleted and one was maintained. The imperfect direct repeats allow us to determine which repeat sequence was lost in the deletion event. There was no apparent selection for which repeat was maintained. Deletion breakpoints lacking direct nucleotide repeat sequences were present in 12 deletion mutations.
Distribution of deletion breakpoints
A non-random distribution of mtDNA deletion breakpoints was observed when the breakpoint locations were mapped on the full-length genome (Fig. 3). Almost every deletion mutation (36/39) had one breakpoint in or within close proximity to the cytochrome b gene. The location of the second breakpoint was over a more dispersed area between nucleotides 1176 and 11025. The D-loop (control) region was the only region of the genome that was consistently conserved, with the OH maintained in all cases. One deletion mutation was contained entirely in the minor arc. In 11 of the deleted genomes, the light-strand origin (OL) was removed. These deleted genomes had an average loss of ∼11 kb of genetic material, whereas the deleted genomes that maintain the OL had an average loss of ∼7 kb.
MtDNA sequence analysis of the muscle tissue homogenates as well as microdissected ETS normal fibers from the 29-year-old monkey demonstrated that the animal was polymorphic, for A and G, at nucleotide 16 103 (Fig. 4). Because this region (control region) was maintained in all of the deletion mutations, the polymorphism allowed us to test the hypothesis that deletion mutations present in a given fiber originated from, and subsequently accumulated from, a single mutation event. The hypothesis would predict that a given deletion mutation obtained from the 29-year-old monkey would contain either an A or a G at nucleotide 16 103 but, unlike the tissue homogenates and ETS normal fibers, not both. When mutant genomes from six separate ETS abnormal fibers from the 29-year-old monkey were amplified and sequenced, this nucleotide position was not polymorphic in the deleted genome. Four of the mtDNA deletion mutations contained an adenine at position 16 103 and two contained the guanine. Full-length mtDNA from the same fibers contain both A and G at position 16 103.
The pathological potential of mtDNA deletion mutations is demonstrated in the mitochondrial myopathies, human diseases caused by mtDNA mutations. In these disorders, the majority of mitochondrial genomes in both muscle and nerve contain single nucleotide mutations (inherited forms) or deletion mutations (sporadic forms). In contrast, aging skeletal muscle contains primarily full-length mitochondrial genomes and there is little evidence indicating single nucleotide mutations accumulate to physiologically significant levels in this tissue. Deletion mutations, however, accrue with age in muscle tissue, clonally expanding within an affected muscle fiber to produce a segmental disruption of cellular processes (COX−/SDH++ phenotype), concomitant fiber atrophy and, ultimately, fiber loss (Wanagat et al., 2001).
We have demonstrated that, in skeletal muscle of aged, non-human primates, every COX−/SDH++ region analysed contained mitochondrial genomes with large deletion mutations. Previous in situ hybridization studies in rhesus monkey skeletal muscle fibers demonstrated (i) deletion mutations accumulate to high abundance within affected regions and (ii) the same hybridization pattern is present within a fiber, a pattern that often differed between fibers. Although these earlier studies determined that mtDNA abnormalities could accumulate to high levels in affected regions of muscle fibers, hybridization analysis limited conclusions regarding the size of the deletion mutation, and the precise location and sequence characteristics of the breakpoints. Through the use of laser capture microdissection of ETS abnormal regions of muscle fibers, combined with PCR and mtDNA sequencing, we identified mitochondrial deletion mutations in every ETS abnormal fiber examined. These mutations resulted in the loss of large regions of the mtDNA (3.68–13.19 kb deleted), including the loss of the light-strand origin of replication in 11 of the deletion products.
Duplication mutations, in which a portion of the mitochondrial genome is represented twice, have been shown to be present in aged human tissues (Bodyak et al., 2001). The partial duplication involving the minor arc could produce the small amplification products (and the loss of light-strand origin) observed with some of the ETS abnormal fibers. This explanation is, however, not consistent with in situ hybridization studies of ragged red rhesus monkey muscle fibers that also identified mtDNAs lacking the light-strand origin of replication (Lee et al., 1998). In addition, primer sets were synthesized to other regions of the genome to screen for duplication mutations. These genome scanning experiments did not identify duplication mutations (data not shown).
The mtDNA deletion breakpoints were distributed non-randomly in the genome. The cytochrome b region of the rhesus mitochondrial genome contained the highest concentration of deletion breakpoints. The region between nucleotides 15 941 and 13 064 represents an area of focal vulnerability to deletion mutations and includes genes for the tRNA glutamic acid, cytochrome b and NADH subunits 5 and 6 (Fig. 3). The cytochrome b‘hotspot’ region was previously described in the aged rat model (Van Tuyle et al., 1996; Cao et al., 2001). The loss of the light-strand origin of replication (OL), located between the genes for tRNA-cysteine and tRNA-asparagine, was a relatively frequent occurrence (11 of the 39 mutant mtDNAs) in the monkey but not in rat deletion mutations. The loss of the OL, while retaining the heavy-strand origin of replication, is also observed in chronic progressive external ophthalmoplegia (CPEO) and Kearns–Sayre syndrome (KSS) patients (Cortopassi & Arnheim 1992). Our studies confirm and extend early in situ hybridization analyses of quadriceps muscle from aged rhesus monkeys, which suggested the loss of OL in fibers exhibiting ETS abnormalities. The deletions that lost the light-strand origin were the largest deletion mutations we characterized. Because of the need to maintain the heavy-strand origin and the predominance of the cytochrome b region as the location of one of the breakpoints, it is perhaps not surprising that all result in a genome that has lost greater than 50% of the mtDNA. The accumulation of mutant genomes lacking the OL may suggest the existence of a secondary light-strand origin of replication site. Alternatively, the unidirectional model of mtDNA replication (Clayton, 1982) has recently been challenged following elucidation of bidirectional mtDNA replication intermediates (Holt et al., 2000; Bowmaker et al., 2003). The investigators proposed that although replication normally initiates at the origin, replication can be initiated elsewhere (Bowmaker et al., 2003), and that in cells which are increasing their mitochondrial copy number these bi-directional intermediates are more prevalent (Holt et al., 2000).
A number of lines of evidence support the clonality of the mtDNA deletion mutations within an ETS abnormal region. First, only a single, smaller than wild-type amplification product was obtained from each ETS abnormal region. Secondly, the amplification of mtDNA from two regions of the same ETS abnormality identified identical deletion mutations. Finally, taking advantage of a mtDNA polymorphism at nucleotide 16 103, we determined that in homogenates and normal fibers this D-loop sequence was present as a mixture of adenine and guanine at position 16 103 (Fig. 4). In contrast, when mutant genomes were amplified and sequenced, this position was singular in nature, with mutant genomes from some fibers containing only guanine whereas others were solely adenine. Thus, although deletion mutations typically differ between fibers, within a given ETS abnormal region, the associated mtDNA deletion mutation is a single species resulting from the accumulation from an initial deletion event.
The most extensively characterized mtDNA abnormality in post-mitotic tissues of aged humans is the ‘common deletion’, mtDNA4977 (Linnane et al., 1990; Cortopassi et al., 1992). This deletion mutation is flanked by 13-bp direct repeats and is found in normally aged humans as well as mitochondrial myopathy patients.An mtDNA common deletion has also been identified in rhesus monkey (Lee et al., 1993, 1994).It is a 5781-bp deletion (rhmtDNA5781) flanked by imperfect direct repeats of 17 and 18 bp.This ‘common’ deletion was observed in muscle tissue homogenates from all rhesus monkeys over 13 years of age (Lee et al., 1994). Fibers 37 and 38 from the 29-year-old monkey in this study and fiber 36 from the 39-year-old monkey had this ‘common’ deletion mutation within the ETS abnormal region.
We have previously examined quadriceps muscle from aged rats and demonstrated that individual muscle fibers, which exhibit the SDH++/COX− phenotype, contain truncated mtDNA (Cao et al., 2001). Molecular analysis of the deletion mutations revealed that the breakpoints were contained entirely in the major arc, and that all maintained both origins of replication. In addition, rat mtDNA deletion breakpoints rarely contained direct repeat sequences. Direct repeat sequences have been identified at the breakpoint junction in humans (Cortopassi & Arnheim, 1990; Linnane et al., 1990; Hattori et al., 1991), with the deletion mutation resulting in the loss of one of the repeats. We found direct repeat sequences in many but not all of the mtDNA deletion breakpoints in rhesus monkey.
The ETS abnormalities in this study were mosaically distributed and segmental. Molecular analysis of these deletion mutations required a technique that permitted sampling of a portion of a single muscle fiber. Laser capture microdissection provided a powerful means of sampling a section of a single cell for subsequent molecular characterization. The mtDNA deletion mutations that accumulated in ETS abnormal fibers of aged rhesus monkeys exhibited characteristics quite different from rat mtDNA deletion mutations. In several rhesus monkey muscle fibers, the deletion event resulted in a mitochondrial genome that lacked the light-strand origin of replication. Additionally, a majority of the deletion breakpoints in the monkey were flanked by direct repeat sequences; direct repeats are also observed in humans, but rarely in rats. The accumulation of mtDNA deletion mutations in aged muscle fibers, occurring in rodents, non-human primates and humans, appears to be a universal phenomenon in mammals. In all of these species, deletion mutations clonally accumulate within regions of the fibers. There are, however, species-specific differences in the region of the genome lost in the deletion event. In addition, the presence of direct repeat sequences in many of the deletion mutations in monkey and humans and their absence in rodents suggest mechanistic differences in deletion mutation formation.
Animals and sample preparation
Rhesus monkeys (Macaca mulatta) were housed at the Wisconsin Regional Primate Research Center (WRPRC) in accordance with the ‘Guide for the Care and Use of Laboratory Animals’ (US Department of Health and Human Services, NIH Publication 86-23) and local institutional animal care committee protocols. Vastus lateralis muscle tissue was obtained at necropsy from two aged (29- and 39-year-old) male monkeys. The maximum lifespan of captive rhesus monkeys is approximately 40 years. The 39-year-old monkey represents the oldest animal from the WRPRC. Muscle samples were embedded in optimal cutting temperature (OCT, Sakura, Torrance, CA, USA) medium, frozen in liquid nitrogen and stored at −80 °C until use. Using a cryostat, 200 consecutive, 10-µm-thick cryosections were cut and placed on slides. Sections were stored at −80 °C until required. The first slide of each series was stained with haematoxylin and eosin (H&E), the second slide for cytochrome c oxidase (COX; Seligman et al., 1968) and the third slide for succinate dehydrogenase (SDH; Dubowitz, 1985) enzyme activities. The 4th−7th slides were stored at −80 °C for laser capture microdissection and mtDNA PCR analysis; the 8th, 9th and 10th slides were stained for H&E, COX and SDH activities, respectively. This staining pattern was repeated throughout the 200 slide sections for each sample.
Histochemistry and microscopy
Muscle fiber sections were examined for COX and SDH enzyme activities using a Nikon E600 microscope at magnifications of 2× and 20×. fibers exhibiting both the COX-negative (COX−) and SDH hyper-reactive (SDH++) phenotypes were followed, at 70-µm intervals, throughout the 2000 µm of cut tissue. COX and SDH phenotypes were recorded along the length of the fibers using digital images (RT Spot Digital Camera, Diagnostic Instruments, Sterling Heights, MI, USA) and MetaVue digital image acquisition and automation software (Universal Imaging Corp., Downington, PA, USA).
Laser capture microdissection
Laser capture microdissection (LCM) was performed as described in Cao et al. (2001). Frozen sections, immediately adjacent to the COX and SDH sections used to identify the abnormal fibers, were stained for SDH activity to identify the fiber during LCM. Sections were dehydrated through an ethanol and xylene series, air dried and immediately used for LCM. Individual, ETS abnormal fiber sections (Fig. 1) were microdissected using a PixCell II laser-capture microscope (Arcturus Bioscience, Inc., Mountainview, CA, USA). Settings for LCM were a laser spot size of 15–30 µm diameter, a pulse power of 30 mV and a pulse width of 50 ms. A thermoplastic-coated CapSure cap was positioned over a single fiber section and the laser aimed at the specific muscle fiber. Caps were secured in 0.6-mL tubes and stored in a desiccator until DNA isolation.
Total DNA was isolated by placing 1 µL of digestion solution (2.0 mg mL−1 proteinase K, 0.5% SDS, 10 mm EDTA) directly on the dissected muscle fiber section on the CapSure lid. The samples were incubated at 37 °C for 30 min in a humidified chamber. The released DNA was recovered with 10 µL water and stored at −20 °C.
Polymerase chain reaction
PCR analysis was performed on 1 µL of the DNA solution. A combination of long extension PCR (LX-PCR) and nested PCR was employed to increase the sensitivity of the reactions. The LX-PCRs were performed (94 °C for 30 s, 60 °C for 30 s, 72 °C for 15 min, 25 cycles) using high-fidelity long-extension TaKaRa LA Taq polymerase (Takara Mirus Corp., Madison, WI, USA) according to standard protocols. These primers were located in the D-loop region and based on the human mitochondrial genome (GenBank accession no. J01415). The PCR products were size-fractionated on a 1% agarose gel. A second LX-PCR was performed, if necessary, under identical conditions with nested primers. The amplification product was fractionated on an agarose gel, the band was excised and the mtDNA was gel purified (Qiagen, Valencia, CA, USA).
Sequencing of deletion breakpoints
The PCR products from abnormal fibers of the 39-year-old monkey were amplified using short extension PCR and TaKaRa EX Taq polymerase (Takara) to produce an amplicon for cloning. The amplification product was purified, ligated into the pGEM®-T Easy vector (Promega), and transformed into JM109 E. coli cells (Promega). The plasmid DNA was purified using Wizard® Plus SV Minipreps DNA Purification System (Promega), and sequenced with T7 and SP6 primers using standard ABI BigDye Terminators protocols (92 °C for 20 s, 45 °C for 30 s, 54 °C for 4 min, 50 cycles). The sequencing reactions were purified using AutoSeq G-50 columns (Amersham Biosciences, Bucks., UK) and sequence analysis performed at the University of Wisconsin Biotechnology Center DNA Sequencing Laboratory.
The mtDNA amplicons from the 29-year-old monkey were gel purified (Qiagen) and the purified sample was directly sequenced using BigDye on an ABI 3700 capillary-based DNA analyser (Applied Biosystems International, Foster City, CA, USA) at the University of Wisconsin-Madison DNA Sequencing Facility using standard protocols (95 °C for 20 s, 45 °C for 30 s, 60 °C for 4 min, 35 cycles). The sequencing reactions were purified using Agencourt magnetic beads (Bioscience Corp., Beverly, MA, USA).
The breakpoint sequences of mtDNA deletion products from both animals were determined by primer walking. In all cases, loss of genetic material predicted by size fractionation on agarose gels corresponded to the locations of breakpoint sequences. All sequence data was aligned using SeqMan (DNASTAR, Madison, WI, USA).
We would like to thank Jody Johnson for her critical review of the manuscript. This work was supported by grants from the National Institutes of Health (RO1 AG11604; PO1 AG11915) and the base grant for the Wisconsin Regional Primate Center (5P51 RR 000167).