The oxidative phosphorylation capacity of human tissues is tightly regulated during development and in response to changing physiological demands (1 ). For example, neuronal tissue, skeletal muscle and spermatozoa each have specific respiratory requirements, some of which can be altered by environmental stimuli. Respiratory efficiency is determined by the interplay between nuclear and mitochondrial genes; mitochondrial DNA (mtDNA) encodes 13 proteins that are all components of the respiratory chain, whereas nuclear DNA encodes the majority of the respiratory chain proteins, all proteins that regulate replication and transcription of mtDNA, as well as proteins necessary for the biogenesis of mitochondria (2 ,3 ). Previous studies have found a correlation between oxidative phosphorylation capacity and mtDNA levels, indicating that mtDNA replication may be regulated in response to different energy demands (4 ).
A possible focal point for regulation of human mtDNA copy number is human mitochondrial transcription factor A (h-mtTFA), a high mobility group (HMG)-box protein that is a key activator of mitochondrial transcription (5 ). It is likely that h-mtTFA also has a role in mtDNA replication, since a transcript from a major mitochondrial promoter is used as an RNA primer for initiation of mtDNA replication (2 ). Recently we have reported that the mouse homolog of mtTFA (m-mtTFA) encodes two protein isoforms, one of which is imported to the nucleus rather than into mitochondria (6 ). The nuclear m-mtTFA isoform is absent from somatic cells, but in male germ cells the nuclear m-mtTFA isoform predominates and levels of the mitochondrial isoform are greatly reduced compared with somatic cells. It has been suggested previously that the nuclear isoform of m-mtTFA functions as a testis-specific transcriptional activator (7 ) or participates in nuclear chromatin condensation (8 ). However, an alternative possibility is that transcriptional interference between the promoters for the transcripts encoding the nuclear and mitochondrial m-mtTFA isoforms serves as a mechanism to down-regulate the mitochondrial m-mtTFA mRNA and protein levels in mouse testis, leading to a decrease in mtDNA copy number during spermatogenesis.
To investigate these possibilities, we have characterized the structure and regulation of h-mtTFA in human somatic tissues and male germ cells. Similarly to the mouse, alternate h-mtTFA transcript isoforms, generated by the use of alternate transcription initiation sites, are present in male germ cells and the levels of h-mtTFA protein are down-regulated during spermatogenesis. However, unlike the mouse, no nuclear mtTFA protein isoform is encoded by the testis-specific transcripts. These findings indicate that mtTFA does not play a critical role in the nucleus and, furthermore, suggest a mechanism for down-regulation of mtDNA during human spermatogenesis with implications for understanding uniparental transmission of mtDNA.
Previous studies of h-mtTFA have been based on cDNA clones isolated from a lymphocyte library, all of which encode a 246 amino acid residue protein with a cleaved mitochondrial import signal located at residues 1-42 (5 ). To investigate h-mtTFA gene structure and regulation in germ cells, we characterized cDNA clones from a human testis library. Thirteen independent clones were isolated; five of these had 5' ends that were truncated, but eight clones were found to have 5' sequences that differed from the widely expressed h-mtTFA transcript of ~2 kb found in somatic cells (Fig. 1 A-C). Sequence characteristics of these cDNAs call into question their protein-coding potential. Testis cDNAs 1-7 (1.1-1.5 kb) contain additional 5' sequence (exon IT) identical to the genomic region immediately upstream of exon I (9 ), which suggests that this class of cDNAs is produced by a testis-specific transcription initiation site located at least 280 bp upstream of the site used in somatic cells. The translation initiation site for h-mtTFA is located at nt 402-404 of exon I (Fig. 1 B), but the additional 5' sequence includes a potential upstream translational initiation codon (CGG
To determine the extent to which cDNA cloning represented transcripts present in vivo, we performed Northern hybridization with isoform-specific probes. A probe containing exons I-VII detects four transcripts on Northern blots (Fig. 2 ); transcripts 1-3 were present in all tissues, whereas transcript 4 corresponds to a heterogeneous group of testis-specific transcripts. Analysis with strand-specific probes demonstrated that all transcripts are detected by antisense, but not by sense, RNA probes (Fig. 3 ).
We have shown previously that a widely expressed mitochondrial m-mtTFA isoform of 25 kDa is present in somatic mouse tissues, whereas a nuclear isoform of 26 kDa is present only in mouse testis (6 ). Western blot analysis of protein extracts from human liver and testis identified only a single isoform of 25 kDa (Fig. 4 A), whose size corresponds to the predicted molecular mass of the mitochondrial form of h-mtTFA (5 ). However, the levels of h-mtTFA were lower in total protein extracts from testis than liver (Fig. 4 A), whereas the total levels of h-mtTFA transcripts were much higher in testis than liver (Fig. 2 ). The difference in immunodetectable levels of h-mtTFA between liver and testis is not due to large amounts of testis-specific proteins that co-migrate with h-mtTFA, and could conceivably interfere with protein transfer or detection, as shown by comparison of Coomassie blue-stained total protein between liver and testis (Fig. 4 B). In addition, the difference in levels of h-mtTFA between liver and testis is not likely to be due to testis-specific post-translational modification that specifically affects immunodetection, since the testis and liver proteins appeared as co-migrating single bands under different fractionation conditions (Fig. 4 A). Instead, the difference between h-mtTFA levels in liver and testis is likely to reflect a true down-regulation of h-mtTFA protein. This conclusion is also supported by measurements of h-mtTFA levels in muscle and different human cell lines, all of which have h-mtTFA transcript and protein levels similar to liver (data not shown).
We have demonstrated that h-mtTFA protein is down-regulated during human spermatogenesis, most likely due to the production of alternate testis-specific h-mtTFA transcript isoforms. Our findings also demonstrate that mtTFA does not play a critical role in the nucleus, and have implications for understanding transmission of mtDNA in human genetics.
Different mechanisms could account for the decreased levels of h-mtTFA protein in testis compared with somatic tissues. One possibility, for which there is no direct evidence at present, is that transcriptional interference between the promoters for the testis-specific transcript isoforms, beginning with exon IT or exon IIT, and the somatic transcript isoforms, beginning with exon I, may lead to decreased levels of the somatic transcript isoform and low levels of h-mtTFA protein in germ cell mitochondria.
Unlike the mouse, several findings suggest that the human testis-specific transcript isoforms containing exon IT or exon IIT do not encode a functional protein. There is an inverse correlation between h-mtTFA transcript and protein levels in testis compared with liver as well as in undifferentiated germ cells compared with differentiated germ cells. The testis-specific transcripts have long 5'-flanking regions, which may prevent translation of the ORF encoding h-mtTFA. Translational regulation plays a prominent role in regulating gene expression during spermatogenesis (11 ) and long 5'-flanking regions have been shown to prevent translation of certain testis-specific transcript isoforms (11 ,12 ).
Surprisingly, several lines of evidence point to the lack of a nuclear isoform of h-mtTFA in human testis: (i) Sequence comparison demonstrated that mouse exon IT, which encodes the amino terminus of the nuclear isoform of m-mtTFA (6 ), is not conserved in the h-mtTFA genomic sequence (9 ). (ii) None of the eight cloned testis-specific cDNAs encodes a h-mtTFA isoform similar to the nuclear m-mtTFA isoform. (iii) Western blot analysis of protein extracts from human testis did not identify any additional h-mtTFA isoforms. (iv) Immunohistochemical staining with a polyclonal antiserum against h-mtTFA demonstrated only cytoplasmic immunoreactivity on tissue sections from testis. The possibility still exists that a nuclear isoform of h-mtTFA may be expressed during a specific developmental stage or certain physiological situations, although we think this is unlikely. A cDNA clone for Xenopus laevis mtTFA has been isolated recently (13 ), demonstrating that the principal mitochondrial form of mtTFA is conserved among vertebrates. Future studies are necessary to determine if the nuclear testis-specific mtTFA isoform is present in other vertebrates besides mouse.
It is likely that h-mtTFA is involved in regulating mtDNA replication, since biochemical studies have demonstrated that replication is linked to transcription of mtDNA and that h-mtTFA is essential for initiation of mitochondrial transcription in vitro (14 ,15 ). Support for the involvement of h-mtTFA in regulation of mtDNA copy number also comes from studies of patients with mitochondrial disorders (3 ). There is a correlation between the levels of h-mtTFA and mtDNA in muscle (16 ,17 ); low levels of h-mtTFA are present in muscle fibers with mtDNA depletion and high levels are found in ragged-red muscle fibers with accumulation of mtDNA (16 ). Patients with infantile mitochondrial myopathy have a depletion of mtDNA and abundant mitochondria in muscle; the finding of low levels of h-mtTFA in muscle from these patients demonstrates that the levels of h-mtTFA correlate with the levels of mtDNA, but not with the abundance of mitochondria (16 ).
This study provides additional support for the hypothesis that h-mtTFA may be involved in regulating mtDNA copy number, since the down-regulation of h-mtTFA correlates with a reduction of mtDNA copy number during human spermatogenesis. We recently have demonstrated that the testis-specific transcript isoforms in the mouse are unlikely to encode the mitochondrial form of m-mtTFA and there are low levels of m-mtTFA in mitochondria isolated from mouse testis (6 ). Others have reported that there is a reduction of mtDNA copy number from ~103 to ~102 copies per cell during spermatogenesis in the mouse (18 ). These findings demonstrate that the down-regulation of the mitochondrial form of mtTFA and mtDNA is conserved in the mouse.
Interestingly, mitochondria are abundant and high levels of mtRNAs are present throughout human spermatogenesis. It has been demonstrated that mitochondria from rat male germ cells have a high oxidative phosphorylation capacity (19 ), perhaps because spermatogenesis requires high ATP production. Because mtDNA copy number is down-regulated, high levels of mtRNAs in human spermatogenic cells are probably caused by stability of existing mitochondrial transcripts, and may ensure that the oxidative phosphorylation capacity is maintained while mtDNA copy number is reduced.
In summary, our findings demonstrate that the nuclear form of mtTFA does not play a critical role during spermatogenesis in mammals. The principal function of the testis-specific transcript isoforms may be to down-regulate the levels of mtTFA in mitochondria during mammalian spermatogenesis and subsequently mtDNA copy number in sperm. The maternal inheritance of mtDNA is highly conserved (20 ), and reduction of mtDNA copy number during spermatogenesis together with other mechanisms (21 ) may prevent paternal transmission of mtDNA. It should be noted that our results do not exclude the possibility that down-regulation of mtTFA during spermatogenesis may be secondary to the down-regulation of mtDNA. However, this seems less likely since the production of alternate transcript isoforms provides a specific mechanism for down-regulating mtTFA during spermatogenesis in mouse and man. Definitive proof for the involvement of mtTFA in regulation of mtDNA copy number during spermatogenesis should be obtained by genetic experiments in the mouse, e.g. by tissue-specific knock-out of the m-mtTFA gene or overexpression of m-mtTFA during mouse spermatogenesis.
A human testis cDNA library in the [lambda]ZAP Express vector (Stratagene) was screened with a h-mtTFA cDNA (5 ) probe by using standard protocols (22 ). Isolated cDNA clones were analyzed by cyclic sequencing with fluorescent labeled sequencing primers and the ABI 373A DNA Sequencer (Applied Biosystems).
Northern blots containing poly(A)+ RNA from multiple human tissues were obtained from Clontech. Total RNA from human liver, testis, sperm and cell lines were electrophoresed in agarose gels containing formaldehyde and blotted to supported nitrocellulose as described (22 ).
Total protein extracts from human liver and testis were separated in 15 or 18% SDS-polyacrylamide gels and analyzed by Western blotting using a rabbit polyclonal antiserum to h-mtTFA as described (16 ).
Enzyme histochemical staining of succinate dehydrogenase activity and immunohistochemical staining of h-mtTFA were performed on cryostat sections of fresh frozen tissue specimens as previously described (16 ).
Paraffin-embedded formalin-fixed tissue specimens were used for in situ hybridization to detect mtDNA and mtRNA as previously described (23 ). A mixture of double-stranded human mtDNA fragments (24 ) labeled with 35S was used as probe.
Three-µm sections of formalin-fixed and paraffin-embedded tissue were mounted on Superfrost/Plus slides, deparaffinated and incubated in 25 mM HCl for 10 min. After rinsing in phosphate-buffered saline (PBS), the sections were incubated in 0.3% Triton X-100 in PBS for 15 min, rinsed in PBS and then treated with pronase E (Sigma), 0.1 mg/ml in 0.05 M Tris-HCl, 50 mM EDTA, pH 7.5 for 5 min. The sections were rinsed in PBS and incubated for 5 min in 0.1 M glycine in PBS and post-fixed in 4% paraformaldehyde in PBS for 5 min. After rinsing in PBS, the sections were acetylated by treatment with 0.25% acetic anhydride in 0.1 M triethanolamine, pH 8.0, for 10 min. After rinsing in water, the sections were dried in air and pre-hybridization was performed in 5* SSPE, 50% formamide, 1% PVP, 10% dextran sulfate, 5* Denhardts solution, 0.5% SDS, denaturated salmon sperm DNA (0.1 mg/ml), 0.1 M dithiothreitol (DTT) for 2 h at 42oC. Hybridization was performed overnight at 55oC in the same solution with the addition of 35S-labeled (107 c.p.m./ml) RNA probes. After hybridization, the sections were rinsed in 4* SSC with 10 mM DTT at 55oC for 1 h and then at 50oC for 30 min. The sections were then washed in 2* SSC with 50% formamide and 10 mM DTT for 20 min at 65oC. After rinsing in buffer (10 mM Tris-HCl, 1 mM EDTA, 0.5 M NaCl, pH 8.0) the sections were treated with RNase A (20 µg/ml) in the same buffer for 30 min at 37oC. After rinsing in RNase buffer and then in 0.1* SSC, the sections were dehydrated and dried in air. Autoradiography was performed with LM-1 emulsion (Amersham) and D19 developer (Kodak).
In vitro transcription with the plasmids pN33, pN59 and pN60 and [[alpha]-35S]UTP was performed to obtain antisense and sense probes corresponding to exons I-VII, exon IIT and exon IT of h-mtTFA, respectively. The plasmid pN33 (a 1911 bp h-mtTFA cDNA clone in the pBK-CMV phagemid obtained from the testis cDNA cloning) was used to generate RNA probes corresponding to exons I-VII of h-mtTFA. Transcripts from pN33 were subjected to partial alkaline hydrolysis to obtain RNA fragments with an estimated average length of 0.3 kb prior to hybridization. The plasmid pN59 was constructed by cloning a 110 bp EcoRI fragment of exon IIT (nucleotides 1-110 of Fig. 1 C) into the corresponding site of the Bluescript KS II+ vector. The plasmid pN60 was constructed by isolating a 270 bp FokI (treated with Klenow polymerase to obtain a blunt end)-EcoRI fragment of exon IT (nucleotides 1-270 of Fig. 1 B), which was cloned into the EcoRI and SmaI sites of the Bluescript KS II+ vector.
Technical assistance by Monica Jacobsson and Monica Jansson is gratefully acknowledged. This study was supported by grant R37 GM-33088-26 from the National Institute of General Medical Sciences to D.A.C. and by a grant from the Swedish Medical Research Council (project # 7122) to A.O. N.-G.L. is a Howard Hughes Medical Institute Physician Postdoctoral Research Fellow. G.S.B. is a Howard Hughes Medical Institute Assistant Investigator.
Human Molecular Genetics Pages
©
Introduction
Results
Cloning of testis h-mtTFA cDNAs
Northern blot analysis of h-mtTFA transcripts
Down-regulation of h-mtTFA protein during spermatogenesis
Discussion
Materials And Methods
Isolation of testis h-mtTFA cDNA clones
Northern blot analysis
Western blot analysis
Enzyme and immunohistochemical staining
In situ hybridization to detect mtDNA and mtRNA
In situ hybridization to detect h-mtTFA transcripts
RNA probes for in situ hybridization
Acknowledgements
References
REFERENCES
This page is maintained by OUP admin. Last updated Fri Jan 31 11:45:11 GMT 1997. Part of the OUP Journals World Wide Web service.Copyright Oxford University Press, 1996