Báo cáo Y học: The nuclear genome is involved in heteroplasmy control in a mitochondrial mutant strain of Drosophila subobscura

Eur. J. Biochem. 269, 998–1005 (2002) (cid:211) FEBS 2002

The nuclear genome is involved in heteroplasmy control in a mitochondrial mutant strain of Drosophilasubobscura

Ge´ raldine Farge, Sylvie Touraille, Samuel Le Goff, Nathalie Petit, Monique Renoux, Fre´ de´ ric Morel and Serge Alziari

Equipe (cid:212)Ge´nome Mitochondrial(cid:213), UMR CNRS 6547, Universite´ Blaise Pascal-Clermont II, Aubie`re, France

these changes, NADH dehydrogenase activity, which is halved in the mutant strain (five subunits of this complex are affected by the mutation), gradually increases and stabilizes near the wild-type activity. A return to a nuclear context is accompanied by the opposite phenomena: progressive increase in heteroplasmy level and stabilization at the value seen in the wild-type strain and a decrease in the activity of complex I. These results indicate that the nuclear genome plays an important role in the control of heteroplasmy level and probably in the production of rearranged genomes.

Keywords: deletion; heteroplasmy; mitochondria; nuclear control; respiratory complexes.

Most (78%) mitochondrial genomes in the studied mutant strain of Drosophila subobscura have undergone a large-scale deletion (5 kb) in the coding region. This mutation is stable, and is transmitted intact to the offspring. This animal model of major rearrangements of mitochondrial genomes can be used to analyse the involvement of the nuclear genome in the production and maintenance of these rearrangements. Successive backcrosses between mutant strain females and wild-type males yield a biphasic change in heteroplasmy level: (a) a 5% decrease in mutated genomes per generation (from 78 to 55%), until the nuclear genome is virtually replaced by the wild-type genome (seven to eight crosses); and (b) a continuous decrease of 0.5% per generation when the nuclear context is completely wild-type. In parallel with

of the multiple deletions [13,14]; however, the mechanisms of these involvements are not yet fully understood.

To date, no mammalian model of mitochondrial genome rearrangement has been elucidated, but different hetero- plasmic mice have recently been obtained by transgenesis [15] or by directly introducing mitochondria into fertilized eggs [16].

We have studied a mutant

regions

Various mutations of the mitochondrial genomes have been correlated with human diseases, first affecting the high energy-consuming tissues such as muscles and the nervous system [1–3]. These mutations generally do not affect all of the mitochondrial genomes, and intact and mutated mitochondrial (mt)DNA can coexist (heteroplasmy) in a given cell or mitochondrial matrix. Disease severity is often greater when the proportion of mutated genomes is higher [3]. These mutations may be point mutations, as in the case of the diseases MERRF (myoclonus epilepsy with ragged red fibers) and MELAS (mitochondrial myopathy, encephalopathy, lactic acidosis and stroke-like episodes) [4,5], or concern larger (duplications and/or deletions) as in Kearns–Sayre syndrome and Pearson syndrome [6–9]. In these cases of substantial rearrange- ments, the mutations are usually sporadic. However, in certain well-described diseases (adPEO), the mutation is hereditary, autosomal and dominant [10,11]. Several differ- ent loci have been identified [11,12]. Various genes are therefore probably involved in the genesis or maintenance

strain of Drosophila (D. subobscura) which is an animal model of these substan- tial rearrangements [17,18]. Two populations of mitochon- drial genomes coexist in this strain: a minority population (20%) whose size (15.9 kb) is equivalent to that of the genomes present in the wild-type strain, and a majority population (80%) of smaller genomes (10.9 kb) which have lost by deletion over 30% of their coding region [19]. The mutation is stable, present in all tissues and is transmitted to the offspring [20]. However, the mutation does not give rise to an evident phenotype, except for a decrease in activity of complex I (40–50%), which is particularly affected by the mutation, and of complex III (20% decrease), but the ATP synthesis capacity and the cellular energy balance are preserved [20,21]. The heteroplasmy is probably intrami- tochondrial [22,23], and is less marked in the oocytes (55–60%) [24,25].

Are the nuclear genes in this mutant strain involved in the production of mutated genomes and in their maintenance at such a high level as in cases of human adPEO? The investigational possibilities allowed by our Drosophila model enable these vital questions to be addressed.

In the work presented here, we crossed virgin females of the mutant strain (H) with males of the wild-type strain (W) for several generations, thus progressively changing the nuclear genome. The mitochondria of the mutant strain were therefore placed in a wild-type nuclear context. Changes in

Correspondence to S. Alziari, Equipe Genome Mitochondrial, UMR CNRS 6547, 63177 Aubie` re Cedex, France. Fax: + 33 04 73 40 70 42,Tel: + 33 04 73 40 74 24, E-mail: serge.alziari@univ-bpclermont.fr Abbreviations: mtDNA, mitochondrial DNA; DTNB, 5,5’-dithio- bis(2-nitrobenzoic acid); MERRF, myoclonus epilepsy with ragged red fibers; MELAS, mitochondrial myopathy, encephalopathy, lactic acidosis and stroke-like episodes; Nbs2, 5,5’-dithiobis(2-nitrobenzoic acid). (Received 7 September 2001, revised 13 November 2001, accepted 12 December 2001)

Nuclear control of heteroplasmy in a Drosophila mutant (Eur. J. Biochem. 269) 999

(cid:211) FEBS 2002

heteroplasmy level, the mitochondrial genome content, and the activity of respiratory complex I were analysed. The results indicate that the heteroplasmy level is controlled, at least in part, by the nuclear genome, and that the biochem- ical phenotype of complex I deficiency disappears when less than 60% of the mitochondrial genomes are mutated.

there was no DNA amplification because the primers were too widely separated (5 kb). Primers, synthesized by Euro- gentec, were homologous with the ND1 and ND5 genes of D. subobscura mtDNA: 5¢-CTCAACCTTTTTGTGATG CAA-3¢ and 5¢-CCACGAAGTATAATTCATCTTTC-3¢ (2080–2101 and 1381–1404: genomic positions according to Volz-Lingenho¨ l [17]).

PCR products were analysed by 1% agarose gel electro-

E X P E R I M E N T A L P R O C E D U R E S

phoresis and visualized by ethidium bromide staining.

Strains

Flies of the D. subobscura strain were used. The hetero- plasmic strain was established from a female collected in the wild [17]. Wild-type lines (W and Greenwich) are standard homoplasmic strains of D. subobscura. Wild and mutant strains were raised on medium standard cornmeal at 19 (cid:176)C as described by [17].

Separation of total RNA Between 50 and 60 flies were ground in 1 mL RNA-PLUSTM at 4 (cid:176)C. One-tenth vol. chloroform was added. After stirring and incubation (5 min at 4 (cid:176)C), the aqueous and organic layers were separated by centrifugation at 12 000 g, 15 min, 4 (cid:176)C. The RNA was precipitated with 1 vol. isopropanol.

Total DNA preparation

Measurement of RNA content

Fractions of total DNA were obtained from 50 to 100 flies, 200 ovaries or 100 larvae, according to the methods described by Be´ ziat et al. [19] and Petit et al. [25].

RNA content was estimated by the Northern technique, using 12 S RNA as internal control. The RNA studied was ND4 (gene involved in deletion), using the probe described in [19]. The signals were analysed by densitometry. The ND4 : 12S ratios were calculated for the different popula- tions and compared.

Isolation of mitochondria

For PCR experiments, DNA was obtained as follows: eight flies were ground in 400 lL of buffer (Tris/ HCl pH 8.2 10 mM, EDTA 1 mM, NaCl 25 mM) with 200 lg proteinase KÆmL)1, and incubated for 1 h at 65 (cid:176)C. After centrifugation, the supernatant was heated for 15 min at 95 (cid:176)C to inhibit the proteinase. One microlitre of this preparation was used for each PCR assay.

Measurement of heteroplasmy

The mitochondria were isolated by differential centrifuga- tion of ground flies in buffer containing 0.22 M sucrose, 0.12 M mannitol, 10 mM Tricine, pH 7.6, 1 mM EDTA [26]. Protein was determined with Bio-Rad reagents using Bradford’s method [27]. For enzymatic assays, mitochon- dria ((cid:25) 1 mgÆmL)1) were sonicated for 6 s at 4 (cid:176)C, then frozen and thawed.

Heteroplasmy level was determined by Southern blotting of DNA fragments obtained after DNA digestion with Msp1, and hybridization with the COIII probe [19]. The signals were analysed by densitometry.

Measurement of biochemical activities

Measurement of mtDNA content

All activities were measured at 28 (cid:176)C and expressed in nmolÆmin)1Æmg)1.

(NADH-ubiquinone

)1Æcm)1)

mtDNA content was determined by Southern blotting of DNA fragments obtained after DNA digestion with EcoRI and HindIII, and hybridization with a nuclear probe (18 S) and a mitochondrial probe (LrRNA) [25]. The signals ratio for each population, analysed by densitometry, was com- pared to that obtained for H.

PCR

Complex I reductase). See also [28]. NADH oxidation was monitored at 340 nm in a buffer containing 35 mM (e (cid:136) 6220 M NaH2PO4, pH 7.2, 5 mM MgCl2, 2.5 mgÆmL)1 BSA, in the presence of 2 lgÆmL)1 antimycin, 2 mM KCN, 97.5 lM ubiquinone, 0.13 mM NADH, 50 lg mitochondrial protein. Only the rotenone-sensitive activity was noted.

)1 cm)1)

For detection of deleted DNA, PCR amplifications were performed in 50 lL reaction mixtures containing 2 mM MgCl2, 300 lM of each dNTP, 300 nM primers, 1 or 1.5 U Red GolstarTM DNA polymerase (Eurogentec) and DNA isolated from whole flies, in reaction buffer obtained from Eurogentec.

Complex IV (cytochrome oxidase). See also [29]. Oxida- tion of partially reduced cytochrome c (reduced OD ) oxidized OD (cid:136) (cid:25) 0.65) was monitored at 550 nm in pH 7.4 buffer containing (e (cid:136) 18 500 M 30 mM KH2PO4, 1 mM EDTA, 56 lM cytochrome c, 5 lg mitochondrial protein.

PCR conditions (Eppendorf Mastercycler Gradient) were as follows: 1 cycle at 94 (cid:176)C, 5 min; five cycles of 94 (cid:176)C for 30 s, 64 (cid:176)C to 59 (cid:176)C (decrease of 1 (cid:176)C/cycle) for 30 s, 72 (cid:176)C for 30 s; 21 cycles of 94 (cid:176)C for 30 s, 59 (cid:176)C for 30 s, 72 (cid:176)C for 30 s; one cycle of 94 (cid:176)C for 30 s, 59 (cid:176)C for 30 s, 72 (cid:176)C for 5 min.

If the 4.9-kb mtDNA deletion was present, the nested PCR product was 719 bp long. In the absence of deletion,

Citrate synthase. See also [30]. Thionitrobenzoic acid (yellow) derived from the reduction of 5,5’-dithiobis (2-nitrobenzoic acid) (Nbs2) with coenzyme A was moni- )1Æcm)1) in 100 mM Tris/ tored at 412 nm (e (cid:136) 13 600 M HCl pH 8 buffer, 2.5 mM EDTA, 37 lM acetyl CoA, 75 lM Nbs2, 300 lM oxaloacetate, 5 lg mitochondrial protein.

1000 G. Farge et al. (Eur. J. Biochem. 269)

(cid:211) FEBS 2002

in different populations. Fig. 2. Distribution of heteroplasmy level H, mutant; FH2 and FH21, 2nd and 21st generation of the population obtained by crossing mutant females H and wild-type males; FR, 14th generation of the population obtained by crossing the FH females (21st generation) with mutant males. Heteroplasmy was estimated as described in Fig. 1. Fig. 1. Change in heteroplasmy level of the different generations of the back-cross (FH): ‘mutant females · wild-type males’. Ten micrograms of total DNA fractions extracted from approximately 100 flies of each generation were digested by MspI, electrophoresed on 1% agarose gel, blotted on to membrane and hybridized with COIII probe. Hetero- plasmy was estimated as described in Experimental procedures.

R E S U L T S

Heteroplasmy level changes during modification of the nuclear context

80% heteroplasmy. Some flies had lower heteroplasmy level (50–60%); none exceeded 90%. The populations from the crosses with wild-type males were also of limited hetero- geneity, and a single peak was seen in all the populations explored: in the F2 generation, the peak was centred around 60–70%, although some individuals had higher hetero- plasmy (the peak is a little less marked in the initial population). In the F21 generation, the peak was centred on 50–60%.

It therefore seems that the trend in the proportion of mutated genomes across the generations concerns all the individuals: there are not several subpopulations with different heteroplasmy peaks.

From 50 generations onwards, in certain individuals, the mutated genomes were no longer detected by Southern blot (data not shown). In certain cases these genomes were always detectable by PCR (data not shown), with primers framing the fusion point (one of the primers is at ND1, the other at ND5, see Experimental procedures). In other cases, amplification was negative, indicating that these genomes had either disappeared totally or were present in very low proportion. Females in which mutated genomes were no longer detectable either by Southern blot or by PCR were crossed with males of the mutant strain, to establish isofemale lines. In the resulting descendants, the mutated genomes were not detectable by PCR (data not shown).

The nuclear context was altered by back-crossing through several generations virgin females of the mutant strain with males of the wild-type strain, giving the (cid:212)FH(cid:213) population. After eight to 10 generations, the nuclear genome was virtually paternal (99.6 and 99.9%, respectively), and hence wild-type. Fig. 1 indicates the heteroplasmy level (mutated genomes as a percentage of all mitochondrial genomes) measured in the different generations. The heteroplasmy level decreased through the generations, but not uniformly: in the first generations (five to seven depending on the experiment), the proportion of mutated genomes decreased from 80% to 50–55% (i.e. an average decrease of 5% per generation). The heteroplasmy level then continued to fall but more slowly: from 50–55% to 30–35% in over 40 generations (30% at generation 52), i.e. an average decrease of 0.5–0.6% per generation. This decrease in heteroplasmy in the second phase is almost 10 times smaller than in the first phase. Identical results were obtained when we used males from the two different wild-type strains. We have not yet obtained a population completely devoid of mutated mitochondrial genomes. The slope of the second part of the curve suggests that this will be achieved by generations 110– 120.

Evolution of heteroplasmy level on further alteration of the nuclear context

Distribution of heteroplasmy level in the different populations

Heteroplasmy level was also determined in individuals ((cid:25) 40) of each generation (Fig. 2). The mutant strain H presents limited heterogeneity with one peak centred on

To study putative effects of a further change in nuclear context, virgin females from generations above F12 (wild- type nuclear context) were back-crossed with males of the mutant strain. These crosses were continued over several generations until the mitochondria were again in a mutant

Nuclear control of heteroplasmy in a Drosophila mutant (Eur. J. Biochem. 269) 1001

(cid:211) FEBS 2002

Fig. 3. Change in heteroplasmy level in the different generations of the back-crosses FH (mutant females · wild-type males) and FR (FH21 females · mutant males).

nuclear context. Heteroplasmy levels in these populations (FR) are shown in Fig. 3. Fig. 2 shows heteroplasmy levels measured in individual flies.

whole flies, and then stabilized. The (cid:212)somatic jump in heteroplasmy(cid:213) was observed irrespective of the nuclear context and always during the phase of intense replication of the mitochondrial genomes.

Evolution of the activities of respiratory complexes I and IV

When this back-cross was performed using females with a low heteroplasmy level (55%), the percentage of mutated genomes again increased in five to seven generations and stabilized around 80%, as in the original mutant popula- tion; an increase of (cid:25) 5% per generation. This mode of variation was therefore identical to that seen in the first phase. The individual measurements show that the popu- lations are homogeneous, with a single peak centred around 80% after five to seven back-crosses, as in the original mutant population.

The progressive return to the mutant nuclear genome context is therefore accompanied by an increase in the proportion of mutated mitochondrial genomes up to the maximum seen in the mutated strain.

Evolution of heteroplasmy level during development in different lines

Complex I (NADH-ubiquinone oxidoreductase) is particu- larly affected by the mutation. In the mutant strain, the peak activity of complex I was 40–50% lower than that measured in the wild-type strains, whatever the tissue [20]. There was a smaller (20%) lowering of the activity of complex III, which is also affected by the mutation. Oxygraphic measurements of the respiratory activities of isolated mitochondria indi- cated a roughly 20–30% decrease with the substrates of complex I (pyruvate + malate or glutamate + malate). The respiration measured with substrates of complex III (a-glycerophosphate) was identical in the wild-type and mutant strains. The change in complex I activity is therefore the main biochemical manifestation of the mutation.

Does the change in nuclear context lead to altered activities of complex I?

We have shown [25,24], that heteroplasmy level is lower in the germinal line. It is 50% in the stage 10 oocytes of the mutant strain, 55% in the stage 14 oocytes, 55–60% in the embryos, and then increases during the larval stages before levelling off around 80% at the third larval stage, after which it remains constant until the death of the flies. This increase corresponds to a phase of marked replication of the mitochondrial genomes [31].

The changes in complex I activity in the Drosophila lines obtained by three FH crosses (FHa, FHb and FHc) are shown in Fig. 5. The activity increased as heteroplasmy level decreased and reached the level measured in the wild- type strain when heteroplasmy was 55–60% (end of the first phase of decrease). The activity remained stable thereafter. The biochemical phenotype therefore appears when around 60% of mitochondrial genomes are mutated, a proportion close to that reported in different pathological cases [32].

Measurements on mitochondria isolated from flies obtained after return to the mutant nuclear context (FR

Using crosses we tested whether the nuclear genome is involved in this increase in heteroplasmy level during development and in the stabilization at 80%. Heteroplasmy was measured in flies, ovaries and the third stage larvae of the different populations as the nuclear context changed. The results are shown in Fig. 4. Whatever the population, heteroplasmy level in the ovaries was always lower than that in whole maternal flies or in whole flies of the next generation. Heteroplasmy level increased at the third larval stage, whatever the nuclear context and heteroplasmy of

Fig. 4. Change in heteroplasmy level in development stages of different populations: H, FH and FR. Heteroplasmy level was measured on total DNA extracted from flies, ovaries and larvae as described in Fig. 1.

1002 G. Farge et al. (Eur. J. Biochem. 269)

(cid:211) FEBS 2002

a compensatory increase in H, considering that in the FH line, where heteroplasmy level is (cid:25) 50%, the biochemical consequences of the mutation are no longer observable.

Fig. 6. Comparison of the cellular mtDNA content (mtDNA/nuclear DNA) in the wild-type (W), mutant (H) and back-cross (FH21 and FR14) lines. Total DNA, extracted from whole flies, was digested with EcoRI and HindIII, electrophoresed on 1% agarose gel, blotted on to membrane and hybridized with nuclear probe (18 S) and mitochon- drial probe (lrRNA). DNA content was estimated as described in Experimental procedures.

However, in the line called FR, obtained by the back- crosses, whose heteroplasmy level is similar to that of the mutant strain ((cid:25) 80%), there was no increase in mitochon- drial genome content, which remained identical to that of the FH strain. A large number of generations may be required for this increase to be measurable.

Mitochondrial transcripts

with heteroplasmy 80%) show that complex I activity again drops to that measured in the mutant strain (data not shown). So the biochemical phenotype seems to be linked to the heteroplasmy level.

The enzymatic activity of complex IV (cytochrome oxidase), which is not affected by the deletion, is identical is not in the mutant and wild-type strains [20]. It significantly different in the mitochondria of FH flies with different heteroplasmy levels (1550 (cid:139) 666, 1592 (cid:139) 523 and 1695 (cid:139) 325 nmolÆmin)1Æmg)1 for W, H and FH, respectively). The same is true for the activity of citrate synthase, a matrix enzyme encoded entirely by the nuclear genome (2600 (cid:139) 615, 2386 (cid:139) 460 and 2530 (cid:139) 367 nmolÆmin)1Æmg)1 for W, H and FH, respectively).

Mitochondrial genome content

When the heteroplasmy level changes in the back-cross experiments, are the transcript concentrations of the genes involved in the deletion also affected? The concentration of the ND4 gene transcript was measured by Northern blot on the RNA extracted from wild-type and mutant strains and in three different FH experiments. These concentrations were estimated by comparison with the 12 S transcript concentrations which were shown to be identical in wild- type and mutant strains [19]. The results are shown in Fig. 7. First, as reported previously, the ND4 transcript concentration in the mutant was diminished by 55%. Second, in the three FH experiments this concentration was identical to that of the mutant until generation three. Third, this concentration increased in generation five to reach the wild-type concentration in generation six. So it appears that the heteroplasmy level decrease is paralleled by an increase in the transcript concentration of the gene concerned by the deletion.

Fig. 5. Variation in complex I activity and heteroplasmy level in the different generations of three FH back-crosses (FHa, FHb, FHc). For complex I, results were expressed as the FH/W activity ratio. Com- plex I activity was measured in the presence of 0.13 mM NADH and 97.5 lM ubiquinone as described in Experimental procedures.

D I S C U S S I O N

Our results show clearly that the percentage of rearranged mitochondrial genomes depends on the nuclear context. It is 80% of the mtDNA in the mutant strain, and decreases

Fig. 6 shows that in the mutant strain, the cells’ mitochon- drial genome content was increased by 70%, compared with the reference wild-type strain. These results have been reported previously [19,20], and were hypothesized as a compensatory effect of the mutation. Identical measure- ments were made on flies from various crosses. The mitochondrial genome content of line FH was lower than that of the mutant strain ((cid:25) 50%), and very close to the low level measured in the wild-type strain. These results point to

Nuclear control of heteroplasmy in a Drosophila mutant (Eur. J. Biochem. 269) 1003

(cid:211) FEBS 2002

If hypotheses (a) and (b) are correct, the mutated genomes should decline continuously until they are fully eliminated, but this is not the case: a new heteroplasmy level of 50% is reached, and as the two types of genomes are at identical concentration, neither is preferentially replicated. Partial duplication of these genomes (fusion of the intact and mutated genomes) could simply explain this propor- tion, as has been shown in man [33]. However, there is no duplication of mtDNA in the mitochondria of the mutant strain [25].

The third hypothesis supposes that there is no further (or less and less) production of mutated molecules from intact genomes, so that heteroplasmy level decreases by dilution of the mutated genomes. But the mutated genomes are nonetheless replicated like the wild-type genomes (no preferential replication) and a new equilibrium occurs at (cid:25) 50% of mutated genomes. The back-cross findings favour this third model: if these mitochondria are again placed in a mutant nuclear context the proportion of mutated genomes increases as described before and equilibrium is reached at (cid:25) 80% mutated genomes. So genes from the mutant nuclear genomes could be directly involved in the production of the deleted molecules.

Several different genes must be involved in this phase: in F1, we obtain heterozygotes of 75–80% heteroplasmy. In F2 this global heteroplasmy level has decreased but we do not observe two different populations, with one of 50% heteroplasmy, which is what we would expect if only one or two genes were implicated (50% and 25% homozygotes, respectively). These findings suggest a multigene system.

uniformly by (cid:25) 5% per generation as the nuclear context becomes progressively wild-type. After five to seven crosses, the nuclear genome is almost entirely wild-type and 50–60% of the mitochondrial genomes remain rearranged. The percentage then decreases much more slowly (0.5% per generation) to (cid:25) 30% after 50 generations. During the gradual transition to the wild-type nuclear context, the activity of complex I (NADH dehydrogenase) increases to that observed in the wild-type strain, and then levels off.

The genes most obviously implicated are those which code for the mitochondrial replication system (SSB genes, gamma polymerase etc.). The replication model proposed by Clayton [34] has been used as a basis for several putative deletion mechanisms [35] by slippage of the replisome. Mutant forms of the gamma polymerase which can induce rearrangements in families presenting AdPEO and ArPEO have been described [36].

The progressive return to a mutant nuclear context leads to a rapid (and uniform in the population) increase in the proportion of mutated genomes. After five to seven generations it levels off at the peak value (80%) seen in the initial mutant population. There is a parallel decrease in the activity of respiratory complex I to the mutant strain value, where it stabilizes. These variations in complex I activity observed in the mutant strain are probably due to a drop in concentration directly related to the genes con- cerned, despite a partially compensating effect of overtran- scription or increased half-life [19].

Other systems differing from those of the mitochondrial genome replisome may also be directly implicated in the deletion mechanism. For example, accumulation of deleted mtDNA is also observed in the premature death of Podospora anserina. It has been shown that mutations in the AS1 gene (coding for a cytosolic ribosomal protein), which is implicated in the fidelity of the translation, were correlated with the mitochondrial genome rearrangements [37]. The genes identified in human diseases such as AdPEO or MNGIE (ANT or thymidine phosphorylase [13,38]), could also be involved.

The change in heteroplasmy level observed in these crosses is biphasic, probably reflecting the occurrence of different phenomena. Moreover, the increase in hetero- plasmy level during development is always observed, whatever the nuclear context.

When the nuclear context is wholly wild-type, a second phase begins: the decrease in the proportion of mutated genomes is still observed, but is then only 0.5% per generation and continues for generations. Several hypo- theses can also explain this progressive and small decrease: (a) one or more elements of the wild-type mitochondrial replication system may have greater affinity for the intact than for the mutated mtDNA. This implies the existence of specific cis sequences in these two mitochondrial genomes [39,40]; (b) the molecules of mutated mtDNA are recog- nized and progressively eliminated by the mtDNA repair system. Such systems are now well characterized ([41–43], reviewed in [44]). Repair by intramolecular recombination with intact molecules could explain the progressive

The first phase of decrease in heteroplasmy level occurs relatively fast, but is far from complete, as 50% of the genomes are still in mutant form at the end of this phase. The replacement of the rearranged genomes by intact genomes is therefore effective but limited. Three hypotheses may be posited to explain this: (a) preferential replication of the intact genomes (or poor recognition of the mutated genomes on replication in a wild-type the nuclear context); (b) progressive elimination of mutated genomes; or (c) arrested production of the mutated genomes enabling heteroplasmy level to be kept constant.

Fig. 7. Comparison of the relative concentration of the mitochondrial transcript ND4 in the mutant (H), wild-type (W) and in different gen- erations of 3 FH back-crosses. The ratios were calculated from the hybridization signals (corrected by 12 S) obtained with the mutant strain (H) or with the FH populations and the wild-type strain (W). The values represent the mean of three measurements.

1004 G. Farge et al. (Eur. J. Biochem. 269)

(cid:211) FEBS 2002

Rowland, L.P. (1989) Mitochondrial DNA deletions in progres- sive external ophtalmoplegia and Kearns–Sayre syndrome. N. Engl. J. Med. 320, 1293–1299.

elimination of the deleted genomes. These enzymes involved in recombination in mitochondria have long been known in yeast (reviewed in [45]), and the same enzymatic activities have recently been reported in man and other organ- isms [46]. Such repair systems eliminate mutated molecules, but they are of limited efficiency, thus possibly explaining the shallow slope of the second phase. In the mutant strain, the genes encoding these mtDNA repair enzymes may have undergone mutation, such that their products are unable to eliminate efficiently the mutated molecules (which will have invaded the mutant’s mitochondria).

8. Poulton, J., Deadman, M.E., Bindoff, L., Morten, K., Land, J. & Brown, G. (1993) Families of mtDNA re-arrangements can be detected in patients with mtDNA deletions: duplications may be a transient intermediate form. Hum. Mol. Genet. 2, 23–30.

9. Rotig, A., Cormier, V., Blanche, S., Bonnefont, J.P., Ledeist, F., Romero, N., Schmitz, J., Rustin, P., Fischer, A., Saudubray, J.M. et al. (1990) Pearson’s marrow–pancreas syndrome. A multisystem mitochondrial disorder in infancy. J. Clin. Invest. 86, 1601–1608.

Another possibility is that these genes are unaffected by the mutation but that the production of mutated molecules exceeds the capacity of the repair systems, which are therefore unable to stop the invasion. In the wild-type nuclear context, when the production of the mutated molecules is arrested, these repair systems could slowly eliminate (0.5% per generation) the deleted genomes.

10. Zeviani, M., Servidei, S., Gellera, C., Bertini, E., DiMauro, S. & DiDonato, S. (1989) An autosomal dominant disorder with multiple deletions of mitochondrial DNA starting at the D-loop region. Nature 339, 309–311.

11. Suomalainen, A., Kaukonen, J., Amati, P., Timonen, R., Haltia, M., Weissenbach, J., Zeviani, M., Somer, H. & Peltonen, L. (1995) An autosomal locus predisposing to deletions of mitochondrial DNA. Nat. Genet. 9, 146–151.

C O N C L U S I O N S

Our results clearly indicate that the nuclear genomes are involved in the control and/or production of the mutated molecules, and that several genes are probably implicated. One potential explanation of the variation in heteroplasmy level observed in the first phase is based on active production of mutated molecules in the mutant strain. This hypothesis can be tested by experiments using back-crosses where the mitochondria of the wild-type strain are placed in a mutant nuclear context. It is then possible to screen for rearrangements identical (or not) to those observed in the mutant strain. These experiments are under way in our laboratory.

12. Kaukonen, J., Zeviani, M., Comi, G.P., Piscaglia, M.G., Peltonen, L. & Suomalainen, A. (1999) A third locus predisposing to multiple deletions of mtDNA in autosomal dominant progressive external ophthalmoplegia. Am. J. Hum. Genet. 65, 256–261. 13. Kaukonen, J., Juselius, J.K., Tiranti, V., Kyttala, A., Zeviani, M., Comi, G.P., Keranen, S., Peltonen, L. & Suomalainen, A. (2000) Role of adenine nucleotide translocator 1 in mtDNA main- tenance. Science 289, 782–785.

14. Spelbrink, J.N., Li, F.Y., Tiranti, V., Nikali, K., Yuan, Q.P., Tariq, M., Wanrooij, S., Garrido, N., Comi, G., Morandi, L. et al. (2001) Human mitochondrial DNA deletions associated with mutations in the gene encoding Twinkle, a phage T7 gene 4-like protein localized in mitochondria. Nat. Genet. 28, 223–231. 15. Zhang, D., Mott, J.L., Chang, S.W., Denniger, G., Feng, Z. & Zassenhaus, H.P. (2000) Construction of transgenic mice with tissue-specific acceleration of mitochondrial DNA mutagenesis. Genomics 69, 151–161.

A C K N O W L E D G E M E N T S

16. Inoue, K., Nakada, K., Ogura, A., Isobe, K., Goto, Y., Nonaka, I. & Hayashi, J.I. (2000) Generation of mice with mitochondrial dysfunction by introducing mouse mtDNA carrying a deletion into zygotes. Nat. Genet. 26, 176–181. This work was supported by grants from the CNRS, the Universite´ Blaise–Pascal and the Association Franc¸ aise contre les Myopathies (AFM).

R E F E R E N C E S

17. Volz-Lingenho¨ hl, A., Solignac, M. & Sperlich, D. (1992) Stable heteroplasmy for a large-scale deletion in the coding region of Drosophila subobscura mitochondrial DNA. Proc. Natl Acad. Sci. USA 89, 11528–11532. 1. Wallace, D.C. (1999) Mitochondrial diseases in man and mouse. Science 283, 1482–1488. 2. Schon, E.A. (2000) Mitochondrial genetics and disease. Trends Biochem. Sci. 25, 555–560. 3. Wallace, D.C. (2000) Mitochondrial defects in cardiomyopathy and neuromuscular disease. Am. Heart J. 139, S70–S85.

18. Alziari, S., Petit. N., Lefai, E., Be´ ziat, F., Lecher, P., Touraille, S., Debise, R. & Morel, F. (1999) An heteroplasmic strain of D. Subobscura: an animal model of mitochondrial DNA rearrangements. In Mitochondrial Diseases. Models and Methods (Lestienne, P., ed.), pp. 197–208. Springer-Verlag, Heidelberg. 19. Be´ ziat, F., Morel, F., Volz-Lingenhol, A., Saint-paul, N. & Alziari, S. (1993) Mitochondrial genome expression in a mutant strain of D. subobscura, an animal model for large scale mtDNA deletion. Nucleic Acids Res. 21, 387–392. 4. Shoffner, J.M., Lott, M.T., Lezza, A.M.S., Seibel, P., Ballinger, S.W. & Wallace, D.C. (1990) Myoclonic epilepsy and ragged red fibers disease (MERRF) is associated with a mitochondrial DNA tRNA lys mutation. Cell 61, 931–937.

20. Beziat, F., Touraille, S., Debise, R., Morel, F., Petit, N., Lecher, P. & Alziari, S. (1997) Biochemical and molecular consequences of massive mitochondrial gene loss in different tissues of a mutant strain of Drosophila subobscura. J. Biol. Chem. 272, 22583–22590. 5. Wallace, D.C., Zheng, X., Lott, M.T., Shoffner, J.M., Hodge, J.A., Kelley, R.I. & Epstein, C.M. (1988) Familial mitochondrial encephalomyopathy (MERRF): genetic, pathophysiological and biochemical characterization of a mitochondrial DNA disease. Cell 55, 601–610.

21. Debise, R., Touraille, S., Durand, R. & Alziari, S. (1993) Biochemical consequences of a large deletion in the mitochondrial genome of a Drosophila subobscura strain. Biochem. Biophys. Res. Commun. 196, 355–362. 6. Holt, I.J., Harding, A.E. & Morgan-Hugues, J.A. (1988) Deletion of muscle mitochondrial DNA in patients with mitochondrial myopathies. Nature 331, 717–719.

22. Lecher, P., Beziat, F. & Alziari, S. (1994) Tissular distribution of heteroplasmy and ultrastructural studies of mitochondria from a Drosophila subobscura mitochondrial deletion mutant. Biol. Cell 80, 25–33. 7. Moraes, C.T., DiMauro, S., Zeviani, M., Lombes, A., Shanske, S., Miranda, A.F., Nakase, H., Bonilla, E., Werneck, L.C., Servidei, S., Nonaka, I., Koga, Y., Spiro, A.J., Brownell, A.K.W., Schmidt, B., Schotland, D.L., Zupanc, M., DeVivo, D.C., Schon, E.A. &

Nuclear control of heteroplasmy in a Drosophila mutant (Eur. J. Biochem. 269) 1005

(cid:211) FEBS 2002

23. Lecher, P., Petit, N., Beziat, F. & Alziari, S. (1996) Localization by ultrastructural in situ hybridization of mitochondrial transcripts in epithelial cells of a Drosophila subobscura deletion mutant. Eur. J. Cell. Biol. 71, 423–427.

35. Shoffner, J.M., Lott, M.T., Voljanec, A.S., Soueidan, S.A., (1989) Spontaneous Costigan, D.A. & Wallace, D.C. Kearns-Sayre/chronic external ophtalmoplegia plus Syndrome associated with a mitochondrial DNA deletion: a slip-replication model and metabolic therapy. Proc. Natl Acad. Sci. USA 86, 7952–7956.

36. Van Goethem, G., Dermaut, B., Lofgren, A., Martin, J.J. & Van Broeckhoven, C. (2001) Mutation of POLG is associated with progressive external ophthalmoplegia characterized by mtDNA deletions. Nat. Genet. 28, 211–212.

24. Lecher, P., Petit, N., Le Goff, S. & Alziari, S. (2000) Quantitative analysis, by ultrastructural in situ hybridization, of mitochondrial genomes and their expression in mid-gut and ovarian cells of a mutant strain of Drosophila subobscura. Biol. Cell 92, 341–350. 25. Petit, N., Touraille, S., Debise, R., Morel, F., Renoux, M., Lecher, P. & Alziari, S. (1998) Developmental changes in heteroplasmy level and mitochondrial gene expression in a Drosophila subobs- cura mitochondrial deletion mutant. Curr. Genet. 33, 330–339. 26. Alziari, S., Stepien, G. & Durand, R. (1981) In vitro Incorporation of [35S]-methionine in mitochondrial proteins of Drosophila mel- anogaster. Biochem. Biophys. Res. Commun. 99, 1–8. 37. Dequard-Chablat, M. & Rotig, A. (1997) Homologous and het- erologous expression of a ribosomal protein gene in Podospora anserina requires an intron. Mol. Gen. Genet. 253, 546–552. 38. Nishino, I., Spinazzola, A. & Hirano, M. (1999) Thymidine phosphorylase gene mutations in MNGIE, a human mitochon- drial disorder. Science 283, 689–692. 27. Bradford, M. (1976) Protein essay reagent. Anal. Biochem. 72,

248–251. 28. Hatefi, Y.

39. Marchington, D.R., Poulton, J., Sellar, A. & Holt, I.J. (1996) Do sequence variants in the major non-coding region of the mitochondrial genome influence mitochondrial mutations associ- ated with disease? Hum. Mol. Genet. 5, 473–479. (1978a) Preparation and Properties of NADH: Ubiquinone Oxidoreductase (Complex I) EC 1.6.5.3,. In Methods in Enzymology (Fleisher, S. & Packer, L., eds), pp. 11–14. Academic press, New York.

29. Errede, B., Kamen, M.D. & Hatefy, Y. (1978) Preparation and properties of complex IV (ferrocytochrome c: oxygen oxidore- ductase EC 1.9.3.1,. In Methods in Enzymology (Fleisher, S. & Packer, L., eds), pp. 40–47. Academic press, New York. 40. Torroni, A., Lott, M.T., Cabell, M.F., Chen, Y.S., Lavergne, L. & Wallace, D.C. (1994) mtDNA and the origin of Caucasians: Identification of ancient Caucasian-specific haplogroups, one of which is prone to a recurrent somatic duplication in the D-loop region. Am. J. Hum. Genet. 55, 760–776.

30. Sheperd, D. & Garland, S. (1969) Citrate synthase from rat liver. In Methods in Enzymology (Lowenstein, J. M., eds), pp. 11–16. Academic press, New York. 41. Shadel, G.S. & Clayton, D.A. (1997) Mitochondrial DNA maintenance in vertebrates. Annu. Rev. Biochem. 66, 409–435. 42. Croteau, D.L., Stierum, R.H. & Bohr, V.A. (1999) Mitochondrial 31. Church, R.B. & Robertson, F.W. (1966) A biochemical study of the DNA repair pathways. Mutat. Res. 434, 137–148. 43. Bohr, V.A. & Anson, R.M. (1999) Mitochondrial DNA repair growth of Drosophila melanogaster. J. Exp. Zool. 162, 337–351. pathways. J. Bioenerg. Biomembr. 31, 391–398. 44. Bogenhagen, D.F. (1999) Repair of mtDNA in vertebrates. Am. J. Hum. Genet. 64, 1276–1281. 32. Lestienne, P., Bouzidi, M., Desguerre, I. & Ponsot, G. (1999) Molecular basis of mitochondrial DNA diseases. In Mitochondrial Diseases. Models and Methods. (Lestienne, P., ed.), pp. 33–58. Springer-Verlag, Heidelberg.

45. Contamine, V. & Picard, M. (2000) Maintenance and integrity of the mitochondrial genome: a plethora of nuclear genes in the budding yeast. Microbiol. Mol. Biol. Rev. 64, 281–315. 33. Poulton, J., Deadman, M.E. & Gardiner, R.M. (1989) Duplica- tions of mitochondrial DNA in mitochondrial myopathy. Lancet 1, 236–239. 34. Clayton, D.A. (1982) Replication of animal mitochondrial DNA. Cell 28, 693–705. 46. Thyagarajan, B., Padua, R.A. & Campbell, C. (1996) Mammalian mitochondria possess homologous DNA recombination activity. J. Biol. Chem. 271, 27536–27543.