Volume 4 Number 9 Volume 4
Number 9
September 1977 September
1977
Nucleic Acids Research
Nucleic Acids
Research
Complex mitochondrial DNA in Drosophila
Dilip M. Shah and Charles H. Langley
Population Genetics Section, Laboratory of Environmental Mutagenesis, National Institute of Environmental Health Sciences, National Institute of Health, Research Triangle Park, NC 27709, USA Received 9 May 1977
ABSTRACT The larval mtDNA isolated from D. virilis, D. simulans and D. melanogaster exists in complex molecular forms in addition to the simple monomeric circular form. -The frequency of circular dimers and oligomers is highly elevated in apparently normal larval tissues. These complex forms of mtDNA are separable on agarose gels. Hind III restriction endonuclease and electron microscopic analyses used in the present study have revealed that circular dimers are simply the circular concatemers of two monomeric circles which are arranged in a head-to-tail structure with no detectable heterologous regions such as insertions or deletions. The electrophoretic patterns of Hind III digested mtDNAs of D. simulans and D. melanogaster (sibling species) are identical and distinguishable from that of a distantly related species, D. virilis.
INTRODUCTION The mitochondrial DNA (mtDNA) of a variety of higher organi.sms occurs as double-stranded circular molecules with contour lengths of approximately 5 pm and a molecular weight of about 107 (1). MtDNA isolated from a number of normal manmalian organs has been shown to contain only a small per centage, 0.1 to 2%, of circular dimers and oligomers (2,3). Based on the analysis of mtDNA from several labs and their own observations with mtDNA of animal thyroids, Matsumoto et al (2) have concluded that an increased frequency of circular dimers is generally associated with genetically and physiologically abnormal cells (tissue culture lines, virus-transformed cells and malignant or pathological tissues) and that they occur only infrequently in normal tissues. They have also advanced the notion that while the frequency of circular dimers is near zero in normal tissues, there exists a small but significant variation which presumably depends on the type of tissue, species and other factors. MtDNA of a number of species of Drosophila has been shown to be in the form of circular molecules with molecular weights ranging from 9.90 to 12.35 x 106 (4,5). Although 50 - 90% of the total egg mtDNA from various species was isolated as open circles in these studies, no complex forms C) Information Retrieval Limited 1 Falconberg Court London W1V 5FG England
2949
Nucleic Acids Research (catenanes, circular dimers and oligomers) of mtDNA were detected. In the present study, we have used restriction endonuclease and electron microscopic analyses to analyze the complexity of larval mtDNA of Drosophila. MATERIALS AND METHODS Three species of Drosophila used in these experiments are 0. melanogaster, D. simulans and D. virilis. Since intraspecific polymorphism in mtDNA is known in several mammalian species (6,7) only inbred lines with homogeneous cytoplasms were utilized. The D. melanogaster line used is cn,bw-Madison. The origin of the cytoplasm of this line is not known. The D. simulans line (D.s. 18) is an isofemale line established from a single, wild-caught female from the Raleigh, NC area. The D. virilis line (D.v. 10) was also established from a single, wild-caught female from southern Japan. All flies were grown on cornmeal media contained in half-pint bottles. The second and third instar larvae were collected from the bottles using 20% sucrose.
Isolation of mitochondrial DNA In a typical experiment, 20 - 25 g of freshly collected larvae were removed into ice-cold mannitol buffer (10 ml/g of larvae) consisting of 0.23 M mannitol, 0.07 M sucrose, 0.01 M Tris-HCL, 0.001 M EDTA, pH 7.6 and then homogenized with a motor-driven, tight-fitting teflon pestle. The homogenate was centrifuged twice at 482 x g for 15 minutes to remove nuclei. The supernatant was then centrifuged at about 11,000 x g for 20 minutes to obtain the mitochondrial pellet. The crude mitochondria were resuspended in mannitol buffer and purified through a step sucrose gradient according to Clayton and Vinograd (8). The mitochondria at the 1.0 - 1.5 M sucrose interface were removed with a pipette, diluted fourfold with mannitol buffer, and pelleted. The mitochondrial pellet was resuspended in 9.5 ml of 0.1 M TrisEDTA, pH 8.0. For lysis, 0.5 ml of 0.1 M Tris-EDTA, pH 8.0 containing 10% sarkosyl was added and allowed to sit at room temperature for 30 minutes. The total volume was then adjusted to 11.3 ml with Tris-EDTA buffer. Then, 13.0 g CsCl and 2 ml of ethidium bromide solution (700 ug/ml in 0.1 M TrisEDTA, pH 8.0) were added. The mixture was then poured in equal volumes into two pollyallomer centrifuge tubes and centrifuged at 44,000 rpm at 200C for 40 - 44 hours. The upper and lower bands of mtDNA were collected and rebanded in CsCl - ethidium bromide gradient. Ethidium bromide and CsCl were removed by dialysis against 1.0 M NaCl, 0.05 M Tris, 0.01 M NA2EDTA, 2950
Nucleic Acids Research pH 8.5 for 24 hours and subsequently against 0.1 M NaCl, 0.05 M Tris, 0.01 M NA2EDTA, pH 8.0 for 24 - 36 hours. When necessary, DNA was concentrated in the dialysis tubing using dry Ficoll (sigma M.W. 400,000).
Hind III restriction endonuclease digestion Hind III was purchased from Miles Labs. Reaction was performed at 37°C in 50 mM NaCl, 6 mM Tris, pH 7.5, 6 mM MgCl2 and 100 pg BSA/ml for two hours. In most cases, 1 pg of DNA in a total volume of 50 pl was used. The enzyme used in a reaction mixture was 3 to 4 times more than was necessary to completely digest 1 ig of lambda DNA in one hour. The reaction was quenched by heating the mixture at 65°C for five minutes.
Gel el ectrophoresi s
Electrophoresis in 0.7 or 1% agarose gels was carried out according to Helling et al (9). Electrophoresis buffer (E buffer) was 40 mM Tris, 20 mM sodim acetate, 2 mM EDTA, pH 7.8. Following electrophoresis, the gels were soaked for 30 minutes in ethidium bromide (1 iig/ml in E Buffer), then illuminated with a short-wave ultraviolet transilluminator (UV products) and photographed through an orange filter on Polaroid type 55 positive/negative film. DNA was recovered from the gels using the freeze-squeeze method of Thuring et al (10). Inter- and intramolecular renaturation Intermolecular renaturation of the denatured lower band mtDNA was accomplished as described by Clayton et al (11) except that reannealing was carried out for 16 hours. Intramolecular renaturation of lower band mtDNA was achieved according to Locker et al (12). In both instances, DNA was mounted for electron microscopy from a hyperphase of 40% formamide onto a hypophase of 17% formamide.
Electrcn microscopy DNA was mounted by standard aqueous and formamide techniques (13). The DNA-protein monolayers were picked up on parlodion-coated copper grids, stained with uranyl acetate and multi-directionally shadowed with a platinum-palladium wire. Grids were examined in a Siemens Elmiskop 1A or Philips EM300 microscope. Magnification was calibrated with a grating replica (2160 lines/nm). To determine the frequency of complex molecules, the molecules were selected at 2951
Nucleic Acids Research random and scored. -Grids were scanned in such a manner as to avoid duplicate counting of the molecules. RESULTS The mtDNAs of D. virilis, D. simulans and D. melanogaster embryos have been previously isolated as circular molecules (4,5). Their molecular weights have been reported to be close to 10 x 106 for D. virilis and 12 x 106 for D. simulans and D. melanogaster. Figure 1 shows the distribution of the lengths of larval mtDNA. We have determined the molecular weights of the circular larval mtDNAs of these species relative to fX DNA by the aqueous and formamide techniques of electron microscopy (see Table 1). The
D. simulans
20F101
a ~~1.0e .U
2.0 n as
4.0
,
,,
6.0
-
12.0
IS
I0
D. melanogaster
20
as 0 0
10
z
t,
.
I
2.0
4.0
6.0
./I
120
19.0
D. virilis
40 20
S~~~~~~~--- I .b£0A
20 Figure 1.
2952
40 6.0 8.0 Length (microns)
10.0
15.0
The Distribution of the lengths of circular mitochondrial and OX DNAs. DNA was mounted by the aqueous technique.
Nucleic Acids Research molecular weights of D. virilis, D. simulans and D. melanogaster mtDNAs 10.15 x 106, 11.86 x 106 and 11.91 x 106, respectively. In this study, mtDNA was isolated by CsCl-ethidium bromide centrifiguration. Lower band mtDNA (covalently closed circular mtDNA) representing only 30 40% of the total mtDNA was analyzed for complex forms of mtDNA. Attempts to obtain higher yields of lower band DNA were unsuccessful. The frequency of complex molecular forms of total mtDNA could not be determined because upper band contained only 20% (or less) of the molecules in circular form. Lower band mtDNA was x-ray nicked before mounting on the electron microscope grids. All grids used in this analysis contained more than 90% of the total molecules in the relaxed form to allow unambiguous scoring of the complex mtDNA (2). Typical electron micrographs of such preparations are shown in Figure 2. Table 2 presents the frequencies of complex forms of mtDNA in the larvae of three Drosophila species. Since only lower band mtDNA was analyzed, the frequencies of complex forms of mtDNA represent the minimal estimates, as discussed previously (3). Circular dimers, oligomers, and linears exceeding the length of a monomeric circle, were also seen in the upper band mtDNA. The percentage of circular dimers and oligomers in D. virilis, D. simulans and D. melanogaster were 7.5, 7.5 and 7.7%, respectively, suggesting little or no variation between species. However, significant variation in the frequency of the catenated forms was observed between species ranging from 0.6% in D. simulans to 1.6% in D. virilis. In previous studies of Drosophila mtDNA which was isolated from the embryos, only a single class of monomeric circles was found. Total mtDNA, including upper and lower bands, was further analyzed by electrophoresis in 0.7% agarose gels. Four distinct bands, (a, b, c, and d) shown in Figure 3, were readily visualized. These bands were separately remean
were
TABLE 1 Molecular weights of monomeric circular mtDNA isolated from Drosophila larvae determined relative to *X DNA
Species D. virilis D. simulans D. melanogaster
Aqueous
Molecular Weight Formamide
10.13(±0.14) x 106 11.84(±0.25) x 106 11.88(±0.21) x 106
10.17(±0.18) x 106 11.88(±0.28) x 106 12.04(±0.33) x 10i6
2953
Nucleic Acids Research
Figure 2. Electron micrographs showing relaxed l-ower band mtDNA molecules. DNA was spread by the aqueous technique. - (a) D. simulans mtDNA (b) 0. virilis mtDNA. covered from the gels and analyzed in the electron microscope (micrographs not shiown). Band A contained circular'dimers, oligomers, and catenated forms of mtDNA; bands B and D were comprised of monomeric circles and linears, respectively. Band C was shown to contain linears whose lengths exceeded the 2954
Nucleic Acids Research TABLE 2 Frequency of complex mtDNA in Drosophila larvae*
Frequency of mtDNA (% - number) No. of
Molecules Species Scored D. virilis 1130 D. simulans 987 956 D. melanogaster
Monomers 90.9 91.9 91.3
Cicular Dimers 6.6 6.7 7.0
Catenated 1.6 0.6 0.9
Higher Oligomers 0.9 0.8 0.7
*Lower band mtDNA containing covalently closed-circular molecules was analyzed.
A B C a
b c
d
Figure 3.
The fluorescent photograph of agarose gel showing four bands: a, b, c and d. The upper and lower band mtDNAs were combined and electrophoresed in 0.7% agarose gel. Migration is from top to bottom. A. D. melanogaster, B. D. simulans, C. D. virilis.
length of a monomeric circle and were presumably the products of the breakage of circular dimers and oligomers. These results also suggest that complex forms of mtDNA constitute a significant fraction of the total mtDNA. In order to gain further insight into the complexity of Drosophila mtDNAs, we have carried out specific cleavage analysis of these DNAs using Hind III restriction endonuclease. Hind III restriction fragments were 2955
Nucleic Acids Research electrophoresed on 1% agarose gels. The resultant patterns are shown in Figure 4. The sizes of the fragments were determined from agarose gels calibrated with X-Hind III fragments and also by electron microscopy. It should be noted that D. virilis mtDNA is cleaved by Hind III in two large fragments and one small fragment. The two large fragments have about same size and are therefore inseparable on the gel. The Hind III fragments of
Figure 4. Gel electrophoresis of Hind III fragments of larval mtDNA and x viral DNA. (1) D. simulans + x, (2) D. simulans, (3) D. virilis + x, (4) U. virilis, (5) D. melanogaster + x, (6) D. melanogaster. 2956
Nucleic Acids Research D. melanogaster mtDNA have been sized by Klukas and Dawid (14). The sizes of Hind III fragments presented in Table 3 for D. melanogaster mtDNA agree with their reported values. The sizes of the restriction fragments (Table 3) in each species add up to the molecular weight of a monomeric circle. This suggests that circular dimers and oligomers are simply the multiple concatemers of the monomeric circles and do not contain major changes in the nucleotide sequences, i.e. deletions, additions, etc. The Hind III restriction patterns of D. simulans and D. melanogaster are indistinguishable from each other which is not unexpected of sibling species. D. virilis, which is a distantly related species, produced a distinct restriction pattern easily recognizable from those of D. simulans and D. melanogaster. In an effort to determine whether circular dimers were composed of two monomeric units joined in head-to-head or head-to-tail fashion, relaxed lower band mtDNA was denatured and allowed to renature in a manner such that only self-renaturation took place. This DNA was immediately mounted and examined in the electron microscope. No double-stranded DNA was seen. Single-stranded monomeric and dimeric circles as well as linears were seen. Clayton et al (11) have shown that circular dimeric mtDNA from human leukemic leukocytes is arranged in a head-to-tail fashion. The rationale of their experiment was that each single strand of a head-to-head circular dimer should be self-complementary and should therefore reanneal spontaneously to form linear duplexes upon self-renaturation. Since they failed to find any linear duplexes, they concluded that circular dimers were arranged in a head-to tail fashion. No linear duplexes were detected in our preparations. It is therefore evident that circular dimers of Drosophila mtDNA are also arranged in a head-to-tail fashion.
TABLE 3 Size (in base pairs)* of Hind III fragments of Drosophila mtDNA
Fragment A B C D
D. virilis EM GEL
6,136 6,136 1,969
6,359±188 6,359±188 2,242±112
-
-
D. simulans D. melanogaster EM GEL GEL EM 7,394 7,735±235 7,394 N/D 4,909 5,172±194 4,909 N/D 4,864±123 4,667 4,667 N/D 530±39 N/D
*The molecular weight of a base pair is assumed to be 660. ND = not determined. 2957
Nucleic Acids Research Inter-molecular renaturation of lower band mtDNA was studied by electron microscopy. The various forms of heteroduplex molecules (Figure 5) as
Figure 5.
2958
Electron micrographs of heteroduplex molecules resulting from inter-molecular renaturation of X-ray nicked lower band mtDNA. (A) lariate with two single-stranded tails, (B) lariate with double-stranded and single-stranded tails, (C) lariate in which a part of the tail is single-stranded, (D) "figure 8" molecule. Arrows indicate the points of attachment of tails.
Nucleic Acids Research visualized in the electron microscope resulted from the reannealing of single-stranded linear and circular molecules of both monomeric and dimeric (as well as oligomeric) duplex circles. The pathways for formation of various forms of heteroduplexes have been discussed in detail by Clayton et al (11). Excepting monomeric circles, the predominent species of DNA was circle with a tail or a lariate. In these lariate forms, circles and tails with single-stranded regions were also seen. Most frequently the tail was single-stranded. Few fused dimers, called "figure 8" molecules by Clayton et al (11), were also seen. Heterologous regions, such as insertions or deletions, were not detected in any of the heteroduplex molecules. This experiment further supports the fact that complex mtDNA constitutes a significant portion of Drosophila larval mtDNA. DISCUSSION From the data presented, it is clear that complex mtDNA is a unique feature of mtDNA within the genus Drosophila. A high frequency of circular dimeric and oligomeric mtDNA found within Drosophila larvae is striking in view of the fact that normal mammalian tissues contain a very low frequency of circular dimers. In previous studies (5,15), mtDNA isolated from Drosophila embryos has been reported to comprise of only monomeric circular molecules, although limits of detection of complex mtDNA in these studies were not estimated. These studies indicate that embryos of Drosophila contain little or no circular dimers and oligomers. The fact that larvae contain a high frequency of these forms suggests that a high frequency of complex forms of mtDNA is associated with actively growing tissues such as those of larvae. The flight muscles of D. melanogaster have giant mitochondria, called sarcosomes. Whether such giant mitochondria have elevated frequency of oligomeric mtDNA remains unknown at the present time. The difficulty of isolating covalently closed circular mtDNA from adult Drosophila makes such study less promising. The renaturation as observed in the electron microscope and restriction endonuclease analysis of larval mtDNA also support the notion that dimeric mtDNA of Drosophila, like vertebrate dimeric mtDNA, is essentially a doublesize copy of the monomeric genome with no detectable insertions or deletions. Moreover, monomeric genomes are arranged in a head-to-tail fashion in a circular dimer. This may be a universal feature of all vertebrate and invertebrate animals. D. simulans and D. melanogaster are two very closely related species 2959
Nucleic Acids Research in that they form viable hybrids, while D. virilis is more distantly related. This is consistent with our observation that Hind III restriction endonuclease patterns of D. simulans and D. melanogaster are indistinguishable from each other, while that of D. virilis can be easily recognized from the others. At present, we are attempting to more clearly define the interspecific differences in mtDNA of Drosophila by heteroduplex analysis and by comparison of the specific cleavage maps using various restri cti on endonucl eases. REFERENCES 1 Borst, P. (1972) Ann. Rev. Biochem. 41, 333-376. 2 Matsumoto, L., Piko, L. and Vinograd, J. (1976) Biochim. Biophys. Acta. 432, 257-266. 3 Clayton, D.A., Smith, C.A., Jordan, J.M., Teplitz, M. and Vinograd, J. (1968) Nature 220, 976-979. 4 Fauron, C.M.-R and Wolstenholme, D.R. (1976) Proc. Natl. Acad. Sci. USA. 73, 3623-3627. 5 Bultmann, H., Zakour, R.A. and Sosland, M.A. (1976) Biochim. Biophys Acta. 454, 21-44. 6 Potter, S.S., Newbold, J.E., Hutchison III, C.A. and Edgell, M.H. (1975) Proc. Nat. Acad. Sci. USA 72, 4492-4500. 7 Upholt, W.B. and Dawid, I.B. (1976), (submitted for publication). 8 Clayton, D.A. and Vinograd, J. (1969) Proc. Natl. Acad. Sci. USA 62, 1077-1084. 9 Helling, R.B., Goodman, H.M. and Boyer, H.W. (1974) J. Virol. 14, 12351244. 10 Thuring, R.W.J., Sanders, J.P.M. and Borst, P. (1975) Anal. Biochem. 66, 213-220. 11 Clayton, D.A., Davis, R.W. and Vinograd, J. (1970) J. Mol. Biol. 47, 137-
153.
12 Locker, J., Rabinowitz, M. and Getz, G.S. (1974) Proc. Natl. Acad. Sci. USA 71, 1366-1370. 13 Davis, R.W., Simon, M. and Davidson, N. (1971) in Methods in Enzymology Vol. XXI pp. 413-428 Academic Press, New York. 14 Klukas, C.K. and Dawid, I.B. (1976) Cell 9, 615-625. 15 Wolstenholme, D.R. and Fauron, C.M.-R. (1976) J. Cell. Biol. 71, 434-448.
2960
Number 9
September 1977 September
1977
Nucleic Acids Research
Nucleic Acids
Research
Complex mitochondrial DNA in Drosophila
Dilip M. Shah and Charles H. Langley
Population Genetics Section, Laboratory of Environmental Mutagenesis, National Institute of Environmental Health Sciences, National Institute of Health, Research Triangle Park, NC 27709, USA Received 9 May 1977
ABSTRACT The larval mtDNA isolated from D. virilis, D. simulans and D. melanogaster exists in complex molecular forms in addition to the simple monomeric circular form. -The frequency of circular dimers and oligomers is highly elevated in apparently normal larval tissues. These complex forms of mtDNA are separable on agarose gels. Hind III restriction endonuclease and electron microscopic analyses used in the present study have revealed that circular dimers are simply the circular concatemers of two monomeric circles which are arranged in a head-to-tail structure with no detectable heterologous regions such as insertions or deletions. The electrophoretic patterns of Hind III digested mtDNAs of D. simulans and D. melanogaster (sibling species) are identical and distinguishable from that of a distantly related species, D. virilis.
INTRODUCTION The mitochondrial DNA (mtDNA) of a variety of higher organi.sms occurs as double-stranded circular molecules with contour lengths of approximately 5 pm and a molecular weight of about 107 (1). MtDNA isolated from a number of normal manmalian organs has been shown to contain only a small per centage, 0.1 to 2%, of circular dimers and oligomers (2,3). Based on the analysis of mtDNA from several labs and their own observations with mtDNA of animal thyroids, Matsumoto et al (2) have concluded that an increased frequency of circular dimers is generally associated with genetically and physiologically abnormal cells (tissue culture lines, virus-transformed cells and malignant or pathological tissues) and that they occur only infrequently in normal tissues. They have also advanced the notion that while the frequency of circular dimers is near zero in normal tissues, there exists a small but significant variation which presumably depends on the type of tissue, species and other factors. MtDNA of a number of species of Drosophila has been shown to be in the form of circular molecules with molecular weights ranging from 9.90 to 12.35 x 106 (4,5). Although 50 - 90% of the total egg mtDNA from various species was isolated as open circles in these studies, no complex forms C) Information Retrieval Limited 1 Falconberg Court London W1V 5FG England
2949
Nucleic Acids Research (catenanes, circular dimers and oligomers) of mtDNA were detected. In the present study, we have used restriction endonuclease and electron microscopic analyses to analyze the complexity of larval mtDNA of Drosophila. MATERIALS AND METHODS Three species of Drosophila used in these experiments are 0. melanogaster, D. simulans and D. virilis. Since intraspecific polymorphism in mtDNA is known in several mammalian species (6,7) only inbred lines with homogeneous cytoplasms were utilized. The D. melanogaster line used is cn,bw-Madison. The origin of the cytoplasm of this line is not known. The D. simulans line (D.s. 18) is an isofemale line established from a single, wild-caught female from the Raleigh, NC area. The D. virilis line (D.v. 10) was also established from a single, wild-caught female from southern Japan. All flies were grown on cornmeal media contained in half-pint bottles. The second and third instar larvae were collected from the bottles using 20% sucrose.
Isolation of mitochondrial DNA In a typical experiment, 20 - 25 g of freshly collected larvae were removed into ice-cold mannitol buffer (10 ml/g of larvae) consisting of 0.23 M mannitol, 0.07 M sucrose, 0.01 M Tris-HCL, 0.001 M EDTA, pH 7.6 and then homogenized with a motor-driven, tight-fitting teflon pestle. The homogenate was centrifuged twice at 482 x g for 15 minutes to remove nuclei. The supernatant was then centrifuged at about 11,000 x g for 20 minutes to obtain the mitochondrial pellet. The crude mitochondria were resuspended in mannitol buffer and purified through a step sucrose gradient according to Clayton and Vinograd (8). The mitochondria at the 1.0 - 1.5 M sucrose interface were removed with a pipette, diluted fourfold with mannitol buffer, and pelleted. The mitochondrial pellet was resuspended in 9.5 ml of 0.1 M TrisEDTA, pH 8.0. For lysis, 0.5 ml of 0.1 M Tris-EDTA, pH 8.0 containing 10% sarkosyl was added and allowed to sit at room temperature for 30 minutes. The total volume was then adjusted to 11.3 ml with Tris-EDTA buffer. Then, 13.0 g CsCl and 2 ml of ethidium bromide solution (700 ug/ml in 0.1 M TrisEDTA, pH 8.0) were added. The mixture was then poured in equal volumes into two pollyallomer centrifuge tubes and centrifuged at 44,000 rpm at 200C for 40 - 44 hours. The upper and lower bands of mtDNA were collected and rebanded in CsCl - ethidium bromide gradient. Ethidium bromide and CsCl were removed by dialysis against 1.0 M NaCl, 0.05 M Tris, 0.01 M NA2EDTA, 2950
Nucleic Acids Research pH 8.5 for 24 hours and subsequently against 0.1 M NaCl, 0.05 M Tris, 0.01 M NA2EDTA, pH 8.0 for 24 - 36 hours. When necessary, DNA was concentrated in the dialysis tubing using dry Ficoll (sigma M.W. 400,000).
Hind III restriction endonuclease digestion Hind III was purchased from Miles Labs. Reaction was performed at 37°C in 50 mM NaCl, 6 mM Tris, pH 7.5, 6 mM MgCl2 and 100 pg BSA/ml for two hours. In most cases, 1 pg of DNA in a total volume of 50 pl was used. The enzyme used in a reaction mixture was 3 to 4 times more than was necessary to completely digest 1 ig of lambda DNA in one hour. The reaction was quenched by heating the mixture at 65°C for five minutes.
Gel el ectrophoresi s
Electrophoresis in 0.7 or 1% agarose gels was carried out according to Helling et al (9). Electrophoresis buffer (E buffer) was 40 mM Tris, 20 mM sodim acetate, 2 mM EDTA, pH 7.8. Following electrophoresis, the gels were soaked for 30 minutes in ethidium bromide (1 iig/ml in E Buffer), then illuminated with a short-wave ultraviolet transilluminator (UV products) and photographed through an orange filter on Polaroid type 55 positive/negative film. DNA was recovered from the gels using the freeze-squeeze method of Thuring et al (10). Inter- and intramolecular renaturation Intermolecular renaturation of the denatured lower band mtDNA was accomplished as described by Clayton et al (11) except that reannealing was carried out for 16 hours. Intramolecular renaturation of lower band mtDNA was achieved according to Locker et al (12). In both instances, DNA was mounted for electron microscopy from a hyperphase of 40% formamide onto a hypophase of 17% formamide.
Electrcn microscopy DNA was mounted by standard aqueous and formamide techniques (13). The DNA-protein monolayers were picked up on parlodion-coated copper grids, stained with uranyl acetate and multi-directionally shadowed with a platinum-palladium wire. Grids were examined in a Siemens Elmiskop 1A or Philips EM300 microscope. Magnification was calibrated with a grating replica (2160 lines/nm). To determine the frequency of complex molecules, the molecules were selected at 2951
Nucleic Acids Research random and scored. -Grids were scanned in such a manner as to avoid duplicate counting of the molecules. RESULTS The mtDNAs of D. virilis, D. simulans and D. melanogaster embryos have been previously isolated as circular molecules (4,5). Their molecular weights have been reported to be close to 10 x 106 for D. virilis and 12 x 106 for D. simulans and D. melanogaster. Figure 1 shows the distribution of the lengths of larval mtDNA. We have determined the molecular weights of the circular larval mtDNAs of these species relative to fX DNA by the aqueous and formamide techniques of electron microscopy (see Table 1). The
D. simulans
20F101
a ~~1.0e .U
2.0 n as
4.0
,
,,
6.0
-
12.0
IS
I0
D. melanogaster
20
as 0 0
10
z
t,
.
I
2.0
4.0
6.0
./I
120
19.0
D. virilis
40 20
S~~~~~~~--- I .b£0A
20 Figure 1.
2952
40 6.0 8.0 Length (microns)
10.0
15.0
The Distribution of the lengths of circular mitochondrial and OX DNAs. DNA was mounted by the aqueous technique.
Nucleic Acids Research molecular weights of D. virilis, D. simulans and D. melanogaster mtDNAs 10.15 x 106, 11.86 x 106 and 11.91 x 106, respectively. In this study, mtDNA was isolated by CsCl-ethidium bromide centrifiguration. Lower band mtDNA (covalently closed circular mtDNA) representing only 30 40% of the total mtDNA was analyzed for complex forms of mtDNA. Attempts to obtain higher yields of lower band DNA were unsuccessful. The frequency of complex molecular forms of total mtDNA could not be determined because upper band contained only 20% (or less) of the molecules in circular form. Lower band mtDNA was x-ray nicked before mounting on the electron microscope grids. All grids used in this analysis contained more than 90% of the total molecules in the relaxed form to allow unambiguous scoring of the complex mtDNA (2). Typical electron micrographs of such preparations are shown in Figure 2. Table 2 presents the frequencies of complex forms of mtDNA in the larvae of three Drosophila species. Since only lower band mtDNA was analyzed, the frequencies of complex forms of mtDNA represent the minimal estimates, as discussed previously (3). Circular dimers, oligomers, and linears exceeding the length of a monomeric circle, were also seen in the upper band mtDNA. The percentage of circular dimers and oligomers in D. virilis, D. simulans and D. melanogaster were 7.5, 7.5 and 7.7%, respectively, suggesting little or no variation between species. However, significant variation in the frequency of the catenated forms was observed between species ranging from 0.6% in D. simulans to 1.6% in D. virilis. In previous studies of Drosophila mtDNA which was isolated from the embryos, only a single class of monomeric circles was found. Total mtDNA, including upper and lower bands, was further analyzed by electrophoresis in 0.7% agarose gels. Four distinct bands, (a, b, c, and d) shown in Figure 3, were readily visualized. These bands were separately remean
were
TABLE 1 Molecular weights of monomeric circular mtDNA isolated from Drosophila larvae determined relative to *X DNA
Species D. virilis D. simulans D. melanogaster
Aqueous
Molecular Weight Formamide
10.13(±0.14) x 106 11.84(±0.25) x 106 11.88(±0.21) x 106
10.17(±0.18) x 106 11.88(±0.28) x 106 12.04(±0.33) x 10i6
2953
Nucleic Acids Research
Figure 2. Electron micrographs showing relaxed l-ower band mtDNA molecules. DNA was spread by the aqueous technique. - (a) D. simulans mtDNA (b) 0. virilis mtDNA. covered from the gels and analyzed in the electron microscope (micrographs not shiown). Band A contained circular'dimers, oligomers, and catenated forms of mtDNA; bands B and D were comprised of monomeric circles and linears, respectively. Band C was shown to contain linears whose lengths exceeded the 2954
Nucleic Acids Research TABLE 2 Frequency of complex mtDNA in Drosophila larvae*
Frequency of mtDNA (% - number) No. of
Molecules Species Scored D. virilis 1130 D. simulans 987 956 D. melanogaster
Monomers 90.9 91.9 91.3
Cicular Dimers 6.6 6.7 7.0
Catenated 1.6 0.6 0.9
Higher Oligomers 0.9 0.8 0.7
*Lower band mtDNA containing covalently closed-circular molecules was analyzed.
A B C a
b c
d
Figure 3.
The fluorescent photograph of agarose gel showing four bands: a, b, c and d. The upper and lower band mtDNAs were combined and electrophoresed in 0.7% agarose gel. Migration is from top to bottom. A. D. melanogaster, B. D. simulans, C. D. virilis.
length of a monomeric circle and were presumably the products of the breakage of circular dimers and oligomers. These results also suggest that complex forms of mtDNA constitute a significant fraction of the total mtDNA. In order to gain further insight into the complexity of Drosophila mtDNAs, we have carried out specific cleavage analysis of these DNAs using Hind III restriction endonuclease. Hind III restriction fragments were 2955
Nucleic Acids Research electrophoresed on 1% agarose gels. The resultant patterns are shown in Figure 4. The sizes of the fragments were determined from agarose gels calibrated with X-Hind III fragments and also by electron microscopy. It should be noted that D. virilis mtDNA is cleaved by Hind III in two large fragments and one small fragment. The two large fragments have about same size and are therefore inseparable on the gel. The Hind III fragments of
Figure 4. Gel electrophoresis of Hind III fragments of larval mtDNA and x viral DNA. (1) D. simulans + x, (2) D. simulans, (3) D. virilis + x, (4) U. virilis, (5) D. melanogaster + x, (6) D. melanogaster. 2956
Nucleic Acids Research D. melanogaster mtDNA have been sized by Klukas and Dawid (14). The sizes of Hind III fragments presented in Table 3 for D. melanogaster mtDNA agree with their reported values. The sizes of the restriction fragments (Table 3) in each species add up to the molecular weight of a monomeric circle. This suggests that circular dimers and oligomers are simply the multiple concatemers of the monomeric circles and do not contain major changes in the nucleotide sequences, i.e. deletions, additions, etc. The Hind III restriction patterns of D. simulans and D. melanogaster are indistinguishable from each other which is not unexpected of sibling species. D. virilis, which is a distantly related species, produced a distinct restriction pattern easily recognizable from those of D. simulans and D. melanogaster. In an effort to determine whether circular dimers were composed of two monomeric units joined in head-to-head or head-to-tail fashion, relaxed lower band mtDNA was denatured and allowed to renature in a manner such that only self-renaturation took place. This DNA was immediately mounted and examined in the electron microscope. No double-stranded DNA was seen. Single-stranded monomeric and dimeric circles as well as linears were seen. Clayton et al (11) have shown that circular dimeric mtDNA from human leukemic leukocytes is arranged in a head-to-tail fashion. The rationale of their experiment was that each single strand of a head-to-head circular dimer should be self-complementary and should therefore reanneal spontaneously to form linear duplexes upon self-renaturation. Since they failed to find any linear duplexes, they concluded that circular dimers were arranged in a head-to tail fashion. No linear duplexes were detected in our preparations. It is therefore evident that circular dimers of Drosophila mtDNA are also arranged in a head-to-tail fashion.
TABLE 3 Size (in base pairs)* of Hind III fragments of Drosophila mtDNA
Fragment A B C D
D. virilis EM GEL
6,136 6,136 1,969
6,359±188 6,359±188 2,242±112
-
-
D. simulans D. melanogaster EM GEL GEL EM 7,394 7,735±235 7,394 N/D 4,909 5,172±194 4,909 N/D 4,864±123 4,667 4,667 N/D 530±39 N/D
*The molecular weight of a base pair is assumed to be 660. ND = not determined. 2957
Nucleic Acids Research Inter-molecular renaturation of lower band mtDNA was studied by electron microscopy. The various forms of heteroduplex molecules (Figure 5) as
Figure 5.
2958
Electron micrographs of heteroduplex molecules resulting from inter-molecular renaturation of X-ray nicked lower band mtDNA. (A) lariate with two single-stranded tails, (B) lariate with double-stranded and single-stranded tails, (C) lariate in which a part of the tail is single-stranded, (D) "figure 8" molecule. Arrows indicate the points of attachment of tails.
Nucleic Acids Research visualized in the electron microscope resulted from the reannealing of single-stranded linear and circular molecules of both monomeric and dimeric (as well as oligomeric) duplex circles. The pathways for formation of various forms of heteroduplexes have been discussed in detail by Clayton et al (11). Excepting monomeric circles, the predominent species of DNA was circle with a tail or a lariate. In these lariate forms, circles and tails with single-stranded regions were also seen. Most frequently the tail was single-stranded. Few fused dimers, called "figure 8" molecules by Clayton et al (11), were also seen. Heterologous regions, such as insertions or deletions, were not detected in any of the heteroduplex molecules. This experiment further supports the fact that complex mtDNA constitutes a significant portion of Drosophila larval mtDNA. DISCUSSION From the data presented, it is clear that complex mtDNA is a unique feature of mtDNA within the genus Drosophila. A high frequency of circular dimeric and oligomeric mtDNA found within Drosophila larvae is striking in view of the fact that normal mammalian tissues contain a very low frequency of circular dimers. In previous studies (5,15), mtDNA isolated from Drosophila embryos has been reported to comprise of only monomeric circular molecules, although limits of detection of complex mtDNA in these studies were not estimated. These studies indicate that embryos of Drosophila contain little or no circular dimers and oligomers. The fact that larvae contain a high frequency of these forms suggests that a high frequency of complex forms of mtDNA is associated with actively growing tissues such as those of larvae. The flight muscles of D. melanogaster have giant mitochondria, called sarcosomes. Whether such giant mitochondria have elevated frequency of oligomeric mtDNA remains unknown at the present time. The difficulty of isolating covalently closed circular mtDNA from adult Drosophila makes such study less promising. The renaturation as observed in the electron microscope and restriction endonuclease analysis of larval mtDNA also support the notion that dimeric mtDNA of Drosophila, like vertebrate dimeric mtDNA, is essentially a doublesize copy of the monomeric genome with no detectable insertions or deletions. Moreover, monomeric genomes are arranged in a head-to-tail fashion in a circular dimer. This may be a universal feature of all vertebrate and invertebrate animals. D. simulans and D. melanogaster are two very closely related species 2959
Nucleic Acids Research in that they form viable hybrids, while D. virilis is more distantly related. This is consistent with our observation that Hind III restriction endonuclease patterns of D. simulans and D. melanogaster are indistinguishable from each other, while that of D. virilis can be easily recognized from the others. At present, we are attempting to more clearly define the interspecific differences in mtDNA of Drosophila by heteroduplex analysis and by comparison of the specific cleavage maps using various restri cti on endonucl eases. REFERENCES 1 Borst, P. (1972) Ann. Rev. Biochem. 41, 333-376. 2 Matsumoto, L., Piko, L. and Vinograd, J. (1976) Biochim. Biophys. Acta. 432, 257-266. 3 Clayton, D.A., Smith, C.A., Jordan, J.M., Teplitz, M. and Vinograd, J. (1968) Nature 220, 976-979. 4 Fauron, C.M.-R and Wolstenholme, D.R. (1976) Proc. Natl. Acad. Sci. USA. 73, 3623-3627. 5 Bultmann, H., Zakour, R.A. and Sosland, M.A. (1976) Biochim. Biophys Acta. 454, 21-44. 6 Potter, S.S., Newbold, J.E., Hutchison III, C.A. and Edgell, M.H. (1975) Proc. Nat. Acad. Sci. USA 72, 4492-4500. 7 Upholt, W.B. and Dawid, I.B. (1976), (submitted for publication). 8 Clayton, D.A. and Vinograd, J. (1969) Proc. Natl. Acad. Sci. USA 62, 1077-1084. 9 Helling, R.B., Goodman, H.M. and Boyer, H.W. (1974) J. Virol. 14, 12351244. 10 Thuring, R.W.J., Sanders, J.P.M. and Borst, P. (1975) Anal. Biochem. 66, 213-220. 11 Clayton, D.A., Davis, R.W. and Vinograd, J. (1970) J. Mol. Biol. 47, 137-
153.
12 Locker, J., Rabinowitz, M. and Getz, G.S. (1974) Proc. Natl. Acad. Sci. USA 71, 1366-1370. 13 Davis, R.W., Simon, M. and Davidson, N. (1971) in Methods in Enzymology Vol. XXI pp. 413-428 Academic Press, New York. 14 Klukas, C.K. and Dawid, I.B. (1976) Cell 9, 615-625. 15 Wolstenholme, D.R. and Fauron, C.M.-R. (1976) J. Cell. Biol. 71, 434-448.
2960
