Fill and Sign the Template Hmrc Tax Declaration Format

(:opyright 0 1990 by the Genetics Society of America Behavior of the [mi-31 Mutation and Conversionof Polymorphic mtDNA Markers in Heterokaryons of Neurospora crussu Alexisann Hawse,* Richard A. Collinst and Frank E. Nargang* *Department of Genetics, University of Alberta, Edmonton, Alberta T6G 2E9, Canada, and ?Departmentof Botany, University of Toronto, Toronto, Ontario M5S lA1, Canada Manuscript received December 29, 1989 Accepted for publication May 24, 1990 ABSTRACT We have examined the behavior of the [mi-?] mitochondrial mutation and two physical mtDNA markers in heterokaryotic culturesof Neurospora crassa. Previous workers showed that a 1.2-kilobase insertion in the larger polymorphic form of EcoRI-5 restriction fragment is a site of high frequency and rapid unidirectional gene conversion. We have confirmed this observation and determined by DNA sequence analysis that theinsertion in the EcoRI-5 fragment correspondsprecisely to anoptional intron that containsa long open reading frame in the NDI gene. Thus, the conversion of the short, intron-lacking, form of EcoRI-5 to the longer, intron-containing, form may be analogous to the unidirectional gene conversion events catalyzed by intron-encoded proteins in other organisms. T h e resolution of two polymorphic forms of the mtDNAEcoRI-9 restriction fragment in our heterokaryons differs from that observed previously and suggests that this locus is not a site of gene conversion in our heterokaryon pair. T h e size polymorphism of the EcoRI-9 fragments is due to atandemly reiterated 78-base-pair sequence which occurs two times in the short form and three times in the long form. One copy of the repeat unit and 66 base pairs following it have been duplicated from the N D 2 gene which is located about 30 kilobases distant on the mtDNA. In contrast to[poky] the mitochondrial mutant, which was completely dominant over wild-type mitochondria in heterokaryons, the [mi-?] mutant was recovered in only seven of twenty heterokaryons after ten cycles of conidiation and subculturing. T h e resolution of the [mi-?] or wild-type phenotype in heterokaryons may depend solely on random factors such as allele input frequency, drift, andsegregation rather than specific dominant or suppressive effects. time for complete takeover by the mutant. Thus, it ARLY studies on a variety of cytoplasmically inlikely that [ p o k y ] is completely dominant towild seems heritedmutants of Neurospora crassa demontype under these conditions. strated that the phenotype of these mutants could The investigations into the behavior of the [ p o k y ] predominant over wild-type strains in heterokaryons, mutant in heterokaryons employed strains of N. crassa or when mitochondria purified from mutant strains carrying polymorphic mtDNA markers. An attempt were microinjected into the hyphal compartments of was made to utilize these markers to correlate converwild-type strains (PITTENGER 1956; GARNJOBST, WILsion of heterokaryons to the mutant phenotype with SON, and TATUM 1965; DIACUMAKOS, GARNJOBST and TATUM 1965; BERTRAND and PITTENGER 1969). Furthe presence of aparticular polymorphic marker ther investigation of this phenomenon showed that (MANNELLAand LAMBOWITZ 1979). The studies were heterokaryons constructed between the [ p o k y ] mutant inconclusive regardingthegenetic location of the [ p o k y ] mutation,butthe analysisof heterokaryon and wild-type would acquire the mutant phenotype mtDNAs revealed unusual behavior for two polymorafter two to eight cycles of subculturing heterokarphic markers. These markers, an insertion of ca. 50 yotic conidia onto fresh minimal medium. The number of cycles required appeared to be unrelated to the base pairs in the EcoRI-9 fragment of the type I initial ratio of conidia used to form the heterokaryon polymorphic form of mtDNA relative to the corre(MANNELLA and LAMBOWITZ 1978). In data combined sponding fragment in the type I1 polymorphic form, from two separate studies, at least 38 of 42 heterokarand a ca. 1200 base pairinsertion in the EcoRI-5 yons constructed between wild type and [ p o k y ] showed fragment of type I1 mtDNA relative to the type I the mutant phenotype (MANNELLAand LAMBOWITZ version, were found to be sites of high frequency 1978; 1979).Since the four thathad not acquired the unidirectional gene conversion.That is, the two inser[ p o k y ] phenotype had been taken through only two to tions were found to spreadrapidly through the mixed five conidial passes (MANNELLAand LAMBOWITZ mtDNA population of the heterokaryons (MANNELLA 1979), it is probable that they had not had sufficient and LAMBOWITZ 1979). E ( ; ~ I I ~ I I C >1 2 6 ti3-72 (Septenlhel-.1990) 64 A. Hawse, R. A. CollinsE.and F. Recently, we constructed heterokaryons to verify the genetic location of t h e [mi-31 mutation (LEMIRE a n d NARGANC 1986). Analysisof these heterokaryons revealed that unlike [ p o k y ] , [mi-)] did not predominantover wild typein all heterokaryons.Furthermore, while we did observe complete conversion to the typeI1 form of EcoRI-5 in heterokaryon mtDNAs, we did not observe the same extent of conversion to the type I form of EcoRI-9 reported in the previous study (MANNELLAa n d LAMBOWITZ 1979). Here we describe further studies of the interaction between [mi-31 a n d wild-type mitochondria, and the conversion of the two polymorphic insertions in heterokaryons examined after varying numbersof cycles of conidial passage. In addition, in an attempt to gain insight into the mechanism(s) involved in the conversion of t h e two insertions, we have determined and analyzed the DNAsequence of theappropriateregions of t h e EcoRI fragments involved. MATERIALS AND METHODS Strains and culture conditions: Strains of N. crassa used in this study were Abbott 12, as described by LEMIRE and NARGANC (1986), 240 (nic-I, al-2, A), and IL-40 (pan-2, [mi-?], A). The 240 strain possesses a wild-type cytoplasm and contains the type I1 polymorphic form of mtDNA (MANNELLAand LAMBOWITZ 1979). The [mi-?] mitochondrial mutation in IL-40 causes a deficiency of cytochrome aa3 (MITCHELL, MITCHELL and TISSIERES 1953; LEMIRE and NARCANG1986). IL-40 and Abbott 12 contain the type I polymorphic form of mtDNA (MANNELLAand LAMBOWITZ 1979). Culture conditions for N. crassa were as described earlier (BERTRANDandPITTENCER1969; DAVISand DE SERRES1970). The differences between the two polymorphic forms of mtDNA can be summarized simply in terms of two restriction fragment length polymorphisms (RFLPs). Type I mtDNA carriesa short form of the EcoRI-5 restriction fragment and a long version of EcoRI-9, relative to the type I1 form of mtDNA. Escherichia coli strains used were derivatives of HB10 1 (MANIATIS, FRITSCHand SAMBROOK 1982) that had been transformed (NORGARD,KEEM and MONAHAN1978) with one of the recombinant plasmids described below. In order to obtain plasmids free of methylation at A or C residues (for use with certain restriction enzymes sensitive to methylation), plasmids were transformed into strain Kur 1226 (dam-,dcm-) obtained from R. KELLN (University of Regina). Strain JM103 (MESSING1983) was utilized for the propagation of M 13 phage for DNA sequence analysis. Single-stranded phage DNA was prepared as described by MESSING(1 983). Heterokaryon construction and analysis:Twenty separate heterokaryons were constructed by superimposing conidia of the pantothenaterequiring IL-40 strain with conidia from the nicotinic acid requiring 240 strain on 50 ml of agar hardened minimal Vogel’s medium in a 250-ml Erlenmeyer flask. The conidia that formed in this flask were considered to representthe first “cycle”of the heterokaryon. An inoculum from these conidia was then placed in a flask of fresh minimal medium and the flask containing the first cycle of the heterokaryon was stored at 4 The conidia that formed in the second flask were considered to be the second O . Nargang cycle of the heterokaryon, and conidia from this flask were used to inoculate another flask for the third cycle. This procedure was continued until 10 cycles of each heterokaryon were obtained. The conidia from each cycle of each heterokaryon were available for the analysisof mtDNA contentand mitochondrial cytochromes. We arbitrarily chose to examine the eighth cycle of each heterokaryon as a starting point. Depending on the results observed, either earlier or later cycles were thenexamined in order to determine at which cycle conversion of the mtDNAs was complete and when the heterokaryon became phenotypically [mi-?]. Isolation of mtDNA: Mitochondria to be used for isolation of mtDNA were purified on either step gradients(NARGANG and BERTRAND 1978) or flotation gradients as described by LAMBOWITZ (1979), except that Tris-HCI was used instead of Tricine-KOH. MtDNA was isolated from flotation gradient purifiedmitochondria as described by NARGANG(1986). LEMIRE and Construction of recombinant plasmidsand isolation of plasmid DNA: To facilitate the isolation of DNA for sequence analysis, each version of the EcoRI-5 and EcoRI-9 fragments was cloned into a bacterial vector. The EcoRI-5 and EcoRI-9 fragmentsfromAbbott12(type I mtDNA) were cloned into pBR325. The EcoRI-9 fragmentfrom strain 240 (type I1 mtDNA) was cloned into pUC19. A clone of the EcoRI-5 fragment from N. crassa strain 74A (type I1 mtDNA) inplasmid pBR322 was kindly provided by H . BERTRAND (University of Guelph). Bacterial plasmid DNA was isolated from CsCI-ethidium bromide gradients as deand SAMBROOK 1982). scribed (MANIATIS, FRITSCH Agarose gel electrophoresis: Agarose gels were made to 0.8% agarose in 100 mM Tris-borate (pH 8.3),2 mM EDTA and contained ethidium bromide at a concentration of 0.5 cLgIm1. DNAsequencing and analysis: DNA sequencing and analysis were as described previously (PANDE, LEMIREand NARGANG1989). Analysis of mitochondrialcytochromes: Cytochrome spectra were obtained by the method of BERTRAND and PITTENCER(1969) using a Shimadzu UV-265recording spectrophotometer. RESULTS Conversion of EcoRI-5 and -9 and dominance of [mi-3]in heterokaryons: In our previous examination into the primary defect in the[mi-3] mutant (LEMIRE a n d NARCANG 1986), we observed that the mutation did not predominate over wild-type mitochondria in heterokaryons to the same extent described for the [ p o k y ] mitochondrial mutant (MANNELLAa n d LAMBOWITZ 1978, 1979). In the present study, we have attempted to confirm this observation, and to examine the rate of resolution of[mi-31 plus wild-type heteroTo this karyons to one of the component phenotypes. end, a set of twenty heterokaryons was constructed MATERIALS between the two strains as described in AND METHODS. Since the component strains carried mtDNAs with different polymorphic markers, examination of the heterokaryon mtDNAs was also used to deevaluate the gene conversion events previously scribed for the insertions in the EcoRI-5 a n d -9 fragbe notedthatmtDNAscarrying ments.Itshould Heterokaryons of Neurospora trassa 65 n wild-type FIGURE2.-Cytochrome spectra of the wild-type strain and the cytochrome aaa deficient [mi-)]strain that are the components of the heterokaryons examinedin this study.Also shown is a spectrum that would be judgedto be a mixtureof thetwo components in the characterization o f heterokaryon types. T h e positions o f the absorption peaks for cytochrome aas (608 nm), cytochrome b (560 nm), and cytochromec (550 nm) are shown. FIGUREI .-RcoRI restrict ion digest patterns for type I and type I 1 nitDNA of Neurospora. The standardnumberingpattern of RcoRl fragments 1 through 10 is derivedfromthetype I 1 form (MANNELLA and LAMROWITZ 1979). In the type I form the pattern o f fragments is identical except for RFLPs giving a smaller EcoRI5 fragment, 5(1). and a larger BcoRI-9 fragment. 9(I). differing forms of these polymorphic restriction fragments were originally described as type I or type I1 mtDNA (MANNELLAand LAMBOWITZ 1978,1979). However, the differences in these mtDNAs can essentially be summarized in terms of restriction fragment length polymorphisms (RFLPs) in the EcoRI-5 and -9 restriction fragments. Thus, onemtDNA (type I)contains a long formof the EcoRI-9 fragment and a short form of the EcoRI-5 fragment. The other mtDNA (type 11) contains the short form of EcoRI-9 and the long form of EcoRI-5. As a starting point in our study the eighth cycle of each heterokaryon was examined for its content of mtDNA and mitochondrial cytochromes. Figure 1 shows an example of the restriction digest patterns expected for the type I and I1 forms of mtDNA that exist in the original component strains of the heterokaryon. Figure 2 shows examples of the mitochondrial cytochrome content observed in thecomponent strains and the heterokaryons. If the heterokaryons exhibited a patternof mitochondrial cytochromesthat was intermediate between the pattern expected for [mi-31 or wild type (“mixture” in Figure 2), or if the mtDNA contained a mixture of the two RFLPs for either the EcoRI-9 or EcoRI-5 fragments, then later cycles of the heterokaryonwere examined. If any trait was resolved by the eighth cycle, then earlier cycles were examined in an attempt to discover at what point resolution had occurred. Our judgmentof resolution of the traits is based on examination of mitochondrial cytochrome spectra for the [mi-3] us. wild-type phenotype, and on the presence of particular mtDNA restrictionfragments on ethidiumbromide stained agarose gels. All the heterokaryons examined had acquired the long form of EcoRI-5 by the eighth cycle. Five were chosen at random for examinationat thesecond cycle of passage and all were found to have fully converted to the long form by this time. One first cycle culture was examined and was also found to be completely resolved to the long form (data notshown). The results observed for each of the heterokaryon cycles examined, with respect to both mitochondrial cytochrome spectrum and type of EcoRI-9 fragment, are summarized in Table 1. By the eighth cycle, 15 heterokaryons possessed mixed populations of the EcoRI-9 fragment, five possessed the long version, A. Hawse, R. A. Collins and F. E. Nargang 66 TABLE 1 EcoRI-9 fragments and cytochrome spectrain heterokaryons 1 through 20 Heterokaryon cycle 2 1 Heterokaryon 1 2 RI-9 " cyt " _ Rl-9 nlix 3 cyt RI-9 wt 5 4 cyt RI-9 cyt RI-9 - - - - - - " " mix mix wt mix wt - - - 5 " - - " - " " " (j - - - - - - 7 - - - " " " " " 8 _ _ _ mix-."- g - - - - - 10 - - - - - - - - - " 11 - - - - - 1 2 - - - - - - - - " - " 13 - - mix wt - - L mi3 - - - - - 14 - Wt 15 - - - - - - - - - - 16 - wt Wt - - - - 17 - - mix - - - L mi3 L - - mi3 - - - - L - Rl-9 cyt wt mi3 - - - - - - - - - L - mi3 - " - - - - L - mix mi3 wt - - wt - - - - - - - - - mix L - mi3 mix - 9 8 cyt - m i x " - - - - - RI-9 mix 3 18 19 20 cyt wt""" 4 7 6 RI-9 mix L mix mix mix mix L - mix mix - wt mix mix L - mix - mix - mix - mix wt L - mix L cyt wt wt mi3 mix RI-9 mix mix L mix wt mix mix mix wt mix mi3 L wt mix wt mix mix wt mix m i 3 wt S wt mix wt S wt mix wt mix L m i 3 - 10 cyt RI-9 cyt wt S S - wt wt mi3 wt mi3 mi3 s wt S wt mi3 wt mi3 wt s S S mi3 wt wt wt wt S wt wt S - - - wt wt wt S S wt wt S mi3 - L - S wt wt wt wt mi3 - T h e "R1-9"columns describe the state of the EcoR1-9 restriction fragments found in the mtDNA of the heterokaryon in a given cycle: "mix" denotes that a mixture of the short and long forms was found, "L"denotes that theEcoRI-9 fragment was entirely the long form, and 5 , " that the fragment was entirely the short form. T h e "cyt" columns describe the characteristics of the mitochondrial cytochrome spectra: "mix" denotesthatacytochromespectrumintermediate between that of wild-type and [mi-3] (see Figure 2) was obtainedfromthe heterokaryon at that cycle, "wt" denotes a wild-type spectrum, and "mi3" denotes a [mi-3] type cytochrome spectrum. Dashes indicate that heterokaryons were not examined at that cycle. and none possessed solely the short form. However, all those heterokaryons listed as having a mix of EcoRI9 fragments at the eighth cycle contained a higher percentage of the short form with the exception of heterokaryon 19. The latter resolved tothe long version and the others to the short form within two additional cycles. Thus, by the tenth cycle 14 heterokaryons contained the short version of EcoR1-9 and six the long form of the fragment. The fact that we observed a mixture of the short and long formsof EcoRI-9at some cycle of passage in all theheterokaryons (except numbers 8 and 20), shows that the input ratios from the two component strains is not grossly skewed in one direction or the other. Furthermore, the fact that all heterokaryon mtDNAs are rapidly converted to contain the long form ofEcoRI-5 demonstrates that efficient mixing of mtDNA molecules must occur, since the conversion process undoubtedly requires physical interaction of mtDNA molecules. Thus, the mixtures of EcoRI-9 molecules do not arise from distinct populations of mitochondria from the input strains that might persist in a sequestered state within the heterokaryon. At the eighth cycle, three heterokaryons possessed a "mixed" cytochrome spectrum while 13 were wildtype andfour were [mi-31. By theninth cycle, the mixed heterokaryons had resolved to [mi-31 so that 13 were wild-type and seven were [mi-31. These data suggest that for those cultures that will become [mi31, thenumber of subculturing cycles required is similar to that required by [ p o k y ] , where up to eight cycles may be needed. However, we cannot rule out the possibility that some cultures, which appeared spectrally wild-type after the tenth cycle might eventually become [mi23]. That is, a cytochrome spectrum with only a small reduction in cytochrome uu3 might be misclassified as a wild-type spectrum. Any [mi-3] mtDNA in such cultures could conceivably result in theculture eventually becoming [mi-31. In this respect, it is difficult to say that any heterokaryon has become established as wild type. However, from the results of both this investigation andthe previous study (LEMIREand NARGANC1986), it appears that some heterokaryotic cultures do become permanently wild type rather than [ m i - 3 ] .This is in contrast to the behavior of heterokaryons constructedbetween [ p o k y ] and wild-type strains, where after maximum a of eight cycles of conidial passage, most if not all heterokaryons exhibit themutantphenotype. Our data do suggest that once [mi-31 begins to become established in aheterokaryoticculture, the culture will almost always become [mz-3] after further cycles. That is, Heterokaryons of Neurospora crassa with the single exception of heterokaryon 18, if a heterokaryon was observed to have a mixed cytochrome spectrum, it eventually became [mi-)].Similarly, if a culture was classified as [mi-31 at any point, it never changed to wild type. In two cases (heterokaryons 4 and 13) a wild-type cytochrome spectrum was observed in the early cycles but later changed to [mi-31. We assume that in these cases either the input ratio of [mi-3] to wild-type mitochondria was relatively low and/orthatthemutant mitochondriadidnot become sufficiently established in the cultures until the later cycles. With respect to the three characters examined in this study, only eight of the 20 heterokaryons could be classified as recombinant types (numbers 3,4,6, 8, 13, 18, 19 and 20) after ten cycles of subculturing. This frequency of recombination is considerably lower thanthatobserved previously in heterokaryons involving the [ p o k y ] mutant (MANNELLA and LAMBOWITZ 1979). DNA sequence analysis of EcoRI-5: Previous DNA sequence analysis of the longform of EcoRI-5 (BURGER and WERNER 1985) revealed the presence of alongopenreadingframe with homology tothe mammalian NDI gene (formerly URF1). T h e latter has been shown to encode a component of the NADH dehydrogenase complex (CHOMYN et al. 1985). The NDI gene of N.crassa was found to be interruptedby a single intron of 11 18base pairs, which contained a separateopenreadingframe(ORF)(BURGERand WERNER 1985). We reasoned that the previously observed ca. 1200-base-pair size difference between the short and long forms of the EcoRI-5 fragment (MANNELLA and LAMBOWITZ 1979) might be accounted for by the absence or presence of this intron, respectively. T o test this possibility, we determined the DNA sequence of the short form of the EcoRI-5 fragment from a ClaI site, 40 base pairs upstream of the predicted splice site, through theregion where the intron exists in the long version of the fragment.Comparison of the sequences revealed that a region corresponding exactly to the predicted intron is absent in the short version of the NDI gene (Figure 3). DNA sequence analysis of EcoRI-9: T o identify the region of the EcoRI-9 fragment that contains the previously described ca. 50 base pair insertion (BERNARD et al. 1975; MANNELLA and LAMBOWITZ 1979), the DNA sequence of both the short andlong EcoRI9 fragments was determined. The sequence of both strands of the long form of the fragment was determined entirely and is shown in Figure 4. T h e sequence of the short form was determined completely on one strand and on both strands in regions where the two forms of the fragment were found to differ,or in any region where compression caused difficulty in interpreting the sequence. The only difference found in 67 A) ND1 Gene (long version) I . . .TTAGCTGAGGCTACTAATATG... _ . _L A E A ' H ND1 Exon2 ORF ND1 lntronic Exon1 Intron . . . TATTTTTTGGAATCAGMTTA ... tE t I E 5oobP ... I 6 ) ND1 Gene (short version) I ND1 Gene 1 . . .TTAGCKA@XTGAATCAGAATTA. . . . . . L A E A E S E L ... FIGURE3.-The ND1 (formerly URFl) gene in Neurospora.A, The long version of the gene containing the intron (BURGERand WERNER1985). The intron contains an ORF with the capacity to encode a 304 amino acid protein. B, The short form of the NDI gene which contains no intron. The intron is absent in the short NDI gene precisely at the sitespreviously predicted (shown by arrows) to be the sites for splicing the intron from the mRNA of the long form of the NDl gene (BURGERand WERNER1985). The nucleotide sequence at theintron/exon boundaries is shown. Amino acids in the ND1 coding sequence at the intron/exon boundaries are given in the one letter code. the two polymorphic forms of the fragmentwas in the number of copies of a 78-base pair repeat. A single sequencing gel which spanned this region in both forms of the fragment, revealed that the larger form of the EcoRI-9 fragment contained three copies of this repeat while the short form contained only two. The three copies of the repeat are present at nucleotides 112 to 189, 190 to 267, and 268 to 345 of Figure 4. The EcoRI-9 fragments were found to be about 38% G+C, in good agreement with the 40% G+C content et al. 1975; reported for N. crassa mtDNA (BERNARD TERPSTRA, HOLTROPand KROON 1977).However, the repeats were found to be only 29% G+C. The EcoRI-9 fragments were analyzed for the presence of reading framesin both strands. The sequence of the adjoining regions of the EcoRI-8 fragment (R. A. COLLINS,unpublished results) was included in the this fragment search so thatORFsextendinginto might bedetected. Only one ORF over 75amino acids long beginning with an ATG codon was found in the long form of the EcoRI-9 sequence. This ORF occurs in the third frameof the top strand andwould be 115 amino acids long starting from the first ATG 4). TheORF codon (nucleotides102-104,Figure extends for an additional 23 aminoacids upstream of the ATG. This ORF spans the direct repeat region so that the corresponding polypeptide in the short form of EcoRI-9 would be only 86 amino acids long. Remarkably, this ORF was found to have a stretch of 47 amino acids identical to the probable ND2 (formerly URF2) protein of Neurospora encoded on the HindIII-9 fragmentof Neurospora mtDNA (DE VRIES et al. 1986). The relationship of the regions of mtDNA involved, the location of the reading frames, and the 0 0 0 0 A. Hawse. R. A. Collins and F. E. Nargang 68 IO 20 - 30 40 50 6.0 70 00 CTGATCAAGG ATTffiCTPRT TTCTTRTATA AACTCPPACR ACCACCTCCA AGGAATAAAC GAATTCTAGTGGATTTTTAT CTTAAGATCA C C T A W A T R GACTAGTTCC TAATCGATTAAAGAATATATTTGAGTTTGT 110 AACTACAATT ATACCCCCCT CATGCRGCAA 90 100 1 30 TGGTGGAGGT TCCTTATTTG 1230 I 50 160 AATRATAG GAACTGTTGT RGGTTTAACA CAGTTTAGAA TTRAAAGATT TTGATGTTAA TATGGGGGGA GTACGTCGTT g T l A T T A T C CTTGACAACATCCAAATTGTGTCAAATCTT RATTTTCTAA 230 270 120 140 220 170 210 100 200 I90 240 GCTTGCATAT AGTACAATCT CTCATCTA AATAATAGGA ACTGTTGTAG GTTTARCRCA GTTTRGAATT MAAGATTGC CGAKGTRTA TCATGTT&W G A G T A G R T ~TTATTATCCT TGACARCATC CAAATTGTGTCAAATCTTAATTTTCTRACG 260 300 250 310 320 TTGCATATAG TACAATCTCT CATCTA AA TAATAGGAAC TGTTGTRGGT TTRACRCRGT TTAGAATTAR AAGATTGCTT ARCGTATATC RTGTTAGAGR GTAGAT$TT ATTATCCTTG ACAACATCCR ARTTGTGTCA ARTCTTAATTTTCTAACGAA 330 340 350 370 360 300 GCATATAGTA CAATCTCTCPl TCTR TTTTATTTTRTTAGCTTTAAGTGT CGTRTATCAT GTTRGAGPGT RGA$AAAR TAARRTAATCGARATTCACA 390 400 TTCCRCACRR RGTGTGGAGT CCACRCAAGC ARGGTGTGTT TCRCACCTCA GGTGTGTTCG 410 420 430 440 470 460 480 A T T I A T A T T T TGGGAGGAGT GARATTACAG GGAGATTAAA AlAGTT* CTGTAAGAAC ATTAAGATGG CGRTTCCACT TAAATATAAA RCCCTCCTCA CTTTAATGTCCCTCTAATTTTATCAAATTAGACATTCTTGT&ATTCTACC GCTAAGGTGA 490 500520 530 510 540560 5501640 I630 CGGCCCCCCC TCCAGCGAAG CTGGPGGGGG TRGCGAGTGA ARGCTTTTAA TAAPGTTGGR TTGCCATGRC CGCTTCGAGC GCCGGGGGGG AGGTCGCTTC GACCTCCCCC ATCGCTCACT TTCGRRAATTATTTCAACCT ARCGGTACTG GCGAATCTCG 590 580 1730 1720 630 570 1710 640 GTTGGTCCCT ATCTTCCTCAGTAGTTTRCT GCTGAGGAAG GTAGAGCRGT CAGCAAAGAG GACGGACCGT AAAGCCATAC CAACCAGGGA TAGPAGGAGT CATCAAATGA CGACTCCTTC CATCTCGTCAGTCGTTTCTC CTGCCTGGCA TTTCGGTATG 1000 1790 bJO670 6bO 720 710 LBO700 690 RTCGRGGGAA TAGTAPACCC CAGCTACCCT ACAGATTTTA ATCTGTAGGG TAGCCTTATG GAATAGAGGR AAAAATTTTC TAGCTCCCTT ATCATTTGGG GTCGATGGGA TGTCTAAAAT TAGACRTCCC ATCGGMTPlCCTTATCTCCTTTTTTRAAAG 7so 740 730 800 CTARCGTAAA TACCTTAATG CCTCCCAGGA GATAAGTATCATTGATGTTA GATTGCATTTATGGAATTAC GGAGGGTCCT CTATTCPTAGTARCTACART TCCTAGTACC TATTTTCACC TPlTAATCACT RGGATCATGG ATAAAAGTGG ATATTAGTGA 010 020 030 840 050 070 860 000 TGTCGCTGAT AAGCTRGCGC AGGTGCGCTG GTGCCTTGRC CCCGGATGGT CGTATTTTTCCTACTRTTGA CGAGGGTATT ACAGCGRCTA TTCGATCGCG TCCRCGCGAC CACGGAACTG GGGCCTRCCA GCATAAAAAG GATGATAACT GCTCCCRTAA 910 900 970 1210 I270 I240 1260 1250 1200 TTCTAGTCTT M G P C T T C T GTPCTTCGCA ACCCCCTCCT TTGCAATTAA GCRRRGGffiG GGGGTAGARG GAGTTTTRTG CGTTTCCTCC CCCCATCTTC CTCAAPATAC W T C A W A T T T C T G M P CATOMGCGTTGGGffiR00AAPCGTTARTT 1290 1300 1310 1320 1330 1350 1340 13bO ffiTATATTAC TTGTTAAACC TCCTCPACAC TAATTAGCGR CCCCTATTCC TCGGGTCGCT GTAAAATTRG TGRRCAGAAG TCATRTAATG AACPATTTGG AGGPGTTGTG ATTAATCGCT GGGGATAAGG AGCCCRGCGA CATTTTRATCACTTGTCTTC 1390 I400 1300 1410 I420 1430 1440 1370 CCTTTGPWG CTCCTTTCAA TATTAGAACR GTMTGTGGT TATCACCRAT A!ATATGTAAG TCIATACCCG GAGACGGCTC G W A C T T C C GAGGPJIAGTT ATRATCTTGT CATTACRCCA ATAGTGGTTA TTATRCATTC AGTTATGGGC CTCTGCCGAG 1450 1460 1470 I400 1490 IS00 1520 I510 TGGPdGIITAG RTTTGAGGTC TGTGGATTTTTATAAAAGAACGTTTGATRA AATAGACAAA GGAGGATTAA TTATGGTTAA PCCTTCTPTCTAAACTCCAG ACRCCTAAAA ATATTTTCTTGCAAACTATTTTATCTGTTTCCTCCTAATT AATACCRATT 1530 1540 1560 1550 1570 1500 1590 1600 AGTTATTATR CCGGACAATA RCTAATTTTA ATTTCATRTTTTTCCTGTAA GTGTCGGTAA CACTTACTRG AGCCGCAAGG TCAATARTRT GGCCTGTTAT TGATTAAAATTAAAGTATRA AAAGGACATT CACAGCCATT GTGAPTGATC TCGGCGTTCC I620 1610 I650 1660 I670 1600 CGTCTATTTT ATTAARCCCC TTGTCTTTTT TAGGATCTTC AGATTTTGAGTTGAAGTTAA TGAGAACTCG AAGGGGTTGA GCPGATAAAA TPATTTGGGG RRCRGRAAAA ATCCTAGAAG TCTAAAACTC WCTTCAATT RCTCTTGPGC TTCCCCAACT 1700 1690 1740 1750 1760 CCTPTGTATT ATCGGAGTCA TGGTGCCACA GAACGATAGG GTCCCCACAA CGAAGTTGTG GGGTCCCCCG ATTTPlGTTCG GGATACATAA TRGCCTCAGT ACCACGGTGT CTTGCTATCC CAGGGGTGTT GCTTCAACAC CCCAGGGGGC TAAATCAAGC 1700 1770 1010 1030 1820 I040 AGGWTTAGA AGGGAACCAA AGGRTCPTPT RAGTTGCAAT T T L T A C A T T T GATGTTGAAG &TCGRTGTAG PiClATTGTATA TCCTTPATCT TCCCTTGGTT TCCTAGTATRTTCAPCGTTAAATATGTAAA CTACRACTTC TAGCTACATC TTTAACATAT 1900 1850 lab0 1070 1910 I920 1000 1090 T G W T T G A G ACTTCGAGTA TTAGGTCCGG ATRATRCTCT ACTTTCCACT CGRCCCCCCC TCACCAATCT TACTTRAACT RCTTTAACTC TGARGCTCAT AATCCAGGCC TATTRTGAGA TGRRAGGTGA GCTGGGGGGG AGTGGTTAGR ATGAAllTGA 1930 1940 1950 1960 1970 1900 1990 2000 CAAATGTTGA GGATCAGCAT TAGCRACTCR TTTAAAATCC TTGTGATCCT TCTTGAAATC TTTCAAATAT TTRAAATCAT GTTTACAACT CCTAGTCGTR ATCGTTGAGT AAATTTTAGG AACACTAGGA AGRACTTTRG RAAGTTTATP AATTTTAGTR 2000 2070 2010 2060 2050 2020 2040 2030 090 CAATATCTAG GAGMRTTTC GRTTTCTCTRGGTAATATGA GTTATAGATCCTCTTTAAAGCTPAAGAGATCCATTATACT 900 1220 990 DAGAGGTGTC AAAGATTCTCCTGATGTTTA TTCTCCPCAG TTTCTRRGAGGPCTACRAAT 1000 ACCAGGCTTT GGCCCACTCC ATTGGGTATT RTCGGTTGAA TGGTCCGAAA CCGGGTGAGG TRACCCRTAR TAGCCAACTT IOIO 1020 1030 IO40 TCTCACCAGC CCCCTTCCCC CAACACCTTT CCCTGATATC CRCTAATAAG AGAGTGGTCG GGGGAAGGGG GTTGTGGAAP GGGACTATAG GTGATTATTC TRCTTTTATC AGATMAAAA ATTTCTTGATTTTGATCACTATTTCCATTTCCATTCGTTG AAAATARTCG TGCTGCLATT ATGAAAATAGTCTACTTTTTTAAAGAACTR AAACTRGTGR TAARGGTAAA GGTAAGCAAC TTTTATTAGCACGACGTTW 2100 2090 2140 2110 2130 2120 TTAATGTGRT ATCTGGGCAA ATAAAAATTA ATCRTAATATARATATTTAC W4lTPCAClA TAGACCCGTT T A T T T T T A A T TRGTATTATRTTTATAAATG 2150 2160 ACCCAATATC TCGRTTTGGC TTTTGTGTCG TGGGTTATAG AGCTRAACCG AAAACPCAGC 2170 2100 2200 2190 1050 IObO 1070 IO80 1090 1100 Ill0 1120 AGTTTGGCGG GCCCTCATTC TATGCGTTCC TTGGGATACG GARTTC TGGAAGTCCT TGGTTATCAACTCAATTTTG TTCGTAGCCT CTATTTTAAG ATRAGCCACC CCGGAGGGGT G G A ~ C T T A T T TCAAACCGCC CGGGAGTARG ATACGCAAGG AACCCTATGC CTTAAG KCTTCAGGA ACCAATAGTTGAGTTAAAAC AAGCATCGGA GATAAAATTC TATTCGGTGG GGCCTCCCCA CCTTGAATAP 1130 1140 1190 1200 1150 IIbO 1170 1100 GAGATTITAG GAAAGGTRAA RTTAGCAGTT AGCTGGATTC TCATGGRAGA ARTCCGATCT TAGTTCTTTG TTGCAAGGAT CTCTAAAATC CTTTCCATTT TAATCGTCAA TCGRCCTAAG AGTACCTTCT TTAGGCTAGR RTCAAGRRAC AACGTTCCTR FIGURE4.-DNA sequence of the long BcoR1-9 fragment. Numbering begins at the EcoRI site adjacent to the EcoRI-8 fragment (see Figure 7 ) . T h e positions of the 7 8 base pair repeats are indicated by square brackets. T h e sequence of the shorter form of the fragment is identical except that it contains only 2 copies of the 7 8 base pair repeat. T h e start (ATG) and stop (TAA) codons of the longest O R F in the sequence are indicated by solid circles above the bases. and NARGANC 1986;this study). The EcoRI-5 fragment of Neurospora mtDNA contains the NDI gene (formerly URFl, BURGER and WERNER 1985). Here we show that the short form lacks the intron present in the NDl gene of the large form, precisely at the splice points predicted previously (BURGER and WERNER 1985). T h e intron of the ND1 gene contains a long ORF of 304 amino acids (BURGER and WERNER 1985;Figure3) andthe rapid conversion tothe longer, intron-containing form is thus reminiscent of the geneconversion events observedin other systems, whereintronencodedproteinspromoteunidirectional conversion of alleles lacking introns into introncontaining alleles. Examples include the omega UACQUIER and DUJON1985; MACREADIEet al. 1985) and al4alpha (WENZLAU et al. 1989;DELAHODDEet al. 1989) systemsinyeast mtDNA, the nuclear rDNA group I intron of Physarum (MUSCARELLAand VOGT 1989), and the td and sunY introns of bacteriophage DISCUSSION T 4 (QUIRK, BELL-PEDERSEN and BELFORT 1989). T h e intronic ORFs in these systems are known to encode In all studies on the behavior of Neurospora mtproteins with double stranded endonuclease activity DNAs in heterokaryons, the complete conversion of and it seems likely that the Neurospora NDl intronic short EcoRI-5 fragments to the longer formhas been observed (MANNELLA and LAMBOWITZ 1979; LEMIRE ORF encodes a similar function. repeat units in EcoRI-9, is shown in Figure 5. At the nucleotide level, a regionof 144 base pairs of the ND2 gene is homologous to a similar region of EcoRI-9 with identity at 141 of those positions (Figure 6). The region of identity begins precisely at the start of the last 78 base pair repeat unit and includes the entire repeat unit plus an additional 66 base pairs. Thus, of the 47 amino acids homologous to the ND2 gene, 25 are encoded within the repeat unit (Fig. 6). We could find no similarity in the sequences, at either the nucleotide or the amino acid levels, outside this short region of identity. Shorter ORFs found in the EcoRI-9 fragment were compared to proteins in release number 20 of the National Biomedical Research Foundation(NBRF) database. The analysis revealed no striking similarities. Heterokaryons of Neurospora crassa 69 78 bp repats y 5y 1000 * 1500 I 2000 f I I I EC0Rl-g (long version) 2000 500 P" FIGURE 5,"Homologybetween the BcoRI-9 fragment and the ND2 gene (formerly URFP, DE VRIES et al. 1986) of N . crassa. The lower right hand corner contains a representation of the circular Neurospora mtDNA with the location of EcoRI fragments indicated. The filled portions of the circle indicate the location of the EcoRI-9 fragment and restriction fragment HindIII-9 which is found at thejunction of the I:'coKI-l and -2 fragments. The N D 2 gene is found within the HindllI-9restriction fragment. Solid lines represent each restriction fragment. The position of the N D 2 gene is shown beneath the HindIII-9 fragment as a cross-hatched box. The numbering of the N D 2 gene sequence is consistent with DE VRIESet al. (1986). The position of the longest ORF in EcoRI-9 is shown below the EcoRI-9 fragment as a cross-hatched box. The region of homology between the two is indicated by the vertical dashed lines. The position of the three 78-base-pair repeats in the type I form of EcoR1-9 is also indicated. The region of homology begins precisely at the start of the third repeat. Repeat 1 Ecm-9 om: Q I T V V G L TGTTGTAGGTTTA TAATMTAGGAACTGTTGTAGGTTTA AGCTGAATCTACGGACTACTAGTCAGTTCTTTTCTTTCCT I I G T V V G L S W I Y G L L V S S F L S L F R I X R L L A Y S T I S H F T G L L A T T T A M T ~ T A A A A G A T T G C L L A Y S T I S B L G F I L mTTTmTTAAUGATTGCTTGCATATAGTATAGTACAATCTCTCATC A I ... mz T Repeat 3 Repeat 2 ~ T T ~ T Q F R I K A R I L S A T T T C L A J V S A G C T C C T ~ T C L S V F T ~ W E E W N Y R E I X I V e n d Q A F I F GGGAGGAGTGAAATTACAGGGAGKTTAAAATAGTTTAA A C A C A M C T G T G G A G T C C T T T ACACAMCTGTAGAGTCTACACAAGCATTTTATATTTT T Q S V E S T Q A F I F ATTTAATACAATA~TATTCTTTTACTAATTTAAATCTCTfi. Y L I Q Y S F S N L N V F .. FIGURE 6.-Homology between the EcoRI-9 fragment (top)and the N D 2 gene (bottom). The long form of EcoRI-9 containing three repeat units is shown i n the figure. The two DNA sequences are aligned closely where they are homologous and farther apart where the sequences differ. Amino acids are shown i n the one letter code. For EcoRI-9, the entire region that encodes the longest ORF in the sequence is shown, except that the nucleotide and amino acid sequences within the 78 base pair repeats are shown only in the third repeat and at the repeat junctions. The sequences within the other repeat units are identical to the third repeat. The location of the repeat units is indicated. Arrows indicate mismatched bases in the region of homology. Only the sequence homologous to the EcoRI-9 fragment, plus a few flanking nucleotides, is shown fox-the N D 2 gene. The N D 2 gene continues for 31 1 codons upstream, and 201 codons downstream of the region shown (DE VRIES et al. 1986). We have shown that thedifference between the two forms of the EcoRI-9 fragment is the numberof copies of a 78-base-pair repeat. One copy of this repeat, plus an additional 66 base pairs of downstream sequence, has been derived from the N D 2 gene, which is located in a distant region of the mitochondrial genome (Figure 7). It appears that either during or following the transfer of sequence from the N D 2 gene, a region of that sequence was duplicated to give rise to the 78base-pair repeats. That originalevent probably resulted in the generation of either two or threecopies. In either case a separate eventmust have occurred to give rise to thetwo separate polymorphic forms of the EcoRI-9 fragment. Sequencesin or near theN D 2 gene ~ A A. Hawse, R. A. Collins and F. E. Nargang 70 II ORF I t m w t \ [mi-31 mutation FIGI‘RE 7.-(;ircuhr representation of the Neurospora mitochondrial genome. The outer circle shows the position of genes discussed in the text.Solid black portions represent exons andwhite portions represent introns. The inner circle shows the EcoRl fragnlents of the mtDNA. The regions designated “A” and “B”,which are indicated above the N D 2 gene regionof the map. represent the sequences which have been duplicated a t other positions in the mtDNA where they are shown a s open boxes on the outer circle. The sm;lllblack rectangle outsidetheouter circle indicates the position of the directly repeated units in the EcoRI-9 fragment. The psition ofthe intronic O K F which encompasses virtually the entire intron of the AID1 gene(BURGERandWERNER1985), andthe position of the [mi-31 mutation in the C O X 1 gene (LEMIREand 1986) are also shown. NARGANG may be sites of high frequency recombination since deletions in various stopper mutants of Neurospora have been shown to begin near the location of this et al. 1980; GROSS,HSIEHand LEVgene (BERTRAND INE 1984; DE VRIES et al. 1986). Furthermore, the amino-terminalportion of the ND2 geneandthe upstreamtRNAmetgene, normally foundatthe junction of the EcoRI-1 and EcoRI-2 fragments, have also been shown to be duplicated at the junction of the EcoRI-1 and EcoRI-6 fragments (ACSTERIRBE, HARTOG,and DE VRIES, 1989; See Fig. 7), although the sequences found in the latter duplication do not include those foundin the EcoRI-9 duplication. Nonetheless, sequences from the ND2 region of the mtDNA appear to behave in a fashion that is in some senses reminiscent of the GC-rich clusters in yeast mtDNA (DE ZAMAROCZYand BERNARDI1986), which are thought to spread through the yeast mitochondrial genome by processes such as transposition and gene and GROSSMAN1985; conversion (BUTOW,PERLMAN WEILLER, SCHUELLER and SCHWEYEN 1989). GC-rich clusters havealso been shown to be sites of both intraand intermolecularrecombination inyeast mtDNA molecules (DE ZAMAROCZY, FAUGERON-FONTY and BERNARDI1983; DIECKMANNand GANDY 1987; CLARK-WALKER 1989). Our results with respect to the resolution of the short and long EcoRI-9 fragments differ from those and LAMROWITZ ( 1 979). In their studies of MANNELLA of “[poky]plus wild-type” heterokaryons, the majority (25 of 34) resolved to the long form of EcoRI-9 and the remainder to the short form. The combined data of the present study and of LEMIREand NARGANC ( 1 986) show that only 9 of 30 “ [ m i - ) ] plus wild-type” heterokaryons resolved to the long form of EcoRI-9, while 21 of 30 resolved to the short version of the fragment. Thus, in our study,a process involving directional conversion of the short to the long form of EcoRI-9 does not seem to have occurred. Interestingly, despite evidence that efficient mixing of mtDNAs occurred (see RESULTS), we recovered only eight recombinant mtDNAs from twenty heterokaryons, while in the study of MANNELLAand LAMBOWITZ ( 1 979), almost all heterokaryon mtDNAs resolved to recombinants. Therefore, a possible explanation for the differences observed in the two sets of studies is that less overall mtDNA recombination has occurred with our heterokaryon pair. Conceivably, the strains usedin the present study could be less efficient in either the promotion of exchange between mtDNA molecules, or a specific gene conversion process, than the strains used in previous studies. Our data show that [mi-31 mitochondria are less dominant over wild-type mitochondria than are the mitochondria of the [poky] mutant. In fact, only seven of twenty heterokaryons acquired the [mi-3] phenotype under conditions of conidial passage similar to those used in at least one of the previous studies where ten of ten heterokaryons became [poky] (MANNELLA and LAMBOWITZ 1978). There are a number of possible explanations for the different behaviorof [mi-31 and [poky] mitochondria in heterokaryons with wildtype mitochondria. First, it is conceivable that small differences in the nuclear genetic backgroundsof the strains involved in the various studies might influence the process of takeover by mutantmitochondria. However, it is difficult to imagine a mechanism whereby particular nuclear gene products would discern differences in mutant and wild-type mitochondria andfavor the propagation of one over the other. Second, it is possible that conversion to the long form of the NDI gene (in EcoRI-5), from the wild-type mtDNA componentof o u r heterokaryons, sometimes extends 4 kilobases into the region of mtDNA where the [mi-31 mutation is found, andco-converts it to the wild-type allele. However, unless co-conversion in Neurospora extends for much greater distances than in other genomes (DUJON, SLONIMSKIand WEILL 1974; JACQUIERand DUJON 1985; WENZLAUet al. 1989; BELL-PEDERSEN et al. 1989), this is unlikely. Third, given thattheparental, wild-type, type I1 mtDNA was the form most frequently recovered, the infrequent takeoverby the [mi-31 phenotype might be Heterokaryons of Neurospora crassa explained if type I1 mtDNA had a replicative advantage over the type I mtDNA which carries the [mi-31 mutation in our heterokaryons.However,alarge difference in the rates of replication between the two mtDNAs is not likely since our data provide evidence that mixtures of the two input formsof mtDNA persist in the heterokaryons for many cycles of subculturing. A fourth possibility is that the mechanism of propagating the heterokaryons could influence the degree of takeover by different mutant mitochondria. For example, if cells that have come to contain predominantly [mi-31 mitochondria produce conidia less efficiently than cells that are predominantly wild-type, possibly due to defects in energy metabolism, then there would bea selective advantagefor wild-type cells during theprocess of conidial passage that would not occur during propagation by continuous growth. However, this seems an unlikely explanation given that conidial passage of heterokaryonsconstructed with the [poky] mutant, which should suffer similar defects in energy metabolism, all became [poky] (MANNELLA and LAMBOWITZ, 1978). Furthermore, it is doubtful that othermethods of propagation are more efficient at producing takeover by cytochrome uu3deficient mutantslike [mi-31,since in a previous search for cytoplasmic mutants from continuously growing cultures, only "poky-like'' mutants, deficient in both cytochromes uu3 and b, were obtained (BERTRAND and PITTENGER 1969). The latter observation suggests a fifth possible explanation for the different behaviorof the [poky] and [mi-?] mutants in heterokaryons. The most likely defect in mutants that lack both cytochromes uu3 and b is a deficiency of mitochondrial proteinsynthesis (COLLINS and BERTRAND 1978). The mutation in [poky] is known to affect mitochondrial protein synthesis and has been characterized as a four base pair deletion in the codingsequencefor the mitochondrial small rRNA (AKINSand LAMBOWITZ 1984).In [mi-?], a missense mutation affects the codingsequence for subunit 1 of cytochrome c oxidase (LEMIRE and NARGANG 1986). Regulationof mtDNA replication in [mi?] and [poky] could be affected differently if a previously postulated mitochondrial repressor of replication was translated in mitochondria (BARATHand KUNTZEL 1972; AKINS and LAMBOWITZ 1984). The production of such a repressor would be less efficient in [poky] than in [mi-31 and [poky] mtDNA would be expected to replicate morerapidly than [mi-)] or wildtype mtDNA. Thus, in heterokaryons, mitochondria containing a higher ratioof [poky] mtDNA molecules mightbeexpectedto have ahigher rate of DNA replication. This model would predict that all mitochondrial mutants that affect mitochondrial protein synthesis in Neurospora would bemoredominant than those that affect single components involved in 71 oxidative phosphorylation. The notion that mutations not affecting mitochondrial protein synthesis are not as strongly dominantas [ p o k y ] might also explain why only two mitochondrialmutantsthat do not affect both cytochromes uug and 6, namely [mi-)] and [exn51, have been described in Neurospora to date. In yeast it has been shown that the frequency of an allele within the mitochondrial population of genes depends on the input frequency, random drift, and random segregation of the mitochondria into daughter cells (BACKER and BIRKY1985). Similarly, in Neurospora mtDNA, these factors may be all that govern the fixing of mutations that affect genes not involved in mitochondrial protein synthesis. Perhaps the most parsimonious interpretation of our results is that the above random factors also determine the frequency of recovery of the [mi-31 mutation in N . crussu heterokaryons. This work was supported by grants from the Natural Sciences and Engineering Research Council of Canada to F.E.N. and R.A.C. A.H. was supported by a postgraduate scholarship from the Alberta Heritage Foundation for Medical Research. We are grateful to P. HASTINGSand E. LEMIRE for helpful discussion. LITERATURE CITED AGSTERIBBE, E., M. HARTOGand H. DE VRIES,1989 Duplication of the tRNAM" and tRNA"Ys genes and of fragments of a gene encoding a subunit of the NADH dehydrogenase complex in Neurosporacrassa mitochondrial DNA. Curr. Genet. 15: 5762. AKINS, R. A., andA. M. LAMBOWITZ, 1984 The [poky] mutant of Neurospora contains a 4-base-pair deletion at the 5' end of the mitochondrial small rRNA.Proc.Natl.Acad. Sci. USA 81: 3791-3795. BACKER, J. S., and C. W. BIRKY, JR., 1985 The origin of mutant cells: mechanism by which Saccharomyces cerevisiae produces cells homoplasmicfornewmitochondrialmutations.Curr. Gene. 10: 7-13. BARATH,Z., and H. KUNTZEL,1972Cooperationofmitochondrial and nuclear genes specifying the mitochondrial genetic apparatus in Neurospora crassa. Proc. Natl. Acad. Sci. USA 69: 1371-1374. BELL-PEDERSEN, D., S. M. QUIRK,M. AUBREY andM. BELFORT, 1989 A site-specific endonuclease and co-conversion of flanking exons associated with the mobile td intron of phage T4. Gene 82: 119-126. BERNARD,U.,A.PUHLER, F. MAYERandH. KUNTZEL, 1985 Denaturation map of the circular mitochondrial genome of Neurospora crassa. Biochim. Biophys. Acta 402: 270-278. BERTRAND, H., and T . H. PITTENGER,1969 Cytoplasmic mutants selectedfromcontinuouslygrowingculturesof Neurospora crassa. Genetics 61: 643-659, BERTRAND, H., R.A. COLLINS, L. L. STOHL,R. R. GOEWERT and A. M. LAMBOWITZ,1980 Deletionmutants of Neurospora crassa mitochondrial DNA and their relationship to the "stopstart" growth phenotype. Proc. Natl. Acad. Sci. USA 77: 60326036. BURGER,G., and S. WERNER,1985ThemitochondrialURFl gene in Neurospora crassa has an intron that contains a novel type of URF. J. Mol. Biol. 1 8 6 231-242. BUTOW,R.A.,P. S. PerlmanandL. I. GROSSMAN,1985 The unusual v a r l gene of yeast mitochondria DNA. Science 228: 1496-1501. 72 A. Hawse, R. A. Collins and F. E. Nargang CHOMYN, A,,P. MARIOTTINI, M. W. J. CLEETER, C.I. RAGAN,A. MATSUNO-YAGI, Y. HATEFI, R.F. DOOLITTLE and G. ATTARDI, 1985 Six unidentified readingframes of humanmitochondrial DNA encode componentsof the respiratory-chain NADH dehydrogenase. Nature 314 592-597. CLARK-WALKER,D., G. 1989 I n vivo rearrangement of mitochondrial DNA in Saccharomyces cerevisiae. Proc. Natl.Acad. Sci. USA 8 6 8847-8851. COLLINS,R. A,, and H. BERTRAND, 1978 Nuclear suppressors of the [ p o k y ]cytotplasmic mutant of Neurospora crassa. 111. Effects on other cytoplasmic nlutants and on ribosome assembly in [poky]. Mol. Gen. Genet. 161: 267-273. DAVIS,R. H., and F. J. DE SERRES,1979 Genetic and microbiological research techniques for Neurospora crassa. Methods En7ymol. 17A:79-143. DELAHODDE, A,, V. GOGUEL, A. M. BECAM,F. CREUSOT, J. PEREA, J. BANROQUES and C. JACQ, 1989 Site-specific DNA endonuclear and RNA tnaturase activities of two homologous intronencoded proteins from yeast mitochondria. Cell 56: 431-441. C. A. BREITENBERGER, DE VRIES,H., B. ALZNER-DEWEERD, D. D. CHANCJ. C. DE JONGE and U. L. RAJBHANDARY, 1986T h e E35 stopper mutant of Neurospora crassa: precise localization of deletion endpoints i n mitochondrial DNA and evidence that the deletedDNA codes for a subunit of NADH dehydrogenase. EMBO J . 5: 779-785. DE ZAMAROCZY,M., and G. BERNARDI, 1986 T h e GC clusters of the mitochondrial genome of yeast andtheirevolutionary origin. Gene 41: 1-22. D E ZAMAROCZY, M., G. FAUGERON-FONTY andG. BERNARDI, 1983 Excision sequences in themitochondrialgenome of veast. Gene 21: 193-202. DIECKMANN, C., and B. CANDY, 1987 Preferential recombination between GC clusters in yeast mitochondrial DNA. EMBO J. 6 4 197-4203, DIACUMAKOS, E. G., L. GARNJOBST and E. L. TATUM, 1965 A cytoplasmic character in Neurospora crassa, the role of nuclei and mitochondria. J.Cell Biol. 26: 427-443. DUJON,B., P. P. SLONIMSKI and L. WEILL, 1974 Mitochondrial genetics IX. A model for recombination and segregation of mitochondrial genomes in Saccharomyces cerevisiae. Genetics 7 8 4 15-437. GARNJOBST, L., J. F. WILSONand E. L. TATUM, 1965 Studies on a cytoplasmic character in Neurospora crassa. J. Cell Biol. 26: 4 13-425. GROSS,S. R.,T-s. HSIEH and P. H. LEVINE, 1984 Intramolecular recombination its a source of mitochondrial chromosome heteromorphism in Neurospora. Cell 38: 233-239. JACQUIER, A,, andB. DUJON,1985 An intron-encoded protein is active i n a gene conversion process that spreads an intron into a mitochondrial gene. Cell 41: 383-394. LAMROWITL, A. M., 1979 Preparation and analysis of mitochondrial ribosomes. Methods Enzytnol. 59: 421-433. LEMIRE,E. G., and F. E. NARGANC, 1986A missense mutation in the 0x2-3 gene of the [mi-3] extranuclear mutant ofNeurospora crassa. J. Biol. Chem. 261: 5610-5615. MACREADIE, I. G., R. M. SCOTT,A . R. ZINN and R. A . BUTOW, of anintron in yeast mitochondria re1985Transposition quires a protein encoded by that intron. Cell 41: 395-402. MANIATIS,T . , E. F.FRITSCH and J. SAMBROOK, 1982Molecular Cloning. A Laboratory Manual. Cold Spring Harbor Laboratory, Cold Spring Harbor, NY. 1978Interaction of MANNELLA, C. A,,and A. M . LAMBOWITZ, wild-type and [poky] mitochondrial DNA in heterokaryons of Neurospora. Biochem. Biophys. Res. Commun. 80: 673-679. C. A,,and A. M . LAMBOWITZ, 1979 Unidirectional MANNELLA, gene conversion associated with two insertions in Neurospora crassa mitochondrial DNA. Genetics 93: 645-654. MESSING,J.,1983 New M I 3 vectors for cloning. Methods Enzymol. 101: 20-78. MITCHELL,M. B., H. K. MITCHELLand A. TISSIERES, 1953 Mendelian and non-Mendelian factorsaffectingthe cytochrome system Neurospora crassa. Proc. Natl. Acad. Sci. USA 39: 606-613. MUSCARELLA, D.D., and V. M. VOGT, 1989 A mobile group I intron in the nuclear rDNA ofPhysarum polycephalum. Cell 56: 443-454. NARGANG, F. E., andH. BERTRAND,1978Nuclearmutantsof Neurospora crassa temperature-sensitiveforthe synthesis of cytochrome aa: