Copyright 0 1993 by the Genetics Society of America

The Isolation of Mutagen-Sensitive nuv Mutants of Aspergillus nidulans and Their Effects on Mitotic Recombination

Fekret Osman, Brian Tomsett and Peter Strike

Department of Genetics and Microbiology, University of Liverpool, P.O. Box 147, Liverpool L69 3BX, England Manuscript received September 25, 1992

Accepted for publication February 26, 1993

ABSTRACT More than 200 mutants of Aspergillus nidulans were isolated as hypersensitive to the monofunctional

alkylating agent MNNG and/or UV-irradiation (designated nuu mutants). Of these, 23 were selected for further characterization. All were markedly hypersensitive to both MNNG and the quasi-UV- mimetic mutagen 4-NQO. The hypersensitive phenotype of each mutant was shown to result from mutation of a single gene. T h e nuu mutants exhibited a diverse range of growth responses on solid media containing various concentrations of MNNG or 4-NQO. This suggested that they represented many nonallelic mutations. Analysis to determine the dominance/recessiveness of the nuu mutations with respect to hypersensitivity revealed that most were fully recessive, although several appeared to be semidominant. A novel system to assay homologous mitotic recombination using simple plating tests was developed. T h e system was exploited to determine the effects of the nuu mutations on mitotic recombination. Of the 23 mutations tested, 10 caused a hypo-recombination phenotype and three a hyper-recombination phenotype, while 10 appeared to have no effect on recombination. T h e hypo-rec effect of one of the mutations, nuv-117, appeared to be semidominant. Transcomplemen- tation analysis between seven of the nuu mutations defined at least six nonallelic loci.

D ESPITE its fundamental importance for a wide variety of biological processes, little is known of

the underlying molecular mechanisms involved in ho- mologous genetic recombination in eukaryotic cells. Much of our knowledge has relied on studies in lower eukaryotes, particularly a few species of fungi that have genetic systems amenable for detecting, analyz- ing and assaying recombination events (WHITEHOUSE 1982; ORR-WEAVER and SZOSTAK 1985; HASTINCS

Studies on the genetic control of homologous re- combination depend on the isolation and character- ization of mutants defective in the recombination pathway(s). DNA metabolic pathways in cells, includ- ing replication, transcription, recombination, repair and mutagenesis, are not distinct biological processes but are overlapping and, in part, under the same genetic control. The isolation and characterization of mutants hypersensitive to DNA damaging agents in many organisms continues to provide a diverse source of mutations defective in these processes, as well as revealing the complex interrelationships between them. Hypersensitive mutants have been a particularly useful source of mutants defective in homologous recombination in both prokaryotes (e.g. , HOWARD- FLANDERS and THERIOT 1966; CLARK 1973) and eu- karyotes (e.g. , HOLLIDAY 1967; BAKER and SMITH 1979; PRAKASH et al. 1980). By far the best studied lower eukaryote with respect to the mutational analy- sis of homologous recombination is the budding yeast

1988).

Genetics 134: 445-454 (June. 1993)

Saccharomyces cerevisiae. A relatively large number of S. cerevisiae mutants affecting DNA repair and/or recombination have been isolated and characterized and many of the corresponding genes have been cloned (HAYNES and KUNZ I98 1 ; GAME 1983; COOPER and KELLY 1987; FRIEDBERC 1988, 1991). Work with filamentous fungi has progressed much more slowly such that only a relatively small number of such mu- tants have been isolated and characterized. In Asper- gillus nidulans only nine UV-sensitive (uvs) mutants have previously been extensively characterized for effects on recombination (reviewed by KAFER and MAYOR 1986).

There may be fundamental differences between DNA repair and recombination in S. cerevisiae and in A. nidulans. Like mammalian cells but unlike S. cere- visiae, germinating A. nidulans cells spend a substantial part of their vegetative cell cycle in G2 (BAINBRIDGE 197 1; BERGEN and MORRIS 1983). During G2 there are two copies of each chromosome existing as sister chromatids, which means there are greater opportu- nities for homologous recombination in haploid cells of A . nidulans than S. cerevisiae. Recombinational re- pair pathways in A. nidulans may therefore be more significant than in S. cerevisiae. This is the case with Schirosaccharomyces pombe, which also spends a large part of its vegetative cell cycle in G2 (PHIPPS, NASIM and MILLER 1985). The pleiotropic phenotypes and epistatic grouping of A. nidulans mutants defective in DNA repair/recombination (KAFER and MAYOR,

446 F. Osman, B. Tomsett and P. Strike

1986) more resemble those of repair/recombination mutants of other eukaryotes than do those of S. cere- visiae, implying that the genetic control of recombi- nation in A. nidulans may be more typical. Indeed, recently cloned S. pombe rad genes thought to be involved in recombination (LEHMAN et al. 199 1; SUB- RAMANI 1991) show no extensive homology to any cloned S. cerevisiae recombination genes.

In a previous paper (OSMAN et al. 1991) we de- scribed the isolation and characterization of an A. nidulans mutant, nuvl l , hypersensitive to N-methyl- N’-nitro-N-nitrosoguanidine (MNNG) and 4-nitroqui- noline-1-oxide (4-NQO), and with effects on meiotic and mitotic recombination. We now describe the iso- lation of a further 23 mutants hypersensitive to these damaging agents and, using simple plating tests, the characterization of their effects on spontaneous mi- totic intragenic recombination.

MATERIALS A N D METHODS

A. nidulans strains and media: A. nidulans strains used in this work carried markers in general use that have been described previously (CLUTTERBUCK 1974). The routine genetic techniques used were modified after PONTECORVO et al. (1 953). Standard media have been described previously (COVE 1966). “Difco” Bacto-Agar (Difco Laboratories, De- troit, Michigan) was used in all media. Minimal base (MB) consisted of a carbon source, salts and trace elements. Min- imal medium (MM) consisted of MB plus a nitrogen source and supplements for auxotrophic markers. Mutagen-con- taining media were made up at pH 6.0.

Isolation of MNNG- and/or UV-sensitive mutants: Mu- tants were isolated from three different parental strains, each of which was phenotypically wild-type with respect to mutagen sensitivity. Conidial (asexual uninucleate haploid spore) suspensions of the three parental strains, B608 (PuAI , wA3, niiAniaD608), CS387 (yA2, luAI, niaD26) and L20 (pabaAI, wAj) , were each treated with MNNG (0.5 pg/ml final concentration) to give less than 10% survival. Master plates were prepared by overlaying mutagenized conidia at a density of approximately 100 surviving conidia per plate onto MM (to eliminate auxotrophic mutants) containing 0.08% sodium deoxycholate (SD) to induce compact colo- nies (MACKINTOSH and PRITCHARD 1963). These plates were incubated at 37“ for 3 days, by which time the muta- genized colonies had grown and sporulated. MNNG-sensi- tive and/or UV-sensitive colonies were identified by replica plating using damp velvet pads. MNNG-sensitive colonies were identified by comparing growth on two replica plates of MM, one of which contained 3.5 pg/ml MNNG. UV- sensitive colonies were scored as showing poor mycelial growth and/or poor sporulation on replica plates preincu- bated for 7 hr and irradiated with 400 J/m’ UV (wavelength 254 nm) in comparison with growth of unirradiated replica colonies.

Potential MNNG-and/or UV-sensitive mutants (desig- nated nuv) were subcultured from the master plates to fresh M M master plates, nine per plate together with the wild- type parental strain, in arrays suitable for 10-point wire replication. These potential hypersensitive mutants were rescreened by wire replication of conidia into MM plates containing 0.75 and 1.0 pg/ml MNNG and 0.5 and 0.75 pg/ml4-NQO. The quasi-UV-mimetic mutagen 4-NQO was

used instead of UV for screening for sensitivity by wire replication since results using UV were ambiguous and inconsistent. Like MNNG, 4-NQO has the advantage that it can be added directly to media. I t gave much clearer results than UV, although there is not 100% correlation between U V and 4-NQO sensitivity. Hypersensitive mutants were taken through two rounds of screening by wire repli- cation, retaining only those that were consistently more sensitive than conidia of the parental strain.

The MNNG, U V and 4-NQO doses used to identify hypersensitive mutants in these replica plating experiments were initially shown to be effective at distinguishing the previously isolated UV- and mutagen-sensitive A. nidulans strains uusB, uvsC and uvsD (KAFER and MAYOR 1986) from wild-type.

Quantifying MNNG and 4-NQO sensitivity: A heavy suspension of conidia was spread onto a M M plate using a sterile glass spreader and incubated for 24 hr at 37” to produce a nonsporulating mycelial “mat.” Mycelial “plugs” were cu t out from the mats using the wide end of a sterile Pasteur pipette. The plugs were transferred in duplicate onto M M plates containing various concentrations of M N N G or 4-NQO and incubated for 48 hr at 3 7 ” . The diameters of the resulting colonies were measured under a stereoscopic microscope to the nearest 0.1 nlm using a micrometer. Radial growth was expressed as percentage of average diameter of colonies on zero-dose plates. This tech- nique avoids the problem of delayed germination found with some mutagen treatments and ensures that the true exponential phase growth rate is measured.

Performing the mitotic recombination assay: The assay measures mitotic recombination between nonoverlapping deletions within the niaD gene. Full details of the assay will be given i n REsuLrs. The mitotic recombination phenotype was determined qualitatively by simply centrally inoculating ;I M M plate containing nitrate as sole nitrogen source with 0.5 pl of a suspension of conidia from the diploid under test. The [dates were incubated at 37’ for 7 days before scoring.

RESULTS

Isolation of nuv mutants: By the end of the mutant isolation program, 18 putative mutagen-hypersensi- tive mutants had been identified from screening about 4,800 mutagenized L20 colonies, over 80 from screening about 8,000 mutagenized B608 colonies, and over 100 from screening about 10,000 mutagen- ized CS387 colonies. T h e mutant strains were desig- nated L20nuv, B608nuv and CS387nuv, respectively, followed by a unique distinguishing number.

After two rounds of screening, 23 of the mutants were retained for further investigation on the basis that they were the most hypersensitive to MNNG: it had been found previously in both S. cerevisiae (BREN- DEL, KHAN and HAYNES 1970) and A. nidulans (KAFER and MAYOR 1986) that hypersensitivity to monofunc- tional alkylating agents was especially pronounced in mutants that affect recombination. The 23 mutants selected were as follows: nine mutants isolated in B608-nuv-1, nuv-2, nuv-4, nun-6, nuv-7, nuv-8, nuv- 10, nuv-12, nuv-20; 12 from CS387-nuv-102, nuv- 103, nuv-107, nuv-109, nuv-110, nuv-11 I , nuv-112, nuv-114, nuv-115, nuv-117, nuv-120, nuv-121; two in

A . nidulans Recombination Mutants 447

the L20 background”nuv-?29 and nuv-??4. All of these mutants were hypersensitive to both M N N G and 4 -NQO.

Mutant nuu allele segregation: Each of the nuv mutant phenotypes was shown to result from mutation of a single gene by sexually crossing each mutant strain with a strain phenotypically wild-type with respect to mutagen-sensitivity. Segregation of the nuv-I0 phe- notype was analyzed by the parasexual cycle, which involved diploid construction and haploidization, since sexual crosses heterozygous for nuv-10 produced no viable hybrid sexual spores. Randomly chosen progeny from each cross were screened for hypersen- sitivity to M N N G and 4 - N Q O by wire inoculation of conidia into mutagen-containing media. In each case there was a 1: 1 mutant/wild-type phenotypic segre- gation with respect to hypersensitivity. Analysis of the sensitivities of the progeny did not reveal any new phenotypic class, all progeny being clearly of one parental type or the other, and with hypersensitivity to M N N G and 4-NQO always segregated together.

Comparison of hypersensitivities of nuv strains to MNNG and 4-NQO: The hypersensitive phenotype of each nuv allele was tested in both B608 and CS387 genetic backgrounds, relative to the parental strains, by measuring the hyphal extension of germinating mycelial plugs on solid minimal media containing var- ious concentrations of M N N G or 4 - N Q O . T o make direct comparisons between the different nuv strains, the determinations were carried out as two large experiments, one for sensitivity to M N N G and the other for sensitivity to 4 - N Q O , in which all the plates at a particular mutagen concentration were made up using the same mutagen-containing media. This was necessary due to the inherent instability of the muta- gens, particularly M N N G , and the variability when making up mutagen-containing media. A typical growth response caused by one of the nuv mutations, nuv-2, is represented graphically in Figure 1. Data on the comparative hypersensitivity to M N N G and 4- N Q O of the different nuv mutants is presented in Figure 2.

The nuv mutants exhibited a variety of growth responses on M N N G and 4 - N Q O relative to the wild- type parental strains, indicating that a diverse range of mutants had been isolated. Some mutants were very hypersensitive to both M N N G and 4-NQO (e.g., nuv-110, nuv-??4). Others displayed relatively greater sensitivity to M N N G than 4-NQO at the con- centrations of mutagen used (e .g . , nuv-l12), and vice versa (e .g . , nuv-6). Another class of mutants displayed relatively low hypersensitivity to both M N N G and 4- N Q O (e .g . , nuv-107).

Dominance-recessiveness of the nuu alleles with respect to hypersensitivity: This was determined by comparing the sensitivity to M N N G and 4-NQO of

homozygous wild-type diploids with that of diploids heterozygous and homozygous for each nuv mutation. As before, to make direct comparisons between the different nuv diploids, the determinations were car- ried out as two large experiments, one for sensitivity to M N N G and the other for sensitivity to 4 -NQO. Diploids were used instead of heterokaryons since some of the mutants may have been defective in nuclear-specific functions. Sensitivity was measured using mycelial plugs inoculated onto M B media con- taining ammonium (selective for diploids) and various concentrations of M N N G and 4 -NQO.

The growth responses of several of the diploid strains are represented graphically in Figure 3. In all cases, except for nun-20, the homozygous nuv diploid was markedly hypersensitive to M N N G and 4 - N Q O relative to the corresponding heterozygous nuv and homozygous wild-type diploids. An interesting result was that the diploid homozygous for nuv-20 did not appear to be hypersensitive to either M N N G or 4- NQO. The homozygous nun-20 diploid was haploid- ized on benlate-containing media (HASTIE 1970) and the haploid segregants tested for hypersensitivity to M N N G and 4 -NQO. All segregants were hypersensi- tive confirming that the nuv-20 hypersensitive phe- notype was haploid-specific.

In most cases the growth response of diploids het- erozygous for the nuv mutations was similar to that of homozygous wild-type diploids, suggesting complete recessiveness of the nuv allele to its wild-type allele. All of the mutations appeared to be recessive except three: nuv-2, appeared to be semi-transdominant for hypersensitivity to both M N N G and 4 -NQO; nuv-8 and nun-1 1 I appeared to be semi-transdominant only for hypersensitivity to 4 -NQO. In these cases the heterozygous diploid was intermediate in hypersensi- tivity relative to that of the corresponding homozy- gous nuv and homozygous wild-type diploids. None of the mutants appeared to be completely dominant.

Establishment of a novel assay for spontaneous mitotic intragenic recombination: Mitotic intragenic recombination between heteroalleles is most conven- iently measured in the case of mutations conferring auxotrophy. This allows rare prototrophic recombi- nants to be selected on medium lacking the appropri- ate nutrient. The mitotic recombination assay devel- oped is based on the two A. nidulans strains, B608 (wA4, puA2, niiAniaD608) and CS387 (yA2, luAI , niaD26), each carrying a non-overlapping heteroal- l e k deletion within the niaD gene on chromosome VIII. The ends of the deletions had been previously determined by fine-structure mapping (TOMSETT and COVE 1979). They are shown schematically in Figure 4, together with the structural organization of the nitrate assimilation genes. The two strains were defi- cient in nitrate reductase, the niaD gene product.

448 F. O m a n . B. Tomsett and P. Strike

f 70 f a

2 3 0 0 I D 2 0 a

10

0

20 25

FIGURE I .-Hypersensitivity to

nuu-2 mutation. The growth re- sponse of germinating mycelia of nuu-2 mutant strains i n comparison wi th parental strains inoculated onto solid media containing various con- centrations of M N N G or 4-NQO. The percentage hyphal growth rep- resents the relative hyphal extension of strains growing in the presence of MNNG or 4-NQO in comparison with controls growing i n the absence of MNNG or 4-NQO. I t should be noted that the linear growth response of multinucleate hyphae to MNNG and 4-NQO was determined in these experiments. not conidial survival.

+m M N N G and 4-NQO caused by the -+-mn*1

10 -

0.0 02 0.4 0.8 Od 1.0 12

4NQo-bmJ

n +1 pglrnl MNNG

I 4 . 5 pglml4NQO

FIGURE 2.-Comparison of hypersensitivity to M N N G and 4-NQO of different nuu mutants and relationship to recombination ability. The data represent the growth responses of germinating mycelia of B608 and B608nuu strains inoculated onto solid media containing either 0.5 pg/ml MNNG or 0.5 pg/ml 4-NQO. The percentage hyphal growth represents the relative hyphal extension of strains growing in the presence of MNNG or 4-NQO in comparison with controls growing in the absence of M N N G or 4-NQO. For completeness sake. the figure includes the data for the nuu-11 mutant, which has been described elsewhere (OSMAN et al. 1991). In the bottom panel, the recombination ability of each mutant is indicated. with the mutants being grouped phenotypically into Rec' (nuu-1 to nuu-329). hypo-rec (nuu-2 to nuu-117) and hyper-rec (nuu-1 I to nuu-334) categories. Recombination ability was assessed by a modification of the simple plate assay. which allows a quantitative assessment of recombination competence to be made (F. OSMAN, unpublished data).

They could not therefore utilize nitrate as sole nitro- gen source. However, commercial agar contains traces of reduced nitrogen, which allows nitrate nonutilizing strains to grow on nitrate-containing solid media with a characteristic residual growth phenotype, namely sparse, nonsporulating mycelial growth. This pheno- type was exploited to qualitatively assay mitotic recom- bination within the niaD gene in the diploid B608/ CS387.

Centrally inoculated B608/CS387 diploid colonies

initially grew with the characteristic residual growth phenotype on nitrate-containing media. After 4-5 days incubation, sporulating subcolonies were ob- served to arise within the sparse mycelial growth, Figure 5 , plates A. As controls, heteroallelic diploids were constructed between each test strain and strains carrying an overlapping deletion (see Figure 4); a diploid was constructed between B608 and B144 (pabaA I ,JiuAl, niiAniaD526), and between CS387 and B3 17 (biA I , niiAniaD564). In plating tests with these

A. nidulans Recombination Mutants 449

100

90

M

lo 1 0.00 0.50 1.00 1.50 2.00 2.50

MNNQ Cmcn. (pglmg

1w

90

M

sa

0.00 0.50 1.00 1.50 2.00 2.50 MNNQ Concn. (puml)

FIGURE 3.-Dominance/reces- siveness of nuu mutations with re- spect to hypersensitivity to MNNG and 4-NQO. This was determined by measuring the growth response of

~ germinating mycelia of diploids het- erozygous and homozygous for the nuu mutations in comparison with the homozygous wild-type parental dip- loid. 10

0

-

0.00 0.25 0.50 0.75 1.00 1.25

" . ( W l m l )

diploids no subcolonies arose, Figure 5 , plates B and notype. All those tested were shown to be diploid by

The subcolonies were genetically analyzed and produced nitrate-utilizing and nitrate-nonutilizing shown to be still diploid and true nitrate-utilizing haploid segregants. Nitrate-utilizing segregants were recombinants. They were subcultured onto nitrate sexually crossed with either nitrate-utilizing or nonu- plates and grew with a normal nitrate-utilizing phe- tilizing strains, and the progeny analyzed. The phe-

C. haploidization on benlate plates (HASTIE 1970). Each

4 nlU-nfaD608 1

4 n l ~ - n i n ~ ~ 2 8 1

notypic ratios of the progeny obtained were consistent with the conclusion that the haploid segregants were true niaD+ recombinants: crosses with nitrate-utilizing strains produced all nitrate-utilizing progeny; the numbers of nitrate utilizers and nonutilizers from crosses with nitrate-nonutilizing strains were not sig- nificantly different ( P = 70-5076) from a 1:l ratio. Preliminary genetic analysis of the haploid segregants from the nitrate-utilizing recombinants (L. IWANEJKO, unpublished data) revealed that they could arise by both reciprocal crossing over and nonreciprocal gene conversion.

Effect of nuv mutations on mitotic recombination: The assay described above was exploited to qualita- tively determine the effect of heterozygous or homo- zygous nuv mutations on spontaneous mitotic recom- bination in B608/CS387 diploids. Each diploid was tested on three separate plates. The results were con- sistent and reproducible. This assay was especially effective at identifying nuv mutations causing a hypo- rec phenotype. Examples of plates from which the results were obtained are shown in Figure 5 , plates D-G. Three classes of nuv mutations were identified on the basis of the mitotic recombination phenotype of the homozygous nuv diploid: rec+ types (nuv-I, nuv-10, nuv-12, nuv-20, nuv-103, nuv-107, nuv-109, nuv-115, nuv-120, nuv-121), hypo-rec types (nuv-2, nuv-4, nuv-6, nuv-7, nuv-8, nuv-110, nuv-112, nuv-114, nuv-117, nuv-329) and hyper-rec types (nuv-102, nuv- I I I, nuv-334). Rec+ types displayed a similar recom- bination phenotype to wild-type diploids. Hyper-rec types produced a several-fold increase in the number of niaD+ recombinants, the effect of nuv-102 being the most marked followed by nuv-I11 and nuv-334. The hypo-rec types produced a several-fold decrease in the number of recombinants compared with wild- type. They could be classified into three groups: dip- loids homozygous for nuv-2, nuv-7, nuv-1 I O , nuv-112 and nuv-329 produced a greater than 10-fold de- crease; nuv-4 and nuv-I14 had an intermediate effect, a less than IO-fold but greater than fourfold reduc- tion; nuv-6, nuv-8 and nuv-117 caused a less than fourfold decrease. The nuv-1 I7 mutation appeared to have a semidominant effect; the heterozygous nuv- I I7 diploid also displayed a hypo-rec phenotype but not to the extent of the homozygous nuv-I17 diploid, Figure 5 , plates F and G. The other nuv mutations affecting recombination appeared to be fully reces-

450 F. Osman, B. Tomsett and P. Strike

cmA-niiA-niaD GENE REGION ON CHROMOSOME Vm FIGURE 4,"Schematic map of the c m A -

niiA-niaD gene cluster on chromosome V l l l I and the niaD and niiAniaD deletions. The

line for the niaD deletion depicts the full 4 ni"ni- 1 extent of the deleted region. The niiAniaD

deletions extend from the niaD gene into

CrnA niaD niiA

-1 the niiA gene and only their right-hand limit is shown.

sive, in each case the heterozygous nuv diploid dis- playing a similar recombination phenotype to the homozygous wild-type diploid.

Complementation analysis with some of the nuv mutations: Complementation analysis is necessary to determine the number of genes defined by the nuv mutations. An initial study was carried out with the previously described mutation nuvl l (OSMAN et al. 1991) and six of the nuv mutations with effects on mitotic recombination from this study: nuv-2, nuv-4, nuv-102, nuv-1 IO, nuv-I14 and nuv-117. Diploids het- erozygous in trans for two nuv mutations were con- structed for all possible pairwise combinations. Allel- ism was determined by testing these heterozygous diploids for hypersensitivity to MNNG and 4-NQO, and for mitotic recombination phenotype using the simple plating assay. The responses were compared with that of the singly heterozygous nuv diploids (as positive controls) and the homozygous nuv diploids (as negative controls). The results are shown in Table 1. The results for trans complementation with respect to hypersensitivity and to recombination phenotype were consistent. They showed that the seven nuv mutations analyzed defined at least six nonallelic loci. The nuv-102 and nuvl17 mutations were noncomple- menting for hypersensitivity, and the nuv-I02/nuv- I I7 diploid had the same hypo-rec phenotype as the homozygous nuv-1 I7 diploid. They thus appeared to be allelic despite their different effects on mitotic recombination. The diploids heterozygous for nuv- I 1 7 and each of the other nuv mutations displayed a hypo-rec phenotype comparable to that of the nuvl17/wild-type diploid. This confirmed the semi- dominant hypo-rec phenotype of nuv-117. Diploids heterozygous with nuv-2 were all hypersensitive to MNNG and 4-NQO compared with wild-type, con- firming its semi-transdominance for this phenotype.

Several of the nuv mutations were assigned to link- age groups (whole chromosomes) by mitotic analysis (MCCULLY and FORBES 1965): nuv-2 to chromosome V; nuv-4 to chromosome 11; nuv-102 and nuv-I14 to chromosome VIII. Despite exhaustive attempts the nuv-117 mutation could not be mapped to a specific chromosome. This was also the case with some previ- ously isolated uvs mutations (EVSEEVA and KAMENEVA 1977). None of the previously characterized muta- tions causing hypersensitivity to DNA damaging agents, the uvs (KAFER and MAYOR 1986) and sag

A

B

D

E

F

G

A . nidulans Reconlbination Mutants 45 1

A

C

D

E

F

G

FIGURE 5 . " A simple assay for mi- totic recombination. Sporulatingsub- colonies arose spontaneously within the characteristically sparse myce- lium of centrally inoculated B608/ CS387 heteroallelic diploid colonies growing on nitrate-containing me- dium (Plates A). These subcolonies were shown to be nitrate-utilizing niaD+ recombinants, and to have arisen either by crossing over or gene conversion. As controls, heteroallelic diploids were constructed between each test strain and a strain carrying an overlapping niaD deletion. In plat- ing tests with these diploids, B608/ B114 (Plate B) and B317/CS387 (Plate C), no nitrate-utilizing subco- lonies arose. Recombinationdefec- tive nuv mutants were identified by comparing the frequency of nitrate- utilizing subcolonies arising from B608/CS387 diploids heterozygous and homozygous for a nuv mutation. Plates demonstrating the recessive hyporec phenotype of nuv-2 and the semi-transdominant hypo-rec pheno- type of nuv-117 are also shown: Plates D, B608nuv2/CS387; Plates E, B608nuv2/CS387nuv2; Plates F, B608/CS387nuvll7; Plates C, B608nuv117/CS387nuvl17.

(SWIRSKI et al. 1988) mutations, mapped to chromo- was analyzed for allelism with the uusl), uusE and uusJ some 11. T h e nuv-4 mutation therefore probably de- mutations also located on chromosome V. Comple- fines a new locus in A. nidulans. T h e nuu-2 mutation mentation analysis was performed by scoring for hy-

452 F. Osman, B. Tomsett and P. Strike

TABLE 1

Trans complementation analysis of nuv mutants

nuv-2 nuv-4 nuu-11 nuv-102 nuv-110 nuu-114

nuv-117 + + + - + + nuv-114 + + + + + nuv-110 i- i- + + nuv-102 + + + nuv-11 + + nuu-4 + Complementation was determined by scoring: (i) sensitivity of

germinating mycelial plugs of heterozygous diploids to growth on solid media containing 2.0 &ml MNNG or 0.75 pg/ml 4-NQO; (ii) the mitotic intragenic recombination phenotype of heterozygous diploids. + = complementation; - = noncomplementation.

persensitivity of diploids to MNNG and 4-NQO. The results (data not shown) showed that nuv-2 was non- allelic to all of these uvs mutations. The nuv-2 muta- tion therefore also defines a new locus in A. nidulans.

DISCUSSION

The mutant isolation program was successful in isolating a large number of mutants hypersensitive to MNNG and/or UV-irradiation. Of the 23 nuv mu- tants chosen for further characterization, all were hypersensitive to both MNNG and 4-NQO and in each case this phenotype appeared to be due to single chromosomal gene mutation. The wide diversity of growth responses to MNNG and 4-NQO exhibited by the nuv mutants was desirable since this meant that they were more likely to be mutations in different genes. Initial complementation analysis seems to bear this out.

From what is known from studies in other model organisms (e.g., FRIEDBERG 1988), mutagen-hypersen- sitive mutants can result from mutations inactivating structural genes for DNA repair activity, or those that affect the regulation of expression of DNA repair genes. For eukaryotic organisms, mutations in genes involved in chromatin structure and regulation of the cell cycle can also produce a hypersensitive phenotype (HARRIS and BOYD 1987; WEINERT and HARTWELL 1988). It is also important to consider that some mutations may cause hypersensitivity due to indirect consequences. For example, sensitivity of cells to MNNG has been previously shown to be affected by intracellular thiol content (SEDGWICK and ROBINS 1980) and pH (DELIC, HOPWOOD and FRIEND 1970). It can also be envisaged that a hypersensitive pheno- type could be due to mutations affecting cell perme- ability, or causing enhanced damage to some essential cellular component other than DNA.

The tests for hypersensitivity of diploids heterozy- gous and homozygous for the nuv mutations relative to wild-type were important for identifying recessive and semi-transdominant nuv alleles. Mutant alleles exhibiting dominance are of particular interest be-

cause they are candidates for specifying genes in- volved in regulation of repair functions. Alternatively they could specify structural genes for repair enzymes, which exhibit a gene dosage effect in diploids. The semidominance of nuv-8 and nuv-Ill to 4-NQO but not to MNNG is most readily explainable if it is postulated that the function encoded by the wild-type genes is a major regulatory or structural component for repair of lesions caused by 4-NQO, but only a minor component for the repair of MNNG-induced lesions. The apparent haploid-specific sensitivity of nuv-20 is interesting but not readily explainable in molecular terms.

A test system has been developed in A. nidulans, which is novel for this organism and which can assay for mitotic intragenic recombination rapidly and ef- fectively using simple plating tests. I t directly meas- ures recombinant subcolonies arising from independ- ent events within the diploids, giving an accurate measure of differences in the frequency of recombi- nation events in different strains. Traditional assays do not give such a direct measure of recombination frequency. They rely on harvesting conidia from het- eroallelic diploid colonies growing on nonselective media and, hence, a recombinant cell can give rise to a clonal population distorting the frequency (JANSEN 1970; FORTUIN 197 1).

In this study we have investigated the effects on mitotic recombination of mutations, which affect the ability to repair general DNA damage. It seems likely that those mutations that do show effects on the intragenic recombination measured in our assay will not only affect recombination in the niaD region but will also affect recombination in other regions of the genome. Among the mutants initially isolated as hy- persensitive to DNA damaging agents, a number of Rec phenotypes were observed. Of the 23 nuv muta- tions characterized in our assay, 10 appeared to have no marked effects on mitotic recombination. These nuv mutations were phenotypically similar to the pre- viously characterized uvsA mutation (JANSEN 1967). Ten nuv mutations appeared to be hypo-rec, similar to the previously characterized uvsC and uvsE muta- tions (FORTUIN 197 1 ; JANSEN 1970). Three appeared to be hyper-rec, as were the uvsB, uvsD, uvsF, uvsH and uvsJ mutations (LANIER, TUVESON and LENNOX 1968; SHANFIELD and KAFER 1969; JANSEN 1970; WRIGHT and PATEMAN 1970; FORTUIN 197 1). nun- I 1 7 had a semidominant hypo-rec phenotype, which could imply a gene dosage effect for its product or a mutation within a regulatory gene or sequences.

The possible defects of the nuv mutations affecting mitotic recombination and the underlying functions of the gene products can to some extent be postulated by extrapolation from what is known from mutational studies in other organisms. Mutations causing a hypo-

A. nidulans Recombination Mutants 453

rec phenotype for spontaneous homologous chromo- somal recombination are the most likely to be defec- tive in structural or regulatory genes directly involved in recombination. For example, the characteristic phe- notype of the S. cerevisiae RAD52 group mutants, which are thought to be defective in recombination processes, is recombination deficiency (HAYNES and KUNZ 1981; GAME 1983; COOPER and KELLY 1987; FRIEDBERC 1988, 199 1). Mitotic hyper-rec mutants could also be defective in primary recombination functions, as is the rad50 mutant of S. cerevisiae (MA- LONE et U L . 1990). However, such a phenotype could also be the result of secondary effects caused by de- fects in genes involved in other aspects of DNA me- tabolism. For example, these mutations could be de- fective in a repair mechanism, which causes the accu- mulation of recombinogenic lesions, o r lesions that induce the expression of recombination enzymes. Some eukaryotic mutants characterized as defective in excision repair or error-prone replicative repair have been shown to be hyper-rec for spontaneous mitotic chromosomal recombination. For example, the rad3 and rad6 mutants of S. cerevisiae, which are deficient in excision repair and error-prone repair, respectively, are also hyper-rec for spontaneous mitotic recombination (KERN and ZIMMERMAN 1978).

Trans complementation analysis between seven of the nuv mutations showed that they defined at least six nonallelic loci. The remaining nuv mutants are currently being assigned to specific chromosomes and tested for complementation between themselves and with the previously characterized uvs and sag mutants. T h e nuv-102 and nuv-117 mutations appear to be noncomplementing for both hypersensitivity and mi- totic recombination. If nuv-102 and nuv-I17 do define a single locus, two allelic mutations have been isolated with very different effects on hypersensitivity and mitotic recombination. Similar differential effects for allelic DNA repair and recombination defective mu- tants are common in both eukaryotes and prokaryotes. For example, whereas most S. cerevisiue rad52 mutants were hypo-rec for mitotic recombination, the leaky rad52-2 mutation displayed a mitotic hyper-rec phe- notype (MALONE et al. 1988). T h e availability of sev- eral different mutants in the same gene is useful in establishing the range of possible phenotypic effects, and in identifying those that are typical for a specific gene. Phenotypic differences between allelic mutants are most likely to arise with genes encoding multi- functional proteins or regulatory functions.

In conclusion, there is little doubt that a set of mutants has been isolated that define genes important in DNA metabolism, with the prospect that some may be involved in regulation and control. The nuv-2 and nuv-4 mutations have already been shown to define new loci in A . nidulans, increasing the number of

genes known to be involved in DNA repair and recom- bination in this organism. The further phenotypic characterization of these mutants together with as- signment to epistatic groups will be important in de- termining their functions. The availability of a large number of mutants should facilitate the cloning of the relevant genes by complementation and, conse- quently, purification and biochemical characterization of the gene products. The correlation of mutant phe- notype to biochemical activity will be the critical cri- terion by which the molecular mechanisms of homol- ogous recombination will be elucidated. The number of mutants now available in A. nidulans make it at- tractive as a complementary system to S. cerevisiae. I t is likely that some of the nuv mutants will define genes of types not previously characterized, although the availability of any homologous genes from two evo- lutionarily distinct fungi should help in the isolation of homologues from higher eukaryotes.

This work was supported by grants from the University of Liverpool and the Wellcome Trust to P.S. and B.T. The helpful advice of B. M. FAULKNER is gratefully acknowledged.

L I T E R A T U R E CITED

BAINBRIDGE, B. W., 1971 Macromolecular composition and nu- clear division during spore gemination in Aspergzllus nidulans. J. Gen. Microbiol. 66 319-325.

BAKER, 8. S., and D. A. SMITH, 1979 The effects of mutagen- sensitive mutants of Drosophila melanogaster in non-mutagen- ized cells. Genetics 92: 291-304.

BERGEN, G. G., and N. R. MORRIS, 1983 Kinetics of the nuclear division cycle of Aspergillus nidulans. J. Bacteriol. 156: 155- 160.

BRENDEL, M., N. A. KHAN and R. H. HAYNES, 1970 Common steps in the repair of radiation and alkylation damage in yeast. Mol. Gen. Genet. 106: 289-295.

CLARK, A. J., 1973 Recombination-deficient mutants of E. coli and other bacteria. Annu. Rev. Genet. 7: 67-86.

CLUTTERBUCK, A. J., 1974 Aspergillus nidulans, pp. 447-510 in Handbook of Genetics, Vol. 1. Bacteria, Bacteriophages and Fungi, edited by R. C. KING. Plenum Press, New York.

COOPER, A. J., and S. L. KELLY, 1987 DNA repair and mutagen- esis in Saccharomyces cerevisiae, pp. 73-1 14 in Enzyme Induction, Mutagen Activation and Carcinogen testing in Yeast, edited by A. WISEMAN. Ellis Horwood, London.

COVE, D. J., I966 The induction and repression of nitrate reduc- tase in the fungus ilspergzllus nidulans. Biochem. Biophys. Acta 113: 51-53.

DELIC, V., D. A. HOPWOOD and E. J. FRIEND, 1970 Mutagenesis by N-methyl-N’-nitro-N-nitrosoguanidine (NTG) in Strepto- myces coelicolor. Mutat. Res. 9: 167-182.

EVSEEVA, G. V. and S. V. KAMENEVA, 1977 Genetic control of sensitivity to mutagenic factors in Aspergillus nidulans. VII. Study of the inheritance of cross-sensitivity to various muta- genic agents in uvs mutants. SOGEB 13: 1332-1336 (from Genetika 13: 198 1-1 987).

FORTUIN, J. J. H., 1971 Another two genes controlling mitotic intragenic recombination and recovery from UV damage in Aspergillus nidulans. 11. Recombination behavior and X-ray sensitivity of uvsD and uvsE mutants. Mutat. Res. 11: 265-277.

FRIEDBERG, E. C., 1988 Deoxyribonucleic acid repair in the yeast Saccharomyces cerevisiae. Microbiol. Rev. 52: 70- 102.

454 F. Osman. B. Tomsett and P. Strike

FRIEDBERG, E. C., processes: new 2310.

GAME,.[. C., 1983

1991 Yeast genes involved in DNA repair looks on old faces. Mol. Microbiol. 5: 2303 -

Radiation-sensitive mutants and DNA repair in yeast, pp. 109-1 37 in Yeast Genetics. Fundamental and Applied Aspects, edited by J. F. T . SPENCER, D. M. SPENCER and A. R. W. SMITH. Springer-Verlag, New York.

HARRIS, J. M., and J. B. BOYD, 1987 Pyrimidine dimers in Dro- sophila chromatin become increasingly accessible after irradia- tion. Mutat. Res. 183: 53-60.

HASTIE, A. C . , 1970 Benlate-induced instability in Aspergillus diploids. Nature 226: 77 1.

HASTINGS, P. J., 1988 Conversion events in fungi, pp. 397-443 in Genetic Recombination, edited by R. KUCHERLAPATI and G . R. SMITH. Am. SOC. Microbiol., Washington, D.C.

HAYNES, P. J., and B. A. KUNZ, 1981 DNA repair and mutagenesis in yeast, pp. 37 1-414 in The Molecular Biology of the Yeast Saccharomyces. L$e cycle and Inheritance, edited by J. STRATH- ERN, E. W. JONES and J. R. BROACH. Cold Spring Harbor Laboratory, Cold Spring Harbor, N.Y.

HOLLIDAY, R., 1967 Altered recombination frequencies in radia- tion-sensitive strains of Ustilago. Mutat. Res. 2: 557-559.

HOWARD-FLANDERS, P., and L. THERIOT, 1966 Mutants of Esch- erichia coli K12 defective in DNA repair and genetic recombi- nation. Genetics 53: 1137-1 150.

JANSEN, G. J. O., 1967 Some properties of the uvsl mutant of Aspergillus nidulans. Asp. News Lett. 8: 20-2 1 .

JANSEN, G. J. O., 1970 Abnormal frequencies of spontaneous mitotic recombination in uvsB and UVSC mutants of Aspergillus nidulans. Mutat. Res. 10: 33-41.

KAFER, E., and 0. MAYOR, 1986 Genetic analysis of DNA repair in Aspergillus nidulans: evidence for different types of MMS- sensitive hyperec mutants. Mutat. Res. 161: 119-134.

KERN, R., and F. K. ZIMMERMANN, 1978 The influence of defects in excision and error- prone repair on spontaneous and induced mitotic recombination in Saccharomyces cerevisiae. Mol. Gen. Genet. 161: 81-88.

LANIER, W. B., R. W. TUVESON and J. E. LENNOX, 1968 A radiation-sensitive mutant of Aspergillus nidulans. Mutat. Res.

LEHMAN, A. R., A. M. CARR, F. Z. WATTS, and J. M. MURRAY, 1991 DNA repair in the fission yeast Schizosaccharomyces pombe. Mutat. Res. 2 5 0 205-210.

MACKINTOSH, M. E., and R. M. PRITCHARD, 1963 The production and replica plating of micro-colonies of Aspergillus nidulans. Genet. Res. 4: 320-322.

MALONE, R. E., B. MONTELONE, C. EDWARDS, K. CARNEY, and M.

5: 23-3 1 .

F. HOEKSTRA, 1988 A reexamination of the role of the RAD52 gene in spontaneous mitotic recombination. Curr. Ge- net. 14: 21 1-223.

MALONE, R. E., T. WARD, S. LIN and J. WARING, 1990 The RAD50 gene, a member of the double strand break repair epistatic group, is not required for spontaneous mitotic recom- bination in yeast. Curr. Genet. 18: 1 1 1-1 16.

MCCULLY, K. S., and E. FORBES, 1965 The use of p-fluorophen- ylalanine with “master strains” of Aspergillus nadulans for as- signing genes to linkage groups. Genet. Res. 6: 352-359.

ORR-WEAVER, T. L., and J. W. SZOSTAK, 1985 Fungal recombi- nation. Microbiol. Rev. 4 9 33-58.

OSMAN, F., C. COTTON, B. TOMSETT and P. STRIKE, 1991 Isolation and characterisation of nuv l l , a mutation affecting meiotic and mitotic recombination in Aspergillus ni- dulans. Biochimie 73: 321-327.

PHIPPS, A,, A. NASIM and D. R. MILLER, 1985 Recovery, repair, and mutagenesis in Schizosaccharomyces pombe. Adv. Genet. 23:

PONTECORVO, G., J. A. ROPER, L. M. HEMMONS, K. D. MACDONALD and A. W. J. BUFTON, 1953 The genetics of Aspergillus ni- dulans. Adv. Genet. 5: 141-238.

PRAKASH, S., L. PRAKASH, W. BURKE and B. MONTELONE, 1980 Effects of the RAD52 gene on recombination in Saccha- romyces cermisiae. Genetics 94: 3 1-50.

SEDGWICK, B.. and P. ROBINS, 1980 Isolation of mutants of Esch- erichia coli with increased resistance to alkylating agents: mu- tants deficient in thiols and mutants constitutive for the adap- tive response. Mol. Gen. Genet. 180: 85-90.

SHANFIELD, B., and E. KAFER, 1969 UV-sensitive mutants increas- ing mitotic crossing over in Aspergillus nidulans. Mutat. Res. 7: 485-487.

SUBRAMANI, S., 199 1 Radiation-resistance in Schizosaccharomyces pombe. Mol. Microbiol. 5: 2311-2314.

SWIRSKI, R. A , , S. G. SHAWCROSS, B. M . FAULKNER and P. STRIKE, 1988 Repair of alkylation damage in Aspergillus nidulans. Mutat. Res. 193: 255-268.

TOMSETT, A. B., and D. J. COVE, 1979 Deletion mapping of the niiA niaD region in Aspergillus nidulans. Genet. Res. 34: 19- 32.

WEINERT, T . A., and L. H. HARTWELL, 1988 The RAD9 gene controls the cell cycle response to DNA damage in Saccharo- myces cerevisiae. Science 241: 3 17-322.

WHITEHOUSE, H., 1982 Genetic Recombination: Understanding the Mechanisms. John Wiley and Sons, Inc., New York.

WRIGHT, P. J., and J. A. PATEMAN, 1970 Ultraviolet-light sensitive mutants of Aspergillus nidulans. Mutat. Res. 9 579-587.

1-23.

Communicating editor: R. H. DAVIS