Noncomplementing Mutants Research Articles

Homologous recombination in a Chinese hamster X-ray-sensitive mutant

We have tested the mutant Chinese hamster cell line xrs-5, which is sensitive to ionizing radiation, for the ability to carry out homologous recombination. In an in vivo assay to detect recombination between two transfected plasmids carrying non-complementing mutants in the neomycin resistance gene, xrs-5 showed a 6-fold reduction in recombination frequency when compared to the parental cell line K1. Extracts prepared from nuclei of the mutant were also tested for their ability to catalyze homologous recombination between the same two plasmids in vitro. Extracts from xrs-5 were found to mediate recombination in this assay at frequencies not significantly different from those obtained with extracts from the parental cell line.

MSW, a yeast gene coding for mitochondrial tryptophanyl-tRNA synthetase.

E569 and E606 are noncomplementing pet mutants of Saccharomyces cerevisiae. Both strains are defective in mitochondrial protein synthesis and as a result exhibit a pleiotropic deficiency in respiratory components that are translated on mitochondrial ribosomes. The wild type gene MSW capable of complementing the protein synthesis defect has been cloned by transformation of one of the mutants with a genomic library of wild type yeast nuclear DNA. The cloned gene has been sequenced and shown to code for a protein with a molecular weight of 42,414 which is 37 and 39% identical to the tryptophanyl-tRNA synthetases of Escherichia coli and Bacillus stearothermophilus, respectively. A strain containing an insertion in the chromosomal copy of MSW was constructed by in situ gene replacement. This mutant fails to charge mitochondrial tryptophanyl-tRNA providing further evidence that MSW is the structural gene for mitochondrial tryptophanyl tRNA synthetase. The existence of another gene coding for the cytoplasmic tryptophanyl-tRNA synthetase is inferred from the observation that mutations in MSW are not lethal but only result in a respiratory deficiency.

Functional interchangeability of DNA replication genes in Salmonella typhimurium and Escherichia coli demonstrated by a general complementation procedure.

Twenty-four genes from Salmonella typhimurium that affect DNA replication were isolated from a lambda-Salmonella genomic library by lysogenic complementation of temperature-sensitive mutants of Salmonella or E. coli, using a new plaque complementation assay. The complementing lambda clones, which make red plaques in this assay, and noncomplementing mutant derivatives, which make uncolored plaques, were used to further characterize the temperature-sensitive Salmonella mutants and to establish the functional similarity of E. coli and Salmonella DNA replication genes. For 17 of 18 E. coli mutants representing distinct loci, a Salmonella gene that complemented the mutant was found. This result indicates that single Salmonella replication proteins are able to function in otherwise all E. coli replication complexes and suggests that the detailed properties of Salmonella and E. coli replication proteins are very similar. The other seven Salmonella genes that were cloned were unrelated functionally to any E. coli genes examined. --As an aid to the derivation of chromosomal mutations affecting some of the cloned genes, a general method was developed for placing a transposon in the Salmonella chromosome in a segment corresponding to cloned DNA. Chromosomal mutations were derived in Salmonella affecting a gene (dnaA) that was cloned by complementation of an E. coli mutant by using the transposon-encoded drug resistance as a selectable marker in local mutagenesis.

Use of lambda unc transducing bacteriophages in genetic and biochemical characterization of H+-ATPase mutants of Escherichia coli.

The eight subunits of the H+-ATPase of Escherichia coli are coded by the genes of the unc operon, which maps between bglB and asnA. A collection of unc mutations were transferred via P1 transduction into a strain in which lambda cI857 S7 was inserted into bglB. The lambda phage was induced, and asnA+ transducing phage that carried unc were selected. Transducing phage carrying mutations in the uncA, B, D, E, and F genes were used for complementation analysis with a collection of unc mutants, including mutants which had been reported previously but not genetically characterized. Some mutations gave a simple complementation pattern, indicating a single defective gene, whereas other mutations gave more complex patterns. Two mutants (uncE105 and uncE107) altered in the proteolipid (omega) subunit of F0 were not complemented by any of the lambda unc phage, even though both mutants had a fully functional F1 ATPase and therefore normal A and D genes. Hence, only limited conclusions can be drawn from genetic complementation alone, since it cannot distinguish normal from abnormal genes in certain classes of unc mutants. The lambda unc phage proved to be essential in characterizing several mutants defective in F0-mediated H+ translocation. The unc gene products were overproduced by heat induction of the lysogenized lambda unc phage to determine whether all the F0 subunits were in the membrane. Two mutants that gave a simple complementation pattern, indicative of one defective gene, did not assemble a three-subunit F0. The uncB108 mutant was shown to lack the chi subunit of F0 but to retain psi and omega. Trace amounts of an altered omega subunit and normal amounts of chi and psi were found in the uncE106 mutant. A substitution of aspartate for glycine at residue 58 of the protein was determined by DNA sequence analysis of the uncE gene cloned from the lambda uncE106 phage DNA. One of the omega-defective, noncomplementing mutants (uncE107) was shown to retain all three F0 subunits. The uncE gene from this mutant was also sequenced to confirm an asparagine-for-aspartate substitution at position 61 (the dicyclohexylcarbodiimide-binding site) of the omega subunit.

Formation and possible functions of alpha-putrescinylthymine in bacteriophage phi W-14 DNA: analysis of bacteriophage mutants with decreased levels of alpha-putrescinylthymine in their DNAs.

The DNA synthesized in the nonpermissive host by the noncomplementing mutants am36 and am42 of bacteriophage phi W-14 contains about half the wild-type level of alpha-putrescinylthymine (putThy) and a correspondingly greater level of thymine. The mechanisms whereby thymine nucleotides are excluded from replicating DNA are functional in both mutants because neither of them incorporates exogenous thymidine into DNA. It is proposed that (i) in wild-type phi W-14, the conversion of hydroxymethyluracil to putThy at the polynucleotide level is sequence specific, but that to thymine is nonspecific; and (ii) in the mutants, the sequence-specific recognition is impaired so that more thymine and less putThy are formed. The thymine-rich DNA can be packaged into phage particles. In the case of am42, the phage particles are morphologically indistinguishable from and have essentially the same polypeptide composition as wild-type particles. However, the DNA molecules they contain are about 11% shorter than those in wild-type phage, am42rev4, a revertant of am42, contains DNA with about 70% of the normal level of putThy; these molecules are about 3% shorter than wild-type DNA. The properties of am42 and am42rev4 are consistent with the suggestion that putThy facilitates the very tight packing of phi W-14 DNA (Scraba et al., Virology 124:152-160, 1983). It also appears that the putThy content of phi W-14 DNA can be reduced by no more than 30% without adversely affecting the production of viable progeny; for example, the burst size of am42rev4 is about 25% of that of the wild type.

Mutagenesis at the ad-3A and ad-3B loci in haploid UV-sensitive strains of Neurospora crassa: VI. Genetic characterization of ad-3 mutants provides evidence for qualitative differences in the spectrum of genetic alterations between wild-type and nucleotide excision-repair-deficient strains

Genetic characterization of ad-3B mutants induced in wild-type and UV-sensitive strains has revealed qualitative differences between the spectra of genetic alterations at the molecular level. Ad-3B mutants induced in the two nucleotide excision-repair-deficient strains upr-1 and uvs-2 (Worthy and Epler, 1973) had significantly lower frequencies of nonpolarized complementation patterns and higher frequencies of noncomplementing mutants than ad-3B mutants induced in the wild-type strain in samples induced by either UV, γ-rays, 4NQO or MNNG. In these same samples ad-3B mutants induced in uvs-4, uvs-5 or uvs-6 did not differ significantly from those induced in the wild-type strain. After ICR-170 treatment, ad-3B mutants induced in the UV-sensitive strains did not differ significantly from those induced in wild-type. The comparisons in the present and previous studies demonstrate that the process of mutation-induction in the ad-3 region is under the control of other loci that not only alter mutant recovery quantitatively (de Serres, 1980; Schüpbach and de Serres, 1981; Inoue et al., 1981a, b) but also qualitatively. These data have important implications for comparative chemical mutagenesis, since the spectrum of genetic alterations produced by a given agent can be modified markedly as a result of defects in DNA repair.

Genetic complementation and enzyme correlates at the locus encoding the last two steps of de novo pyrimidine biosynthesis in Drosophila melanogaster.

Eight mutations of the rudimentary-like (r-l) locus were isolated following mutagenesis with ethylmethanesulfonate and inter se crosses revealed three basic complementation groups, using the wing phenotype as an index of complementation. One group consists of three entirely noncomplementing mutants that each specify severe reductions in levels of both r-l-encoded enzymes, orotate phosphoribosyltransferase (OPRTase) and orotidylate decarboxylase (ODCase). The other two groups consist of complementing mutants, such that any member of one group fully complements all members of the other group. One of these groups consist of two mutants that each specify severely reduced OPRTase, but normal ODCase. The other group consists of three mutants that specify severe OPRTase and ODCase reductions in homoallelic flies, but that appear to contribute OPRTase in certain heteroallelic genotypes. It is concluded that the reciprocal and complementing enzymatic phenotypes of mutants in these two groups account for most instances of genetic trans complementation among r-l mutants. These findings are discussed relative to extant information on OPRTase and ODCase in animals and an hypothesis is developed that the r-l locus encodes a single polypeptide product that contains both enzyme activities.

Recombination and linkage between structural and regulatory genes of herpes simplex virus type 1: study of the functional organization of the genome.

Phenotypic and genetic properties of 12 markers in structural and regulatory functions of herpes simplex virus type 1 were characterized, and their recombination and segregation behavior was investigated and interpreted with reference to available information on their physical locations. The markers were: (i) ts markers in a structural glycoprotein (tsB5) and in alpha (immediate early; tsLB2, tsc75) or beta (early, delayed early; tsB1) functions with regulatory effects; together with (ii) plaque morphology (syn), phosphonoacetate resistance (Pr), and thymidine kinase (TK) phenotypes; and (iii) electrophoretically distinct variants of glycosylated (glycoprotein C, gpC; ICP10) and non-glycosylated [VP(13-14), VP23] structural and nonstructural [ICP(47-48)] polypeptides. Mean two-factor recombination frequencies ranged from 2% (for noncomplementing mutants tsLB2 and tsc75) to 35 to 40% (for unlinked markers) and were influenced by the relative contributions of parental viruses to the mixed infection. Even with control of this variable, standard deviations of mean measures of recombination frequency ranged from a minimum of 14% (with n greater than or equal to 10) to 65% (with n = 3) of mean values; no recombination frequencies higher than 55% were observed. Differences in mean two-factor recombination frequencies between a small number of loosely linked markers were, therefore, not reliable measures of real differences in linkage. Measurements of the segregation of unselected markers among recombinant progeny were, therefore, used as measures of linkage. These experiments (i) established a linkage group for markers in the long unique region of the genome additional to, but consistent with, existing physical data, i.e., TK-syn-tsB5-(tsB1.Pr)-[gpC.VP(13-14)]; (II) identified markers, e.g., ICP(47-48), linked to regulatory mutations (tsLB2, tsc75) in redundant DNA sequences; and (iii) used the segregation of these regulatory mutations and linked markers among unselected progeny to demonstrate the linkage groups: Pr-syn-TK-tsc75-ICP(47-48), [VP(13-14).gpC]-Pr-syn-TK, and TK-tsc75-[VP(13-14).gpC]. These results were most simply explained if bi- or intermolecular recombination occurred between circular molecules or molecules catenated "head-to-tail" and were incompatible with intermolecular recombination as the mechanism of isomerization of herpes simplex virus DNA.

Genetics of Gracilaria tikvahiae (Rhodophyceae). VI. Complementation and linkage analysis of pigmentation mutants

Pigmentation mutants of Gracilaria tikvahiae that differed in several ways from the brownish colour of wild type were characterized genetically by complementation and recombination. Thirty cistrons with multiple alleles at several loci were identified. Many of the cistrons were genetically linked with one or more other loci, and the beginnings of a linkage map is presented. Examples of intragenic recombination between noncomplementing mutants were also encountered.

Isolation and preliminary characterization of temperature-sensitive mutants of Newcastle disease virus.

Temperature-sensitive (ts) mutants of Newcastle disease virus have been isolated and characterized genetically (complementation), biochemically (RNA synthesis) and biologically (fusion from within and hemadsorption). Fifteen of these mutants have been divided into five complementation groups. Groups A (five mutants) and E (one mutant) are ts for RNA synthesis (RNA-) as well as for the other functions. Group B contains four RNA+ mutants of which one is ts for fusion, one for hemadsorption and two for neither function. Group C contains one RNA+ mutant which is a poor cell fuser. Group D contains two RNA+ mutants which are ts for fusion. In addition, two noncomplementing mutants (group BC) fail to complement both group B and group C mutants while exhibiting complementation with mutants in groups A, D, and E.

Genetic and morphological study of aggregation in the cellular slime mold Polysphondylium violaceum.

A system for genetic analysis in the cellular slime mold P. violaceum has been developed. Two growth-temperature-sensitive mutants were isolated in a haploid strain and used to select rare diploid heterozygotes arising by spontaneous fusion of the haploid cells. A recessive mutation to cycloheximide resistance in one stain enables selection of segregants, which often appear to be aneuploid.-Aggregation-defective (ag-) mutants having a wide range of phenotypes were isolated in both temperature-sensitive strains after nitrosoguanidine treatment, and complementation tests were performed between pairs of these mutants. Of 380 diploids isolated, 32 showed defective aggregation and were considered to contain 2 noncomplenting ag- mutations. Among noncomplementing mutants interallelic complentation is common. Noncomplementing mutants fall into 4 complementation groups, and those within each complementation group are phenotypically similar. Statistical analysis of the results suggests that the number of complementation units involved in aggregation is about 50.

Arg-7 mutant X wild-type crosses in Chlamydomonas reinhardi: study of the enzyme produced in diploid strains.

The arg-7 locus is the structural gene for the argininosuccinate lyase (ASL). Interallelic complementation was previously found to occur between several mutants of the locus: this is indicative for the homomultimeric nature of ASL. Two complementing (arg-7-5 and arg-7-7) and two non-complementing (arg-7-1 and arg-7-6) mutants of the arg-7 locus were crossed to the pab-2 strain (which is wild-type for the arg-7 locus). In each cross, heterozygote phenotypically wild-type strains were isolated; their diploid pattern was demonstrated by various criteria: mating type, cell volume, nuclear size. The four heterozygotes were compared to the haploid wild-type and in some experiments, to the diploid strain arg-1 X pab-2 homozygous for the arg-7 locus. No difference was found in growth rate and in the Michaelis constant values for ASL. The specific activity of the enzyme produced in the heterozygotes was about 50 percent of the activity found in haploid or diploid wild-type. The heat sensitivity of ASL was also investigated in the different strains: two (containing the complementing mutations arg-7-5 and arg-7-7) of the four heterozygotes produce ASL varieties different from the wild-type enzyme as far as the thermolability is concerned. These results suggest that hybrid ASL can be formed by interaction between the products of wild-type and mutant genes. A clear dominance of the wild-type allele is expected only when the mutant allele has no product of the gene: this could be the case for arg-7-1 and arg-7-6.

The product of the ade1: gene in Schizosaccharomyces pombe: a bifunctional enzyme catalysing two distinct steps in purine biosynthesis.

The assignment of the known ade genes to steps in purine biosynthesis in Schizosaccharomyces pombe has been completed with the demonstration that an ade3 mutants lacks FGAR amidotransferase, ade1A mutants lack GAR synthetase and ade1B mutants lack AIR synthetase. A comparison of enzyme activity with map position for ade1 mutants shows that (1) complementing ade1A mutants lack GAR synthetase but posses wild type amounts of AIR synthetase, (2) complementing ade1B mutants lack AIR synthetase but posses variable amounts of GAR synthetase, (3) non-complementing mutants lack both activities. In wild type strains the two activities fractionate together throughout a hundred-fold purification. Hence the ade1 gene appears to code for a bifunctional enzyme catalysing two distinct steps in purine biosynthesis. The two activities are catalysed by two different regions of the polypeptide chain which can be altered independently by mutation. Gel filtration studies on partially purified enzymes from wild type and various complementing mutant strains, indicate that the bifunctional enzyme is a multimer consisting of between four and six sub-units of 40,000 daltons each. GAR synthetase activity is associated with both the monomeric and multimeric forms but AIR synthetase is only associated with the multimer. A comparison of enzyme levels between diploids and their original complementing haploid strains suggests that complementation is due to hybrid enzyme formation.

Decapsidation of polyoma virus mutants

The decapsidation of several temperature-sensitive mutants of polyoma virus was compared with that of wild-type by infecting mouse embryo cells at low multiplicity simultaneously with 131I-labeled mutant and 125I-labeled wild-type virus. Analysis of the cytoplasm revealed that the noncomplementing mutant ts-3 virutally failed to give rise to the subviral particles which are characteristic of wild-type virus decapsidation. Correspondingly, little or no DNA from the infecting ts-3 particles was found in the cell nucleus. In contrast, the late mutants ts-10 and ts-1260 were decapsidated to a greater extent than wild-type virus. The results obtained with these three mutants can be accounted for on the basis of mutational changes in structural polypeptides, conferring either increased or decreased stability to the viral particle. A decapsidation pattern similar to that of wild-type virus was registered for the early mutant ts-a. These data support a model for Py decapsidation which we have proposed (1).

Reversion of presumptive frameshift and base-pair substitution mutants of Saccharomyces cerevisiae by five carcinogens

Genetic characterization of ad-3B mutants induced in wild-type and UV-sensitive strains has revealed qualitative differences between the spectra of genetic alterations at the molecular level. Ad-3B mutants induced in the two nucleotide excision-repair-deficient strains upr-1 and uvs-2 (Worthy and Epler, 1973) had significantly lower frequencies of nonpolarized complementation patterns and higher frequencies of noncomplementing mutants than ad-3B mutants induced in the wild-type strain in samples induced by either UV, γ-rays, 4NQO or MNNG. In these same samples ad-3B mutants induced in uvs-4, uvs-5 or uvs-6 did not differ significantly from those induced in the wild-type strain. After ICR-170 treatment, ad-3B mutants induced in the UV-sensitive strains did not differ significantly from those induced in wild-type. The comparisons in the present and previous studies demonstrate that the process of mutation-induction in the ad-3 region is under the control of other loci that not only alter mutant recovery quantitatively (de Serres, 1980; Schüpbach and de Serres, 1981; Inoue et al., 1981a, b) but also qualitatively. These data have important implications for comparative chemical mutagenesis, since the spectrum of genetic alterations produced by a given agent can be modified markedly as a result of defects in DNA repair.

Revertants and secondary arom-2 mutants induced in non-complementing mutants in the arom gene cluster of Neurospora crassa.

Extensive genetical and biochemical studies have been performed with revertants and secondary arom-2 mutants induced in two different primary non-complementing mutants which map within the arom gene cluster of Neurospora crassa. These studies indicate that mutant M54 but not M25 can revert by super-suppressor mutations in unlinked genes, thus confirming previous evidence that M54 contains a nonsense codon. At least three new super suppressors of M54 have been detected. All four super suppressors (including one previously detected) when combined with M54 result in high levels of all five of the arom enzymic activities in the form of arom multienzyme complexes very similar to (but not necessarily identical with) that in wild type (WT).-Evidence has also been obtained that the two non-complementing mutants can yield revertants which appear to result from true back mutations and produce arom aggregates essentially indistinguishable from that of WT. In addition, M25, but not M54, when plated on quinic acid yields revertants (secondary mutants) some of which are phenotypically indistinguishable from arom-2 primary mutants and others of which, although also mapping within the arom-2 gene, exhibit unusual properties. Genetic evidence indicates that the M25 secondary mutants are localized within the arom-2 gene, but that they arise from mutational events more complex than ones resulting in single base pair changes in the M25 codon.-The recovery of secondary arom-2 mutants as revertants of non-complementing arom mutants provides strong evidence, independent of earlier recombination data, that non-complementing arom mutants are located within the arom-2 structural gene of the arom gene cluster. In addition, the occurrence and characteristics of these secondary arom-2 mutants provide strong evidence, independent of the results with nonsense suppressors, that the arom gene cluster is transcribed, beginning with the arom-2 gene, as a single polycistronic messenger ribonucleic acid (mRNA) molecule which is subsequently translated into the arom multienzyme complex.

Purification and Properties of a Mutant Bifunctional Enzyme from the HisB Gene of Salmonella typhimurium

Abstract A distinct change occurs in the molecular size of the enzyme, histidinol phosphate phosphatase, specified by nonsense mutants with lesions near the middle of hisB. The hisB gene of Salmonella typhimurium controls two different activities, imidazoleglycerolphosphate dehydratase and histidinol phosphate phosphatase, both of which reside in a single protein. The phosphatase activity from mutants with lesions distal to this transition point appears in or near the void volume of Sephadex G-200, whereas that from mutants with lesions proximal to the transition point produce a polypeptide with little or no activity. One mutant with its lesion near this transition point was chosen for further study. The phosphatase enzyme was purified from strain TR691, a non-complementing UGA hisB278 mutant. Two forms of the enzyme were found by preparative polyacrylamide gel electrophoresis. One form had a molecular weight of 38,000 and was a dimer of 18,000 molecular weight polypeptide chains as shown by electrophoresis in sodium dodecyl sulfate. The other form of the enzyme had a molecular weight of 47,000 and was a dimer of dissimilar chains with molecular weights of 38,000 and 11,500. The pH optima for catalytic activity of both forms was 7.0 in the absence of MnCl2 and 6.5 in the presence of 1.0 mm MnCl2. The Km for histidinol phosphate was 0.83 mm and was unchanged by the addition of Mn2+, although the velocity at the pH optimum increased about 30%. No difference in the Ki between the two forms could be detected towards the competitive inhibitor, histidinol. The mutant enzymes had no imidazoleglycerolphosphate dehydratase activity and did not cleave p-nitrophenylphosphate. It is possible that the 38,000 molecular weight form is produced by limited proteolysis of the 47,000 molecular weight form. Preliminary experiments using trypsin and subtilisin NOVO did not show this conversion; however, the 11,500 molecular weight component appeared to be very sensitive to this treatment.

Genetic analysis of purple adenine ( ad-3) mutants induced by methyl methanesulfonate in Neurospora crassa

The mutagenicity of methyl methanesulfonate (MMS) was studied in a genetically marked two-component heterokaryon of Neurospora crassa. Types of genetic alterations detectable in this system are ( I) point mutations in the ad-3 A and ad-3 B loci; ( 2) multilocus (chromosome) deletions in the ad-3 region, and ( 3) recessive lethal mutations in the whole genome. Study of the inactivation kinetics of the heterokaryotic and homokaryotic conidial fractions has made it possible to distinguish between nuclear and cytoplasmic inactivation. Forward mutations in the ad-3 region induced by MMS in the heterokaryotic fraction of conidia were obtained by a direct method with the following results: ( I) The overall ad-3 forward mutation frequency increases in proportion to the 1.91 power of the concentration of MMS. ( 2) The forward mutation frequency of point mutations at the ad-3 A and ad-3 B loci increases in proportion to the 1.68 power of the concentration. ( 3) The forward mutation frequency of chromosome deletions in the ad-3 region increases more than exponentially with increasing concentrations of MMS. ( 4) After treatment for 300 min with 20 m M MMS, 15.5% of the ad-3 mutations are multilocus deletions. Tests for genotype and allelic complementation of the point mutations showed that ( I) the ratio between ad-3 B and ad-3 A mutants was 1.75, ( 2) 52.1% of the ad-3 B mutants showed allelic complementation, with 39.2% non-polarized and 12.9% polarized complementation patterns and 47.9% noncomplementing mutants, and ( 3) both the ratio between point mutations in the ad-3 A and ad-3 B loci and the spectrum of complementation patterns among the ad-3 B mutants were independent of MMS concentration.

Isolation and preliminary characterization of temperature-sensitive mutants of influenza virus.

Isolation of temperature-sensitive (ts) mutants was attempted from the WSN strain of influenza A virus which was grown and assayed in MDBK cells. After growth of wild-type virus in the presence of 5-fluorouracil, 15 ts mutants were selected for which the ratio of plaquing efficiency at 39.5 C to that at 33 C was 10(-3) or less. In pairwise crosses of ts mutants, recombination and complementation were either very efficient or undetectable. It is suggested, therefore, that the viral genome consists of physically discrete units and recombination occurs as an exchange of these units. All 15 mutants have been assigned with certainty into five recombination groups. Three mutants are suspected to be double mutants. Any two complementing mutants always recombined with each other, and noncomplementing mutants did not recombine. In physiological tests, mutants showed diverse patterns of functional defects at the nonpermissive temperature. However, it was not always possible to correlate these physiological defects with the results of genetic characterization.

Mutagen specificity in Neurospora crassa

Tests for mutagen specificity were made on spontaneous and by X-ray, UV, 2-aminopurine (AP), nitrous-acid (NA), ethyl methanesulfonate (EMS), and mesodiepoxybutane (DEB) induced ad-3 mutants of Neurospora crassa induced in wild-type strains OR 74A and 74-OR23-1A. Differences were found between the percentages of leaky ad-3A mutants and leaky ad-3B mutants as a function of mutagenic origin as well as the ratios of ad-3B mutants to ad-3B mutants. Ad-3B mutants were further characterized by means of tests for allelic complementation with a tester set of 17 strains that define a linear complementation map consisting of 17 complons. Complementation tests were made on a total of 564 ad-3B mutants. Of the 107 different complementation patterns detectable with this tester set, 52 different patterns were detected among the 530 ad-3B mutants that could be mapped. Marked differences were found in the percentages of mutants with nonpolarized or polarized complementation patterns and noncomplementing mutants as well as the average complon coverage for all of the complementing mutants in each sample. Mutants with nonpolarized patterns have an average complon coverage of 4.9 and mutants with polarized patterns of 13.4. Evidence for “hot spotting” was found in the NA- and EMS-induced samples where many of the mutants had the same type of nonpolarized complementation patterns. The highest percentage of mutants with different nonpolarized complementation patterns was found among DEB- and UV-induced mutants. The striking difference between the characteristics of the ad-3 mutants in each of the 7 samples was interpreted as evidence for mutagen specificity.