Reciprocal Crossover Research Articles

High-frequency homologous recombination between duplicate chromosomal immunoglobulin mu heavy-chain constant regions.

Homologous recombination was used in a previous study to correct a 2-base-pair deletion in the third constant domain (Cmu3) of the haploid chromosomal mu gene in a mutant hybridoma cell line by transfer of a pSV2neo vector bearing a subfragment of the normal Cmu region (M.D. Baker, N. Pennell, L. Bosnoyan, and M.J. Shulman, Proc. Natl. Acad. Sci. USA 85:6432-6436, 1988). In these experiments, both gene replacement and single reciprocal crossover events were found to restore normal, cytolytic 2,4,6-trinitrophenyl-specific immunoglobulin M production to the mutant cells. In the cases of single reciprocal recombination, the structure of the recombinant mu gene is such that the normal Cmu region, in its correct position 3' of the expressed 2,4,6-trinitrophenyl-specific heavy-chain variable region, is separated from the mutant Cmu region by the integrated vector sequences. I report here that homologous recombination occurs with high frequency between the duplicate Cmu regions in mitotically growing hybridoma cells. The homologous recombination events were easily detected since they generated hybridomas that were phenotypically different from the parental cells. Analysis of the recombinant cells suggests that gene conversion is the most frequent event, occurring between 60 and 73% of the time. The remaining events consisted of single reciprocal crossovers. Intrachromatid double reciprocal recombination was not detected. The high frequency of recombination, the ability to isolate and analyze the participants in the recombination reactions, and the capacity to generate specific modifications in the immunoglobulin Cmu regions by gene targeting suggest that this system will be useful for studying mammalian chromosomal homologous recombination. Moreover, the ability to specifically modify the chromosomal immunoglobulin genes by homologous recombination should facilitate studies of immunoglobulin gene regulation and expression and provide a more convenient of engineering specifically modified antibody.

High-Frequency Homologous Recombination between Duplicate Chromosomal Immunoglobulin μ Heavy-Chain Constant Regions

Homologous recombination was used in a previous study to correct a 2-base-pair deletion in the third constant domain (Cmu3) of the haploid chromosomal mu gene in a mutant hybridoma cell line by transfer of a pSV2neo vector bearing a subfragment of the normal Cmu region (M.D. Baker, N. Pennell, L. Bosnoyan, and M.J. Shulman, Proc. Natl. Acad. Sci. USA 85:6432-6436, 1988). In these experiments, both gene replacement and single reciprocal crossover events were found to restore normal, cytolytic 2,4,6-trinitrophenyl-specific immunoglobulin M production to the mutant cells. In the cases of single reciprocal recombination, the structure of the recombinant mu gene is such that the normal Cmu region, in its correct position 3' of the expressed 2,4,6-trinitrophenyl-specific heavy-chain variable region, is separated from the mutant Cmu region by the integrated vector sequences. I report here that homologous recombination occurs with high frequency between the duplicate Cmu regions in mitotically growing hybridoma cells. The homologous recombination events were easily detected since they generated hybridomas that were phenotypically different from the parental cells. Analysis of the recombinant cells suggests that gene conversion is the most frequent event, occurring between 60 and 73% of the time. The remaining events consisted of single reciprocal crossovers. Intrachromatid double reciprocal recombination was not detected. The high frequency of recombination, the ability to isolate and analyze the participants in the recombination reactions, and the capacity to generate specific modifications in the immunoglobulin Cmu regions by gene targeting suggest that this system will be useful for studying mammalian chromosomal homologous recombination. Moreover, the ability to specifically modify the chromosomal immunoglobulin genes by homologous recombination should facilitate studies of immunoglobulin gene regulation and expression and provide a more convenient of engineering specifically modified antibody.

Molecular analysis of recombination events in Drosophila.

The locations of crossover junctions and gene conversion tracts, isolated in the rosy gene of Drosophila melanogaster, were determined using DNA sequencing and denaturing gradient gel electrophoresis. Frequent DNA sequence polymorphisms between the parental genes served as unselected genetic markers. All conversion tracts were continuous, and half of the reciprocal crossover events had conversion tracts at the crossover junction. These experiments have also identified the sequence polymorphisms responsible for altered gene expression in two naturally occurring rosy variants.

MMS-induced primary aneuploidy and other genotoxic effects in mitotic cells of Aspergillus

The possibility of more than 1 target for genotoxic effects of methyl methanesulphonate (MMS) was investigated, using mitotic test systems of the fungus Aspergillus. Haploid and diploid strains were exposed, either as dormant conidia or during mitosis, and analysed for induced aneuploidy and effects on genetic segregation. MMS treatment of haploid strains resulted in dose-dependent increases of stable mutants with altered phenotypes and semi-stable unbalanced aberrations (presumably duplications). In addition, but only in dividing cells, MMS induced unstable aneuploids. These mostly were hyperhaploid with few extra chromosomes and could be identified by comparison with standard disomic phenotypes. When well-marked diploids were treated 3 types of effect could be distinguished, using genetic and phenotypic criteria: (1) Clastogenic and mutagenic effects which caused dose-dependent increases of partial aneuploids with various abnormal phenotypes. These showed secondary genetic segregation of all types and produced euploid normal sectors by eliminating damaged chromosome segments. In addition, but only in dividing nuclei, MMS induced 2 types of segregation: (2) Reciprocal crossing-over at high frequency, recognisable as half or quarter colonies of mutant colour and in some cases as ‘twin spots’ (i.e., complementary pairs); (3) Trisomics and other aneuploids which showed characteristic phenotypes and expected segregation of markers: the types recovered indicate random malsegregation of chromosomes (occasional deviations resulted from coincidence with induced crossing-over). These results suggest that MMS may have 2 (or more) targets for genotoxic effects: DNA, as evident from induced mutations and aberrations, and from induced recombination in dividing cells; some non-DNA target (nucleotide or protein) essential for nuclear division and susceptible to alkylation, resulting in malsegregation and primary aneuploidy.

A reexamination of the role of the RAD52 gene in spontaneous mitotic recombination

The RAD52 gene is required for much of the recombination that occurs in Saccharomyces cerevisiae. One of the two commonly utilized mutant alleles, rad52-2, increases rather than reduces mitotic recombination, yet in other respects appears to be a typical rad52 mutant allele. This raises the question as to whether RAD52 is really necessary for mitotic recombination. Analysis of a deletion/insertion allele created in vitro indicates that the null mutant phenotype is indeed a deficiency in mitotic recombination, especially in gene conversion. The data also indicate that RAD52 is required for crossing-over between at least some chromosomes. Finally, examination of the behavior of a replicating plasmid in rad52-1 strains indicates that the frequency of plasmid integration is substantially reduced from that in wild type, a conclusion consistent with a role for RAD52 in reciprocal crossing-over. Analysis of recombinants arising in rad52-2 strains suggests that this allele may result in the increased activity of a RAD52-independent recombinational pathway.

Apparent gene conversion in an Escherichia coli rec+ strain is explained by multiple rounds of reciprocal crossing-over.

Gene conversion, the non-reciprocal transfer of sequence information between homologous DNA sequences, has been reported in lower eukaryotes, mammals and in Escherichia coli. In an E. coli rec+ strain, we established a plasmid carrying two different deleted neo genes (neoDL and neoDR) in an inverted orientation and then selected for homologous recombination events that had reconstructed an intact neo+ gene. We found some plasmids that had apparently experienced intramolecular gene conversion. Further evidence, however, suggests that they are products of multiple rounds of reciprocal crossing-over, apparently involving two plasmid molecules. First, most of the Neo+ clones contained multiple types of Neo+ plasmids, although the frequency of producing the neo+ clones was low. Second, all the neo+ clones also contained, as a minority, one particular form of dimer, which can be formed by reciprocal crossing-over between neoDL of one plasmid molecule and neoDR of another plasmid molecule. Third, in reconstruction experiments, we cloned and purified this dimer and transferred it back into the rec+ cells. The dimer gave rise to clones containing multiple types of neo+ recombinant monomers, including those apparent gene conversion types, and containing only few molecules of this dimer plasmid.

Nitrogenase promoter-lacZ fusion studies of essential nitrogen fixation genes in Bradyrhizobium japonicum I110.

DNA fragments containing either the nifD or nifH promoter and 5' structural gene sequences from Bradyrhizobium japonicum I110 were fused in frame to the lacZ gene. Stable integration of these nif promoter-lacZ fusions by homologous double reciprocal crossover into a symbiotically nonessential region of the B. japonicum chromosome provided an easy assay for the effects of potential nif regulatory mutants. The level of beta-galactosidase activity expressed from these two nif promoter-lacZ fusions was assayed in bacteroids of B. japonicum I110 wild type and Fix mutants generated by transposon Tn5 mutagenesis and identified in the accompanying paper. No nif-positive regulatory mutants were identified from among an array of Fix- mutants in which Tn5 was inserted 9 kilobase pairs upstream of the nifDK operon and within the 18-kilobase-pair region separating the nifDK and nifH operons. This result indicates that there are no genes in these regions involved in the regulation of nitrogenase structural gene expression. Interestingly, the level of beta-galactosidase activity expressed from the nifH promoter was twice that expressed from the nifD promoter, suggesting that the normal cellular level of the nifH gene product in bacteroids is in a 2:1 ratio with the nifD gene product instead of in the 1:1 stoichiometry of the nitrogenase enzyme complex.

Role of homology and pathway specificity for recombination between plasmids and bacteriophage lambda.

To determine the minimum amount of homology required for efficient recombination in Escherichia coli, we measured recombination frequencies between bacteriophage lambda and pBR322 derivatives containing lambda DNA fragments of various sizes by assaying for phages that could transduce the bla and ori genes of pBR322. Efficient recombination required about 40 bp of homology; increases in homology above 40 bp resulted in proportionate increases in recombination, while decreases below 40 bp resulted in precipitous decreases in recombination. The recA+ gene stimulated recombination over the entire range of homologies tested. Restriction enzyme digests of several recombinant DNA molecules indicated that they contained the complete plasmid DNA inserted in the lambda genome as expected for a reciprocal crossover. Analysis of recombination frequencies in different recombination-deficient mutant strains indicated that the formation of lambda-plasmid cointegrates by homologous recombination proceeded predominantly by the RecBC pathway and very inefficiently, if at all, by the RecE and RecF pathways.

Tests which distinguish induced crossing-over and aneuploidy from secondary segregation in Aspergillus treated with chloral hydrate or γ-rays

A system of tests with the ascomycete Aspergillus nidulans was devised that can detect 3 primary effects of genotoxic agents: (1) increases in mitotic crossing-over; (2) induced aneuploidy; and (3) clastogenic effects which cause chromosomal imbalance. Conidia of a new diploid tester strain, heterozygous for 4 recessive markers which alter conidial color, are treated and plated onto nonselective media. In cases of induced crossing-over, large color segments are found in normal green colonies, frequently adjacent to reciprocal twin segments. In contrast, both malsegregation and chromosome breakage produce unbalanced types which grow poorly and segregate further. Cases with yellow segregants are replated and their secondary diploid sectors tested for markers which are located on both chromosome arms in coupling with yA. Induced aneuploidy can be distinguished from chromosome breakage by the pattern of marker segregation. Any aneuploid type will produce euploid sectors solely by segregation of whole chromosomes; trisomic colonies ( yA/yA/+) will show 1:2 ratios for yellow (homozygous yA) to parental green ( yA/+) sectors and have characteristic phenotypes. Other induced unbalanced types, if heterozygous for deletions or aberrations, may produce yellow diploid sectors by secondary crossing-over as well as by nondisjunction and such cases show unique patterns of genetic segregation and non-predictable phenotypes. As a complementary test, haploid strains are treated, and induced abnormally growing types are replated and classified by phenotype. Aneuploids are unstable and produce many normal sectors, and some of these disomic or trisomic types can be visually identified. In contrast, induced deletions are lethal, and duplications or ‘morphological’ mutants show much more stable abnormal phenotypes. This test system was used to characterize the primary effects of γ-rays and chloral hydrate. Results and evidence were as follows: (1) A dose-dependent increase of color segments resulting from reciprocal crossing-over was found after treatment of dividing nuclei in germinating diploid conidia with γ-rays, but not with chloral hydrate. (2) Highly aneuploid and polyploid types were induced in diploid and haploid germinating conidia by chloral but not to any significant extent by γ-rays. (3) γ-Rays caused a dose-dependent increase of abnormally growing colonies when dormant or germinating diploid conidia were treated. These colonies produced secondary euploid sectors by spontaneous nondisjunction and frequently also by crossing-over, which provided evidence for induced semidominant and recessive lethal mutations of many types. γ-Irradiation of haploid conidia induced mainly abnormal colonies with stable or near-stable phenotypes (i.e., mutations and aberrations) and only very few unstable disomics were found.

Staphylococcal enterotoxin A is encoded by phage.

The gene for staphylococcal enterotoxin A (entA), in two wild-type strains, is carried by related temperate bacteriophages. Hybridization analysis of DNA from entA-converting phage PS42-D and its bacterial host suggests that this phage integrates into the bacterial chromosome by circularization and reciprocal crossover (the Campbell model) and that the entA gene is located near the phage attachment site. DNA from three of eight staphylococcal strains that did not produce enterotoxin A and seven wild-type enterotoxin A-producing (EntA+) strains had extensive homology to the entA-converting phage PS42-D DNA, although there was a high degree of restriction-fragment length polymorphisms. At least one EntA+ strain did not produce detectable viable phage after induction. These data indicate that a polymorphic family of Staphylococcus aureus phages (some of which may be defective) can carry the entA gene.

Genetic effects of 5-azacytidine in Saccharomyces cerevisiae

The base analog 5-azacytidine induced a variety of genetic and epigenetic effects in different organisms. It was tested in two diploid strains of the yeast Saccharomyces cerevisiae to study the induction of point mutation, mitotic reciprocal crossing-over, mitotic gene conversion (strain D7) and mitotic aneuploidy (strain D61.M). It was used on cells growing in its presence for 4–5 generations. There was a strong induction of both types of mitotic recombination and point mutation. However, there was no induction of mitotic chromosomal malsegregation under the same conditions.

Yeast transformation: a model system for the study of recombination.

DNA molecules that integrate into yeast chromosomes during yeast transformation do so by homologous recombination. We have studied the way in which circular and linear molecules recombine with homologous chromosomal sequences. We show that DNA ends are highly recombinogenic and interact directly with homologous sequences. Circular hybrid plasmids can integrate by a single reciprocal crossover, but only at a low frequency. Restriction enzyme digestion within a region homologous to yeast chromosomal DNA greatly enhances the efficiency of integration. Furthermore, if two restriction cuts are made within the same homologous sequence, thereby removing an internal segment of DNA, the resulting deleted-linear molecules are still able to transform at a high frequency. Surprisingly, the integration of these gapped-linear molecules results in replacement of the missing segment using chromosomal information. The final structure is identical to that obtained from integration of a circular molecule. The integration of linear and gapped-linear molecules, but not of circular molecules, is blocked by the rad52-1 mutation. Consideration of models for plasmid integration and gene conversion suggests that RAD52 may be involved in the DNA repair synthesis required for these processes. Implications of this work for the isolation of integrative transformants, fine-structure mapping, and the cloning of mutations are discussed.

Chromosome replacement in mixed populations of compound-2L; free-2R and standard strains of Drosophila melanogaster

Crosses between compound-2L; free-2R (free-arm) and standard strains of Drosophila melanogaster produce two classes of inviable aneuploid hybrids in equal proportions: monosomic 2L and trisomic 2L. The lethal period for monosomics occurs during embryogenesis while the trisomics survive to late pupae. Since the hybrids are inviable, standard and free-arm strains within a mixed population remain genetically isolated. Genetic isolation in the absence of mating isolation offers an extreme example of unstable equilibrium. Relative fitness data indicate that an unstable equilibrium will be established between free-arm and standard strains at a ratio of 2.5∶1. Indeed, in three cage experiments established at initial ratios of 3∶1, free arms to standards, laboratory (Oregon R) or native (Okanagan S) standard strains were completely replaced in approximately 100 days by free-arm lines derived either from laboratory or from native genetic background. In contrast, one cage established at an initial ratio of 4∶1 failed to show replacement and for 92 days remained at approximately the initial ratio. Subsequent genetic analysis of flies removed from this cage identified the presence of an anomalous strain through which genetic information was transferred reciprocally between the free-arm and standard lines. The second chromosomes carried by this strain consisted of a free-2R and a standard second on the right arm of which was attached a duplication for all of 2L. While the origin of the 2L·2R+2L chromosome was uncertain, genetic and cytological examinations revealed that it represented the reciprocal crossover product expected from an exchange that generated a F(2R). Additional crosses disclosed that the transmission frequency of the asymmetrical pair of second chromosomes, as well as their right-arm crossover products, was disproportionately in favor of the short arm. Since unequal transmission was invariably greater from female parents, this phenomenon was viewed as further evidence in support of the drag hypothesis.

Lateral element cross connections of the synaptonemal complex and their relationship to chiasmata in rat spermatocytes.

Observations are presented in support of the hypothesis that at meiotic prophase a reciprocal crossover is accompanied by a crossover of the lateral elements of the synaptonemal complex, SC (Moens, 1974). Rat spermatocyte nuclei in developmental stage VII (Clermont, 1972) of the seminiferous epithelium cycle, but not pachytene nuclei in stages I to VI, were found to have SC modifications in the form of a cross connection between the lateral elements. In structure these crossover elements, CO elements, resemble the lateral element. It is found in a variety of positions, usually more or less perpendicular to the SC but also slanted or parallel along the central element or detached from the SC. Reconstructions of entire nuclei indicate an average of one such CO element per SC and a nonrandom distribution of CO elements among the SCs. Because the crossing-over of lateral elements produces a 180 degree twist or removes a 180 degree twist, the pattern of SC coiling was examined. Coiling starts in early pachytene prior to CO element formation. At stage VII one nucleus had a total of 78 coils, all counter clockwise, and another nucleus had 97 such 180 degree coils. It is noted that if SC coils are associated with the process of crossing-over, then the regulation of crossover distribution such as chiasma position interference has an explanation in the structure and behavior of the SC.

Mitotic crossing over and nondisjunction in translocation heterozygotes of Aspergillus.

To analyze mitotic recombination in translocation heterozygotes of A. nidulans two sets of well-marked diploids were constructed, homo- or heterozygous for the reciprocal translocations T1 (IL;VIIR) or T2 (IL;VIIR) and heterozygous for selective markers on IL. It was found that from all translocation heterozygotes some of the expected mitotic crossover types could be selected. Such crossovers are monosomic for one translocated segment and trisomic for the other and recovery depends on the relative viabilities of these unbalanced types. The obtained segregants show characteristically reduced growth rates and conidiation dependent on sizes and types of mono- and trisomic segments, and all spontaneously produce normal diploid sectors. Such secondary diploid types either arose in one step of compensating crossing over in the other involved arm, or--more conspicously--in two steps of non-disjunction via a trisomic intermediate.--In both of the analyzed translocations the segments translocated to IL were extremely long, while those translocatated from IL were relatively short. The break in I for T1 (I;VII) was located distal to the main selective marker in IL, while that of T2 (1;VIII) had been mapped proximal but closely linked to it. Therefore, as expected, the selected primary crossover from the two diploids with T2 (I;VIII) in coupling or in repulsion to the selective marker, showed the same chromosomal imbalance and poor growth. These could however be distinguished visually because they spontaneously produced different trisomic intermediates in the next step, in accordance with the different arrangement of the aneuploid segments. On the other hand, from diploids heterozygous for T1 (I;VII) mitotic crossovers could only be selected when the selective markers were in coupling with the translocation; these crossovers were relatively well-growing and produced frequent secondary segregants of the expected trisomic, 2n + VII, type. For both translocations it was impossible to recover the reciprocal crossover types (which would be trisomic for the distal segments of I and monosomic for most of groups VII or VIII) presumably because these were too inviable to form conidia.--In addition to the selected segregants of expected types a variety of unexpected ones were isolated. The conditions of selection used favour visual detection of aneuploid types, even if these produce only a few conidial heads and are not at a selective advantage. For T2 (I;VIII) these "non-selected" unbalanced segregants were mainly "reciprocal" crossovers of the same phenotype and imbalance as the selected ones. For T1 (I;VII) two quite different types were obtained, both possibly originating with loss of the small VII-I translocation chromosome. One was isolated when the selective marker in repulsion to T1 (I;VII) was used and, without being homo- or hemizygous for the selective marker, it produced stable sectors homozygous for this marker...

Irradiation uv des cellules polaires de l'oeuf chez Drosophila melanogaster III. Étude de la recombinaison mitotique induite chez le mâle

The UV induction of genetic recombination in Drosophila males was studied by irradiating polar cap cells in eggs with doses ranging from 750 to 1200 erg/mm 2. About 1000 males (heterozygous for genes on the second and third chromosomes) derived from the irradiated eggs were screened and 33 of these males showed certain disturbances in the segregation of the marker genes employed in the study. The data permit the following conclusions to be drawn: (1) Clusters of induced mitotic crossovers appear very early; they are sometimes large in size and in fact all the descendants of a male may be exclusively of the crossover class. In most instances, however, the cluster size is greatly reduced. (2) The clusters include either two reciprocal crossover classes or only one or two nonreciprocal classes; in particular, when double crossovers are observed, there is no reciprocity. Several hypotheses to explain the origin of crossovers and the absence of reciprocity are presented and discussed; it appears that the latter is not due to the elimination of one of the crossovers. It is suggested that UV irradiaton induces chromosomal lesions which, at the time of DNA replication, lead to mitotic recombination. The latter could originate through triplication of a small chromosomal segment.

A MODEL FOR SYNAPSIS IN FEMALES HETEROZYGOUS FOR THE CHERRY SUB-LOCUS DUPLICATION IN DROSOPHILA.

IN Drosophila the white-eye locus consists of five subloci in the fourth of which JUDD (1964) has reported a duplication (herein indicated as 4 4). The allele cherry ( wCh) is at this sublocus. The duplication ordinarily carries no mutant allele and by itself causes no change in eye color. It is detectible by virtue of the fact that it enhances zeste (z, an eye color mutation at the left end of the X). The identification and location of the duplication involved the employment of W“ (apricot at sublocus 3) and wrh (cherry, at sublocus 4) as well as zeste and outside markers close to the white locus. A testcross of a female heterozygous for the duplication in one X (red in expression) and w‘ wch in the other (a combination which is white) yields F, females which are predominantly either red or white. However, as a result of crossing over between wa and wCh, JUDD occasionally detected F2 females that were either apricot or cherry. These are referred to as “exceptions.” Some of the apricot exceptions had the duplication (five in a total of about 71,000 females) and some did not (21 in the total); none of the cherry exceptions (six in the total) had it. If we show all the elements at the white locus (1 to 5), then the above F2 exceptions would be as listed in Table 1 (only the X derived from the female parent is shown, the X from the male is wa wCh). In order to account for the above results, JUDD assumes that pairing in the heterozygous female may be of two kinds, as shown in Figure 1 (in which the arrow indicates the crossover point). Crossing over between Wa and wCh in type I pairing would give apricot with the duplication ( wa 4 4), and in type 11, apricot with no duplication ( wa 4). The reciprocal crossover products would be formed in each case. In type I pairing they would be 3 wch, the cherry exceptions. In type 11, they would be 3 4 wCh. This

THE EFFECT OF ULTRAVIOLET LIGHT ON RECOMBINATION IN YEAST.

tion, diploid segregants which are homozygous for one or more of these genes. PONTECORVO, ROPER, HEMMONS, MACDONALD and BUFTON (1953) and PONTECORVO and =FER (1958) have shown that such segregations of homozygous diploids in Aspergillus niduluns are the results of either of two mechanisms: mitotic crossing over, as first discovered in Drosophilu melunoguster by STERN (1936) , in which crossing over occurs at the four-strand stage as in meiosis but the centromeres divide as in mitosis; or nondisjunction of sister chromatids at mitosis without crossing over. In untreated A. niduluns the frequency of mitotic crossing over is less than meiotic crossing over by a factor of about lo4 (PRITCHARD 1960). The frequency of mitotic nondisjunction is even rarer than the frequency of mitotic crossing over (but the relative frequencies of these two mechanisms depends upon the selective method used [PONTECORVO et al. 19531). Various treatments given to induce somatic segregation have increased the frequency of segregation. FRATELLO, MORPURGO and SERMONTI (1960) found that in Aspergillus nidulans increasing doses of ultraviolet light (UV) and nitrogen mustard increased the frequency of mitotic nondisjunction but had no effect on mitotic crossing over. However, HOLLIDAY (1961) working with Ustilugo maydis obtained a 90-fold increase over the spontaneous rate of mitotic crossing over with UV treatment. A further important difference in mitotic crossing-over in different organisms has been found when intragenic recombination is analysed. PRITCHARD ( 1960), working with the ad-20 gene in A. niduluns, found the reciprocal products, the 4- allele and the double mutant, in the same diploid nucleus in the proportion expected from the mechanism of reciprocal mitotic crossing over. ROMAN and JACOB (1958), on the other hand, working with a diploid heteroallelic for is-1 in Saccharomyces, did not find the reciprocal double mutant and concluded that the recombination was not the result of a reciprocal crossover (see also KAKAR 1963). Spontaneous mitotic crossing over in Aspergillus nidulans and spontaneous and induced mitotic crossing over in Ustilugo maydis are so rare that in general only segregants for homozygosis in one arm of one chromosome are found in any one nucleus. In Saccharomyces cerevisiue, on the other hand, a high frequency of simultaneous homozygosis of unlinked genes on different chromosomes after UV treatment was described by WILKIE and HAWTHORNE (1961 ) . These results

THE CORRELATION EFFECT FOR A HISTIDINE LOCUS OF NEUROSPORA CRASSA

OR different organisms rare wild type segregants have been observed in crosses F involving independently isolated mutants which all show the same phenotypic effect. In all cases in which markers on one or both sides of these mutational sites were present the segregants which are wild type with respect to the isolocaP markers are associated with outside marker combinations which in the crossing over theory would be assigned to 1, 2, and even 3 crossovers between and outside the isolocal markers. The correlation between these “recombinations” is so high that the occurrence of wild type segregants cannot be explained by statistically independent crossovers. The first report of this kind was given by DEMEREC (1928) who described the phenomenon as reverse mutation associated with crossing over in heterozygotes of Drosophila virilis. That neither statistically independent crossovers nor clusters of (reciprocal) crossovers can explain the results has most clearly been shown for Neurospora crassa by M. B. MITCHELL (1955 a,b), who found that of four asci containing a wild type segregant for the isolocal markers none contained the reciprocal double mutant. Further results have been published for Neurospora by GILES (1955), ST. LAWRENCE (1956), and DE SERRES (1956), for Aspergillus nidulans by PRITCHARD (1955) and for phages by STREISINGER and FRANKLIN (1956). Irregular asci have also been found in yeast by LINDEGREN (1955) and ROMAN (1956). In this situation several points of view can be or have been suggested. Agreement among all workers in the field seems now to exist for the statement that there occur regions of “intimate” (or “effective”) pairing statistically distributed along the chromosome. In these regions events take place which result in the irregular segregation for the isolocal markers and the correlated appearance of the described 1, 2 and 3 fold “recombinants”. Concerning the events within sucli a region of intimate pairing two different attitudes to the problem can be stated.