Blocked Recombination Along the Mating-Type Chromosomes of Neurospora tetrasperma Involves Both Structural Heterozygosity and Autosomal Genes

Abstract

The Neurospora tetrasperma mating-type chromosomes have been shown to be structurally heterozygous by reciprocal introgression of these chromosomes between N. tetrasperma and N. crassa. This structural heterozygosity correlates with both a previously described recombination block and cytologically visible unpaired chromosomes at pachytene. Genes on the autosomes are also implicated in blocking recombination.

THE filamentous fungus Neurospora tetrasperma is self-fertile with a vegetative thallus (mycelium) that is heterokaryotic for mating type (pseudohomothallic). Sexual development in N. tetrasperma has been genetically and developmentally reprogrammed so that two haploid nuclei of opposite mating type (mat A and mat a) are delivered into each of the four ascospores (Dodge 1927; Raju and Perkins 1994). To accomplish this nuclear packaging, crossing over is suppressed in the mating-type bivalent (but not in the autosomes), ensuring that mat A and mat a will segregate at the first division of meiosis. Recombination is blocked for >120 MU, which is ∼90% of this chromosome (linkage group I) and ∼12% of the total genome (Gallegos et al. 2000). The two homologs of the mating-type bivalent are seen to remain unpaired at pachytene over most of its length. Chiasmata near the ends hold the homologs together and ensure proper disjunction at anaphase I. Spindles at the second division are precisely repositioned pairwise to ensure that two nuclei of opposite mating type are enclosed in each ascospore, rendering it self-fertile. Pseudohomothallism, as exemplified in N. tetrasperma, leads to sustained but not absolute inbreeding, with profound consequences for population structure and evolution (Powell et al. 2001).

Because recombination is blocked on the mating-type chromosomes of N. tetrasperma, the regions linked to mat A and mat a have undergone substantial sequence divergence (Merino et al. 1996). Sequences in the region of suppressed crossing over exhibit heteroallelism (heterozygosity), even when sibling nuclei from the same wild-type isolate are compared. In contrast, the autosomes recombine freely and sibling nuclei of opposite mating type from each wild-type isolate exhibit nearly complete homoallelism (homozygosity) along the autosomes. N. tetrasperma mating-type chromosomes resemble the sex chromosomes in animals in failing to recombine over most of their length. The structure, function, and evolution of fungal mating-type loci and chromosomes is an area of current interest (e.g., Fraser et al. 2004; Hood et al. 2004). Three questions arise from these observations:

1. Why has the crossover block come to be extended over most of the chromosome? A recombination block in the short region between mat and centromere would suffice to provide the mechanism for pseudohomothallism by maintaining first-division segregation of mating type and the subsequent packaging of opposite mating-type nuclei into individual ascospores.

2. What are the consequences of an extended recombination block for a haploid organism in the context of free recombination in the autosomes? The mating-type chromosome should be subject to Muller's ratchet, resulting in accumulation of alleles of reduced fitness as expected for a nonrecombining chromosome in an otherwise recombining genome (Muller 1964).

3. What is the cellular basis of blocked recombination?

This study is concerned with the third question.

A likely explanation for the cellular basis of the recombination block would be structural heterozygosity between the mat a and mat A mating-type chromosomes of N. tetrasperma, involving rearrangements such as inversions. However, blocked recombination might be under some other genetically controlled mechanism based on particular gene(s), the mating-type chromosome, and/or the autosomes. To test these alternatives, the mating-type chromosomes were introgressed reciprocally between N. tetrasperma and N. crassa. The interval introgressed (ro-10–un-18) encompasses most of the linkage group (Figure 2, A and B) (Perkins et al. 2001). Classical genetic markers (mutations) were followed in the crosses reported here and were present both within the recombination block (leu-3, mep, mat, the centromere, cyh-1, and al-2) and outside it (ro-10 and un-18) (Gallegos et al. 2000).

Figure 2.

Crossing over between mat A and mat a mating-type chromosomes (linkage group I) in conspecific crosses of N. crassa (A) and N. tetrasperma (B), and after reciprocal introgression from one species into the other (C). (A) N. crassa. Numbers along the linkage group are distances between N. crassa markers, expressed as map units (approximately equal to percentage recombination). Parentheses around ro-10 and cyt-21 indicate equivocal gene order. This map is based on Perkins et al. (2001) and D. J. Jacobson (unpublished results). (B) N. tetrasperma. Recombination rates are expressed as the ratio of the number of recombinants over total progeny tested. Thick lines and ratios in boldface type are data from cross P556 mat A × P581 mat a, except those marked with ^‡, which are results of a selfing cross (P556 mat A × P556 mat a) (Gallegos et al. 2000). The recombination block extends between the markers nit-2 on the left and arg-13 on the right. Tetrad analysis confirmed that 100% of meioses underwent crossing over in the interval between nit-2 and cyt-21 (marked with *) and 86% underwent crossing over between arg-13 and un-18 (marked with ^†). Other ratios are from previous studies of N. tetrasperma (data from D. D. Perkins, unpublished results, and G. S. Saenz and D. O. Natvig, unpublished results). (C) Reciprocal introgression crosses between N. crassa and N. tetrasperma. Crosses are described in Figure 1. First hybrid cross: the fractions indicate the number of crossovers scored over the total f₁ hybrid progeny. Series 1 and 2: the fractions indicate the number of crossovers scored over the total number of progeny in the sixth backcross. Series 3 and 4: the fractions indicate the number of homoallelic recessive progeny over the total progeny scored in the sixth backcross. The percentage of homoallelics is the same as the distance in conventional map units (Howe 1963) (see text).

Nearly complete sterility between N. crassa and N. tetrasperma required the use of a hybrid strain, C4,T4 (FGSC 1778), in the initial bridging cross (Metzenberg and Ahlgren 1969). The N. tetrasperma strains used to construct the C4,T4 hybrid were different from those used in this study. The cross DJ1307, therefore, has been designated the first hybrid cross of the present introgression series (Figure 1). This cross had very low fertility and fecundity. The programmed recombination block of N. tetrasperma broke down upon initial hybridization with N. crassa. Recombination along the mating-type chromosomes was evident even in the small number of viable progeny available (Figure 2C ). Four subsequent series of introgressions were necessary for a completely reciprocal design. Pedigrees and selection strategies are shown in Figure 1. Mating types from N. tetrasperma and N. crassa are designated as mat A^T or mat a^T and mat A^C or mat a^C, respectively. Recombination rates were measured among progeny of the sixth backcross (Figure 2C).

Figure 1.

Flow chart of crosses and selection strategy for reciprocal introgression of mating-type chromosomes between N. crassa and N. tetrasperma. Classical genetic markers along the mating-type chromosome (mutations and mating-type alleles) were followed in the crosses (see Perkins et al. 2001 for explanation of markers). For crosses, haploid parental genotypes are indicated above and below the thick line, which also indicates the zygote genotype. The N. crassa or N. tetrasperma origin of the mat allele is indicated by a superscript (C or T, respectively); mutant alleles for morphological and auxotrophic markers are shown in italics with the corresponding wild-type alleles indicated by +; resistance or susceptibility is indicated by superscript (R or S, respectively) for the methylpurine resistance or cycloheximide resistance-1 loci. The cross number, where appropriate, is indicated by DJxxxx. Strain numbers are given in parentheses to the right of genotypes. FGSC, Fungal Genetics Stock Center; DJ, strain derived from a Jacobson cross. DJ1243-2 is a methylpurine-resistant strain of N. tetrasperma (see Gallegos et al. 2000) and DJ1687-82 is a strain of N. crassa with a novel marker combination developed for this study. The bridging cross (DJ1292) initiated all series. Even though strain C4,T4 is itself a hybrid between N. crassa and N. tetrasperma, the progeny from cross DJ1307 were considered the first hybrid strains (f₁) for series 1, 2, and 4. Series 3 required the opposite mating type of N. crassa (mat a^C) and different markers than were available from DJ1307. These were introduced in crosses DJ1700 and DJ1725, which again employed progeny from the bridging cross (DJ1258). Progeny from DJ1725 were considered the first hybrid strains (f₁) for series 3. Progeny from the sixth backcross generation of each series were analyzed for recombination of the markers indicated (Figure 2).

Recombination in introgression series:

Series 1—the mat a^T chromosome introgressed into N. crassa:

No recombination block was evident after six generations (Figure 2C, series 1). Recombination frequencies along the mating-type chromosomes in this introgression were very similar to those of wild-type N. crassa (compare Figure 2C with 2A). Progeny viability was normal and allele ratios were near 1:1.

Series 2—the mat A^T chromosome introgressed into N. crassa:

In contrast, crossing over remained blocked between mat and al-2 in this opposite orientation (Figure 2C, series 2). The <1% recombination in this interval was reminiscent of wild-type N. tetrasperma. However, even after six generations, crosses of mat A^T × mat a^C showed a high proportion of aborted, unpigmented ascospores. Viability was low among black ascospores, and allele ratios were highly skewed among those that germinated, for example, 14 mat A to 85 mat a progeny. These factors indicated general developmental problems in the crosses, which perhaps confounded measurements of recombination.

Series 3 and 4—the N. crassa mating-type chromosomes introgressed into N. tetrasperma:

By the fourth backcross into N. tetrasperma, the crosses in both series produced normal four-spored asci (Figure 2C, series 3 and 4). Genes centromere-distal to crossovers undergo second-division segregation in one-half of crossover asci. Because of ascus programming in N. tetrasperma, such genes become homoallelic in ascospores, while genes proximal to the crossover remain heteroallelic (Raju and Perkins 1994). Therefore, recombination was scored as the number of homoallelic recessive progeny over the total progeny. This proportion of homoallelic ascospores, as a percentage, is the same as the distance in map units between the centromere and the marker locus (Howe 1963).

Series 3—the mat a^C chromosome introgressed into N. tetrasperma:

Recombination was blocked between leu-3 and cyh-1 (Figure 2C, series 3). However, 14% recombination (3/21 progeny) occurred in the interval between nit-2 and leu-3, an area normally within the recombination block of N. tetrasperma (Figure 2B). The leftmost marked interval in series 3, therefore, was more comparable to the normal 17% recombination in N. crassa (Figure 2A).

Series 4—the mat a^C chromosome introgressed into N. tetrasperma:

In the opposite orientation, the recombination block was also reestablished from mat to al-2 (Figure 2C, series 4). Recombination between the distal markers (inside and outside the recombination block) on both chromosome arms was lower than expected, compared to normal N. tetrasperma (Figure 2B). Cytological observation of pachytene chromosomes in the sixth backcross of mat A^C into N. tetrasperma showed the typically long unpaired central chromosome segment and the short paired terminal segments (N. B. Raju, personal communication), similar to Figure 4 in Gallegos et al. (2000).

Mating-type chromosome structure:

In the N. crassa background, recombination is normal between the mat a^T and mat A^C chromosomes (Figure 2C, series 1). This demonstrates that these two chromosomes are colinear. Normal recombination is defined in N. crassa by mat a^C × mat A^C; therefore, it follows that mat a^T must also be colinear with mat a^C (although, obviously, recombination cannot be tested by directly crossing mat a^T × mat a^C strains of the same mating type).

In the N. crassa background, recombination is blocked in the opposite orientation—between the mat A^T and mat a^C chromosomes (Figure 2C, series 2). This strongly suggests that mat A^T is structurally different from both mat A^C and mat a^C chromosomes. This was not seen in the earlier backcrosses because of low fertility and fecundity (data not shown).

In the N. tetrasperma background, recombination is blocked from leu-3 to cyh-1 in mat a^C × mat A^T, suggesting that these chromosomes are structurally different (Figure 2C, series 3). Recombination was blocked between mat and al-2 even in the first hybrid cross (DJ1725, 0/27 recombinants). However, the mat a^C and mat A^T chromosomes did recombine in the nit-2–leu-3 interval, a region still within the N. tetrasperma recombination block, contrary to expectations when structural differences extend left of mating type.

Other factors controlling the recombination block in N. tetrasperma:

Introgression series 4 further indicates that a structurally unique mat A^T chromosome cannot be the only explanation for the recombination block in N. tetrasperma. If blocked recombination was due solely to differences in chromosome structure, the expectation would be that the colinear mating-type chromosomes would recombine freely in the N. tetrasperma background, as was demonstrated for mat A^C and mat a^T in the N. crassa background (series 1). The original hybrid cross of mat A^C × mat a^T (DJ1307) did show recombination (Figure 2C, F₁ hybrid progeny), but during the six backcrosses of mat A^C to N. tetrasperma the recombination block was reestablished despite the absence of a mat A^T chromosome (Figure 2C, series 4).

These results show that the mating-type chromosomes of N. tetrasperma are structurally different from each other. Structural heterozygosity could, of itself, cause the recombination block between those chromosomes. However, complete genetic control of the recombination block is likely more complicated than simple differences in chromosome structure. Blocked recombination was reestablished during backcrossing to N. tetrasperma, even when the mating-type chromosomes were colinear (mat A^C × mat a^T). Genetic background must, therefore, have a critical role in this qualitative regulation of recombination. The data also suggest that genes affecting recombination are present both on the mating-type chromosomes and on the autosomes.

Involvement of autosomal genes raises the possibility that differences in mating-type chromosome structure in N. tetrasperma may have occurred concomitantly with blocked recombination, rather than causing it. Determining the evolutionary history of the mating-type chromosome structures would address issues of cause and effect. When did the structure of mat A^T and mat a^T diverge? Was it coincident with the origin of pseudohomothallism or subsequent to it? The genetic data reported here leave these and other questions unanswered. An assembled genome sequence of N. tetrasperma, which includes both mating types, will likely provide the definitive answers by enabling both mat A^T and mat a^T chromosome structures to be compared with each other and with those of N. crassa. This might also explain why the recombination block is much larger than necessary for pseudohomothallism. Continued genetic work will then be needed to identify the genes that affect recombination in the N. tetrasperma background.