Escape From Repeat-Induced Point Mutation of a Gene-Sized Duplication in Neurospora crassa Crosses That Are Heterozygous for a Larger Chromosome Segment Duplication

Corresponding author: Durgadas P. Kasbekar, Centre for Cellular and Molecular Biology, Uppal Rd., Hyderabad 500 007, India., kas{at}ccmb.ap.nic.in (E-mail)

Communicating editor: J. J. LOROS

	ABSTRACT

TOP ABSTRACT MATERIALS AND METHODS RESULTS DISCUSSION LITERATURE CITED

In Neurospora crassa the ability of an ectopic gene-sized duplication to induce repeat-induced point mutation (RIP) in its target gene was suppressed in crosses that were heterozygous for another larger chromosome segment duplication. Specifically, the frequency of RIP in the erg-3 gene due to a 1.3-kb duplication was reduced if the chromosome segment duplications Dp(IIIR > [I;II]) AR17, Dp(VIR > IIIR) OY329, or Dp(IVR > VII) S1229 were present in either the same or the other parental nucleus of the premeiotic dikaryon. We suggest that the larger duplications act as sinks to titrate the RIP machinery away from the smaller duplication. In contrast, RIP efficiency was relatively unaffected in comparably unproductive interspecies crosses with N. intermedia and N. tetrasperma. These findings offer a novel explanation for the observed persistence of the transposable element Tad in only a subset of Neurospora strains.

REPEAT-INDUCED point mutation (RIP) is a unique mutational process that occurs in the sexual cycle of Neurospora crassa during the dikaryotic stage between fertilization and karyogamy. As a result of RIP, duplicated DNA sequences in the otherwise haploid nuclei suffer multiple G:C to A:T transition mutations and methylation of many of the remaining cytosine residues (for reviews see SELKER 1990 ; IRELAN and SELKER 1996 ). RIP has been used to ascertain the null phenotype of cloned genes. In these studies, the duplications are produced by transformation of the cloned DNA and its insertion into ectopic chromosomal locations. Such duplications are typically only a few kilobases in size. But RIP can also occur in much larger duplications that are obtainable as segregants from crosses heterozygous for translocation chromosomes (PERKINS et al. 1997 ). Crosses heterozygous for large chromosome segment duplications (segmental aneuploidy) are characteristically barren; i.e., only a few exceptional asci produce a few viable ascospores. The efficiency of RIP appeared to be reduced in large duplications and when it did occur, the mutagenesis and methylation seemed milder than that typically induced by gene-sized duplications (PERKINS et al. 1997 ). Since both large (e.g., >100 kb) and small (e.g., <10 kb) duplications can induce RIP and serve as its substrates, it was of interest to determine whether the ability of a small duplication to induce RIP in its target gene was affected by the presence of a larger chromosome segment duplication in the same cross. In this article we examine whether induction of RIP in the erg-3 gene by an ectopically integrated 1.3-kb fragment of erg-3, designated Dp 1.3^ec, is affected in crosses that are also heterozygous for the much larger chromosome segment duplications Dp(IIIR > [IR; IIR]) AR17, Dp(VIR > IIIR) OY329, and Dp(IVR > VII) S1229.

The erg-3 gene is located in LGIII and encodes the ergosterol biosynthetic enzyme sterol C-14 reductase (ELLIS et al. 1991 ; PAPAVINASASUNDARAM and KASBEKAR 1994 ). RIP-induced null mutants of erg-3 are viable but have altered sensitivities to isoflavonoids and to the steroidal glycoside -tomatine (SENGUPTA et al. 1995 ; PRAKASH et al. 1999 ). Whereas the wild type is resistant to isoflavonoids and sensitive to tomatine, erg-3 mutants are resistant to tomatine and sensitive to isoflavonoids. Additionally, the colonies generated from erg-3 mutant ascospores exhibit a characteristic slow growth morphology on Vogel's-sorbose agar medium, thereby making them easy to score by mere inspection of plates under a dissection microscope (NOUBISSI et al. 2000 ).

	MATERIALS AND METHODS

TOP ABSTRACT MATERIALS AND METHODS RESULTS DISCUSSION LITERATURE CITED

Strains:
The wild-type N. crassa strains 74-OR23-1 matA [Fungal Genetics Stock Center (FGSC) no. 987] and OR8-1 mata (FGSC no. 988); the mutant strains erg-3 mata (FGSC no. 2725), dow erg-3 matA (FGSC no. 7243), col-18 matA (FGSC no. 8283), and col-18 mata (FGSC no. 8284); the translocation strains T(VIR > IIIR) OY329, matA (FGSC no. 3670) and T(VIR > IIIR) OY329, mata (FGSC no. 3671); and the duplication strains Dp(IVR > VII) S1229 matA (FGSC no. 264) and Dp(IVR > VII) S1229 mata (FGSC no. 265) were obtained from the FGSC, University of Kansas Medical Center, Kansas City, KS 66103.

T(VIR > IIIR) OY329 [also referred to as T(OY329)] is an insertional translocation in which a segment of VIR is inserted into IIIR; the translocated segment includes the wild-type allele of the VIR marker colonial-18 (col-18; PERKINS et al. 1997 ). The viable progeny from crosses between a T(OY329) strain and a normal sequence strain include those that contain a duplication of the VIR segment, designated Dp(VIR > IIIR) OY329. Dp(VIR > IIIR) OY329 strains are stably barren in crosses.

The translocation strain T(IIIR >[I;II])AR17, mata (Perkins lab stock no. xx-366 = FGSC no. 1463) was provided by David D. Perkins (Stanford University). This strain is a complex insertional translocation involving IR, IIR, and IIIR. One-third of the viable progeny from crosses of this strain with a normal sequence strain contain a duplication of a distal IIIR segment, designated Dp(IIIR >[I;II])AR17. The duplication covers the IIIR marker downy (dow) but not erg-3 (PERKINS 1997 ). (There is an error in the literature; Fig 2 of PERKINS et al. 1997 mistakenly suggests that the duplication covers erg-3.) The dow and erg-3 loci are separated by 10% crossover distance (PERKINS et al. 1982 ). Although Dp(IIIR >[I;II])AR17 strains are stably barren in crosses they do nevertheless produce a sufficient number of viable ascospores for meaningful analysis (PERKINS et al. 1997 ).

View larger version (20K): In this window In a new window Download PPT slide   Figure 1. Generation of Dp(AR17); Dp1.3ec double duplication strains by crossing over in the interval between the hph and T(AR17) insertions into linkage group I. All strains are erg+ and, hence, tomatine sensitive. The drawing of T(AR17) is simplified by ignoring involvement of linkage group II, which is irrelevant to the present experiment. In this and the following figure, segments of linkage group I are shown as solid lines and those of linkage group III as dotted.

View larger version (26K): In this window In a new window Download PPT slide   Figure 2. Progeny produced in a cross between T(AR17) and a normal-sequence dow erg-3 strain. Note that progeny with the dow phenotype represent one-third of all viable crossover products and that progeny that are phenotypically tomatine sensitive and barren are produced only if there has been a crossover in linkage group III between the translocation breakpoint and erg-3. Conventions as in Fig 2.

The strain ad-3B cyh-1 mata^m1 (FGSC no. 4564) was provided by Ramesh Maheshwari (Indian Institute of Science, Bangalore, India). It is an adenine auxotroph and contains a defective mat allele and is useful for constructing heterokaryons that can be maintained on adenine-less medium. The col-18 nuclei were maintained in (col-18 + ad-3B cyh-1 mata^m1) heterokaryons that were used in crosses.

The construction of the Dp 1.3^ec hph mata and Dp 1.3^ec hph matA strains has been described by PRAKASH et al. 1999 . These strains contain the wild-type allele at the erg-3 locus and also duplication of a 1.3-kb HindIII fragment from erg-3 that is inserted as a single copy into LGI linked to mata. The duplicated segment does not encode a functional sterol C-14 reductase but serves to target RIP to the erg-3 gene. The duplicated fragment is marked by the bacterial hph gene, which encodes the enzyme hygromycin phosphotransferase that when expressed confers resistance to the antibiotic hygromycin B. These strains have a tomatine-sensitive and hygromycin-resistant phenotype.

The N. intermedia standard reference strain Shp-1 matA (FGSC no. 3416) was provided by Ramesh Maheshwari and the N. tetrasperma reference strain 85 matA (FGSC no. 1240) was obtained from the FGSC.

Growth conditions:
Crossing and maintenance of the Neurospora strains was essentially as described by DAVIS and DE SERRES 1970 . Antibiotic resistance was scored by streaking conidia onto 1.5% agar plates containing Vogel's N medium plus "sorbose" (0.05% fructose, 0.05% glucose, and 2% sorbose) and supplemented with the antibiotic. The antibiotics tested were -tomatine (Sigma, St. Louis) at 90 µg/ml made from a 25 mg/ml stock solution in DMF and hygromycin B (Sigma) 200 µg/ml made from a 100 mg/ml aqueous stock solution. After an overnight incubation at 30° on tomatine-supplemented medium, growth can be observed of only the erg-3 mutant strains (SENGUPTA et al. 1995 ). Only strains expressing the hph gene could grow on hygromycin medium. The ad-3B cyh-1 mata^m1 strain was grown on Vogel's-sucrose medium supplemented with adenine (0.5 mg/ml).

Ascospore collection:
Crosses were performed in petri dishes. Ascospores were collected by washing the lids with 1 ml water. The frequency of RIP can differ greatly between early vs. late collected ascospores (SINGER et al. 1995 ). Therefore we pooled the early and late ascospores before determining RIP frequencies. Typically ascospores began to be shot within 16–18 days in fertile crosses whereas in barren crosses the first ascospores were seen only after 21 days. For fertile crosses the spores were collected at regular intervals for up to 31 days, and the spores from all the collections of each cross were pooled. For barren crosses usually only one collection was feasible, after 31 days; thereafter the number of additional spores was negligible.

Scoring for erg-3 mutant segregants:
RIP-induced erg-3 mutants can be distinguished from the wild type by the characteristic slow growth phenotype of their ascospore-derived colonies on Vogel's-sorbose agar (PRAKASH et al. 1999 ; NOUBISSI et al. 2000 ). Therefore, to score for erg-3 mutants among the segregants from a cross we merely counted all colonies with the mutant growth morphology. In most cases we subsequently confirmed their mutant phenotype on tomatine medium. Likewise most of the wild-type colonies were scored merely on the basis of their normal growth phenotype. In each cross a significant number (50) of "normal growers" was tested to confirm that they all indeed possessed the tomatine-sensitive wild-type phenotype.

	RESULTS

TOP ABSTRACT MATERIALS AND METHODS RESULTS DISCUSSION LITERATURE CITED

Construction of Dp 1.3^ec hph; Dp(AR17) strains:
A cross was performed between the strains T(IIIR > [IR; IIR]) AR17 mata [henceforth referred to as T(AR17)] and Dp 1.3^ec hph matA to produce progeny that should include Dp (IIIR > [IR; IIR]) AR17 segregants [henceforth referred to as Dp(AR17)] that also contain the Dp 1.3^ec hph transgene (Fig 1). Although both parental strains are erg-3⁺ and therefore have a tomatine-sensitive phenotype, erg-3 mutants can be generated in this cross due to the induction of RIP in the Dp 1.3^ec hph nucleus. These mutants could be distinguished by the characteristic slow growth phenotype of the ascospore-derived colonies on Vogel's-sorbose agar medium (see NOUBISSI et al. 2000 for a figure) and also on the basis of their resistance to tomatine. Of 67 segregants examined, 17 (25.4%) were mutant in erg-3. This frequency is the same as that reported previously for the recovery of erg-3 mutants from a cross between Dp 1.3^ec hph mata and the wild-type strain 74-OR23-1 matA (PRAKASH et al. 1999 ). Of the remaining 50 tomatine-sensitive segregants, 30 (60%) were hygromycin resistant, indicating that they also contained the Dp 1.3^ec hph transgene. Each of the hygromycin-resistant, tomatine-sensitive segregants was crossed with the wild-type strains 74-OR23-1 matA or OR8-1 mata. These crosses are referred to as series A. Twenty-five of the series A crosses were determined to be fertile and 5 (involving segregant nos. A13, A17, A30, A40, and A52) were barren. The proportion of barren crosses (5/30) is <1/3, presumably because the segregants were first screened for the hygromycin-resistance phenotype, so a crossover between the Dp 1.3^ec hph transgene and the T(AR17) breakpoint on LGI would be necessary for them to also possess Dp(AR17) (Fig 1). Ascospore production was estimated by eye to be 100- to 1000-fold lower in the barren crosses than in the fertile crosses. These results suggested that the segregants producing the barren crosses ("barren segregants") were Dp 1.3^ec hph; Dp(AR17) double duplication strains.

That four of the barren segregants were indeed Dp 1.3^ec hph; Dp(AR17) double duplication strains was confirmed by the recovery of downy (dow) mutants among their progeny (Table 1). The dow locus is covered by Dp(AR17) and these mutations are presumed to have resulted from RIP in the large duplication. The dow mutation frequencies in these four barren crosses were comparable with the 4.7% frequency reported by PERKINS et al. 1997 . No dow mutants were found among the 35 progeny examined from the cross parented by the barren segregant A30 and we did not attempt to scale up this cross further to obtain additional progeny. No dow mutants were found among any of the segregants examined from the fertile crosses (data not shown).

From the barren cross between segregant A13, mata and 74-OR23-1 matA we initially obtained 44 progeny (all tomatine-sensitive) of which 22 were hygromycin resistant. Twenty of the hygromycin-resistant progeny and their 22 hygromycin-sensitive sibs were crossed with the wild-type strains 74-OR23-1 matA or OR8-1 mata. These crosses are referred to as series B. Twenty-eight crosses were fertile and 14 were barren. Of the fertile segregants, 10 were mata and 5 of the barren segregants (including B40 and B41, both hygromycin sensitive and the latter with a dow phenotype) were matA. These 15/42 segregants represented crossovers between the mat locus and a T(AR17) breakpoint on LGI. Such loose linkage is consistent with the assignment of the T(AR17) breakpoint to LGIR (PERKINS 1997 ).

The two hygromycin-sensitive and barren segregants, B40 and B41, represented Dp(AR17) matA strains. They were crossed with T(AR17) mata. Progeny from these crosses can be either Dp(AR17) or T(AR17) strains (see PERKINS et al. 1997 for an explanatory figure). One cross yielded 2 dow segregants out of 11 examined, and the other, 3 out of 20. One of these dow segregants was fertile and thus represented a T(AR17), dow strain.

RIP in erg-3 is reduced in crosses involving Dp 1.3^ec hph; Dp(AR17) strains:
The erg-3 mutation frequencies were determined for all 5 barren crosses and 18 of the 25 fertile crosses of series A and the results are summarized in Table 2. The crosses involving the barren segregants A13 and A17 were scaled up to obtain larger numbers of progeny. In the fertile crosses the erg-3 mutation frequencies ranged between 1.5 and 27.8%. For the barren crosses parented by the four confirmed double duplication strains the range was between <0.07 and <0.56%. The clean separation of the two ranges allows us to conclude that the ability of Dp 1.3^ec to induce RIP in erg-3 is suppressed in nuclei that contain Dp(AR17). In fact, the frequency of RIP in erg-3 was even lower than in dow (compare entries for segregants A13, A17, A40, and A52 in Table 1 and Table 2).

Control crosses were made between Dp 1.3^ec hph and the N. intermedia standard reference strain Shp-1 matA and the N. tetrasperma standard reference strain 85 matA. Interspecies crosses in Neurospora are usually quite unproductive; in fact these crosses were even less productive than the barren crosses described above and yielded 1000-fold fewer ascospores than the fertile crosses of series A. Of 90 segregants examined from Dp 1.3^ec hph x Shp-1 matA, three (3.3%) were mutant in erg-3, and in Dp 1.3^ec hph x 85 matA the erg-3 mutation frequency was 15 out of 80 (18.8%). Thus, even though the productivity of the interspecies crosses was poor, their RIP efficiencies were within the range of the fertile crosses. This argues against the possibility that suppression of RIP in the crosses involving segmental aneuploidy is a trivial consequence of their poor productivity and instead implicates a role for Dp(AR17) in this effect. On the basis of these results we can conclude that the relatively smaller gene-sized duplication Dp 1.3^ec tends to be ignored by the RIP machinery in a nucleus that also contains the large chromosome segment duplication Dp(AR17).

RIP efficiency is restored in the fertile segregants from a barren cross:
Of the hygromycin-resistant segregants examined in series B (see above), 8 were fertile and 12 were barren. We examined the progeny of 7 fertile and 10 barren segregants and the results are summarized in Table 3. It can be seen that the Dp 1.3^ec hph transgene regains the ability to induce erg-3 mutations in the fertile crosses and continues to be ignored by the RIP machinery in the barren crosses. These results allow us to conclude that the ability of Dp 1.3^ec hph to engage in RIP is restored subsequent to its segregation from Dp(AR17).

Dp(AR17) suppresses induction of RIP by Dp 1.3^ec even in the trans configuration:
The Dp(AR17) and Dp 1.3^ec hph; Dp(AR17) segregants identified in the series A and B crosses were used to examine whether Dp(AR17) affected the induction of RIP by Dp 1.3^ec when the two duplications were in different parental nuclei of the premeiotic dikaryon (i.e., in trans). The results of these crosses (Table 4) suggest that Dp(AR17) can suppress the induction of RIP by Dp 1.3^ec even in trans. However, this suppression was not always as severe as when the duplications were in the same nucleus.

We also performed crosses that were homozygous for the large duplication and either homozygous or heterozygous for the small duplication. Interestingly the productivity of the Dp(AR17) homozygous crosses was comparable to that of the heterozygous crosses. Of a total of 2778 segregants examined from such crosses none was mutant in erg-3; therefore, the frequency of RIP in erg-3 was <0.04%.

Crossover between the T(AR17) breakpoint in LGIII and erg-3:
A T(AR17) mata x dow erg-3 matA cross can produce progeny with either the dow phenotype or with a tomatine-sensitive and barren phenotype if there has been a crossover in LGIII between the translocation breakpoint and erg-3 and such segregants represent one-third of the viable crossover products (Fig 2). Of 81 segregants examined from this cross, one was dow; thus, the crossover frequency was (1 x 3)/ 81, i.e., 3.7%. In a second experiment 309 segregants were examined from the cross T(AR17) mata x erg-3 matA and 5 were tomatine sensitive and barren. In this experiment the crossover frequency was (5 x 3)/309, i.e., 4.8%. The latter determination may be more accurate because it was based on the examination of more segregants.

Construction of Dp 1.3^ec hph; Dp(OY329) strains:
To determine if other chromosome segment duplications, besides Dp(AR17), also suppress RIP in Dp 1.3^ec, we examined crosses that were heterozygous for the duplication Dp(VIR > IIIR) OY329 [also referred to as Dp(OY329)]. We constructed Dp(OY329) Dp 1.3^ec hph double duplication strains and determined the frequency of erg-3 mutants among progeny parented by these strains.

A cross was performed between the Dp 1.3^ec hph mata and col-18 matA strains. The frequency of RIP-induced erg-3 mutants in this cross was 11/165 (6.7%). Thirty-two of 78 progeny (41%) had the colonial mutant phenotype (col^-) and, of 27 col^- segregants examined, 12 (44.4%) were hygromycin resistant. The col^-, hygromycin-resistant phenotype represented the Dp 1.3^ec hph; col-18 progeny.

Another set of crosses was performed between the col-18 and T(OY329) strains. Dp(OY329) has been reported to cover col-18 (PERKINS 1997 ), so the ratio of col^- to col⁺ progeny was expected to be 1:2. The col⁺ segregants should include both the T(OY329) and Dp(OY329), col-18⁺/col-18 progeny, and the latter should be distinguishable by their barrenness and ability to yield col-18 progeny in crosses with the wild type. Surprisingly, the observed ratio of col^- to col⁺ segregants from col-18 matA x T(OY329) mata was 54:25 and from col-18 mata x T(OY329) matA it was 56:21. We do not have a simple explanation for this discrepancy between the expected and observed segregation frequencies (but we consider a potentially interesting and testable hypothesis in the DISCUSSION). We did, however, identify one col⁺ segregant that was barren in a cross with Dp 1.3^ec mata and this cross yielded 20 col^- segregants out of 52 progeny examined. Thus this col⁺ segregant represented a bona fide Dp(OY329), col-18⁺/col-18 matA strain. The erg-3 RIP frequency in the cross between this strain and Dp 1.3^ec mata was 3/172 (1.7%), which was comparable with the frequencies in crosses where Dp(AR17) was present in trans (Table 4).

The confirmed Dp(OY329), col-18⁺/col-18 matA strain was crossed with two Dp 1.3^ec hph; col-18 mata strains (designated 19 and 25). Both crosses were barren and, as expected, they produced col⁺ and col^- segregants in approximately 1:1 ratios (20:27 and 10:11). Of the col⁺ segregants, 9 from the cross parented by 19 and 5 from the one parented by 25 were also resistant to hygromycin. These 14 hygromycin-resistant col⁺ strains were presumably the Dp 1.3^ec hph; Dp(OY329), col-18⁺/col-18. double duplication progeny. And the segregants with the hygromycin-resistant col^- phenotype were their Dp 1.3^ec hph; col-18 euploid siblings (Fig 3).

View larger version (17K): In this window In a new window Download PPT slide   Figure 3. Generation of Dp(OY329); Dp1.3ec double duplication strains. All strains are erg+ and, hence, tomatine sensitive. See text for details of Dp(OY329), col-18+/col-18 and Dp1.3ec hph; col-18 strain construction. Segments of linkage group VI are shown as dotted lines and those of linkage groups I and III as solid.

RIP in erg-3 is reduced in crosses involving Dp 1.3^ec hph; Dp(OY329) strains:
Nine col⁺ putative double duplication segregants from the cross parented by strain 19 and four from the cross parented by 25 (see above) were crossed with the wild-type strains 74-OR23-1 matA or OR8-1 mata. And seven col^- segregants (two and five from the crosses parented by 19 and 25, respectively) were also crossed with the wild type as controls. These crosses are referred to as series C.

Surprisingly, six of the col⁺ segregants (all derived from 19) were fertile and only seven were barren (Table 5). The six fertile col⁺ segregants did not yield any col^- progeny (20 progeny were examined for each cross), which confirmed that the fertile col⁺ segregants were not Dp(OY329) strains, but they did segregate the Dp 1.3^ec hph transgene. We discuss later the possible origin of these col⁺ nonduplication strains. The remaining seven barren col⁺ segregants all segregated col^- progeny (data not shown), which confirmed that they were indeed Dp(OY329) strains. The col^- segregants were all fertile, which confirmed that they were nonduplication strains. Table 5 summarizes the erg-3 RIP frequencies in the barren and fertile crosses of series C.

As can be seen in the table, in most of the barren crosses the erg-3 mutation frequency was very low. In contrast, we could recover erg-3 mutants from all the fertile crosses, although there was considerable variation in their RIP frequencies. These results allow us to conclude that, like Dp(AR17), Dp(OY329) also can suppress the induction of RIP by Dp 1.3^ec.

RIP is reduced in crosses involving Dp(IVR > VII) S1229:
The duplication Dp(IVR > VII) S1229 [henceforth designated as Dp(S1229)] has been described by PERKINS 1997 and strains bearing this duplication were available from the FGSC. This duplication is stable in crosses and segregates 1:1 in progeny from Dp x N. Crosses were performed between the strains Dp(S1229) matA and Dp 1.3^ec mata and between Dp(S1229) mata and Dp 1.3^ec matA. These crosses were barren and ascospores could be harvested only after 2 months. No erg-3 mutants were found among the 88 segregants examined from the first cross and the 137 segregants from the latter cross. Thus the erg-3 RIP frequencies were, respectively, <1.13% and <0.73%. These frequencies are comparable with the frequencies where Dp(AR17) was present in trans (Table 4).

The hygromycin-resistant segregants from these crosses must contain the transgene, and they could be either Dp(S1229) Dp 1.3^ec double duplication strains (barren), or Dp 1.3^ec normal sequence strains (fertile). Ten hygromycin-resistant segregants were crossed with the wild-type strains 74-OR23-1 matA or OR8-1 mata. After 43 days two crosses could unambiguously be scored as fertile, six as barren, and two were intermediate in ascospore productivity. Of the six unambiguously barren crosses, only one (designated D9) produced a sufficient number of ascospores to permit a meaningful estimate of RIP frequencies; one had to be discarded due to contamination; and the other four together yielded only 22 progeny (none were mutant in erg-3). The RIP frequencies in the five crosses that could be analyzed (series D) are summarized in Table 6. As can be seen in the table the barren cross showed a very low erg-3 mutation frequency relative to the two fertile crosses. This was consistent with the idea that Dp(S1229) suppresses the induction of RIP by Dp 1.3^ec in erg-3.

The erg-3 mutation frequency in the two crosses producing an intermediate number of ascospores was much lower than in the fertile crosses.

	DISCUSSION

TOP ABSTRACT MATERIALS AND METHODS RESULTS DISCUSSION LITERATURE CITED

We have examined whether the ability of a small duplication to induce RIP in its target gene is affected by the presence of a larger chromosome segment duplication in the same cross. For this we first constructed strains that contained both the large chromosome segment duplication, Dp(AR17), and a smaller gene-sized duplication, Dp 1.3^ec. Assuming that dow and erg-3 represent comparable targets for RIP, the frequency of dow and erg-3 mutants among the progeny from crosses made with such double duplication strains provides estimates, respectively, of RIP efficiency in the large and small duplications. We found that the presence of the Dp(AR17) duplication suppressed the ability of the smaller gene-sized duplication to induce RIP in its target gene. In fact, the induction of RIP in erg-3 was even lower than in dow. Suppression was evident even when the two duplications were in different nuclei of the premeiotic dikaryon. Dp(AR17) was initially chosen because we imagined that its linkage to erg-3 would increase the sensitivity of our tests. But in view of the observation that Dp(AR17) suppresses RIP even in trans, linkage does not appear to be germane to this effect.

Like Dp(AR17), the duplications Dp(OY329) and Dp(S1229) also suppressed the induction of RIP by Dp 1.3^ec, both in cis and in trans. Two crosses in the experiments with Dp(S1229) produced an intermediate number of ascospores (Table 6). These two segregants could not be unambiguously designated as either duplication or euploid strains because this experiment lacked a marker for Dp(S1229) like the dow and col-18 markers for Dp(AR17) and Dp(OY329), respectively. But even if we assume the devil's advocate position that the "intermediate" segregants represent the Dp(S1229) Dp 1.3^ec double duplication strains, the erg-3 mutation frequency in the two crosses was much lower than that in the fertile crosses. Therefore these results do not negate the conclusion that Dp(S1229) reduces RIP in Dp 1.3^ec. Overall our results suggest that any large duplication can suppress a smaller duplication's ability to induce RIP in its target. It is conceivable that large duplications act as sinks to titrate the RIP machinery from the dikaryotic cell. Since a considerable proportion of nuclei in standard laboratory strains harbor rearranged chromosomes (PERKINS and KINSEY 1993 ), studies of RIP in such genetic backgrounds can potentially be confounded by the generation of cryptic duplications in a subset of the progeny.

The frequency of col⁺ to col^- progeny from the col-18 x T(OY329) crosses was expected to be 2:1 but the observed frequency was 1:2. This result would normally have led one to question whether col-18 is covered by Dp(OY329). But the recovery of one bona fide Dp(OY329), col-18⁺/col-18 matA segregant with the col⁺ phenotype confirmed that Dp(OY329) does in fact cover col-18. So how might one explain the discrepant segregation ratios? One possibility is that the col-18 locus may be deleted from a subset of nuclei of the T(OY329) parents. Thus these strains may effectively be heterokaryons, in which nuclei with the active col-18⁺ allele ensure their col⁺ phenotype. It is well known that chromosome segment duplications often break down during vegetative growth by loss of one copy of the duplicated segment, and this loss occurs more frequently from the translocated position than from the normal position (see PERKINS 1997 for a review). Similar processes might have led to the loss of the col-18⁺ allele (as well as that of a nearby essential locus) from a subset of T(OY329) nuclei. In this case some of the col^- progeny from col-18 x T(OY329) might represent the Dp(OY329), col-18/col-18 genotype. Moreover, progeny inheriting the modified translocation chromosomes might be inviable. Both these effects could contribute to the discrepant phenotypic ratios. This hypothesis predicts that a cross between the T(OY329) and Dp(OY329), col-18⁺/col-18 strains will yield col^- segregants that are not products of RIP or gene conversion events.

Another unexpected finding was that six of nine col⁺ segregants examined from Dp(OY329), col-18⁺/col-18 matA (#19) x Dp 1.3^ec hph; col-18 mata were non-Dp(OY329) strains. This was surprising because all the col⁺ segregants were expected to be genotypically Dp(OY329), col-18⁺/col-18 (Fig 3). The generation of non-Dp(OY329) col⁺ progeny suggests that the col-18 allele on the nontranslocation LGVI was gene converted to col-18⁺ using the duplication-borne allele as template. Since only one of the two crosses examined exhibited such gene conversion, this event may represent a "jackpot." Thus the conversion event possibly occurred either during the vegetative growth of Dp(OY329), col-18⁺/col-18 (#19) or in the premeiotic dikaryon stage between fertilization and karyogamy.

Breakdown of chromosome segment duplications during vegetative growth restores euploidy (PERKINS 1997 ). Only 5 of the 23 barren crosses examined in series A, B, C, and D yielded any erg-3 mutants. If the mutations did not depend on a prior breakdown of the large duplications we would expect some of these mutants to display a barren phenotype. Of seven erg-3 mutants examined, six were clearly fertile. One appeared to be barren, but none of its progeny were dow (A. BHAT, unpublished results). Thus it will be necessary to examine additional erg-3 mutants before we can assert that erg-3 can be RIPed in the presence of the larger duplication.

RIP frequencies in the fertile crosses of series A, C, and possibly D were more variable than in those of series B. Such unexplained variability in RIP frequencies is not without precedent (e.g., Table 2 in KINSEY et al. 1994 ) and merits further investigation.

Might chromosome rearrangements have sheltered active copies of Tad from RIP?
It has been suggested that RIP might serve to protect the genome against the proliferation of transposable elements (SELKER 1990 ). Transposable elements can be regarded as gene-sized duplications, but they also have the potential to generate segmental aneuploidy via homologous recombination between unlinked copies followed by segregation of the resulting translocation chromosome with normal chromosomes in meiosis. Our results suggest that duplications generated in this way would protect the transposable elements from destruction by RIP and the residual fertility of the duplication-bearing strains might provide a virtually ineradicable source of active elements through successive generations. Tad, an active LINE-like Neurospora transposon was discovered in the Adiopodoume strain of N. crassa; it is noteworthy that translocations were observed to be unusually frequent in crosses involving this strain (reported by KINSEY and HELBER 1989 as a personal communication from David Perkins). KINSEY et al. 1994 had suggested several factors that might account for Tad's survival in Adiopodoume, and recent results from our laboratory (NOUBISSI et al. 2000 ) indicate that this strain even possesses dominant RIP suppressors. The results presented here suggest that segmental duplications also may have contributed to Tad's RIP-free passage in the preceding generations. The translocations in the Adiopodoume strain might represent elements of those ancestral duplications (KASBEKAR 1999 ).

	ACKNOWLEDGMENTS

We thank David D. Perkins for his interest and guidance right from the conception of these experiments through to the satisfactory redrawing of the figures, Ramesh Maheshwari for the N. intermedia strains, and J. Gowrishankar and Esteban Temporini for reading the manuscript. K. Aparna analyzed the interspecies cross with N. intermedia. Felicite K. Noubissi contributed some of the data in Table 2. The Fungal Genetics Stock Center waived charges for supplying strains. A.B. was supported by a Council of Scientific and Industrial Research-University Grants Commission Junior Research Fellowship.

Manuscript received December 15, 1999; Accepted for publication December 19, 2000.

	LITERATURE CITED

TOP ABSTRACT MATERIALS AND METHODS RESULTS DISCUSSION LITERATURE CITED

DAVIS, R. H. and J. F. DE SERRES, 1970  Genetic and microbiological research techniques for Neurospora crassa.. Methods Enzymol. A 17:79-143.

ELLIS, S. W., M. E. ROSE, and M. GRINDLE, 1991  Identification of a sterol mutant of Neurospora crassa deficient in 14,15 reductase activity. J. Gen. Microbiol. 137:2627-2630[Medline].

IRELAN, J. T. and E. U. SELKER, 1996  Gene silencing in filamentous fungi: RIP, MIP and quelling. J. Genet. 75:313-324.

KASBEKAR, D. P., 1999  A Tad of dexterity: did the Neurospora transposon use chromosome rearrangements to evade repeat-induced point mutation in Adiopodoume? (Commentary). J. Biosci. 24:413-415.

KINSEY, J. A. and J. H. HELBER, 1989  Isolation of a transposable element from Neurospora crassa.. Proc. Natl. Acad. Sci. USA 86:1929-1933[Abstract/Free Full Text].

KINSEY, J. A., P. W. GARRETT-ENGELE, E. B. CAMBARERI, and E. U. SELKER, 1994  The Neurospora transposon Tad is sensitive to repeat-induced point mutation (RIP). Genetics 138:657-664[Abstract].

NOUBISSI, F. K., K. MCCLUSKEY, and D. P. KASBEKAR, 2000  Repeat-induced point mutation (RIP) in crosses with wild-isolated strains of Neurospora crassa: evidence for dominant reduction of RIP. Fungal Genet. Biol. 31:91-97[Medline].

PAPAVINASASUNDARAM, K. G. and D. P. KASBEKAR, 1994  The Neurospora crassa erg3 gene encodes a protein with sequence homology to both yeast sterol C-14 reductase and chicken lamin B receptor. J. Genet. 73:33-41.

PERKINS, D. D., 1997  Chromosome rearrangements in Neurospora and other filamentous fungi. Adv. Genet. 36:239-397[Medline].

PERKINS, D. D. and J. A. KINSEY, 1993  New chromosome rearrangements from conidia of Neurospora wild type OR23-1VA. Fungal Genet. Newsl. 40:67-68.

PERKINS, D. D., A. RADFORD, D. NEWMEYER, and M. BJORKMAN, 1982  Chromosomal loci of Neurospora crassa.. Microbiol. Rev. 46:426-570[Free Full Text].

PERKINS, D. D., B. S. MARGOLIN, E. U. SELKER, and S. D. HAEDO, 1997  Occurrence of repeat induced point mutation in long segmental duplications of Neurospora. Genetics 147:125-136[Abstract].

PRAKASH, A., S. SENGUPTA, K. APARNA, and D. P. KASBEKAR, 1999  The erg-3 (sterol 14, 15 reductase) gene of Neurospora crassa: generation of null mutants by repeat-induced point mutation and complementation by proteins chimeric for human lamin B receptor sequences. Microbiology 145:1443-1451[Medline].

SELKER, E. U., 1990  Premeiotic instability of repeated sequences in Neurospora crassa.. Annu. Rev. Genet. 24:597-613.

SENGUPTA, S., T. BHAVANI PRASANNA, and D. P. KASBEKAR, 1995  Sterol 14, 15 reductase (erg-3) mutants switch the phenotype of Neurospora crassa from sensitivity to the tomato saponin -tomatine to sensitivity to the pea phytoalexin pisatin. Fungal Genet. Newsl. 42:71-72.

SINGER, M. J., E. A. KUZMINOVA, A. THARP, B. S. MARGOLIN, and E. U. SELKER, 1995  Different frequencies of RIP among early vs. late ascospores of Neurospora crassa.. Fungal Genet. Newsl. 42:74-75.

This article has been cited by other articles:

				D. D. Perkins, M. Freitag, V. C. Pollard, L. A. Bailey-Shrode, E. U. Selker, and D. J. Ebbole Recurrent Locus-Specific Mutation Resulting From a Cryptic Ectopic Insertion in Neurospora Genetics, February 1, 2007; 175(2): 527 - 544. [Abstract] [Full Text] [PDF]

				M. Vyas, C. Ravindran, and D. P. Kasbekar Chromosome Segment Duplications in Neurospora crassa and Their Effects on Repeat-Induced Point Mutation and Meiotic Silencing by Unpaired DNA Genetics, March 1, 2006; 172(3): 1511 - 1519. [Abstract] [Full Text] [PDF]

				M. E. Hood, M. Katawczik, and T. Giraud Repeat-Induced Point Mutation and the Population Structure of Transposable Elements in Microbotryum violaceum Genetics, July 1, 2005; 170(3): 1081 - 1089. [Abstract] [Full Text] [PDF]

				A. Bhat, R. Tamuli, and D. P. Kasbekar Genetic Transformation of Neurospora tetrasperma, Demonstration of Repeat-Induced Point Mutation (RIP) in Self-Crosses and a Screen for Recessive RIP-Defective Mutants Genetics, July 1, 2004; 167(3): 1155 - 1164. [Abstract] [Full Text] [PDF]

				A. Bhat, F. K. Noubissi, M. Vyas, and D. P. Kasbekar Genetic Analysis of Wild-Isolated Neurospora crassa Strains Identified as Dominant Suppressors of Repeat-Induced Point Mutation Genetics, July 1, 2003; 164(3): 947 - 961. [Abstract] [Full Text] [PDF]