Altering a Gene Involved in Nuclear Distribution Increases the Repeat-Induced Point Mutation Process in the Fungus Podospora anserina

Corresponding author: Robert Debuchy, Bâtiment 400, Université Paris-Sud, F-91405 Orsay Cedex, France., debuchy{at}igmors.u-psud.fr (E-mail)

Communicating editor: P. J. PUKKILA

	ABSTRACT

TOP ABSTRACT MATERIALS AND METHODS RESULTS DISCUSSION LITERATURE CITED

Repeat-induced point mutation (RIP) is a homology-dependent gene-silencing mechanism that introduces C:G-to-T:A transitions in duplicated DNA segments. Cis-duplicated sequences can also be affected by another mechanism called premeiotic recombination (PR). Both are active over the sexual cycle of some filamentous fungi, e.g., Neurospora crassa and Podospora anserina. During the sexual cycle, several developmental steps require precise nuclear movement and positioning, but connections between RIP, PR, and nuclear distributions have not yet been established. Previous work has led to the isolation of ami1, the P. anserina ortholog of the Aspergillus nidulans apsA gene, which is required for nuclear positioning. We show here that ami1 is involved in nuclear distribution during the sexual cycle and that alteration of ami1 delays the fruiting-body development. We also demonstrate that ami1 alteration affects loss of transgene functions during the sexual cycle. Genetically linked multiple copies of transgenes are affected by RIP and PR much more frequently in an ami1 mutant cross than in a wild-type cross. Our results suggest that the developmental slowdown of the ami1 mutant during the period of RIP and PR increases time exposure to the duplication detection system and thus increases the frequency of RIP and PR.

THREE homology-dependent gene-silencing mechanisms affecting the sexual cycle in Euascomycete fungi have been identified: meiotic silencing by unpaired DNA (SHIU et al. 2001 ), repeat-induced point mutation (RIP), and methylation induced premeiotically (reviewed by COGONI 2001 ). A fourth mechanism, resulting in gene loss rather than silencing, is based on premeiotic recombination (PR) between cis-duplicated sequences and deletion of the interstitial sequence (reviewed by SELKER 1990 ). RIP appears to be a defense genome mechanism whereby multicopy sequences within one nucleus are affected by C-to-T transitions. Although RIP-like mutations are present in almost all identified fungal transposable elements (reviewed by DABOUSSI and CAPY 2003 ), functional machinery for RIP has been demonstrated only in Neurospora crassa (CAMBARERI et al. 1989 ), Podospora anserina (HAMANN et al. 2000 ; GRAIA et al. 2001 ), Magnaporthe grisea (IKEDA et al. 2002 ), and Leptosphaeria maculans (IDNURM and HOWLETT 2003 ). RIP characteristics have been extensively investigated in N. crassa (reviewed by SELKER 1990 , SELKER 2002 ). RIP operates in a pairwise fashion on linked and unlinked DNA repeats, but is more efficient on linked duplications. While RIP is frequently observed in N. crassa, only two RIP events have been reported so far in P. anserina, suggesting that RIP is rare in this species (HAMANN et al. 2000 ; GRAIA et al. 2001 ). A single gene (RID) implicated in RIP has been identified in N. crassa. Mutations in this gene result in a complete loss of RIP (FREITAG et al. 2002 ). RID encodes a DNA methyltransferase homolog, assumed to be involved in C-to-T transition. In N. crassa, RIP is associated with PR between tandem repeats (SELKER et al. 1987 ; IRELAN et al. 1994 ). PR of tandem repeats has been also described in P. anserina (PICARD et al. 1987 ; COPPIN-RAYNAL et al. 1989 ) but RIP has not been investigated during these studies. PR occurs between fertilization and meiosis in P. anserina (PICARD et al. 1987 ) and in N. crassa (SELKER et al. 1987 ), and the occurrence of RIP has also been established between these two stages in N. crassa (reviewed by SELKER 1990 ). Development from fertilization to meiosis requires tightly regulated nuclear movements (ZICKLER et al. 1995 ). The relationship between homology-dependent gene-silencing mechanisms and the early steps of the sexual cycle, notably nuclear distribution, has not yet been investigated.

The P. anserina ami1 gene was previously shown to be required for nuclear distribution at different stages of the fungal life cycle (GRAIA et al. 2000 ). The predominant phenotype of an ami1 mutant is male sterility, due to faulty nuclear migration into male cells (microconidia). ami1 encodes an ortholog of ApsA of A. nidulans (FISCHER and TIMBERLAKE 1995 ). ApsA is assumed to be a positional regulatory protein because the corresponding mutants display nuclear misplacement rather than impaired nuclear migration (reviewed by FISCHER 1999 ). ApsA complements the male sterility of the ami1-1 mutant, suggesting that ApsA/AMI1 may have similar molecular functions. Recently, a genetic interaction was demonstrated between apsA and nud F, a gene involved in the dynein pathway, providing the first evidence of a connection between the putative regulatory protein and the cytoplasmic dynein pathway (EFIMOV 2003 ). However, the effect of AMI1/ApsA alterations on the sexual cycle have not been described.

Analyses of the phenotypes of a partial deletion of ami1 show that nuclei are abnormally distributed all along the fruiting-body development in this ami1 homozygous cross. Unexpected interactions between ami1, RIP, and PR were also identified. The frequency of progeny affected by RIP and PR was highly increased in ami1 homozygous crosses and this phenotype correlates with a developmental retardation at the early steps of the sexual cycle. These results suggest that RIP and PR phenotypes may be due to an increased exposure time to detection of duplicated sequences during the mutant sexual cycle.

	MATERIALS AND METHODS

TOP ABSTRACT MATERIALS AND METHODS RESULTS DISCUSSION LITERATURE CITED

Plasmids:
pCBSMR1 and pPAH1 contain the target genes for measuring the frequency and nature of inactivating events in P. anserina. pCBSMR1 is a derivative of pCB1004 (CARROLL et al. 1994 ) and the hph gene of this vector was used as a target gene. It encodes a hygromycin phosphotransferase determining the resistance to hygromycin, which serves as the selectable marker upon transformation. pCBSMR1 also contains the SMR1 gene from the mat– mating-type idiomorph (ARNAISE et al. 2001A ). The pPAH1 plasmid is based on pUC18 with a 3.9-kb PstI genomic fragment from P. anserina containing the pah1 gene (ARNAISE et al. 2001B ). pah1 encodes a homeodomain protein and the inactivation of this target gene in P. anserina results in colonial growth (ARNAISE et al. 2001B ).

p4-4-5ble contains a partial deletion in the ami1 gene and was used for replacing the resident ami1⁺ gene of P. anserina through homologous recombination. The construction of p4-4-5ble is presented in Fig 1.

View larger version (19K): [in this window] [in a new window] [Download PPT slide]   Figure 1. Construction of p4-4-5Ble. The p4-5 plasmid (A), which contains a 4.6-kb EcoRI fragment from the ami1 coding sequence (GRAIA et al. 2000 ), has been digested with KpnI and BsaAI. This fragment of ami1 has been replaced by a 1.5-kb KpnI-BsaAI fragment of pPable (COPPIN and DEBUCHY 2000 ) containing a phleomycin resistance gene (CALMELS et al. 1991 ) under control of the gpd promoter of P. anserina (RIDDER and OSIEWACZ 1992 ), resulting in p4-4-5ble (B).

Cloning of the complete ami1 gene required the use of pUN121, a low-copy vector with a positive selection for recombinant (NILSSON et al. 1983 ). An 8-kb fragment containing the ami1 gene was amplified by PCR with the high-fidelity DNA polymerase PfuTurbo (Stratagene, Amsterdam) between primers N1 (5'-CCCTTAAGGTATACGCTGTCCTTTGACACCAAG-3') and X1 (5'-GATGATCAGATCTGTAACCACGCAAATCAAGAC-3') from cosmid N10 (PICARD et al. 1991 ) and cloned in the SmaI site of pUN121, resulting in pUN121ami1. The sequence of ami1 from pUN121ami1 indicated that two mutations had been introduced during the amplification process, leading to the replacement of methionine 939 by a valine and deletion of a glutamine in a polyglutamine track at position 1790. The ami1 gene of pUN121ami1 was noted as ami1^M939V,Q1790, but these mutations did not impair the function of ami1 in complementation tests.

P. anserina growth conditions, transformation, genetic analysis, and crosses:
P. anserina is a Euascomycete whose life cycle and general methods for genetic analysis have been described (ESSER 1974 ). Transformations were performed as previously described (BERTEAUX-LECELLIER et al. 1995 ). All strains are derived from the wild-type S strain. Asci contain four twinned binucleate ascospores (RAJU and PERKINS 1994 ). Thus, three ascospores per ascus are sufficient to determine genetic segregations. A few asci contain three binucleate and two uninucleate ascospores, which produce homokaryotic strains. Most of the binucleate ascospores contain mat+ and mat– nuclei and yield self-fertile mycelia. Homokaryotic strains are self-sterile and must be crossed with a complementary strain. Three kinds of crosses were used in this study. If mating partners produced fertile microconidia, they were cultured in the same petri dish and microconidia were spread with sterile water from one strain over the other (spermatization). If mating partners were male sterile, either small agar blocks from a culture of each parent were deposited side by side on the same petri dish and mating occurred after growing at the junction of compatible mycelia (side-by-side culture) or the mycelium of one strain was fragmented in a mixer and spread on the mycelium of the compatible strain.

Construction of the ami1⁺ pah1 strains:
While examining the interaction between the homeobox gene pah1 (ARNAISE et al. 2001B ) and ami1 (GRAIA et al. 2000 ), we observed that fertilization of a double-mutant ami1-1 pah1 by wild-type microconidia resulted in weak sporulation and perithecia without beak (our unpublished result). This phenotype has been used for screening hundreds of putative candidates with a knockout of ami1 in a pah1 context. The deletion of the pah1 gene and its replacement by the hph gene in an ami1⁺ mat+ strain has been described previously (ARNAISE et al. 2001B ). The resulting ami1⁺ mat+ pah1 primary transformant was crossed with an ami1⁺ mat– strain, and the ami1⁺ mat– pah1 and ami1⁺ mat+ pah1 strains were identified in the progeny. The ami1⁺ smr1::ura5 SMR1^ec+ hph^ec+ pah1 strain has been constructed as follows. The disruption of the SMR1 mating-type gene by ura5 has been described previously (the resulting ami1⁺ smr1::ura5 strain was referred to as smr1-r1 in the original article; ARNAISE et al. 1997 ). SMR1 belongs to the mat– idiomorph (DEBUCHY et al. 1993 ) and its disruption results in a complete arrest of the sexual cycle (ARNAISE et al. 1997 ). This defect is complemented by the SMR1 gene of pCBSMR1 (ARNAISE et al. 2001A ). The ami1⁺ smr1::ura5 strain has been crossed with a ami1⁺ mat+ strain containing pCBSMR1 and the progeny have been analyzed for a ami1⁺ smr1::ura SMR1^ec+ hph^ec+ strain. This latter strain has been crossed with ami1⁺ mat+ pah1 to obtain ami1⁺ smr1::ura5 SMR1^ec+ hph^ec+ pah1.

Construction of the ami1 strains:
The plasmid p4-4-5ble has been introduced by transformation in an ami1⁺ mat+ pah1 strain and examination of 600 transformants for weak sporulation and perithecia without beak allowed us to identify 81 candidates that displayed this phenotype. The ami1-1 mutant was male sterile (GRAIA et al. 2000 ), suggesting that the partial deletion of ami1 should have a similar phenotype. We retained eight male-sterile strains among the 81 transformants for further examinations. The replacement of a part of ami1 by ble was tested by PCR with two primer pairs. Each pair of primers contain one primer in the ble gene, KB4 (5'-GAAGGCTTTAATTTGCAAGC-3') or KB3 (5'-GTGGAAGGGAAGGGATGCTC-3'), and the other primer, KB1 (5'-CCTGCCAGTCGGATCCACGA-3') or KB2 (5'-GGGCATCCGCTTGCCATAGC-3'), is targeted to the genomic sequences flanking the 4.6-kb EcoRI fragment, which contains either the wild-type or the deleted ami1 gene. PCR fragments of the expected size were obtained with the two pairs of primers for 3 transformants, indicating that the disrupted gene of the p4-4-5ble plasmid has integrated at its homologous locus by a double crossing over, resulting in the replacement of the wild-type ami1 gene by the disrupted copy. The primary transformants were purified by crossing with a tester strain of opposite mating type and mat+ phleomycin-resistant strains were recovered. The replacement of ami1 was tested by Southern blotting in 2 transformants. Genomic DNA was digested with EcoRI and probed with the 8-kb fragment containing the ami1 gene. An ami1⁺ mat+ strain gave three bands at 22, 7.6, and 4.6 kb. The 4.6-kb fragment should be missing in a ami1 strain and replaced by two hybridizing bands of 2 and 1.1 kb. Only one transformant displayed the expected pattern for a ami1 mat+ strain. The procedure described for the construction of ami1 mat+ was followed for the construction of ami1 mat– and ami1 smr1::ura5 SMR1^ec+ hph^ec+ strains after the transformation of ami1⁺ mat– pah1 and ami1⁺ smr1::ura5 SMR1^ec+ hph^ec+ pah1 strains by plasmid p4-4-5ble. Each of these ami1 strains was constructed independently, since the ami1 gene is physically tightly linked to the mating-type locus.

Construction of the ami1⁺ strains containing the hygromycine resistance transgene:
The SMR1^ec+ hph^ec+ transgenes of ami1 smr1::ura5 SMR1^ec+ hph^ec+ were introduced after successive crosses into an ami1⁺ smr1::ura5 strain resulting in ami1⁺ smr1::ura5 SMR1^ec+ hph^ec+. We checked by Southern blotting that the SMR1^ec+ hph^ec+ transgenes were not altered in the latter strain after the successive crosses.

Construction of the strains containing the pah1 transgene:
The pPAH1 plasmid was introduced by transformation in an ami1⁺ mat+ pah1 strain, giving rise to the ami1⁺ mat+ pah1 pah1^ec+ strain (ARNAISE et al. 2001B ). Then the pah1^ec+ transgene was introduced by crossing in a ami1 mat– pah1 strain, resulting in ami1 mat– pah1 pah1^ec+.

Construction of the strains for the complementation test:
Plasmids pUN121ami1 and pUChygro (RUPRICH-ROBERT et al. 2002 ) were cotransformed in ami1 mat+. Hygromycin-resistant (Hyg^R) strains were screened for restoration of male fertility, yielding ami1^M939V,Q1790 ami1 mat+. This strain was crossed with ami1 mat– pah1 pah1^ec+ and the ami1^M939V,Q1790 ami1 mat– pah1 pah1^ec+ strain was identified by restoration of the male fertility. The ami1^M939V,Q1790 transgene from ami1^M939V,Q1790 ami1 mat+ was transferred successively by crossing in ami1 mat– and then in ami1 mat+ pah1, resulting in ami1^M939V,Q1790 ami1 mat+ pah1, which was identified by restoration of male fertility.

DNA procedures:
Genomic DNA for PCR or hybridization experiments was extracted according to the CTAB method (ROGERS and BENDICH 1988 ). Hybridizations were performed as described previously (CHURCH and GILBERT 1984 ). The probes were prepared using the Megaprime DNA labeling system (Amersham Biosciences). The 1988-bp pah1 probe was obtained by PCR with SA13 (5'-GGCTACCGACGTGGTCAAGC-3') and SA17 (5'-TCCTCTCAAAACGCCGCAG) on wild-type genomic DNA. The 1039-bp hph probe was obtained by PCR with KB13 (5'-CCCAAGCATCAAATGAAAAAGCC-3') and KB14 (5'-CATTTCCTTTGCCCTCGGACGAGTGC-3') on the pCBSMR1 plasmid. To sequence pah1 transgenes, we first amplified their coding sequence from genomic DNA by PCR with SA13 and SA17 or SA13 and SA8 (5'-CAGACATGGAGAGCGTGTG-3'), and the PCR product was purified and sequenced with one of these primers: SA8, SA13, SA1 (5'-CCGTTCGTGAGAGGATAGCG-3'), SA7 (5'-GTTCATGCACCGTCATACCG-3'), or KB16 (5'-GCGAGAAGCTTGCAGGTCTTC-3'). To sequence hph transgenes, we first amplified their coding sequence with KB13 and KB14, and the PCR product was purified and sequenced with one of these primers: KB13, KB14, hph-s-27 (5'-CCACTAGTCTGTCGAGAAGTTTCTGATC-3'), or hph-a-997 (5'-GGACTAGTGGCGTCGGTTTCCACTATCG-3').

	RESULTS

TOP ABSTRACT MATERIALS AND METHODS RESULTS DISCUSSION LITERATURE CITED

ami1 strains display abnormal nuclear distribution and retardation during the sexual cycle:
P. anserina is a heterothallic fungus with two mating-type idiomorphs, mat+ and mat– (COPPIN et al. 1997 ). The ami1-1 mutant is tightly linked to the mat+ idiomorph and no ami1-1 mat– strain was obtained previously (GRAIA et al. 2000 ). We constructed ami1 mat+ and ami1 mat– strains by replacing part of the ami1 coding sequence with the ble gene in mat+ and mat– strains (see MATERIALS AND METHODS). The partial deletion within ami1 left a 3' end encoding 539 residues of the 1882 deduced from the entire ami1 gene. Both mutant strains produced anucleate microconidia, as does the ami1-1 mutant, resulting in male sterility. This defect is complemented by the ami1 gene from pUN121ami1 (data not shown). The absence of functional microconidia in an ami1 mutant strain does not preclude mating, because the mycelium can also act as a donor of nuclei to the female organs of opposite mating type (GRAIA et al. 2000 ). Mating between ami1 strains was obtained by spreading fragmented mycelium over a sexually compatible strain. When the same protocol was used with wild-type strains, ascospore projection occurred 4 days after fertilization. In a ami1 mat+ x ami1 mat– cross, ascospore projection was delayed, beginning the seventh day after fertilization. To determine precisely when delay occurred, we did a detailed cytological analysis of the fruiting-body development in both ami1 and wild-type homozygous crosses. In wild-type crosses, parental nuclei do not fuse immediately after fertilization: they divide several times in the female cells before mat+ and mat– nuclei migrate into specialized binucleated ascogenous hyphae. After a few divisions, the tip cell of the ascogenous hyphae differentiates into a hook-shaped cell, the crozier, in which the two nuclei undergo coordinate mitosis yielding, after septal formation, two uninucleate basal and lateral cells and a binucleate ascus mother cell (Fig 2A). Karyogamy and meiosis occur in the ascus mother cell and, after a postmeiotic mitosis, ascospores are delineated (for details see ZICKLER et al. 1995 ). In a wild-type perithecium, croziers and the first meiotic-prophase asci appear 48 hr after fertilization (Fig 2A), and by 72 hr, perithecia contain croziers, asci at different stages of meiosis, and a few asci with just membraned ascospores. Ascospore projection starts 4 days after fertilization. In homozygous ami1 perithecia, the first croziers and prophases occur 72 hr after fertilization, indicating at least a 24-hr developmental delay compared to that of wild type. The first ascospores were projected 4–5 days later. When compared to wild type, this indicates that meiosis and spore formation are also delayed by 2–3 days: in wild type, ascospore projection starts 2 days after crozier formation. These data suggest that mutant development was delayed at all stages of the sporulation process. In parallel with this delay, we observed abnormal nuclear distribution before and during crozier formation (Fig 2B and Fig C). Moreover, nuclear distribution is also very irregular during both meiotic and postmeiotic divisions (Fig 2D), leading to the formation of abnormal ascospores (Fig 2E) when compared to wild type (Fig 2F). Each developmental step is likely delayed and/or defective because of nuclear misplacement, and probably only random movement of nuclei, which brings them to the right position, will allow further development. Progeny of ami1 homozygous crosses are also consequently reduced. A self-fertile heterokaryotic mycelium issued from an ascospore containing ami1 mat+ and ami1 mat– nuclei yielded 70,000 ascospores per petri dish, while a wild-type ami1⁺ mat–/ami1⁺ mat+ strain produced 470,000 ascospores in similar conditions.

View larger version (102K): [in this window] [in a new window] [Download PPT slide]   Figure 2. Asci and ascospore formation in wild type and ami1. (A) Progressive crozier (arrowhead) and ascus development (length of the three arrows indicates the progression from karyogamy to pachytene) on an ascogenous hypha in a wild-type perithecium (already published in ZICKLER et al. 1995 ). (B–E) ami1. (B) Abnormal crozier formation. Some croziers are plurinucleated (arrowhead), some contain only one nucleus (thin arrow), and some are anucleated (thick arrow). Note also that the cell below the empty crozier shows three nuclei. (C) Example of abnormal distribution of nuclei during crozier formation. Some cells are plurinucleated (arrowheads) while others contain one (small arrow) or no nuclei (large arrow). (D) Postmeiotic mitosis with all four spindles formed in the ascus center when, in wild type, the fourspindles are aligned in two well-separated pairs. Large arrow points to a group of chromosomes; small arrow indicates a spindle pole body. (E) Four ascospores: two are binucleated (thick arrows), one contains four nuclei (arrowhead), and one is anucleated (thin arrow, bottom). The top ascus contains only abnormal spores (thin arrow, top). (F) Two wild-type asci. The top ascus contains four binucleate ascospores (large arrows). The bottom ascus (not complete) contains one binucleate and two uninucleate ascospores (small arrows; already published in ZICKLER et al. 1995 ). Bar, 5 µm.

ami1 enhances loss of function of transgenes:
To assay the frequencies of transgene loss of function, we used the hph and pah1 genes, which encode a hygromycin phosphotransferase (CARROLL et al. 1994 ) and a P. anserina homeodomain protein (ARNAISE et al. 2001B ), respectively. The structure of pah1 and hph ectopic integrations were analyzed by Southern blotting: they contained at least three copies of pah1 (Fig 3) and two copies of hph (data not shown). The multiple copies of pah1 remained associated and identical to the parental pattern in 19 independent ascospores resulting from an ami1⁺ mat– pah1 x ami1⁺ mat+ pah1 pah1^ec+ cross (data not shown). The multiple copies of hph remained associated and identical to the parental pattern in 16 independent ascospores resulting from a cross between ami1⁺ smr1::ura5 SMR1^ec+ hph^ec+ and ami1⁺ mat+ (data not shown). During this analysis, we found that one copy of hph was missing in one strain. However, all ascospores issued from the same ascus also displayed the loss of one hph copy, indicating unequivocally that this loss of one hph copy was a premeiotic deletion event rather than a segregation event. These data indicated that the multiple copies of each transgene were and remained genetically linked in an ami1⁺ background.

View larger version (75K): [in this window] [in a new window] [Download PPT slide]   Figure 3. Analysis of pah1 sequence rearrangements in the progeny from a ami1 homozygous cross. Six asci (designated by their reference numbers 9, 13, 14, 29, 35, and 37) from the ami1 mat– pah1 pah1ec+ x ami1 mat+ pah1 cross (Table 3, type I) were analyzed by Southern blot and compared to the parental hybrization pattern (lane P). The four strains issued from the four ascospores of each ascus are indicated as a, b, c, and d. DNA was digested with BstBI, which cuts once within the pah1 gene, transferred to nylon membranes, and probed with a pah1 fragment corresponding to the coding sequence to detect ectopic pah1 transgenes and their rearrangements. All strains issued from the four ascospores of each ascus display identical rearrangements indicating that they occurred premeiotically. For one strain (14d), DNA extraction has failed; however, this does not preclude the analysis of the ascus (see MATERIALS AND METHODS). The pah1 transgenes were amplified from genomic DNA of each one of these strains and sequenced. In asci 13, 14, 29, and 37, identical C:G-to-T:A transitions were found in the four strains issued from the four ascospores of the same ascus, indicating that they occurred premeiotically. The patterns of mutation were specific to each ascus and the number of RIP mutations for asci 13, 14, 29, and 37 were 15, 56, 4, and 19, respectively. Strains from asci 9 and 35 did not display any mutation.

We next analyzed the transgene stabilities in a ami1 genetic background. The inactivation of hph was tested in a ami1 smr1::ura5 SMR1^ec+ hph^ec+ x ami1 mat+ cross. The progeny of an ami1⁺ smr1::ura5 SMR1^ec+ hph^ec+ x ami1⁺ mat+ cross was used as a control. We observed the expected tetrads with first- and second-division segregation of the hph marker, but unexpectedly both crosses exhibited a third type of tetrads consisting only of hygromycin-sensitive (Hyg^S) ascospores. Moreover, the frequency of these Hyg^S tetrads was higher in the ami1 homozygous cross than in the ami1⁺ cross. The increase of Hyg^S tetrads in a ami1 homozygous cross cannot be calculated confidently at the first day of ascospore projection, but Hyg^S tetrads frequency is seven times higher in the ami1 homozygous cross than in the ami1⁺ cross at the fifth day (Table 1).

To further test the increased loss of function of a transgene in a ami1 background, we used pah1, a P. anserina homeobox gene, whose deletion results in colonial growth (ARNAISE et al. 2001A ). The loss of function of the pah1 transgenes was scored in a cross between ami1 mat– pah1 pah1^ec+ and ami1 mat+ pah1 strains and compared with the reference ami1⁺ mat– pah1 x ami1⁺ mat+ pah1 pah1^ec+ cross. All strains were deleted for the endogenous pah1 gene to avoid complementation of transgene loss of function. Both crosses produced three types of tetrads: those displaying first- or second-division segregation of the pah1 transgene and tetrads containing four ascospores giving rise to colonial growth (Table 2). The frequency of these tetrads, which was probably due to a loss of function of the pah1 transgene, was 18 times higher in ami1 than in ami1⁺ homozygous crosses at the first day of projection. The frequency of tetrads with four colonial strains increased in an ami1⁺ cross and their frequencies were similar after the fourth day of ascospore projection in ami1 and ami1⁺ homozygous crosses.

We tested whether this phenotype resulted from ami1 mutation per se by complementation experiments with the ami1^M939V,Q1790 gene from plasmid pUN121ami1. A total of 80 tetrads resulting from a cross between ami1^M939V,Q1790 ami1 mat+ pah1 and ami1^M939V,Q1790 ami1 mat– pah1 pah1^ec+ strains were analyzed for loss of function of pah1^ec+ at the first day of ascospore projection. Of these tetrads, 5% contained four ascospores germinating with a colonial growth, a frequency similar to the frequency obtained in a wild-type cross (see Table 2). This complementation test confirmed that the loss of function is due to ami1.

ami1 increases both RIP and premeiotic recombination frequency:
In P. anserina, two transgene inactivation mechanisms have been described: RIP (HAMANN et al. 2000 ; GRAIA et al. 2001 ) and PR (PICARD et al. 1987 ; COPPIN-RAYNAL et al. 1989 ). To determine which mechanism accounted for the increase of transgene inactivation in ami1 homozygous crosses, we analyzed the molecular events leading to the loss of function of the transgene by sequencing and Southern blotting.

The pah1 ectopic locus was analyzed in 12 ascospores from independent asci in a ami1 homozygous cross on the first and second day of projection. We amplified the pah1 transgenes from the genomic DNA and the PCR product was sequenced. Nine strains obtained from the 12 ascospores displayed transgene mutations, all restricted to C:G-to-T:A transitions (Table 3, type I). For 8 strains among the 9 RIPed strains, RIP mutations corresponded to superimposed T and C or G and A peaks on the electrophoregram, suggesting that at least two copies of pah1 have been amplified. Southern blot analysis of genomic DNA from the 12 strains indicated that 11 strains have undergone transgene deletion or rearrangements. To test whether identical mutations and deletion or rearrangements could be found in the other ascospores from the same tetrad, all ascospores of 6 of the 12 tetrads were also examined. All ascospores of each tetrad displayed identical mutations (data not shown) and identical Southern blot patterns (Fig 3), indicating that these events occurred premeiotically, as expected for RIP and PR. Previously, PR was found only on a 40-kb tandem duplication in P. anserina (COPPIN-RAYNAL et al. 1989 ). We show here that duplications obtained with smaller inserts can also be the target of PR. In some cases, PR resulted in the loss of one or several bands, suggesting that a deletion of transgenes has occurred (Fig 3, asci 29, 35, and 37), while other strains diplayed simultaneous loss of parental bands and appearance of new bands, suggesting either that RIP has altered the restriction site used for the digestion of the genomic DNA or that a rearrangement, possibly combined with a deletion, has occurred (Fig 3, asci 13 and 14). Strains from asci 9 and 35 (Fig 3) did not display any RIP mutations in the pah1 coding region but nevertheless showed differences in parental hybridization pattern. RIP or PR were also found in the pah1 ectopic locus in 5 of 10 strains growing with a wild-type phenotype and issued from a ami1 homozygous cross (Table 3, type II). Therefore, 75% (9/12) of colonial tetrads and 20% (2/10) of tetrads with wild-type growth displayed RIP. As colonial and wild-type growing tetrads represented 74 and 26% of the progeny on the first day of ascospore projection (Table 2), the frequency of RIP in the progeny at this stage can be estimated at 60%. A similar rationale indicates that PR affected 75% of the progeny of a ami1 homozygous cross at the first day of ascospore projection. As reference, we analyzed the molecular events occurring at the pah1 ectopic locus in the progeny of a ami1⁺ homozygous cross. Three strains issued from three different tetrads with genetic evidence of loss of pah1 function contained a rearranged pah1 sequence and one of them has undergone RIP (Table 3, type III). In contrast, the 19 analyzed strains exhibiting a wild-type growth phenotype did not display RIP or rearrangements (Table 3, type IV). The frequency of RIP and PR in an ami1⁺ cross is therefore 1.3 and 4%, respectively.

View this table: [in this window] [in a new window]   Table 3. Molecular characterization of the events leading to the loss of function of pah1 and hph transgenes in ami1 and ami1+ homozygous crosses

Furthermore, we have analyzed the hph marker in the progeny obtained from a ami1 homozygous cross and from a reference ami1⁺ cross. In both crosses, all nine tested strains issued from tetrads with exclusively Hyg^S ascospores contained a RIPed hph sequence, and six of them also showed PR (Table 3, types V and VII). A single Hyg^R strain issued from a ami1 homozygous cross was analyzed (Table 3, type VI). It displayed one RIP mutation. Three Hyg^R strains issued from an ami1⁺ cross showed no RIP and no PR (Table 3, type VIII).

These results demonstrate that the high efficiency of transgene silencing in a ami1 cross resulted from a dramatic increase of both PR and RIP efficiency. In a ami1 cross, 60% of tetrads were affected by RIP in the pah1 gene as compared to 1.3% in the ami1⁺ cross on the first day of ascospore production, thus a 46-fold increase. PR affected 75% of the tetrads in a ami1 cross while 4% of the tetrads were affected by this event in a ami1⁺ cross, indicating a 19-fold increase in a ami1 background. Although the sampling was less extensive for the loss of function of the hph gene, RIP and PR were clearly increased in a ami1 cross as compared to a wild-type cross. The frequency of RIP mutations cannot be calculated accurately, since the exact number of transgene copies in each strain is not known. Only an overestimation of RIP frequency can be calculated. The average RIP nucleotide mutation found in an ami1⁺ context is 1.8 and 2% for pah1 and hph genes, respectively. The average RIP nucleotide mutation in a ami1 context is 3 and 1% for pah1 and hph genes, respectively.

	DISCUSSION

TOP ABSTRACT MATERIALS AND METHODS RESULTS DISCUSSION LITERATURE CITED

We show here that RIP, which was considered rare in P. anserina (GRAIA et al. 2001 ), can be easily detected on linked multiple sequences at the end of ascospore projection in a ami1⁺ strain. It is thus surprising that only two RIP events have been thus far detected in P. anserina (HAMANN et al. 2000 ; GRAIA et al. 2001 ), although transformation has been available in this fungus for 15 years. Three features may explain why RIP has gone undetected in P. anserina for such a long time. First, we have been unable to detect RIP on unlinked copies of the pah1 gene (data not shown), even in a ami1 background, which clearly increases the probability of RIP. This observation suggests that RIP on unlinked copies is very inefficient in P. anserina, reducing the chance to detect a RIP event. Second, after the introduction of a transgene in P. anserina, any transformant not displaying an active transgene is usually discarded. Finally, most of the genetic work on P. anserina is performed on five ascospore asci, which are produced predominantly at the beginning of ascospore projection, at a time of minimal RIP efficiency.

Previous work indicated that the ami1 gene was involved in nuclear distribution and positioning (GRAIA et al. 2000 ). This study shows that it is also involved during the sexual cycle. Interestingly, as ami1 was found to increase RIP and PR efficiency, it raises the question of the relation between nuclear distribution, RIP, and PR. Abnormal nuclear positioning per se might trigger high RIP and PR, suggesting that the level of RIP and PR might be directly controlled by developmental events. Several lines of evidence indicate that the increase of RIP and PR in ami1 homozygous crosses is an indirect consequence of abnormal nuclear distribution. We have demonstrated that development is delayed in ami1 homozygous crosses, notably before meiosis, which marks the end of RIP and PR occurrence. This correlation suggests that the developmental delay of a ami1 homozygous cross increases the exposure of duplicated sequences to RIP and PR, resulting in a greater frequency of these events in the progeny. According to this rationale, we predict that any mutations that slow down initial fruiting-body development should increase RIP and PR efficiency. Accordingly, the high efficiency of RIP in N. crassa may be attributed, at least partially, to a specific developmental feature. In N. crassa, the first ascospores were found 9–10 days after fertilization (SINGER et al. 1995 ), in contrast to 4 days after fertilization in wild-type P. anserina. This developmental delay in N. crassa may have similar effects on RIP as did the ami1 mutation in P. anserina, resulting in a high proportion of RIPed progeny.

In contrast to the dramatically increased frequency of asci affected by RIP in a ami1 background, the frequency of RIP mutation is not modified by ami1. The previously reported average RIP frequency in P. anserina was 1% (GRAIA et al. 2001 ), similar to the RIP frequency found in this study, in either ami1⁺ or ami1 backgrounds. RIP frequency may be limited by the fact that PR might result in the loss of the duplicated sequences required for RIP. According to this hypothesis, the strains that have conserved the parental multiple-band pattern should display a very high mutation rate in ami1 crosses; however, these strains do not show a higher mutation rate than strains that have undergone PR (see Table 3). Another explanation can account for the absence of mutation increase in ami1, in agreement with a model proposed by SELKER 1990 . The detection of duplication may take place during the G₁ phase, which could be lengthened as a consequence of development delay in a ami1 cross. Consequently, the number of progeny affected by RIP is expected to increase, as observed in a ami1 cross. The mutation process may occur during the nuclear division, the length of which should be unaffected by development delay. If there are no more nuclear divisions during the early development stages of the fruiting body in a ami1 homozygous cross than there are in a wild-type cross, no increase of mutation is expected in a ami1 cross.

If ami1 increases RIP and PR by a developmental retardation, it should be possible to find some conditions in which fruiting-body development was delayed and RIP/PR were increased. In P. anserina, as in N. crassa (SINGER et al. 1995 ), RIP is increased in late-expelled ascospores relative to early ascospores. These late-expelled ascospores in a wild-type cross could be considered as phenocopies of the early-projected ascospores in a ami1 homozygous cross. In both cases, detection of transgene duplication is increased while RIP mutation efficiency remains unaffected. In fact, the parental nuclei yielding late-expelled ascospores are likely to correspond to croziers observed at 72 hr in the wild-type cross and thus could have been submitted to a more prolonged detection of duplication. This modulation of the RIP efficiency between early- vs. late-expelled ascospores may allow duplications to escape RIP in early progeny while they would be inactivated in late progeny, as proposed by SINGER et al. 1995 . Such variation in RIP efficiency could favor the selection of beneficial duplication in early progeny and the inactivation of detrimental duplication in late progeny. This mechanism and the very low efficiency of RIP on unlinked duplication lead us to predict that RIP has not retarded the creation of new genes through genomic duplication in P. anserina, in contrast to its effect in N. crassa (GALAGAN et al. 2003 ). The comparative genomic analysis from N. crassa and P. anserina will thus be interesting and may allow us to assess more precisely the impact of various RIP modalities on genome evolution of the two fungi.

	ACKNOWLEDGMENTS

We are greatly indebted to Fatima Graïa and Véronique Berteaux-Lecellier for the gift of p4-5 and for communicating the ami1 unpublished sequence data. We are grateful to Marguerite Picard for her constant interest in this work. We thank our colleagues for helpful discussions. K.B. was supported by scholarship no. 00302 of the Ministère de la Recherche et des Nouvelles Technologies.

Manuscript received September 3, 2003; Accepted for publication January 28, 2004.

	LITERATURE CITED

TOP ABSTRACT MATERIALS AND METHODS RESULTS DISCUSSION LITERATURE CITED

ARNAISE, S., R. DEBUCHY, and M. PICARD, 1997  What is a bona fide mating-type gene? Internuclear complementation of mat mutants in Podospora anserina. Mol. Gen. Genet. 256:169-178.[Medline]

ARNAISE, S., D. ZICKLER, S. LE BILCOT, C. POISIER, and R. DEBUCHY, 2001a  Mutations in mating-type genes of the heterothallic fungus Podospora anserina lead to self-fertility. Genetics 159:545-556.[Abstract/Free Full Text]

ARNAISE, S., D. ZICKLER, C. POISIER, and R. DEBUCHY, 2001b  pah1: a homeobox gene involved in hyphal morphology and microconidiogenesis in the filamentous ascomycete Podospora anserina.. Mol. Microbiol. 39:54-64.[CrossRef][Medline]

BERTEAUX-LECELLIER, V., M. PICARD, C. THOMPSON-COFFE, D. ZICKLER, and A. PANVIER-ADOUTTE et al., 1995  A nonmammalian homolog of the PAF1 gene (Zellweger syndrome) discovered as a gene involved in caryogamy in the fungus Podospora anserina.. Cell 81:1043-1051.[CrossRef][Medline]

CALMELS, T., M. PARRICHE, H. DURAND, and G. TIRABY, 1991  High efficiency transformation of Tolypocladium geodes conidiospores to phleomycin resistance. Curr. Genet. 20:309-314.[CrossRef][Medline]

CAMBARERI, E. B., B. C. JENSEN, E. SCHABTACH, and E. U. SELKER, 1989  Repeat-induced G-C to A-T mutations in Neurospora.. Science 244:1571-1575.[Abstract/Free Full Text]

CARROLL, A. M., J. A. SWEIGARD, and B. VALENT, 1994  Improved vectors for selecting resistance to hygromycin. Fungal Genet. Newsl. 41:22.

CHURCH, G. M. and W. GILBERT, 1984  Genomic sequencing. Proc. Natl. Acad. Sci. USA 81:1991-1995.[Abstract/Free Full Text]

COGONI, C., 2001  Homology-dependent gene silencing mechanisms in fungi. Annu. Rev. Microbiol. 55:381-406.[CrossRef][Medline]

COPPIN, E. and R. DEBUCHY, 2000  Co-expression of the mating-type genes involved in internuclear recognition is lethal in Podospora anserina.. Genetics 155:657-669.[Abstract/Free Full Text]

COPPIN, E., R. DEBUCHY, S. ARNAISE, and M. PICARD, 1997  Mating types and sexual development in filamentous ascomycetes. Microbiol. Mol. Biol. Rev. 61:411-428.[Abstract]

COPPIN-RAYNAL, E., M. PICARD, and S. ARNAISE, 1989  Transformation by integration in Podospora anserina. III. Replacement of a chromosome segment by a two-step process. Mol. Gen. Genet. 219:270-276.[Medline]

DABOUSSI, M.-J. and P. CAPY, 2003  Transposable elements in fungi. Annu. Rev. Microbiol. 57:275-299.[CrossRef][Medline]

DEBUCHY, R., S. ARNAISE, and G. LECELLIER, 1993  The mat– allele of Podospora anserina contains three regulatory genes required for the development of fertilized female organs. Mol. Gen. Genet. 241:667-673.[CrossRef][Medline]

EFIMOV, V. P., 2003  Roles of NUDE and NUDF proteins of Aspergillus nidulans: insights from intracellular localization and overexpression effects. Mol. Biol. Cell 14:871-888.[Abstract/Free Full Text]

ESSER, K., 1974 Podospora anserina, pp. 531–551 in Handbook of Genetics, edited by R. C. KING. Plenum, New York/London.

FISCHER, R., 1999  Nuclear movement in filamentous fungi. FEMS Microbiol. Rev. 23:39-68.[CrossRef][Medline]

FISCHER, R. and W. E. TIMBERLAKE, 1995  Aspergillus nidulans apsA (anucleate primary sterigmata) encodes a coiled-coil protein required for nuclear positioning and completion of asexual development. J. Cell Biol. 128:485-498.[Abstract/Free Full Text]

FREITAG, M., R. L. WILLIAMS, G. O. KOTHE, and E. U. SELKER, 2002  A cytosine methyltransferase homologue is essential for repeat-induced point mutation in Neurospora crassa.. Proc. Natl. Acad. Sci. USA 99:8802-8807.[Abstract/Free Full Text]

GALAGAN, J. E., S. E. CALVO, K. A. BORKOVICH, E. U. SELKER, and N. D. READ et al., 2003  The genome sequence of the filamentous fungus Neurospora crassa.. Nature 422:859-868.[CrossRef][Medline]

GRAIA, F., V. BERTEAUX-LECELLIER, D. ZICKLER, and M. PICARD, 2000  ami1, an orthologue of the Aspergillus nidulans apsA gene, is involved in nuclear migration events throughout the life cycle of Podospora anserina.. Genetics 155:633-646.[Abstract/Free Full Text]

GRAIA, F., O. LESPINET, B. RIMBAULT, M. DEQUARD-CHABLAT, and E. COPPIN et al., 2001  Genome quality control: RIP (repeat-induced point mutation) comes to Podospora.. Mol. Microbiol. 40:586-595.[CrossRef][Medline]

HAMANN, A., F. FELLER, and H. D. OSIEWACZ, 2000  The degenerate DNA transposon Pat and repeat-induced point mutation (RIP) in Podospora anserina.. Mol. Gen. Genet. 263:1061-1069.[CrossRef][Medline]

IDNURM, A. and B. J. HOWLETT, 2003  Analysis of loss of pathogenicity mutants reveals that repeat-induced point mutations can occur in the Dothideomycete Leptosphaeria maculans.. Fungal Genet. Biol. 39:31-37.[CrossRef][Medline]

IKEDA, K., H. NAKAYASHIKI, T. KATAOKA, H. TAMBA, and Y. HASHIMOTO et al., 2002  Repeat-induced point mutation (RIP) in Magnaporthe grisea: implications for its sexual cycle in the natural field context. Mol. Microbiol. 45:1355-1364.[CrossRef][Medline]

IRELAN, J. T., A. T. HAGEMANN, and E. U. SELKER, 1994  High frequency repeat-induced point mutation (RIP) is not associated with efficient recombination in Neurospora. Genetics 138:1093-1103.[Abstract]

NILSSON, B., M. UHLEN, S. JOSEPHSON, S. GATENBECK, and L. PHILIPSON, 1983  An improved positive selection plasmid vector constructed by oligonucleotide mediated mutagenesis. Nucleic Acids Res. 11:8019-8030.[Abstract/Free Full Text]

PICARD, M., R. DEBUCHY, J. JULIEN, and Y. BRYGOO, 1987  Transformation by intergration in Podospora anserina. II. Targeting to the resident locus with cosmids and instability of the transformants. Mol. Gen. Genet. 210:129-134.[CrossRef]

PICARD, M., R. DEBUCHY, and E. COPPIN, 1991  Cloning the mating types of the heterothallic fungus Podospora anserina: developmental features of haploid transformants carrying both mating types. Genetics 128:539-547.[Abstract]

RAJU, N. B. and D. D. PERKINS, 1994  Diverse programs of ascus development in pseudohomothallic species of Neurospora, Gelasinospora, and Podospora.. Dev. Genet. 15:104-118.[CrossRef][Medline]

RIDDER, R. and H. D. OSIEWACZ, 1992  Sequence analysis of the gene coding for glyceraldehyde-3-phosphate dehydrogenase (gpd) of Podospora anserina: use of homologous regulatory sequences to improve transformation efficiency. Curr. Genet. 21:207-213.[CrossRef][Medline]

ROGERS, S. O., and A. J. BENDICH, 1988 Extraction of DNA from plant tissues, pp. 1–10 in Plant Molecular Biology Manual, edited by S. B. GELVIN, R. A. SCHILPEROORT and D. P. S. VERMA. Kluwer Academic Publishers, Dordrecht, The Netherlands.

RUPRICH-ROBERT, G., D. ZICKLER, V. BERTEAUX-LECELLIER, C. VELOT, and M. PICARD, 2002  Lack of mitochondrial citrate synthase discloses a new meiotic checkpoint in a strict aerobe. EMBO J. 21:6440-6451.[CrossRef][Medline]

SELKER, E. U., 1990  Premeiotic instability of repeated sequences in Neurospora crassa.. Annu. Rev. Genet. 24:579-613.[CrossRef][Medline]

SELKER, E. U., 2002  Repeat-induced gene silencing in fungi. Adv. Genet. 46:439-450.[Medline]

SELKER, E. U., E. B. CAMBARERI, B. C. JENSEN, and K. R. HAACK, 1987  Rearrangement of duplicated DNA in specialized cells of Neurospora.. Cell 51:741-752.[CrossRef][Medline]

SHIU, P. K., N. B. RAJU, D. ZICKLER, and R. L. METZENBERG, 2001  Meiotic silencing by unpaired DNA. Cell 107:905-916.[CrossRef][Medline]

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.

ZICKLER, D., S. ARNAISE, E. COPPIN, R. DEBUCHY, and M. PICARD, 1995  Altered mating-type identity in the fungus Podospora anserina leads to selfish nuclei, uniparental progeny and haploid meiosis. Genetics 140:493-503.[Abstract]

This Article Abstract Full Text (PDF) Alert me when this article is cited Alert me if a correction is posted Services Similar articles in this journal Similar articles in PubMed Alert me to new issues of the journal Download to citation manager Citing Articles Citing Articles via Google Scholar Google Scholar Articles by Bouhouche, K. Articles by Arnaise, S. Search for Related Content PubMed PubMed Citation Articles by Bouhouche, K. Articles by Arnaise, S.