Characterization of SLFL1, a pollen-expressed F-box gene located in the Prunus S locus

The S locus and its flanking regions in the genus Prunus (Rosaceae) contain four pollen-expressed F-box genes. These genes contain the S locus F-box genes with low allelic sequence polymorphism genes 1, 2, and 3 (SLFL1, SLFL2, and SLFL3) as well as the putative pollen S gene, named the S haplotype-specific F-box protein gene (SFB). As much less information is available on the function of SLFLs than that of SFB, we analyzed the SLFLs of six S haplotypes of sweet cherry (Prunus avium) in this study. Genomic DNA blot analysis and the isolation of SLFL1 showed that the SLFL1 gene in a functional self-incompatible S ³ haplotype is deleted and only a partial sequence resembling SLFL1 is left in the S ³ locus region, suggesting that SLFL1 by itself is not directly involved in either the GSI reaction or pollen-tube growth. Genomic DNA blot analysis showed that there was no substantial modification or mutation in SLFL2 and SLFL3. A phylogenic analysis of F-box genes in the rosaceous S locus and its border regions showed that Prunus SLFLs were more closely related to maloid S locus F-box brothers than to Prunus SFBs. The functions of SLFLs and the evolution of self-incompatibility in Prunus are discussed based on these results. Electronic supplementary material  The online version of this article (doi:) contains supplementary material, which is available to authorized users. The nucleotide sequence data reported appear in the DDBJ, EMBL, and GenBank Nucleotide Sequence Databases under the accession numbers, AB360339, AB360340, AB360341, and AB360342, for SLFL1-S ¹ , SLFL1-S ² , SLFL1-S ⁵ , and SLFL1-S ⁶ , respectively.     

Genomic organization of the S-locus region of Brassica

To gain some insights into the structure of the S-locus and the mechanisms that have kept its diversity, a 75-kb genomic fragment containing the self-incompatibility (S) locus region was isolated from the S12-haplotype of Brassica rapa and compared with those of other S-haplotypes. The region around the S determinant genes was highly polymorphic and filled with S-haplotype-specific intergenic sequences. The diverse genomic structure must contribute to the suppression of recombination at the S-locus.     

Evolution of the S-locus region in Arabidopsis relatives

The S locus, a single polymorphic locus, is responsible for self-incompatibility (SI) in the Brassicaceae family and many related plant families. Despite its importance, our knowledge of S-locus evolution is largely restricted to the causal genes encoding the S-locus receptor kinase (SRK) receptor and S-locus cysteine-rich protein (SCR) ligand of the SI system. Here, we present high-quality sequences of the genomic region of six S-locus haplotypes: Arabidopsis (Arabidopsis thaliana; one haplotype), Arabidopsis lyrata (four haplotypes), and Capsella rubella (one haplotype). We compared these with reference S-locus haplotypes of the self-compatible Arabidopsis and its SI congener A. lyrata. We subsequently reconstructed the likely genomic organization of the S locus in the most recent common ancestor of Arabidopsis and Capsella. As previously reported, the two SI-determining genes, SCR and SRK, showed a pattern of coevolution. In addition, consistent with previous studies, we found that duplication, gene conversion, and positive selection have been important factors in the evolution of these two genes and appear to contribute to the generation of new recognition specificities. Intriguingly, the inactive pseudo-S-locus haplotype in the self-compatible species C. rubella is likely to be an old S-locus haplotype that only very recently became fixed when C. rubella split off from its SI ancestor, Capsella grandiflora.     

The number, age, sharing and relatedness of S-locus specificities in Prunus.

In gametophytic self-incompatibility systems, many specificities (different 'lock-and-key' combinations) are maintained by frequency-dependent selection for very long evolutionary times. In Solanaceae, trans-specific evolution (the observation that an allele from one species may be more closely related to an allele from another species than to others from the same species) has been taken as an argument for the very old age of specificities. In this work, by determining, for the first time, the age of extant Prunus species, we show that this reasoning cannot be applied to Prunoideae. Furthermore, since our sample size is large (all S-RNase encoding the female component and SFB encoding the male component GenBank sequences), we were able to estimate the age of the oldest Prunus specificities. By doing so, we show that the lower variability levels at the Prunus S-locus, in comparison with Solanaceae, is due to the younger age of Prunus alleles, and not to a difference in silent mutation rates. We show that the ancestor to extant Prunus species harboured at least 102 specificities, in contrast to the maximum of 33 observed in extant Prunus species. Since the number of specificities that can be maintained in a population depends on the effective population size, this observation suggests a bottleneck in Prunus evolutionary history. Loss of specificities may have occurred during this event. Using only information on amino acid sites that determine specificity differences, and a simulation approach, we show that a model that assumes closely related specificities are not preferentially lost during evolution, fails to predict the observed degree of specificity relatedness.     

Recombination, gene conversion, and identity-by-descent at three loci

We investigate the probabilities of identity-by-descent at three loci in order to find a signature which differentiates between the two types of crossing over events: recombination and gene conversion. We use a Markov chain to model coalescence, recombination, gene conversion and mutation in a sample of size two. Using numerical analysis, we calculate the total probability of identity-by-descent at the three loci, and partition these probabilities based on a partial ordering of coalescent events at the three loci. We use these results to compute the probabilities of four different patterns of conditional identity and non-identity at the three loci under recombination and gene conversion. Although recombination and gene conversion do make different predictions, the differences are not likely to be useful in distinguishing between them using three locus patterns between pairs of DNA sequences. This implies that measures of genetic identity in larger samples will be needed to distinguish between gene conversion and recombination.     

Recombination activating gene 1 product alone possesses endonucleolytic activity

Two lymphoid-specific proteins, RAG1 and RAG2, are required for the initiation of the V(D)J recombination in vitro. The V(D)J cleavage that is mediated by RAG proteins at the border between the coding and signal sequences results in the production of a hairpin at the coding end and a double-stranded break at the signal end. Two hairpin coding ends are re-opened, modified, and sealed; whereas, the signal ends are directly ligated. Here I report that only RAG1 can carry out a distinct endonucleolytic activity in vitro using an oligonucleotide substrate that is tethered by a short single-stranded DNA. The purified RAG1 protein alone formed a nick at the near position to the recombination signal sequence. This endonucleolytic activity was eliminated by immunoprecipitation using the RAG1-specific antibody, and required the 3'-hydroxy group. All of the RAG1 mutants that were incapable of the nick and hairpin formation in the V(D)J cleavage analysis also showed this new endonucleolytic activity. This suggests that the nicking activity that was observed might be functionally different from the nick formation in the V(D)J cleavage.     

Characterization of the S-locus region of almond (Prunus dulcis): analysis of a somaclonal mutant and a cosmid contig for an S haplotype

Almond has a self-incompatibility system that is controlled by an S locus consisting of the S-RNase gene and an unidentified "pollen S gene." An almond cultivar "Jeffries," a somaclonal mutant of "Nonpareil" (S(c)S(d)), has a dysfunctional S(c) haplotype both in pistil and pollen. Immunoblot and genomic Southern blot analyses detected no S(c) haplotype-specific signal in Jeffries. Southern blot showed that Jeffries has an extra copy of the S(d) haplotype. These results indicate that at least two mutations had occurred to generate Jeffries: (1) deletion of the S(c) haplotype and (2) duplication of the S(d) haplotype. To analyze the extent of the deletion in Jeffries and gain insight into the physical limit of the S locus region, approximately 200 kbp of a cosmid contig for the S(c) haplotype was constructed. Genomic Southern blot analyses showed that the deletion in Jeffries extends beyond the region covered by the contig. Most cosmid end probes, except those near the S(c)-RNase gene, cross-hybridized with DNA fragments from different S haplotypes. This suggests that regions away from the S(c)-RNase gene can recombine between different S haplotypes, implying that the cosmid contig extends to the borders of the S locus.     

Recombination and gene conversion in a 170-kb genomic region of Arabidopsis thaliana

Arabidopsis thaliana is a highly selfing plant that nevertheless appears to undergo substantial recombination. To reconcile its selfing habit with the observations of recombination, we have sampled the genetic diversity of A. thaliana at 14 loci of approximately 500 bp each, spread across 170 kb of genomic sequence centered on a QTL for resistance to herbivory. A total of 170 of the 6321 nucleotides surveyed were polymorphic, with 169 being biallelic. The mean silent genetic diversity (pi(s)) varied between 0.001 and 0.03. Pairwise linkage disequilibria between the polymorphisms were negatively correlated with distance, although this effect vanished when only pairs of polymorphisms with four haplotypes were included in the analysis. The absence of a consistent negative correlation between distance and linkage disequilibrium indicated that gene conversion might have played an important role in distributing genetic diversity throughout the region. We tested this by coalescent simulations and estimate that up to 90% of recombination is due to gene conversion.     

Recombination at the Rp1 locus of maize.

Summary The Rp1 locus of maize determines resistance to races of the maize rust fungus (Puccinia sorghi). Restriction fragment length polymorphism markers that closely flank Rp1 were mapped and used to study the genetic fine structure and role of recombination in the instability of this locus. Susceptible progeny, lacking the resistance of either parent, were obtained from test cross progeny of several Rp1 heterozygotes. These susceptible progeny usually had non-parental genotypes at flanking marker loci, thereby verifying their recombinational origin. Seven of eight Rp1 alleles (or genes) studied were clustered within about 0.2 map units of each other. Rpl ^G, however, mapped from 1–3 map units distal to other Rp1 alleles. Rp5 also mapped distally to most Rp1 alleles. Other aspects of recombination at Rp1 suggested that some alleles carry duplicated sequences, that mispairing can occur, and that unequal crossing-over may be a common phenomenon in this region; susceptible progeny from an Rp1 ^A homozygote had recombinant flanking marker genotypes, and susceptible progeny from an Rp1 ^DlRp1 ^F heterozygote showed both possible nonparental flanking marker genotypes.     

Industrial yeasts display tandem gene iteration at the CUP1 region.

The gene copy number at the CUP1 locus and the resistance level to external copper was directly correlated in five wild-type commercial Saccharomyces strains. An increased copy number of the CUP1 gene leads to increased accumulation of chelatin mRNA, which codes for a low-molecular-weight, copper-binding protein. The enhanced production of this rapidly inducible protein mediates resistance of the cell to copper. Industrial yeasts exhibit homologies to the amplified copper resistance repeat unit found in laboratory strains. However, the extent of tandem iteration is strain dependent, and the repetitious unit is either 1.7 or 1.5 kilobases in length compared with the 2.0-kilobase unit in laboratory strains. Strain 522 (Montrachet) contains two chromosome VIII segments distinguishable by their numbers of repeat units (2 and 11) and the size of the units (1.5 and 1.7 kilobases). Distillers yeast 513 carries a 1.5-kilobase repeat unit on each homologous chromosome, although they contain nine and five iterations, respectively.     

Analysis of the S-locus structure in Prunus armeniaca L. Identification of S-haplotype specific S-RNase and F-box genes

The gametophytic self-incompatibility (GSI) system in Rosaceae has been proposed to be controlled by two genes located in the S-locusan S-RNase and a recently described pollen expressed S-haplotype specific F-box gene (SFB). However, in apricot (Prunus armeniaca L.) these genes had not been identified yet. We have sequenced 21kb in total of the S-locus region in 3 different apricot S-haplotypes. These fragments contain genes homologous to the S-RNase and F-box genes found in other Prunusspecies, preserving their basic gene structure features and defined amino acid domains. The physical distance between the F-boxand the S-RNase genes was determined exactly in the S ₂-haplotype (2.9kb) and inferred approximately in the S ₁-haplotype (< 49kb) confirming that these genes are linked. Sequence analysis of the 5 flanking regions indicates the presence of a conserved region upstream of the putative TATA box in the S-RNase gene. The three identified S-RNase alleles (S ₁, S ₂ and S ₄) had a high allelic sequence diversity (75.3 amino acid identity), and the apricot F-box allelic variants (SFB₁, SFB₂ and SFB₄) were also highly haplotype-specific (79.4 amino acid identity). Organ specific-expression was also studied, revealing that S ₁- and S ₂-RNases are expressed in style tissues, but not in pollen or leaves. In contrast, SFB ₁ and SFB ₂ are only expressed in pollen, but not in styles or leaves. Taken together, these results support these genes as candidates for the pistil and pollen S-determinants of GSI in apricot.     

Isolation of S-locus F-box alleles in Prunus avium and their application in a novel method to determine self-incompatibility genotype

This study characterises a series of 12 S-locus haplotype-specific F-box protein genes (SFB) in cherry (Prunus avium) that are likely candidates for the pollen component of gametophytic self-incompatibility in this species. Primers were designed to amplify 12 SFB alleles, including the introns present in the 5′ untranslated region; sequences representing the S-alleles S ₁ , S ₂ , S ₃ , S ₄ , S ₄ ′, S ₅ , S ₆ , S ₇ , S ₁₀ , S ₁₂ , S ₁₃ and S ₁₆ were cloned and characterized. [The nucleotide sequences reported in this paper have been submitted to the EMBL/GenBank database under the following accession numbers: PaSFB ₁ (AY805048), PaSFB ₂ (AY805049), PaSFB ₃ (AY805057), PaSFB ₄ (AY649872), PaSFB ₄ ′ (AY649873), PaSFB ₅ (AY805050), PaSFB ₆ (AY805051), PaSFB ₇ (AY805052), PaSFB ₁₀ (AY805053), PaSFB ₁₂ (AY805054), PaSFB ₁₃ (AY805055), PaSFB ₁₆ (AY805056).] Though the coding regions of six of these alleles have been reported previously, the intron sequence has previously been reported only for S ₆ . Analysis of the introns revealed sequence and length polymorphisms. A novel, PCR-based method to genotype cultivars and wild accessions was developed which combines fluorescently labelled primers amplifying the intron of SFB with similar primers for the first intron of S-RNase alleles. Intron length polymorphisms were then ascertained using a semi-automated sequencer. The convenience and reliability of this method for the determination of the self-incompatibility (SI) genotype was demonstrated both in sweet cherry cultivars representing alleles S ₁ to S ₁₆ and in individuals from a wild population encompassing S-alleles S ₁₇ to S ₂₂ . This method will greatly expedite SI characterisation in sweet cherry and also facilitate large-scale studies of self-incompatibility in wild cherry and other Prunus populations.     

Fbh1 Limits Rad51-Dependent Recombination at Blocked Replication Forks

Controlling the loading of Rad51 onto DNA is important for governing when and how homologous recombination is used. Here we use a combination of genetic assays and indirect immunofluorescence to show that the F-box DNA helicase (Fbh1) functions in direct opposition to the Rad52 orthologue Rad22 to curb Rad51 loading onto DNA in fission yeast. Surprisingly, this activity is unnecessary for limiting spontaneous direct-repeat recombination. Instead it appears to play an important role in preventing recombination when replication forks are blocked and/or broken. When overexpressed, Fbh1 specifically reduces replication fork block-induced recombination, as well as the number of Rad51 nuclear foci that are induced by replicative stress. These abilities are dependent on its DNA helicase/translocase activity, suggesting that Fbh1 exerts its control on recombination by acting as a Rad51 disruptase. In accord with this, overexpression of Fbh1 also suppresses the high levels of recombinant formation and Rad51 accumulation at a site-specific replication fork barrier in a strain lacking the Rad51 disruptase Srs2. Similarly overexpression of Srs2 suppresses replication fork block-induced gene conversion events in an fbh1Δ mutant, although an inability to suppress deletion events suggests that Fbh1 has a distinct functionality, which is not readily substituted by Srs2.Homologous recombination (HR) is often described as a double-edged sword: it can maintain genome stability by promoting DNA repair, while its injudicious action can disturb genome stability by causing gross chromosome rearrangement (GCR) or loss of heterozygosity (LOH). Both GCR and LOH are potential precursors of diseases such as cancer, and consequently there is need to control when and how HR is used.A key step in most HR is the loading of the Rad51 recombinase onto single-stranded DNA (ssDNA), which forms a nucleoprotein filament (nucleofilament) that catalyzes the pairing of homologous DNAs and subsequent strand invasion (32). This is a critical point at which recombination can be regulated through the removal of the Rad51 filament (60). Early removal can prevent strand invasion altogether, freeing the DNA for alternative processing. Later removal may limit unnecessary filament growth, free the 3′-OH of the invading strand to prime DNA synthesis, and ultimately enable ejection of the invading strand, which is important for the repair of double-strand breaks (DSBs) by synthesis-dependent strand annealing (SDSA). SDSA avoids the formation of Holliday junctions that can be resolved into reciprocal exchange products (crossovers), which may result in GCR or LOH if the recombination is ectopic or allelic, respectively.One enzyme that appears to be able to control Rad51 in the aforementioned manner is the yeast superfamily 1 DNA helicase Srs2 (42). In Saccharomyces cerevisiae, Srs2 is recruited to stalled replication forks by the SUMOylation of PCNA, and there it appears to block Rad51-dependent HR in favor of Rad6- and Rad18-dependent postreplication repair (1, 2, 35, 50, 53, 58). In vitro Srs2 can strip Rad51 from ssDNA via its DNA translocase activity (31, 62) and therefore probably controls HR at stalled replication forks by acting as a Rad51 disruptase. In accord with this, chromatin immunoprecipitation analysis has shown that Rad51 is enriched at or near replication forks in an srs2 mutant (50). Srs2 also plays an important role in crossover avoidance during DSB repair, where it is thought to promote SDSA by both disrupting Rad51 nucleofilaments and dissociating displacement (D) loops (20, 27).Srs2 is conserved in the fission yeast Schizosaccharomyces pombe (19, 43, 63) and has a close relative in bacteria called UvrD, which can similarly control HR by disrupting RecA nucleofilaments (61). However, an obvious homologue in mammals has not been detected. Recently, two mammalian members of the RecQ DNA helicase family, BLM and RECQL5, were shown to disrupt Rad51 nucleofilaments in vitro (11, 25), although in the case of BLM, this activity appears to be relatively weak (5, 55). Nevertheless these data have led to speculation that both BLM and RECQL5 might perform a function similar to that of Srs2 in vivo (6). Certainly mutational inactivation of either helicase results in elevated levels of HR and genome instability, with an associated increased rate of cancer (23, 25). However, BLM and RECQL5 are not the only potential Rad51 disruptases in mammals; a relative of Srs2 and UvrD called FBH1 was recently implicated in this role by genetic studies of its orthologue in S. pombe and by its ability to partially compensate for the loss of Srs2 in S. cerevisiae, which, unlike S. pombe, lacks an FBH1 orthologue (15). FBH1 is so named because of an F box near its N terminus—a feature that makes it unique among DNA helicases (28). The F box is important for its interaction with SKP1 and therefore the formation of an E3 ubiquitin ligase SCF (SKP1-Cul1-F-box protein) complex (29). The targets of this complex are currently unknown. In S. pombe, mutations within Fbh1''s F-box block interaction with Skp1 and prevent Fbh1 from localizing to the nucleus and forming damage-induced foci therein (57). Fbh1''s role in constraining Rad51 activity in S. pombe is evidenced by the increase in spontaneous Rad51 foci and accumulation of UV irradiation-induced Rad51-dependent recombination intermediates in an fbh1Δ mutant (47). Moreover, loss of both Fbh1 and Srs2 in S. pombe results in a synergistic reduction in cell viability, and like Srs2, Fbh1 is essential for viability in the absence of the S. pombe RecQ family DNA helicase Rqh1, which processes recombination intermediates (47, 48). In both cases the synthetic interaction is suppressed by deleting rad51, suggesting that Fbh1 works in parallel with Srs2 and Rqh1 to prevent the formation of toxic recombination intermediates. In yeast, Rad51-mediated recombination is dependent on Rad52 (Rad22 in S. pombe), which is believed to promote the nucleation of Rad51 onto DNA that is coated with the ssDNA binding protein replication protein A (RPA) (18, 32). Intriguingly, the genotoxin sensitivity and recombination deficiency of a rad22 mutant are suppressed in a Rad51-dependent manner by deleting fbh1 (48). This suggests that Fbh1 and Rad22 act in opposing ways to modulate the assembly of the Rad51 nucleofilament. Although current data indicate a role for Fbh1 in controlling HR, the only evidence so far that Fbh1 limits recombinant formation is in chicken DT40 cells, for which a modest increase in sister chromatid exchange has been noted when FBH1 is deleted (30).Here we present in vivo evidence suggesting that Fbh1 does indeed act as a Rad51 disruptase, which is dependent on its DNA helicase/translocase activity. We confirm predictions that this activity works in opposition to Rad22 for the loading of Rad51 onto DNA and show that Fbh1''s modulation of Rad51 activity, while not essential for limiting spontaneous direct-repeat recombination, is critical for preventing recombination at blocked replication forks. Finally, we highlight similarities and differences between Fbh1 and Srs2, based on their mutant phenotypes and relative abilities to suppress recombination when overexpressed. Overall our data affirm that Fbh1 is one of the principal modulators of Rad51 activity in fission yeast and therefore may play a similar role in vertebrates.     

Recombination between paralogues at the Rp1 rust resistance locus in maize

Rp1 is a complex rust resistance locus of maize. The HRp1-D haplotype is composed of Rp1-D and eight paralogues, seven of which also code for predicted nucleotide binding site-leucine rich repeat (NBS-LRR) proteins similar to the Rp1-D gene. The paralogues are polymorphic (DNA identities 91-97%), especially in the C-terminal LRR domain. The remaining family member encodes a truncated protein that has no LRR domain. Seven of the nine family members, including the truncated gene, are transcribed. Sequence comparisons between paralogues provide evidence for past recombination events between paralogues and diversifying selection, particularly in the C-terminal half of the LRR domain. Variants selected for complete or partial loss of Rp1-D resistance can be explained by unequal crossing over that occurred mostly within coding regions. The Rp1-D gene is altered or lost in all variants, the recombination breakpoints occur throughout the genes, and most recombinant events (9/14 examined) involved the same untranscribed paralogue with the Rp1-D gene. One recombinant with a complete LRR from Rp1-D, but the amino-terminal portion from another homologue, conferred the Rp1-D specificity but with a reduced level of resistance.     

Recombination block in the Spore killer region of Neurospora

Spore killers Sk-2K and Sk-3K are chromosomal meiotic drive factors in Neurospora. In heterozygous crosses, ascospores not containing the Spore killer die. Sk-2K and Sk-3K, which differ in killing specificity, were found to be associated with suppression of recombination in a centromere-spanning region of linkage group III, and investigation of that recombination block is reported here. The block covers a region that is normally 30 to 40 map units long. A locus (r(Sk-2)) conferring resistance to Sk-2K maps to the left end of the recombination block. Recombination is normal in r(Sk-2) X Sk sensitive but blocked in Sk-2K X r(Sk-2); so the block does not depend upon killing. By selective plating, SkK stocks carrying genetic markers within the block were obtained at frequencies on the order of 10(-5) or 10(-6). Since this tight block is far beyond what has been observed for genetic reduction of recombination, a structural basis is assumed. No evidence of chromosome rearrangement was obtained. Crosses homozygous for Sk-2K show normal crossing-over and map order for the flanking markers cum and his-7 and three included markers (acr-7, acr-2, and leu-1). Results would be consistent with a divergence of sequence great enough to interfere with homologous pairing.