The chromatin assembly factor subunit FASCIATA1 is involved in homologous recombination in plants

Abstract

DNA replication in cycling eukaryotic cells necessitates the reestablishment of chromatin after nucleosome redistribution from the parental to the two daughter DNA strands. Chromatin assembly factor 1 (CAF-1), a heterotrimeric complex consisting of three subunits (p150/p60/p48), is one of the replication-coupled assembly factors involved in the reconstitution of S-phase chromatin. CAF-1 is required in vitro for nucleosome assembly onto newly replicated chromatin in human cells and Arabidopsis thaliana, and defects in yeast (Saccharomyces cerevisiae) affect DNA damage repair processes, predominantly those involved in genome stability. However, in vivo chromatin defects of caf-1 mutants in higher eukaryotes are poorly characterized. Here, we show that fasciata1-4 (fas1-4), a new allele of the Arabidopsis fas1 mutant defective in the p150 subunit of CAF-1, has a severe developmental phenotype, reduced heterochromatin content, and a more open conformation of euchromatin. Most importantly, homologous recombination (HR), a process involved in maintaining genome stability, is increased dramatically in fas1-4, as indicated by a 96-fold stimulation of intrachromosomal HR. Together with the open conformation of chromatin and the nearly normal expression levels of HR genes in the mutant, this result suggests that chromatin is a major factor restricting HR in plants.

Figure 1.

The fas1-4 Mutant. (A) Visible phenotypes of C24 wild-type plants and fas1-4 at the stage of seed set. Mutants (right) are considerably smaller than C24 wild-type plants (left). (B) to (E) Magnifications show the reduced size and developmental abnormalities in leaves (B), inflorescences (C), siliques (D), and flowers (E). Bars = 1 cm in (A) to (D) and 1 mm in (E). (F) Structure of the FAS1 gene as deduced from published cDNA and genomic sequences. Exons are represented by arrows; the position of the T-DNA insertion in the sixth exon in the mutant FAS1 gene is indicated (top). The FAS1 construct (bottom) used for complementation by Agrobacterium tumefaciens–mediated transformation of the mutant constitutes the entire gene, including the upstream region. Small boxes indicate T-DNA borders. (G) Toluidine blue–stained cross sections through mature leaves like those shown in (B).

Figure 2.

HR Is Stimulated in fas1-4. (A) The HR reporter of line N1IC4 651. The line harbors two complementary half genes (UG, US) of the GUS reporter gene as an inverted repeat separated by a functional hygromycin resistance gene (Hyg^R). The open arrows indicate the orientation of the repeats. The black arrows symbolize the cauliflower mosaic virus promoter driving the GUS gene. ICHR generates an intact GUS gene. (B) Quantitative analysis of ICHR in fas1-4, the wild type, and complemented mutants. The number of GUS-positive sectors in N1IC4 651, fas1-4×N1IC4 651, and three independent complementation lines (fas1-4/C1×651, fas1-4/C2×651, and fas1-4/C3×651) was counted and the data normalized and plotted. On average, fas1-4 seedlings showed a 96-fold (minimum, 85-fold; maximum, 108-fold) increase in the number of blue sectors over the reporter line N1IC4 651. The number of recombination events in mutants complemented with the FAS1 gene is comparable to that of the N1IC4 651 reporter line. Data are means of three experiments, and error bars show sd. (C) and (D) Typical patterns of staining for in vivo GUS expression obtained for fas1-4×N1IC4 651 (C) and N1IC4 651 (D). (E) Close-up view of a fas1-4×N1IC4 651 seedling showing the narrow true leaves and the global growth reduction of the mutant. (F) An N1IC4 651 seedling with well-developed true leaves is shown for comparison.

Figure 3.

fas1-4 Causes Loss of Heterochromatin. (A) and (B) Images of DAPI-stained wild-type (A) and fas1-4 (B) nuclei of mature leaves visualized by confocal microscopy. The chromocenters in the majority of wild-type nuclei have sharply defined boundaries. By contrast, the boundaries in fas1-4 nuclei are more blurred. (C) The size of chromocenters is reduced in fas1-4. Nuclei from wild-type and fas1-4 leaf cells were prepared, stained with DAPI, and spread. The staining intensity in chromocenters and the remaining nuclear area was quantified by digital imaging, and the ratios were calculated. Because chromocenters represent mainly heterochromatin and the remaining staining represents euchromatin, values are expressed as the ratio of heterochromatin to euchromatin. Data are means of three experiments, and error bars show sd.

Figure 4.

Methylation of Centromeric Heterochromatic DNA Is Unaffected in fas1-4. Genomic DNA isolated from wild-type and fas1-4 leaves was digested with the methylation-sensitive enzyme HpaII, its methylation-insensitive isoschizomer MspI, and the methylation-insensitive enzyme DraI. All three restriction enzymes release a similar fragment from the Arabidopsis centromeric repeat. After separation by gel electrophoresis, the DNA was blotted onto nylon membranes and hybridized with a 180-bp centromeric repeat fragment (Vongs et al., 1993).

Figure 5.

fas1-4 Affects Chromatin Conformation. (A) Chromatin is globally more sensitive to DNase I in fas1-4. Nuclei were isolated from N1IC4 651 and fas1-4×N1IC4 651 and treated with DNase I for the times indicated (in minutes; U = untreated control). DNA was then extracted and separated by agarose gel electrophoresis. (B) Scheme of Arabidopsis chromosomes 2, 3, 4, and 5. The positions of the RFLP markers (numbers in open circles) in pARMS2, pARMS5, and pARMS7, the positions of markers determined by sequence alignment (numbers in gray circles), and chromosomal landmark markers are shown. The centromere is shown as a closed bar. (C) Euchromatic regions in fas1-4 are hypersensitive to DNase I. Nuclei from N1IC4 651 and fas1-4×N1IC4 651 mixed with nuclei from Col-0 were digested with DNase I for the times indicated (in minutes; U = untreated control). The DNA was extracted, digested with EcoRI, separated by gel electrophoresis, blotted onto a nylon membrane, and probed with ARMS2, ARMS5, and ARMS7. All chromosomal loci are considerably more sensitive to DNase I in fas1-4×N1IC4 651 than in N1IC4 651. (D) The recombination reporter locus is hypersensitive to DNase I in fas1-4. The membrane used in (C) was reprobed with a GUS gene fragment.

Figure 6.

Effects of fas1-4 on Nuclear DNA Content, Homolog Pairing, and HR Gene Expression. (A) The frequency of homolog pairing remains unaffected in fas1-4. Examples of 2C leaf nuclei of the C24 wild type from a single-point pairing analysis by FISH. The scheme of chromosome 3 (top) indicates the position of the BAC sequences used for analysis. The left panel shows simultaneous homologous pairing of two distant segments (F18C1 and MGL6), and the right panel shows complete separation. Bars = 5 μm. (B) Flow cytometric DNA content histogram of C24 wild type (left) and fas1-4 (right) nuclei. During sorting, the number of nuclei was plotted against DNA content. Although most wild-type leaf nuclei have a content of 2C and 4C, this ratio is shifted toward 4C and 8C in fas1-4 leaf nuclei. Correspondingly, the cell cycle value, indicating the mean number of endoreduplication cycles per nucleus (Barow and Meister, 2003), is shifted from 0.716 in C24 to 1.419 in fas1-4, indicating duplication of the DNA content in fas1-4 cells. (C) Relative transcript levels of representative genes involved in HR in fas1-4. The results of a quantitative RT-PCR analysis after quantification and normalization to wild-type expression levels are shown for Arabidopsis RAD50, MRE11, and RAD51. The data are means of three experiments, and error bars show sd.