INCURVATA2 encodes the catalytic subunit of DNA Polymerase alpha and interacts with genes involved in chromatin-mediated cellular memory in Arabidopsis thaliana

Abstract

Cell type-specific gene expression patterns are maintained by the stable inheritance of transcriptional states through mitosis, requiring the action of multiprotein complexes that remodel chromatin structure. Genetic and molecular interactions between chromatin remodeling factors and components of the DNA replication machinery have been identified in Schizosaccharomyces pombe, indicating that some epigenetic marks are replicated simultaneously to DNA with the participation of the DNA replication complexes. This model of epigenetic inheritance might be extended to the plant kingdom, as we report here with the positional cloning and characterization of INCURVATA2 (ICU2), which encodes the putative catalytic subunit of the DNA polymerase alpha of Arabidopsis thaliana. The strong icu2-2 and icu2-3 insertional alleles caused fully penetrant zygotic lethality when homozygous and incompletely penetrant gametophytic lethality, probably because of loss of DNA polymerase activity. The weak icu2-1 allele carried a point mutation and caused early flowering, leaf incurvature, and homeotic transformations of sepals into carpels and of petals into stamens. Further genetic analyses indicated that ICU2 interacts with TERMINAL FLOWER2, the ortholog of HETEROCHROMATIN PROTEIN1 of animals and yeasts, and with the Polycomb group (PcG) gene CURLY LEAF. Another PcG gene, EMBRYONIC FLOWER2, was found to be epistatic to ICU2. Quantitative RT-PCR analyses indicated that a number of regulatory genes were derepressed in the icu2-1 mutant, including genes associated with flowering time, floral meristem, and floral organ identity.

Figure 1.

Leaf and Flower Phenotypic Traits of the icu2-1 Mutant. (A) and (B) Excised cotyledons (left) and leaves. (C) and (D) Scanning electron micrographs of the central region of the adaxial surface of third node leaf laminae (obtained as described in Serrano-Cartagena et al., 2000). (E) Flowers. (A) and (C) The En-2 wild type. (B), (D), and (E) icu2-1/icu2-1 plants. A patch of epidermal cells with reduced size is highlighted in (D). Arrows in (E) indicate partial homeotic transformations of sepals into carpels (yellow) and petals into stamens (red). Bars = 1 mm in (A), (B), and (E) and 100 μm in (C) and (D).

Figure 2.

Positional Cloning and Structural Analysis of the ICU2 Gene. (A) Map-based strategy followed for the cloning of the icu2-1 mutation. A mapping population of 1000 F2 plants, derived from icu2-1/icu2-1 × Landsberg erecta (Ler) and icu2-1/icu2-1 × Columbia (Col-0) crosses (the icu2-1 mutation was in an En-2 genetic background) was used to delimit ICU2 to a region of 90 kb, which encompassed three overlapping transformation-competent artificial chromosome (Liu et al., 1999) clones. Several of the candidate genes were amplified and sequenced, allowing us to develop and use for further analyses seven previously unidentified single nucleotide polymorphism (SNP) markers (SNPk8a10-a and SNPk21h1-b to SNPk21h1-g; see Table 2), which were polymorphic between En-2 and either Ler or Col-0. This reduced the candidate region to 50 kb, all whose putative transcription units (http://mips.gsf.de/proj/thal/db/index.html) were PCR amplified and sequenced. Only the At5g67100 gene, which encodes the catalytic subunit of DNA polymerase α, was found to carry a point mutation in the icu2-1 mutant. The number of informative recombinants identified is indicated in parentheses. In the representation of the structure of the ICU2 gene, exons are indicated by boxes, introns by lines between boxes, and T-DNA (icu2-2) and Ds (icu2-3) insertions by triangles. cM, centimorgan. (B) Predicted amino acid sequence of the catalytic subunit of DNA polymerase α in Arabidopsis. The domains named as I to VII in Wang et al. (1989) and A to E in Miyazawa et al. (1993) are boxed. The region assumed to interact with primase is bordered by arrowheads (Biswas et al., 2003). The putative HP1 binding pentamere (MIR domain) is indicated by a dotted line, the position of the icu2-1 mutation by an asterisk, and the icu2-2 (T-DNA) and icu2-3 (Ds) insertions by triangles. Residues shaded in black and gray indicate the identities and similarities, respectively, found after the alignment of the sequence of the Arabidopsis DNA polymerase α catalytic subunit (GI 15240200) with those of Homo sapiens (8393995), Mus musculus (6679409), D. melanogaster (217344), Caenorhabditis elegans (32565317), O. sativa (6015010), S. pombe (6018683), and Neurospora crassa (32416196). The sequences were aligned using the ClustalX 1.5b program (Thompson et al., 1997) and shaded with BOX SHADE 3.21 (http://www.ch.embnet.org/software/BOX_form.html). The gaps generated by the alignment have been removed. (C) and (D) Models for the three-dimensional structure of the protein products of the En-2 wild-type (C) and icu2-1 (D) alleles of the ICU2 gene. The residue affected by the mutation is in green, and the HP1 binding domain is in blue. The three-dimensional models were obtained using the 3D-JIGSAW (Bates et al., 2001; http://www.bmm.icnet.uk/∼3djigsaw/) and RasMol 2.7 (Sayle and Milner-White, 1995; http://www.umass.edu/microbio/rasmol/) programs.

Figure 3.

Other Phenotypic Traits of icu2 Alleles and Suppression of Some of Them by ag and ft Mutations. (A) and (B) Rosettes from a phenotypically wild-type transgenic line carrying the pG-ICU2 wild-type transgene in an icu2-1/icu2-1 background, demonstrating the phenotypic rescue of icu2-1 by the wild-type allele of At5g67100 (A) and an icu2-3/icu2-1 heterozygote (B), which is indistinguishable to the icu2-1/icu2-1 homozygotes. (C) to (E) Dissected siliques from selfed Col-0 (C) and heterozygous icu2-2/ICU2 ([D]; in a Col-0 genetic background) plants. Arrows in (D) indicate abnormal seeds that are likely to be aborted or unfertilized ovules. One of the latter is magnified in (E) as seen with Nomarski optics and did not display any embryonic structure, which was clearly visible in the morphologically normal seeds. (F) to (H) Suppression of some of the phenotypic traits of the icu2-1 mutation by loss-of-function mutations in FT and AG. (F) and (G) Early flowering is suppressed in icu2-1/icu2-1;ft-1/ft-1. (H) Leaf incurvature is suppressed in icu2-1/icu2-1;ag-1/ag-1 double mutant plants. Images were taken 21 ([A], [B], [G], and [H]), 40 ([C] to [E]), and 35 (F) d after sowing. Bars = 2 mm in (A), (B), (G), and (H), 1 mm in (C) and (D), 0.1 mm in (E), and 5 cm in (F).

Figure 4.

Genetic Interactions between Mutations Affecting ICU2 and Genes Involved in Chromatin-Mediated Cellular Memory. Rosettes are shown for the wild-type En-2 (A) and single ([B] to [D] and [H]) and double mutants ([E] to [G] and [I]). All plants shown were homozygous for the indicated mutations. Images were taken 21 d after sowing. Bars = 2 mm.

Figure 5.

Genetic Interactions between Mutations Affecting ICU2 and FAS1. Plants are shown for single ([A], [B], and [E]) and double ([C], [D], and [F]) mutants. A magnification of the lateral view of (C) is highlighted in (D). All plants shown were homozygous for the indicated mutations, and all the mutants are in an En-2 background. Images were taken 10 d ([A] to [D]) and 2 months ([E] and [F]) after sowing. Bars = 1 mm.

Figure 6.

QRT-PCR Analyses. Quantifications are shown for the transcript levels of ICU2 ([A] and [B]) in assorted tissues of the wild-type En-2 (A) and rosettes and shoot apices of En-2 and the icu2-1 mutant (B), as well as for those of the indicated genes in leaves (C) or whole rosettes ([D] and [E]) of the icu2-1 mutant ([C] and [D]) and some other single and double mutants (E). All plants were homozygous for the mutations indicated. Plant material was collected 21 d after sowing, with the roots being removed before the extraction of rosette RNA. All data were referred (using the 2^−ΔΔC_T method) to those obtained for the En-2 wild type, to which a value of 1 was given.

Figure 7.

ChIP Assay. The histone methylation and acetylation patterns of the promoter (AG P) and second intron (AG 2I) of AG in the En-2 wild type and the icu2-1 mutant (lanes headed as icu2) were tested. ChIP duplex PCR was used to amplify the ORNITHINE TRANSCARBAMILASE (OTC) gene and regions of the AG gene. Numbers below gel lanes indicate the ratio of the intensity of AG products compared with that of OTC products after normalization with input product intensity (Schubert et al., 2006).

Figure 8.

GST Pull-Down Assay Demonstrating Direct Interaction between the TFL2 and ICU2 Proteins. The carboxylic half of the wild-type ICU2 protein (ICU2-ct) and that of its mutated version (ICU2-1-ct) were translated in vitro in the presence of ³⁵S-Met. GST-TFL2, but not GST, was able to interact with the full-length ICU2-ct as well as with a truncated ICU2-1-ct protein (indicated by an asterisk). GST-TFL2 did not interact with the full-length ICU2-1-ct. The input lane shows the signal from 10% of the amount of in vitro–translated ICU2-ct and ICU2-1-ct proteins present in each of the remaining samples.

Figure 9.

AG and TFL2 Spatial Expression Patterns in the icu2-1 Mutant. (A) to (D) GUS staining of leaves of transgenic plants carrying the pAG-I:GUS construct in En-2 (A) and icu2-1/icu2-1 ([B] to [D]) backgrounds. (E) to (H) Confocal micrographs of leaves from transgenic plants carrying the gTFL2:GFP construct in En-2 (E) and icu2-1/icu2-1 ([F] to [H]) backgrounds. All plants shown were homozygous for the indicated mutations. Images were taken 21 d after sowing. Bars = 1 mm in (A) and (B) and 100 μm in (C) to (H).