Genetic Control of Developmental Changes Induced by Disruption of Arabidopsis Histone Deacetylase 1 (AtHD1) Expression

Corresponding author: Z. Jeffrey Chen, Department of Soil and Crop Sciences, Texas A&M University, College Station, TX 77843-2474., zjchen{at}tamu.edu (E-mail)

Communicating editor: J. A. BIRCHLER

	ABSTRACT

TOP ABSTRACT MATERIALS AND METHODS RESULTS DISCUSSION LITERATURE CITED

Little is known about the role of genetic and epigenetic control in the spatial and temporal regulation of plant development. Overexpressing antisense Arabidopsis thaliana HD1 (AtHD1) encoding a putative major histone deacetylase induces pleiotropic effects on plant growth and development. It is unclear whether the developmental abnormalities are caused by a defective AtHD1 or related homologs and are heritable in selfing progeny. We isolated a stable antisense AtHD1 (CASH) transgenic line and a T-DNA insertion line in exon 2 of AtHD1, resulting in a null allele (athd1-t1). Both athd1-t1 and CASH lines display increased levels of histone acetylation and similar developmental abnormalities, which are heritable in the presence of antisense AtHD1 or in the progeny of homozygous (athd1-t1/athd1-t1) plants. Furthermore, when the athd1-t1/athd1-t1 plants are crossed to wild-type plants, the pleiotropic developmental abnormalities are immediately restored in the F₁ hybrids, which correlates with AtHD1 expression and reduction of histone H4 Lys12 acetylation. Unlike the situation with the stable code of DNA and histone methylation, developmental changes induced by histone deacetylase defects are immediately reversible, probably through the restoration of a reversible histone acetylation code needed for the normal control of gene regulation and development.

PLANT development is plastic and affected by genetic, epigenetic, and environmental factors. Vegetative and reproductive (inflorescence) development is initiated at shoot apical meristems and/or axillary meristems that can be induced by internal and external signals (BERNIER 1986 ; WALBOT 1996 ; BLEECKER and PATTERSON 1997 ; MEYEROWITZ 1997 ). The molecular mechanisms underlying the plastic nature of plant development are largely unknown. Both genetic and epigenetic changes may contribute to the developmental plasticity of plants (WALBOT 1996 ; MEYEROWITZ 1997 , MEYEROWITZ 2002 ; SRINIVAS 2000 ; HABU et al. 2001 ; MARTIENSSEN and COLOT 2001 ).

Epigenetic regulation is a major aspect of gene control by which heritable changes in gene expression occur without an alteration in DNA sequence. Changes in chromatin structure may affect accessibility of promoter elements to the transcriptional machinery and thus affect transcription. Modifications on core histones and their associated covalent bonds are known as the "histone code" (JENUWEIN and ALLIS 2001 ). Changes in the histone code may facilitate fine tuning of gene expression in response to developmental programs or changes in environmental signals. Disruption of histone deacetylases results in growth and developmental abnormalities and aging in yeast cells (IMAI et al. 2000 ). These developmental changes are associated with down- or upregulation of several hundred genes (BERNSTEIN et al. 2000 ; ROBYR et al. 2002 ), suggesting that histone deacetylases are key regulators of eukaryotic development. Histone acetylation and deacetylation may also play a role in gene expression and development in animals. For example, mouse histone deacetylase 1 (HD1) is a growth factor-inducible gene (BARTL et al. 1997 ). Histone hyper-acetylation plays a role in establishing stable states of differential gene activity during gastrulation in Xenopus (ALMOUZNI et al. 1994 ).

The Arabidopsis genome contains 18 putative HDs (HDAs or HDACs) and 12 putative histone acetyltransferases (HATs) distributed among all five chromosomes (ARABIDOPSIS GENOME INITIATIVE 2000; PANDEY et al. 2002 ). There are four classes of histone deacetylases in plants (LUSSER et al. 2001 ). First, the RPD3-like protein is the major histone deacetylase in yeast and mammals. Mutations in RPD3 affect 500 genes in yeast (BERNSTEIN et al. 2000 ; ROBYR et al. 2002 ). Arabidopsis thaliana HD1 (AtHD1; GenBank accession no. AAB66486), also known as AtHDA19 (PANDEY et al. 2002 ), is a putative homolog of yeast RPD3 and the major histone deacetylase (HD1) in humans and mice (RUNDLETT et al. 1996 ; TAUNTON et al. 1996 ; BARTL et al. 1997 ; LUSSER et al. 1997 ). This class consists of at least two isoforms (HD1B-I and -II) in maize embryos (ROSSI et al. 1998 ; LECHNER et al. 2000 ) and four genes (HDA6, -7, -9, and -19) in Arabidopsis. One of the maize genes, ZmRpd3 or HD1B-II (LECHNER et al. 2000 ), complements a yeast rpd3 null mutant (ROSSI et al. 1998 ). Second, on the basis of DNA sequences, HDA- and HOS-like HDs are related to RPD3 but have different specific activities in deacetylating histones (VIDAL and GABER 1991 ; DE RUBERTIS et al. 1996 ; RUNDLETT et al. 1996 , RUNDLETT et al. 1998 ). Eight Arabidopsis genes in this category are predicted and some of them may be distinct from other members in this group (PANDEY et al. 2002 ). Third, HD2 is a plant-specific histone deacetylase (LUSSER et al. 1997 ) localized in the nucleolus. At least two isoforms exist in maize and four genes in Arabidopsis (LUSSER et al. 1997 ; DANGL et al. 2001 ). Fourth, a NAD-dependent HD (SIR2-like protein) forms a newly discovered class of HDs (IMAI et al. 2000 ; LANDRY et al. 2000 ). In yeast, the deacetylation by SIR2 is NAD dependent and possibly coupled to ADP ribosylation (TANNER et al. 2000 ). Arabidopsis has two SIR2-like genes (PANDEY et al. 2002 ).

The role of histone acetylation and deacetylation in plant gene regulation is poorly understood (VERBSKY and RICHARDS 2001 ; LI et al. 2002 ). Transgenic plants treated with propionic or butyric acid, chemical inhibitors of histone deacetylases, display increased levels of DNA methylation and epigenetic variegation (TEN LOHUIS et al. 1995 ). HC toxin, the host-selective toxin of the maize fungal pathogen Cochliobolus carbonum, inhibits histone deacetylases in host plants (BROSCH et al. 1995 ). Blocking deacetylation by sodium butyrate or trichostatin A derepresses silent rRNA genes subject to nucleolar dominance (CHEN and PIKAARD 1997 ). In a genetic screen for auxin-insensitive mutants, MURFETT et al. 2001 identified mutagenized plants with enhanced expression of gusA and hptII transgenes. Further analysis indicated that several of these mutations were in AtHDA6, a presumed histone deacetylase, suggesting that this gene is important for transcriptional regulation of the promoters controlling these transgenes. Indeed, AtHDA6 is required to maintain DNA methylation patterns induced by double-stranded RNA (AUFSATZ et al. 2002 ). Overexpression of OsHDAC1, a putative histone deacetylase 1 gene in rice, is correlated with the induction of OsHDAC1, increased growth rate, and altered plant morphology (JANG et al. 2003 ). Antisense-mediated downregulation of the Arabidopsis genes AtHD1 or AtHD2, which putatively encode histone deacetylases, results in a variety of abnormal phenotypes in early and late stages of Arabidopsis development (WU et al. 2000A , WU et al. 2000B ; TIAN and CHEN 2001 ).

Downregulation of histone deacetylation induces pleiotropic effects on Arabidopsis development (TIAN and CHEN 2001 ), suggesting that histone deacetylation directly or indirectly affects the expression of many genes in regulatory networks (FINNEGAN 2001 ). However, it is not known whether the abnormal phenotypes observed in the antisense AtHD1 plants result from the disruption of single gene or related homologs in the AtHD multi-gene family. Moreover, it is unclear whether disruption of histone deacetylation induces other epigenetic lesions and whether the induced phenotypic changes are heritable in the absence of the original AtHD1 defect. Finally, the acetylation code is reversible and dynamic because core histones can be acetylated or deacetylated through the activities of histone acetyltransferases (HATs) or histone deacetylases (HDs, HADs, HDACs) during growth and development (ALLFREY et al. 1964 ), whereas the code of DNA and histone methylation is stable because no active demethylation pathway has been identified (JENUWEIN and ALLIS 2001 ). However, both the stable and reversible codes can be stored in chromatin and propagated through meiosis. Thus, compared to the stable code, the reversible histone acetylation code may contribute differently to gene regulation and development. In this study, we show that defects induced by expressing antisense AtHD1, or by T-DNA insertion mutagenesis of AtHD1, result in similar developmental pleiotropy. Unlike the situation with ddm1 (decrease in DNA methylation) mutants, defects in AtHD1 do not induce cumulative epigenetic lesions after four to five generations of selfing. Furthermore, genetic analyses indicate that disruption of developmental programs, AtHD1 expression, and histone acetylation is immediately corrected in AtHD1/athd1-t1 heterozygous plants probably through restoration of the reversible histone acetylation code.

	MATERIALS AND METHODS

TOP ABSTRACT MATERIALS AND METHODS RESULTS DISCUSSION LITERATURE CITED

Plant materials:
Constitutive antisense histone deacetylase (CASH) transgenic plants were described previously (TIAN and CHEN 2001 ). All plants were grown in vermiculite mixed with 10% soil in a growth chamber with growth conditions of 22°/18° (day/night) and 16 hr of illumination per day, except as noted otherwise. Seeds from CASH plants were germinated in Murashige/Skoog medium (Sigma, St. Louis) in the presence of 15–50 µg/ml of kanamycin. After 2 weeks the plants were transferred to soil and grown in a growth chamber. The T-DNA insertion line was grown without kanamycin selection except as noted otherwise. All photographs, except those taken with the imaging system in a scanning electron microscope, were taken using a Nikon N-900 digital camera or the CCD system of a Nikon SMZ-100 fluorescence microscope.

Scanning electron microscopy:
Scanning electron microscopy was performed using a modified protocol (MURAI et al. 2002 ) with a Hitachi s570 scanning electron microscope (Tokyo) at an accelerating voltage of 5–15 kV. Tissue samples were fixed in a solution containing 3% (v/v) formaldehyde/glutaraldehyde and 0.1 M sodium cacodylate buffer (pH 7.4) at room temperature for 13 hr, washed three times with 0.2 M sodium cacodylate buffer (pH 7.4), dehydrated through an ethanol series, critical point dried with CO₂, and sputter coated with gold before viewing by scanning electron microscopy.

Nucleic acid preparation and analysis:
RNA and DNA were isolated from at least five leaves of each plant at the same developmental stages as previously described (CHEN and PIKAARD 1997 ). For DNA and RNA blot analyses, hybridization was performed following the method of CHURCH and GILBERT 1984 . The DNA and RNA blots were washed in 2x SSC and 0.2x SDS at 65° for 30–60 min and hybridization signals were detected using a digital imaging system or exposure to X-ray film or to a PhosphorImager.

PCR-based reverse genetic approach to identify T-DNA insertions in the AtHD1 gene:
We used a PCR-based approach similar to that described by KRYSAN et al. 1996 to identify Arabidopsis (ecotype Ws) mutants containing a T-DNA insertion in or near the AtHD1 gene. Pooled samples of DNA from 1000, 100, and 20 plants from the Feldmann T-DNA insertion library (FELDMANN and MARKS 1987 ; FORSTHOEFEL et al. 1992 ) were successively assayed for insertions, followed by assay of individual plants from the pool of 20 mutant plants. The zygosity of a mutant allele was determined using forward and reverse primers specific to the AtHD1 gene and one primer specific to the AtHD1 gene in combination with either a T-DNA left- or right-border primer. AtHD1 primer sequences were 5'-GCACTAGTGGCGCGCCATGGATACTGGCGGCAA-3' and 5'-GCAGATCTATTTAAATCGCCTGCTCCGCCCCACC-3'. T-DNA primer sequences were left border, 5'-GATGCACTCGAAATCAGCCAATTTTAGAC-3'; right border, 5'-TCCTTCAATCGTTGCGGTTCTGTCAGTTC-3'. PCR was carried out using 0.5 unit ExTaq (Takara, Berkeley, CA) DNA polymerase with a robocycler (Stratagene, La Jolla, CA) using one cycle of 95° for 5 min, 30–36 cycles of 94° for 40–60 sec, 56° for 1 min, and 72° for 3 min, plus an extension cycle of 72° for 10 min followed by 4° on hold. We used 0.24 µM of each primer and 0.2 mM of each dNTP and either 100 ng DNA (for screening superpools of 1000 plants) or 20 ng DNA (for screening pools of 100, 20, or individual plants) in a 50-µl final reaction volume.

T-DNA junction sequence identification in antisense transgenic lines:
T-DNA junction sequences in the CASH lines were identified using a modification of a procedure previously described (ZHOU et al. 1997 ). Briefly, 1 µg genomic DNA and 1 µg pBluescript plasmid DNA were separately digested at 37° overnight in a 40-µl reaction containing 10 units of PstI. The digested genomic DNA solution was extracted with phenol/chloroform/isoamylalcohol (25:24:1; Fisher) and precipitated using 2 volumes of ethanol and 1/10 vol of 3 M sodium acetate (pH 5.2). The linear form of pBluescript plasmid was purified using a Gelpure kit (GeneMate; ISC Bioexpress, Kaysville, UT). For DNA ligation, 250 ng of digested genomic DNA and 180 ng of digested pBluescript plasmid DNA were mixed in a 20-µl solution containing 3 units of T4 DNA ligase (Promega, Madison, WI) and incubated at 16° overnight.

Two consecutive PCR reactions were performed using three different primers. T3 primer (5'-AAT TAA CCC TCA CTA AAG GG-3') was derived from an endogenous sequence of the pBluescript plasmid. TR2 (5'-GAT GGG GAT CAG ATT GTC GTT TC-3') and TR3 (5'-GTC GTT TCC CGC CTT CAG TTT A-3') were two nested primers derived from the T-DNA sequence close to the right border. The first PCR reaction was performed using TR2 and T3 primers and 5 µl of ligation solution as template. After purification using a QIAquick PCR purification kit (QIAGEN, Chatsworth, CA), an aliquot of 2 µl PCR product from the first reaction was added to the second PCR reaction containing TR3 and T3 primers. Both PCR reactions were performed for 35 cycles with a program of 30 sec at 94° for denaturation, 30 sec at 52° for annealing, and 3 min at 72° for extension, followed by a final extension at 72° for 8 min. The products of the second PCR amplification were subjected to electrophoresis through a 1.0% agarose gel. The band containing the DNA fragment of interest was excised from the gel and the DNA was purified using a DNA purification kit (GeneMate). The purified PCR product was cloned into the plasmid pGEM T-easy (Promega) and sequenced using an ABI Prism Big Dye terminator cycle sequencing reaction kit (PE Applied Biosystems). The small fragment that appeared in one-step reverse transcriptase (RT)-PCR was cloned into pGEM T-easy (Promega). The plasmid DNA was isolated using a QIAprep spin miniprep kit (QIAGEN) and sequenced.

RT-PCR:
RT-PCR was carried out using 500 ng of total RNA prepared from leaf tissues. SuperScript one-step RT-PCR was performed using the Platinum Taq system (Invitrogen, San Diego) according to the manufacturer's instructions. The primers used for detecting AtHD1 transcripts were AtHD1-R, 5'-GCT TAC AAC AAC AAC AAC TCC AGA AAC TT-3' and AtHD1-F, 5'-AGA AAG CCA GAG AGA GAG AGA GAG ATC AT-3'. For detecting CYC2b transcripts, we used the following primers: Cyc2b-F, 5'-TCG GTG TAG AGA TGA AGA GAC AGA-3' and Cyc2b-R, 5'-GCA ACT AAA CCA ACA AGC TGA AGC-3'.

Strand-specific first-strand cDNAs were synthesized using 500 ng of total RNA and Omniscript reverse transcriptase (QIAGEN). AtHD1-R and AtHD1-F primers were used to synthesize sense and antisense strands of AtHD1, respectively. RT was performed at 37° for 60 min and the enzyme was then inactivated by incubation at 93° for 5 min. For antisense AtHD1 detection, following an initial denaturation step, 35 cycles of PCR were performed using the program of 94° for 30 sec, 57° for 30 sec, and 72° for 1.5 min with a final extension of 10 min at 72°. A 5-µl aliquot of PCR products was resolved by electrophoresis through a 1% agarose gel and subjected to DNA blot analysis. For sense AtHD1 analysis, PCR was performed using the same conditions as described above, except that 20 cycles were used. Hybridization was performed using a full-length AtHD1 cDNA as a probe. The actin gene Act2 (AN et al. 1996 ) was used as an internal control for quantification. Relative intensities of individual DNA fragments were measured using a PhosphorImager (Molecular Dynamics, Sunnyvale, CA).

Western blot analysis:
Western blot analysis was carried out according to TIAN and CHEN 2001 . In brief, crude protein and histone extracts were prepared and subjected to electrophoresis through 8 and 15% SDS polyacrylamide gels. Immunoblots were prepared and probed with antisera against the N-terminal portion of AtHD1 (TIAN and CHEN 2001 ) or with a site-specific antibody against histone H4 Lys12 (Upstate Biotechnology, Lake Placid, NY) and developed by enhanced chemiluminescence (ECL; Amersham, Buckinghamshire, UK).

	RESULTS

TOP ABSTRACT MATERIALS AND METHODS RESULTS DISCUSSION LITERATURE CITED

Isolation of a stable antisense AtHD1 transgenic plant:
AtHD1 is constitutively expressed throughout plant development, although a slightly high level of expression is detected in reproductive tissues. Expressing CASH in transgenic Arabidopsis plants results in the reduction of tetra-acetylated histone H4 and a variety of developmental abnormalities (TIAN and CHEN 2001 ). Many lines showed variable phenotypes and deceased in early developmental stages. To understand better the AtHD1 effects, we isolated a CASH126 line that displayed consistent phenotypes in selfing progeny. CASH126 plants displayed phenotypes in the vegetative and flowering stages in the fourth generation (Fig 1) similar to the "pinhead" phenotype observed in Argonaute mutants (MOUSSIAN et al. 1998 ; LYNN et al. 1999 ; MOREL et al. 2002 ). The plants had defective shoot apical meristems and grew slowly in vitro. After the plants were transferred to soil, cotyledonary leaves became bleached and yellow. The first pair of true leaves was very small, narrowly elongated, and lacked chlorophyll (Fig 1A and Fig B). Starting with the second pair of true leaves, leaf development appeared normal except for a delay of the transition from adult vegetative to inflorescence development (Fig 1C and Fig D). CASH126 plants developed additional true leaves, probably resulting from a developmental transition initiated from axillary meristems. Vegetative development could continue up to 4–5 weeks under long-day (16/8 hr of day/night) conditions. Eventually inflorescences developed from lateral meristems similar to those of wild-type plants. Although some inflorescence branches were sterile (Fig 1E), seeds that could be harvested and germinated and the resulting plants showed developmental abnormalities similar to those observed in plants of previous generations.

View larger version (103K): [in this window] [in a new window] [Download PPT slide]   Figure 1. Development and inheritance of phenotypes in a stable CASH126 plant. Only the plants in the fourth generation are shown. (A) A 7-day-old seedling displayed a defective SAM. (B) The defective SAM in the 10-day-old seedling developed a pair of lobed true leaves. (C) A lateral SAM was initiated after 3 weeks. (D) The transition from vegetative to flowering development was delayed. (E) In the flowering stage, the plants display low fertility. About 75% of siliques were small and contained few viable seeds.

One explanation for the abnormal phenotypes in CASH126 plants would be that insertion of the transgenes into the genome disrupted a locus important for plant development, such as a homeotic gene controlling flowering. The endogenous AtHD1 is a single-copy gene located on chromosome 4 (ARABIDOPSIS GENOME INITIATIVE 2000; TIAN and CHEN 2001 ). We analyzed the copy number of transgenes in CASH126 plants using DNA blot analysis and detected two fragments of the transgenes (Fig 2A and Fig B), indicating that these transgenic plants contain two copies of the AtHD1 transgene. We further used a PCR method (ZHOU et al. 1997 ) to identify the transgene insertion sites in the genome in CASH126. As expected, two fragments were amplified from DNA of transgenic plants. DNA sequencing results indicated that one of the transgene insertion sites was located 150 bp downstream of the CYC2b stop codon of chromosome 4. The insertion did not affect CYC2b RNA accumulation; we detected an approximately equal amount of CYC2b transcripts in the control and CASH126 plants (Fig 2C). The other insertion site remains unknown, as the sequenced fragment did not match Arabidopsis sequence in the database.

View larger version (36K): [in this window] [in a new window] [Download PPT slide]   Figure 2. Detection of AtHD1 transgenes in the genome and expression of the transgene and endogenous AtHD1 in CASH126 plants. (A) Map of the transgene construction indicating the nptII and antisense AtHD1 (asAtHD1) genes located between the left and right T-DNA borders (LB and RB, respectively). (B) DNA blot containing genomic DNA digested with BamHI and EcoRI was hybridized with the nptII probe. Two copies of the transgene containing nptII and AtHD1 were detected in CASH126 plants (lanes 2 and 3), whereas no transgene was present in the control plant (lane 1). (C) One copy of the asAtHD1 transgene is inserted into the 3' untranslated region (UTR) of the CYC2b gene, as shown in the top diagram. RT-PCR analysis indicated that CYC2b expression was similar in control (lane 2) and CASH126 (lane 3) plants. DNA size markers are shown in lane 1. PCR product amplified from a genomic DNA is shown in lane 4. (D) Strand-specific RT-PCR analysis in CASH126 plants. (Top) asAtHD1 was highly expressed in CASH 126 plants (lanes 3–5) and undetectable in control plants (ecotype Columbia, lane 1, and ecotype Ws, lane 2). (Middle) RT-PCR and DNA blot analyses were performed to detect the expression levels of sense AtHD1 (sAtHD1). The strand-specific RT-PCR product was resolved by electrophoresis through a 1.0% agarose gel and transferred onto a Hybond-N+ membrane (Amersham Pharmacia). The blot was hybridized with an AtHD1 full-length cDNA probe. The relative expression levels (%) of sense AtHD1 (sAtHD1) transcripts were estimated using Act2 as an internal control.

We used strand-specific RT-PCR to determine the expression levels of antisense and endogenous AtHD1 genes in transgenic and control plants. Antisense AtHD1 (asAtHD1) transcripts were abundant in each of three CASH126 plants that were derived from single-seed descent, but absent in control plants (Fig 2D, lanes 1 and 2). To determine the expression levels of the endogenous sense AtHD1 (sAtHD1) transcript, we performed a semiquantitative RT-PCR analysis using Act2 (AN et al. 1996 ) as an internal control. sAtHD1 and Act2 transcripts were amplified and subjected to DNA blot analysis, and the relative expression ratios of sAtHD1 and Act2 transcripts were calculated in control and CASH126 plants. The results indicated that endogenous sense AtHD1 transcripts were greatly reduced in the CASH126 plants, ranging from 5 to 60% of the level detected in the control plants. It is notable that sAtHD1 RNA levels were inversely correlated with the amount of asAtHD1 transcripts (Fig 2D: compare lanes 3–5). Overexpressing the antisense AtHD1 gene reduced the accumulation of endogenous AtHD1 transcripts, suggesting that the penetrance of developmental abnormalities is correlated with the level of sAtHD1 transcript.

A T-DNA insertion in AtHD1 results in an AtHD1 null mutation:
Phenotypic abnormalities in transgenic plants may fluctuate because of variable levels of transgene expression (Fig 2D). Moreover, although downregulation of AtHD1 in antisense transgenic plants results in a variety of developmental abnormalities, it is unclear whether these phenotypic changes result from disrupted expression of AtHD1. The expression of antisense AtHD1 may also affect the expression of other genes, because there are several AtHD1 homologs (PANDEY et al. 2002 ). We therefore identified an Arabidopsis (ecotype Ws) mutant line that contains a T-DNA insertion in the AtHD1 gene (athd1-t1; Fig 3A). DNA blot analysis indicated only one T-DNA insertion in the genome (Fig 3B). Using a PCR-based reverse genetic approach (KRYSAN et al. 1996 ) and DNA sequence analysis, we determined that the T-DNA was inserted into exon 2 of AtHD1 (Fig 3A). This insertion caused a null mutation of AtHD1; no AtHD1 expression was detected in homozygous insertion lines (lane 7, Fig 3C). Furthermore, Western blot analysis using antibodies against the N terminus of AtHD1 (TIAN and CHEN 2001 ) indicated that, in the homozygous mutant line, the level of AtHD1 protein was greatly reduced (Fig 3D, lane 4). Only a small amount of protein (5% that of the wild type) was detected in homozygous plants, most likely resulting from cross-reaction of the antibodies to other AtHD1 homologs. For example, AtHD1 and AtHD6 share 69 and 84% of amino-acid sequence identity and similarity in the N termini, respectively. Alternatively, the low level of AtHD1 detected could be due to residual expression of athd1-t1; however, a different-sized protein would be observed because the T-DNA was inserted in the coding sequence (Fig 3A). Taken together, the data suggest that insertion of T-DNA into exon 2 of AtHD1 generated an AtHD1 null mutation.

View larger version (30K): [in this window] [in a new window] [Download PPT slide]   Figure 3. A null mutation (athd1-t1) is generated by T-DNA insertion into the AtHD1 gene. (A) AtHD1 contains seven exons. The T-DNA is inserted into exon 2. Primers used in RT-PCR for detecting AtHD1 expression are shown above the first exon and below the last exon. Primers used to amplify the nptII gene in RT-PCR reactions (C) and the nptII fragment used for DNA blot analysis (B) are shown below and above the T-DNA diagram, respectively. Striped box indicates the location of the AtHD1 fragment used for DNA blot analysis in Fig 7. (B) DNA blot analysis shows a single T-DNA insertion in the athd1-t1 line. (C) RT-PCR analysis indicates that nptII gene expression is detected in the athd1-t1 plants (lanes 1 and 2), but absent in wild-type plants (ecotype Columbia, lane 3; Ws, lane 4). T-DNA insertion produces null alleles in the AtHD1 locus. AtHD1 expression was detected in Ws (lane 6) but absent in the athd1-t1 line (lane 7). Act2 was amplified as a positive internal control. (D) AtHD1 production is reduced in the athd1-t1 plants. Crude protein extracts (25 µg) were subjected to electrophoresis though an SDS-polyacrylamide gene and immunoblotted onto Immobilon-P (Millipore, Bedford, MA) or Hybond-ECL (Amersham) membranes. The membrane was probed with antibodies against the N terminus of AtHD1. A band cross-reacting with anti-AtHD1 antibodies in athd1-t1 plants was reduced to only a trace amount (lane 4) compared to the band detected in the control plant (lane 3). A protein loading control is shown in an 8% SDS-PAGE stained with Coomassie Blue (lanes 1 and 2).

CASH126 and athd1-t1 plants display similar developmental abnormalities:
To determine whether CASH126 and athd1-t1 plants show similar changes in early developmental stages, we grew them side by side in a growth chamber under short-day conditions (8/16 hr of day/night). The leaves of both CASH126 and athd1-t1 plants were slightly chlorotic and showed disrupted radial symmetry (Fig 4A–D), probably resulting from defective shoot apical meristems as previously described (Fig 1). A prominent phenotype was the somewhat left-handed "twist" of the longitudinal axis of rosette leaves in athd1-t1 plants (Fig 4B and Fig F). This twist did not occur in other parts of the plants and was different from the previously described "lefty" mutants (THITAMADEE et al. 2002 ). This left-handed twist was not obvious in the leaves of CASH126 plants (Fig 4C). However, under close examination, the leaves of CASH126 plants were also twisted (Fig 4G), although the orientation of the distortion was not consistent.

View larger version (158K): [in this window] [in a new window] [Download PPT slide]   Figure 4. CASH126 and athd1-t1 plants show similar abnormal developmental phenotypes. CASH126 and athd1-t1 plants were grown side by side in a growth chamber under short-day (A–H) or long-day (I–L) conditions. (A) Wild-type plant (Ws) 3 weeks after germination. (B) Homozygous athd1-t1/athd1-t1 plants 3 weeks after germination. (C) CASH126 plants 3 weeks after germination. (D) Wild-type plant (Col, Columbia) 3 weeks after germination. (E) Mature inflorescence branches in athd1-t1 and Ws. (F) Rosette leaves of Ws and athd1-t1 plants. (G) Rosette leaves of CASH126 and control plants. (H) Florescence branches of CASH126 and Columbia plants. (I) Enlarged picture of mildly defective flowers in athd1-t1 plants. (J) Plant morphology of athd1-t1 and CASH126 plants. (K and L) Enlarged pictures of inflorescence branches showing severely defective flower organs (K) and mildly abnormal flowers (L) of CASH126 plants.

Both wild-type and AtHD1 disruption lines flowered at approximately the same time, suggesting that vegetative development initiated from axillary meristems had little effect on flowering time under short-day conditions. In a separate experiment we grew plants under long-day conditions (16/8 hr of day/night). As with the CASH lines, athd1-t1 lines flowered 2–5 days late (31.9 ± 1.1 days, n = 43) compared to the wild-type plants (28.1 ± 2.3 days, n = 42). Although severe phenotypes in the vegetative stage were often observed under short-day conditions (Fig 4, A–D, F, G), similar abnormalities in the reproductive stage were observed under both long- and short-day conditions (Fig 4E, Fig H, I–L).

CASH126 and athd1-t1 plants developed abnormal flower structures, including reduced numbers of petals (Fig 4I, Fig K, and Fig L), split flowers (Fig 4K), and sterility (Fig 4E and Fig H). The abnormal flower phenotypes in athd1-t1 plants (Fig 4I) were not as severe as those of CASH126 plants (Fig 4J), but were more uniform. CASH126 lines, however, displayed a range of penetrance in developmental abnormalities, ranging from severe (Fig 4K) to mild (Fig 4L).

CASH126 and athd1-t1 plants were partially sterile and had low seed set. To investigate the cause of sterility, we examined the flower structures of athd1-t1 plants using light and scanning electron microscopy. Flower morphology of athd1-t1 plants was irregular. Many flowers had missing petals and sepals (Fig 5B, Fig C, and Fig E) compared to the typical "crucifer-shaped" flowers of wild-type plants (Fig 5A and Fig D). Some flowers had fewer than six stamens, and some of these were fused (Fig 5F). The stamens were short (Fig 5E). As a result, pollen would have difficulty reaching the "tall" stigma. The incompatibility between "impotent" male stamens and tall female stigmas is likely associated with sterility because, compared with wild-type stigmas that were completely covered with pollen grains (Fig 5G), only a few pollen grains could reach the stigmas of homozygous athd1-t1 plants (arrows in Fig 5H). As a result, siliques of homozygous athd1-t1 plants were short and contained few seeds (Fig 5M and Fig O), whereas siliques of wild-type plants were long and contained many fully developed seeds (Fig 5M and Fig N). Pollen grains of athd1-t1 plants were apparently normal (Fig 5K and Fig L). Although structural incompatibility between stigma and stamen is likely related to sterility, we do not rule out other possibilities, such as biological incompatibility between pollen and stigma interactions during pollination in athd1-t1 plants, which may also cause sterility.

View larger version (83K): [in this window] [in a new window] [Download PPT slide]   Figure 5. Abnormal flower development and sterility in athd1-t1 plants. (A) Wild-type (Ws) plants have a typical "crucifer" flower with four petals, six stamens, and a regular stigma. (B and C) Transformation of the "crucifer" shape into an irregular form in athd1-t1 plants resulting from fused petals (B) and missing flower parts (C). (D) Normal flower morphology of wild-type plants revealed by scanning electron microscopy. (E and F) Short stamens and missing petals (E) and fused stamens (F) in athd1-t1 plants. (G) The stigma of the wild-type flower is completely covered with pollen grains. (H) Only a few pollen grains (arrows) are found on the stigma of athd1-t1 plants. (I and J) Dehisced anthers showing normal pollen development in wild-type (I) and athd1-t1 (J) plants. (K and L) Pollen grains in Ws (K) and athd1-t1 (L) anthers. (M). Short and immature siliques (left and middle) developed in athd1-t1 plants compared to the normal silique in the control (right). (N) Normal seed development in Ws. (O) Abortive seed development in athd1-t1.

The pleiotropic phenotype resulting from AtHD1 defects is immediately restored in heterozygous plants and displays Mendelian segregation:
The developmental defects of CASH and athd1-t1/athd1-t1 plants are consistently observed in selfing generations. However, it is unknown whether the phenotypic abnormalities were dependent on the continued deficiency in AtHD1 expression and core histone acetylation profiles. To address this, we made F₁ hybrids between wild-type (Ws) and homozygous athd1-t1/athd1-t1 plants (Fig 6A) and examined the phenotypes in the resulting F₁ hybrids and F₂ siblings. In all the F₁ plants examined, most of the developmental abnormalities, including rosette leaf morphology and fertility, were reversed to those of wild-type plants, except that F₁ plants still developed slightly shorter siliques. These results indicate that most of the developmental defects induced by the homozygous mutants are immediately corrected in AtHD1/athd1-t1 heterozygotes. Indeed, the expression level of AtHD1 in the heterozygous plants was equal to that of wild-type plants (Fig 6B). Furthermore, the acetylation level of histone H4 Lys12 was increased approximately threefold in the athd1-t1/athd1-t1 homozygous mutants compared to the wild-type plants (Fig 6C). H4 Lys12 is one of the primary deacetylation sites targeted by RPD3 in yeast (VIDAL and GABER 1991 ; RUNDLETT et al. 1996 ). The H4 Lys12 level was reversed to that of the wild-type plants in the heterozygous plants, implying that acetylation at some specific sites is responsible for the changes in developmental abnormalities and gene expression in these plants. In the F₂ progeny, the wild-type and abnormal phenotypes segregated 3:1, cosegregating with the level of AtHD1 expression in these plants. These data indicate that the developmental abnormalities are dependent on disruption of AtHD1 expression and of some specific acetylation sites (e.g., H4 Lys12) in the athd1-t1 lines.

View larger version (75K): [in this window] [in a new window] [Download PPT slide]   Figure 6. (A) Developmental abnormalities in rosette leaves and siliques of athd1-t1/athd1-t1 homozygous plants were restored in F1 (AtHD1/athd1-t1) progeny. The F1 hybrid, heterozygous for the athd1-t1 locus, was generated by crossing a wild-type Ws plant with a homozygous athd1-t1 plant. (Right) Genomic DNA was digested with EcoRI and SacII, subjected to electrophoresis through a 1% agarose gel, and blotted onto a Hybond + membrane. The blot was hybridized with a DNA fragment containing exons 1 and 2 of AtHD1 (see Fig 3A). The F1 plant contained a normal AtHD1 allele and an athd1-t1 allele. (B) Equal quantities of AtHD1 transcripts were detected by RT-PCR in the F1 and Ws plants, but no AtHD1 transcript was detected in the athd1-t1/athd1-t1 homozygous plants. RT-PCR amplification of Act2 was used as an internal control. (C) Western blot indicates that histone H4 Lys12 acetylation is increased in the homozygous athd1-t1 plants. Hybridization intensities of a nonspecific (n.s.) protein detected by the anti-H4 Lys12 antibody were used as internal controls to calculate the relative ratio of H4 Lys12 acetylation accumulation, which is shown in a histogram below the Western blot.

Does athd1-t1 induce additional epigenetic changes?
The ddm1 mutant induces epigenetic lesions in addition to those observed in the original ddm1 mutant (STOKES et al. 2002 ). To determine whether athd1-t1 acts as an epigenetic modulator, we examined phenotypic changes in the plants derived from single-seed descent of one athd1-t1/athd1-t1 homozygous plant. Consistent with the results in CASH plants, within four generations of selfing we did not observe abnormal phenotypes in these plants in addition to developmental abnormalities observed in the original homozygous plants (Fig 7). Phenotypic abnormalities were less severe under long-day conditions than under short-day conditions, suggesting that the penetrance of phenotypes is dependent on day length. AtHD1 expression was not detected in the plants after four generations of selfing. These data, together with the immediate restoration of normal development in AtHD1/athd1-t1 heterozygous plants, suggest that developmental changes induced by disruption of AtHD1 expression are reversible and dependent on AtHD1 expression and histone deacetylation.

View larger version (141K): [in this window] [in a new window] [Download PPT slide]   Figure 7. athd1-t1/athd1-t1 plants showed constant developmental abnormalities during four generations (G1–G4) of selfing. The plants were grown side by side in the same growth chamber. No AtHD1 expression was detected by RT-PCR in the homozygous athd1-t1/athd1-t1 plants obtained in four generations of selfing.

	DISCUSSION

TOP ABSTRACT MATERIALS AND METHODS RESULTS DISCUSSION LITERATURE CITED

Heritable changes of developmental abnormalities induced by AtHD1 disruption:
Blocking histone deacetylation induces a wide range of developmental changes. The abnormal development includes defective shoot apical meristems (SAM), irregular trichomes and cellular patterns, late flowering, abnormal inflorescences and flowers, and aborted seeds. These developmental abnormalities are stable in the presence of AtHD1 disruption after four to five generations of selfing. Moreover, a T-DNA insertion into the AtHD1 gene (athd1-t1) results in a mutant line (Ws ecotype) that shows developmental abnormalities similar to those of CASH126 plants (Columbia ecotype), confirming that AtHD1 plays an important role in reprogramming developmental processes. The affected plants develop through initiating axillary meristems and additional changes that ensure the completion of a life cycle, although they have to overcome structural and developmental incompatibilities resulting from irregularly orchestrated patterns and tempos of organogenesis. Histone deacetylation and the resulting effects on gene regulation may contribute to the fundamental and dynamic process of developmental plasticity in plants (WALBOT 1996 ; MEYEROWITZ 1997 ). Disruption of AtHDs may directly affect expression of genes such as superman (TIAN and CHEN 2001 ) in specific developmental stages. Alternatively, the disruption may induce a series of changes in regulatory networks. Complex phenotypes such as defective apical shoot meristems and abnormal flowers could be some of these pleiotropic effects. It will be interesting to identify target genes whose expression is affected in the athd1-t1 and antisense AtHD1 transgenic lines.

Developmental abnormalities induced by disruption of histone deacetylation are different from those induced by DNA methylation defects (VONGS et al. 1993 ; FINNEGAN et al. 1996 ; RONEMUS et al. 1996 ; GENDREL et al. 2002 ; STOKES and RICHARDS 2002 ). First, the abnormal development in the athd1-t1 homozygous plants can be immediately corrected in the heterozygous state. Second, no additional visual abnormalities accumulate in the selfing progeny of athd1-t1/athd1-t1 homozygous plants. The data suggest that the dynamic and reversible process of acetylation and deacetylation provides a spatial and temporal code for gene activation and silencing. Consistent with this hypothesis, histone deacetylation may affect the expression of genes in response to environmental (e.g., light, day length) and developmental (e.g., homeotic genes) changes. Indeed, both antisense AtHD1 transgenic and athd-t1 homozygous lines show variable flowering time and severity of developmental abnormalities under short- and long-day conditions.

AtHD1 is a member of a multi-gene family that may diverge in functions. For example, AtHD6, a RPD3-like homolog, is involved in the release of transgene silencing (MURFETT et al. 2001 ) and is associated with the maintenance of DNA methylation patterns of transgenes (AUFSATZ et al. 2002 ). However, recessive mutations in AtHDA6 do not induce visible abnormal phenotypes (MURFETT et al. 2001 ). Antisense AtHD2 lines (WU et al. 2000A , WU et al. 2000B ) show less severe developmental abnormalities than antisense AtHD1 lines do (TIAN and CHEN 2001 ). Significantly, although histone deacetylases are encoded by a multi-gene family and share sequence homology (PANDEY et al. 2002 ), other AtHD homologs cannot compensate for the loss of AtHD1 activity in the athd1-t1 lines, suggesting that AtHD1 is a key member of the histone deacetylase gene family. Unlike the tetra-acetylated histone H4 induced by overexpressing antisense AtHD1 (TIAN and CHEN 2001 ), hyper-acetylation of core histones in the athd1-t1 line is restricted to a specific set of lysine residues including H4 Lys12. These data suggest that specific patterns of histone acetylation may be established by AtHD1, consequently controlling the expression of a subset of genes during development.

The role of a histone code in genetic and epigenetic regulation:
Chromatin-based gene regulation in eukaryotes is controlled by a chromatin code (JENUWEIN and ALLIS 2001 ) that can be further classified as stable or reversible. The stable code includes DNA and histone methylation (RICHARDS and ELGIN 2002 ). Once the stable code is established, it is difficult to alter unless through a reset of developmental programming during meiosis, because no active demethylation pathway has been identified (JENUWEIN and ALLIS 2001 ; LI 2002 ; RICHARDS and ELGIN 2002 ). As a result, changing a stable code results in epigenetic inheritance or variation (STOKES et al. 2002 ; STOKES and RICHARDS 2002 ). Moreover, alteration in the stable code may serve as an epigenetic "modulator" that induces changes at other loci, independent of the original chromatin conformation. Indeed, in the ddm1 genetic background, epigenetic lesions are induced in a genomic region containing multiple repeats associated with disease resistance genes, presumably because they are vulnerable in response to changes in DNA methylation (STOKES and RICHARDS 2002 ). Moreover, these epialleles may be affected not only by demethylation, but also by allelic interactions among loci or within a locus (STOKES et al. 2002 ), a phenomenon similar to meiotic transvection (ARAMAYO and METZENBERG 1996 ) or paramutation (HOLLICK et al. 1997 ). Methylation-associated epialleles have also been reported in mutants with defective flower structure (e.g., SUPERMAN) or a delay in flowering time (e.g., FWA), which may result from aberrant DNA methylation patterns (JACOBSEN and MEYEROWITZ 1997 ; KAKUTANI 1997 ; SOPPE et al. 2000 ).

The reversible code (e.g., histone acetylation and deacetylation) is heritable (JENUWEIN and ALLIS 2001 ) but the action of the code is dependent on the respective biochemical activities that set the code. The code is dynamic and reversible during any developmental stage and is independent of meiosis. Although the connection between a specific acetylation code and gene regulation remains to be elucidated, the heritable and reversible nature of developmental abnormalities observed in AtHD1 disruption lines is likely associated with changes in histone acetylation (e.g., H4 Lys12 acetylation) and methylation. It is conceivable that under normal conditions, genes, including homeotic genes (e.g., superman; TIAN and CHEN 2001 ) responsible for plant development, may be controlled by the reversible code of acetylation and deacetylation. However, the acetylation code is reversible and dependent on histone deacetylases or other factors in the chromatin control. As soon as the proper code is restored, controls of regulatory networks and developmental programs return to normal, independent of reprogramming through meiosis. The concept of a stable or reversible code is also supported by the results obtained from a recent study indicating that the loss of histone H4 Lys16 acetylation in ddm1 homozygous plants is compensated in DDM1/ddm1 heterozygous plants, whereas DNA and histone H3 Lys9 methylation remain unchanged (SOPPE et al. 2002 ).

The relationship between reversible and stable codes is largely unknown. Acetylation and deacetylation may act on active or inactive chromatin as a competitor for histone methylation sites (JENUWEIN and ALLIS 2001 ). For example, histone acetyltransferases and methyltransferases may compete for histone H3 Lys9 to establish an active or inactive form of chromatin (JENUWEIN and ALLIS 2001 ; LITT et al. 2001 ). Histone deacetylation and DNA methylation may also be interdependent (SELKER 1998 ; SOPPE et al. 2002 ). Moreover, a putative histone deactylase (AtHDA6) is needed to enhance DNA methylation induced by double-stranded RNA (AUFSATZ et al. 2002 ). Finally, some DNA methyltransferases (e.g., CMT) or histone deacetylases (e.g., HD2) were identified only in plants (LUSSER et al. 1997 , LUSSER et al. 2001 ; HENIKOFF and COMAI 1998 ; LINDROTH et al. 2001 ), suggesting that plants may control gene regulation via both general and unique chromatin pathways. It will be interesting to test how the reversible code of histone acetylation controlled by AtHD1 interacts with the biochemically stable code of DNA and histone methylation and affects the plastic nature of plant development.

	ACKNOWLEDGMENTS

We thank Eric J. Richards for insightful discussions and Gary E. Hart and Karel Riha for critical suggestions on improving the manuscript, Joe Wang for assistance in photography, Hyeon-Se Lee and Jenny Lee for helpful discussions, and Mike Pendleton in the Microscopy and Imaging Center at Texas A&M University for providing excellent technical services. Work in S.B.G.'s laboratory was supported by grants (99-75715 and 99-75930) from the National Science Foundation Plant Genome Research Program. Research in the Chen laboratory was supported by a grant (00-77774) from the National Science Foundation Plant Genome Research program and the Texas Agricultural Experiment Station.

Manuscript received April 6, 2003; Accepted for publication May 23, 2003.

	LITERATURE CITED

TOP ABSTRACT MATERIALS AND METHODS RESULTS DISCUSSION LITERATURE CITED

ALLFREY, V. G., R. FAULKNER, and A. E. MIRSKY, 1964  Acetylation and methylation of histones and their possible role in regulation of RNA synthesis. Proc. Natl. Acad. Sci. USA 51:786-794.[Free Full Text]

ALMOUZNI, G., S. KHOCHBIN, S. DIMITROV, and A. P. WOLFFE, 1994  Histone acetylation influences both gene expression and development of Xenopus laevis.. Dev. Biol. 165:654-669.[Medline]

AN, Y. Q., J. M. MCDOWELL, S. HUANG, E. C. MCKINNEY, and S. CHAMBLISS et al., 1996  Strong, constitutive expression of the Arabidopsis Act2/Act8 actin subclass in vegetative tissues. Plant J. 10:107-121.[Medline]

Analysis of the genome sequence of the flowering plant Arabidopsis thaliana.. (2000) Nature 408:796-815.[Medline]

ARAMAYO, R. and R. L. METZENBERG, 1996  Meiotic transvection in fungi. Cell 86:103-113.[Medline]

AUFSATZ, W., M. F. METTE, J. VAN DER WINDEN, M. MATZKE, and A. J. MATZKE, 2002  HDA6, a putative histone deacetylase needed to enhance DNA methylation induced by double-stranded RNA. EMBO J. 21:6832-6841.[Medline]

BARTL, S., J. TAPLICK, G. LAGGER, H. KHIER, and K. KUCHLER et al., 1997  Identification of mouse histone deacetylase 1 as a growth factor-inducible gene. Mol. Cell. Biol. 17:5033-5043.[Abstract]

BERNIER, G., 1986  The flowering process as an example of plastic development. Symp. Soc. Exp. Biol. 40:257-286.[Medline]

BERNSTEIN, B. E., J. K. TONG, and S. L. SCHREIBER, 2000  Genomewide studies of histone deacetylase function in yeast. Proc. Natl. Acad. Sci. USA 97:13708-13713.[Abstract/Free Full Text]

BLEECKER, A. B. and S. E. PATTERSON, 1997  Last exit: senescence, abscission, and meristem arrest in Arabidopsis.. Plant Cell 9:1169-1179.[Medline]

BROSCH, G., R. RANSOM, T. LECHNER, J. D. WALTON, and P. LOIDL, 1995  Inhibition of maize histone deacetylases by HC toxin, the host-selective toxin of Cochliobolus carbonum.. Plant Cell 7:1941-1950.[Abstract]

CHEN, Z. J. and C. S. PIKAARD, 1997  Epigenetic silencing of RNA polymerase I transcription: a role for DNA methylation and histone modification in nucleolar dominance. Genes Dev. 11:2124-2136.[Abstract/Free Full Text]

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

DANGL, M., G. BROSCH, H. HAAS, P. LOIDL, and A. LUSSER, 2001  Comparative analysis of HD2 type histone deacetylases in higher plants. Planta 213:280-285.[Medline]

DE RUBERTIS, F., D. KADOSH, S. HENCHOZ, D. PAULI, and G. REUTER et al., 1996  The histone deacetylase RPD3 counteracts genomic silencing in Drosophila and yeast. Nature 384:589-591.[Medline]

FELDMANN, K. A. and M. D. MARKS, 1987  Agrobacterium-mediated transformation of germinating seeds of Arabidopsis thaliana: a non-tissue culture approach. Mol. Gen. Genet. 208:1-9.

FINNEGAN, E. J., 2001  Is plant gene expression regulated globally? Trends Genet. 17:361-365.[Medline]

FINNEGAN, E. J., W. J. PEACOCK, and E. S. DENNIS, 1996  Reduced DNA methylation in Arabidopsis thaliana results in abnormal plant development. Proc. Natl. Acad. Sci. USA 93:8449-8454.[Abstract/Free Full Text]

FORSTHOEFEL, N. R., Y. WU, B. SCHULZ, M. J. BENNETT, and K. A. FELDMANN, 1992  T-DNA insertion mutagenesis in Arabidopsis: prospects and perspectives. Aust J. Plant Physiol. 19:353-366.

GENDREL, A. V., Z. LIPPMAN, C. YORDAN, V. COLOT, and R. A. MARTIENSSEN, 2002  Dependence of heterochromatic histone H3 methylation patterns on the Arabidopsis gene DDM1. Science 297:1871-1873.[Abstract/Free Full Text]

HABU, Y., T. KAKUTANI, and J. PASZKOWSKI, 2001  Epigenetic developmental mechanisms in plants: molecules and targets of plant epigenetic regulation. Curr. Opin. Genet. Dev. 11:215-220.[Medline]

HENIKOFF, S. and L. COMAI, 1998  A DNA methyltransferase homolog with a chromodomain exists in multiple polymorphic forms in Arabidopsis. Genetics 149:307-318.[Abstract/Free Full Text]

HOLLICK, J. B., J. E. DORWEILER, and V. L. CHANDLER, 1997  Paramutation and related allelic interactions. Trends Genet. 13:302-308.[Medline]

IMAI, S., C. M. ARMSTRONG, M. KAEBERLEIN, and L. GUARENTE, 2000  Transcriptional silencing and longevity protein Sir2 is an NAD-dependent histone deacetylase. Nature 403:795-800.[Medline]

JACOBSEN, S. E. and E. M. MEYEROWITZ, 1997  Hypermethylated SUPERMAN epigenetic alleles in Arabidopsis.. Science 277:1100-1103.[Abstract/Free Full Text]

JANG, I. C., Y. M. PAHK, S. I. SONG, H. J. KWON, and B. H. NAHM et al., 2003  Structure and expression of the rice class-I type histone deacetylase genes OsHDAC1–3: OsHDAC1 overexpression in transgenic plants leads to increased growth rate and altered architecture. Plant J. 33:531-541.[Medline]

JENUWEIN, T. and C. D. ALLIS, 2001  Translating the histone code. Science 293:1074-1080.[Abstract/Free Full Text]

KAKUTANI, T., 1997  Genetic characterization of late-flowering traits induced by DNA hypomethylation mutation in Arabidopsis thaliana.. Plant J. 12:1447-1451.[Medline]

KRYSAN, P. J., J. C. YOUNG, F. TAX, and M. R. SUSSMAN, 1996  Identification of transferred DNA insertions within Arabidopsis genes involved in signal transduction and ion transport. Proc. Natl. Acad. Sci. USA 93:8145-8150.[Abstract/Free Full Text]

LANDRY, J., A. SUTTON, S. T. TAFROV, R. C. HELLER, and J. STEBBINS et al., 2000  The silencing protein SIR2 and its homologs are NAD-dependent protein deacetylases. Proc. Natl. Acad. Sci. USA 97:5807-5811.[Abstract/Free Full Text]

LECHNER, T., A. LUSSER, A. PIPAL, G. BROSCH, and A. LOIDL et al., 2000  RPD3-type histone deacetylases in maize embryos. Biochemistry 39:1683-1692.[Medline]

LI, E., 2002  Chromatin modification and epigenetic reprogramming in mammalian development. Nat. Rev. Genet. 3:662-673.[Medline]

LI, G., T. C. HALL, and R. HOLMES-DAVIS, 2002  Plant chromatin: development and gene control. Bioessays 24:234-243.[Medline]

LINDROTH, A. M., X. CAO, J. P. JACKSON, D. ZILBERMAN, and C. M. MCCALLUM et al., 2001  Requirement of CHROMOMETHYLASE3 for maintenance of CpXpG methylation. Science 292:2077-2080.[Abstract/Free Full Text]

LITT, M. D., M. SIMPSON, M. GASZNER, C. D. ALLIS, and G. FELSENFELD, 2001  Correlation between histone lysine methylation and developmental changes at the chicken beta-globin locus. Science 293:2453-2455.[Abstract/Free Full Text]

LUSSER, A., G. BROSCH, A. LOIDL, H. HAAS, and P. LOIDL, 1997  Identification of maize histone deacetylase HD2 as an acidic nucleolar phosphoprotein. Science 277:88-91.[Abstract/Free Full Text]

LUSSER, A., D. KOLLE, and P. LOIDL, 2001  Histone acetylation: lessons from the plant kingdom. Trends Plant Sci. 6:59-65.[Medline]

LYNN, K., A. FERNANDEZ, M. AIDA, J. SEDBROOK, and M. TASAKA et al., 1999  The PINHEAD/ZWILLE gene acts pleiotropically in Arabidopsis development and has overlapping functions with the ARGONAUTE1 gene. Development 126:469-481.[Abstract]

MARTIENSSEN, R. and V. COLOT, 2001  DNA methylation and epigenetic inheritance in plants and filamentous fungi. Science 293:1070-1074.[Abstract/Free Full Text]

MEYEROWITZ, E. M., 1997  Genetic control of cell division patterns in developing plants. Cell 88:299-308.[Medline]

MEYEROWITZ, E. M., 2002  Plants compared to animals: the broadest comparative study of development. Science 295:1482-1485.[Abstract/Free Full Text]

MOREL, J. B., C. GODON, P. MOURRAIN, C. BECLIN, and S. BOUTET et al., 2002  Fertile hypomorphic ARGONAUTE (ago1) mutants impaired in post-transcriptional gene silencing and virus resistance. Plant Cell 14:629-639.[Abstract/Free Full Text]

MOUSSIAN, B., H. SCHOOF, A. HAECKER, G. JURGENS, and T. LAUX, 1998  Role of the ZWILLE gene in the regulation of central shoot meristem cell fate during Arabidopsis embryogenesis. EMBO J. 17:1799-1809.[Medline]

MURAI, K., S. TAKUMI, H. KOGA, and Y. OGIHARA, 2002  Pistillody, homoeotic transformation of stamens into pistil-like structures, caused by nuclear-cytoplasm interaction in wheat. Plant J. 29:169-181.[Medline]

MURFETT, J., X. J. WANG, G. HAGEN, and T. J. GUILFOYLE, 2001  Identification of Arabidopsis histone deacetylase hda6 mutants that affect transgene expression. Plant Cell 13:1047-1061.[Abstract/Free Full Text]

PANDEY, R., A. MULLER, C. A. NAPOLI, D. A. SELINGER, and C. S. PIKAARD et al., 2002  Analysis of histone acetyltransferase and histone deacetylase families of Arabidopsis thaliana suggests functional diversification of chromatin modification among multicellular eukaryotes. Nucleic Acids Res. 30:5036-5055.[Abstract/Free Full Text]

RICHARDS, E. J. and S. C. ELGIN, 2002  Epigenetic codes for heterochromatin formation and silencing: rounding up the usual suspects. Cell 108:489-500.[Medline]

ROBYR, D., Y. SUKA, I. XENARIOS, S. K. KURDISTANI, and A. WANG et al., 2002  Microarray deacetylation maps determine genome-wide functions for yeast histone deacetylases. Cell 109:437-446.[Medline]

RONEMUS, M. J., M. GALBIATI, C. TICKNOR, J. CHEN, and S. L. DELLAPORTA, 1996  Demethylation-induced developmental pleiotropy in Arabidopsis.. Science 273:654-657.[Abstract]

ROSSI, V., H. HARTINGS, and M. MOTTO, 1998  Identification and characterisation of an RPD3 homologue from maize (Zea mays L.) that is able to complement an rpd3 null mutant of Saccharomyces cerevisiae.. Mol. Gen. Genet. 258:288-296.[Medline]

RUNDLETT, S. E., A. A. CARMEN, R. KOBAYASHI, S. BAVYKIN, and B. M. TURNER et al., 1996  HDA1 and RPD3 are members of distinct yeast histone deacetylase complexes that regulate silencing and transcription. Proc. Natl. Acad. Sci. USA 93:14503-14508.[Abstract/Free Full Text]

RUNDLETT, S. E., A. A. CARMEN, N. SUKA, B. M. TURNER, and M. GRUNSTEIN, 1998  Transcriptional repression by UME6 involves deacetylation of lysine 5 of histone H4 by RPD3. Nature 392:831-835.[Medline]

SELKER, E. U., 1998  Trichostatin A causes selective loss of DNA methylation in Neurospora. Proc. Natl. Acad. Sci. USA 95:9430-9435.[Abstract/Free Full Text]

SOPPE, W. J., S. E. JACOBSEN, C. ALONSO-BLANCO, J. P. JACKSON, and T. KAKUTANI et al., 2000  The late flowering phenotype of fwa mutants is caused by gain-of-function epigenetic alleles of a homeodomain gene. Mol. Cell 6:791-802.[Medline]

SOPPE, W. J. J., Z. JASENCAKOVA, A. HOUBEN, T. KAKUTANI, and A. MEISTER et al., 2002  DNA methylation controls histone H3 lysine 9 methylation and heterochromatin assembly in Arabidopsis.