The Arabidopsis transcription factor NAI1 is required for enhancing the active histone mark but not for removing the repressive mark on PYK10, a seedling–specific gene upon embryonic-to-postgerminative developmental phase transition

Abstract

We have recently shown that the expression onset of a seedling-specific gene, PYK10, occurs in a cell-by-cell manner upon the transition from the embryonic to the postgerminative phase and during embryogenesis in seed maturation regulator mutants such as lec1, and implicated epigenetic mechanisms in the process. Here, the role of the NAI1 transcription factor required for PYK10 expression in the developmental switching of PYK10 was investigated. The cell-by-cell onset of PYK10-EGFP in lec1 embryo was still observed in the nai1 background, but at greatly reduced levels. Decreases in the level of the repressive histone mark, H3K27 trimethylation observed upon the transition to the postgeminative phase normally occurred in nai1. However, concomitant increases in the level of the active mark, H3K4 trimethylation observed in wild type was significantly compromised in nai1. These results indicate that the switching of PYK10 upon developmental phase transition involves 2 separable steps of chromatin state change.

Abbreviations

DAF

days after flowering

DAG

days after germination

H3K27me3

histone H3 lysine 27 trimethylation

H3K4me3

histone H3 lysine 4 trimethylation

LAFL

transcription factors LEC1, ABI3, FUS3 and LEC2.

Upon the developmental transition from the embryonic to the postgermitive seedling phase, gene expression programs are dramatically reorganized. Postgerminative gene expression programs that only operate after seed imbibition/germination are tightly repressed during embryo formation and maturation, and programs specific for embryonic development become tightly repressed after the transition to the postgerminative phase. The mechanisms for the activation and repression of embryonic phase-specific genes, particularly those involving the LAFL (transcription factors LEC1, LEC2, FUS3 and ABI3) seed maturation master regulators have been extensively studied.^1-11 However, our knowledge remains limited concerning how postgerminative genes are repressed during the embryonic phase and derepressed upon the phase transition. Our recent study¹² together with previous studies by others^13-17 revealed that LAFL factors also play active roles in the maintenance of repressed states of the postgerminative genes. From comparative transcriptome analyses between the wild type and lafl mutant embryos, we identified a set of seedling-specific genes related to ER-body function that are tightly repressed in the wild type, but expressed at high levels in these mutants during embryogenesis.¹² Among these genes, PYK10 encoding ER-body-localized β-glucosidase showed a remarkably high level of upregulation in the mutants relative to the wild type and was therefore chosen as a marker for the postgeminative gene expression program that is tightly repressed during embryogenesis. Analyses using promoter-EGFP/CFP reporter genes revealed that the expression of PYK10 and other ER-body-related genes begins in a stochastic and cell-by-cell manner during germination in the wild type and during embryogenesis in lafl mutants.¹² Clusters of cells expressing the reporter were also observed along cell files. In agreement with these observations, the degree of ER-body development differed from cell to cell during embryo development in the mutants and during seed germination in the wild type, indicating that the developmental transition take places on a cell-by-cell basis. Furthermore, the level of a repressive epigenetic mark, histone H3-Lys27 trimethylation (H3K27me3), at the PYK10 locus was high during embryogenesis, but was significantly reduced in early seedlings.¹² These observations led us to postulate that tight repression and its release upon the developmental transition of the postgeminative genes, as represented by PYK10, is controlled by epigenetic mechanisms.¹² In addition, in transgenic lines carrying 2 independent reporters, PYK10-EGFP and PYK10-CFP, cells expressing only EGFP or CFP were observed together with a substantial number of cells expressing both reporters in developing lafl embryos as well as in germinating wild-type embryos. These results led us to postulate that the expression state of PYK10 is limited not only by the local chromatin state but also factors acting in trans, which may facilitate the change in the chromatin states.¹²

NAI1, a basic-helix-loop-helix (bHLH) transcription factor, has been identified as being required for the expression of PYK10 and the formation of ER-bodies.¹⁸ No visible phenotypes beside the ER-body defects have been detected in nai1 mutants under normal growth conditions.¹⁹ As reported, PYK10 transcript levels were greatly reduced in the nai1 seedlings compared to those in the wild type seedlings (Fig. 1A). Thus, a question is raised about how NAI1 is involved in the switching of PYK10 upon the developmental transition from the embryonic to the postgeminative phase. Examination of the transcriptome data comparing 8 days-after-flowering (DAF) wild type and lec1 or lec2 embryos¹² and quantitative RT-PCR revealed high levels of upregulation of NAI1 in the mutant embryo (Fig. 1B).

Figure 1.

Effects of the nai1 mutation on PYK10 expression and of lec1/lec2 on NAI1 expression. (A, B) Real-time RT-PCR quantification of PYK10 and NAI1 transcript levels in wild-type (ecotype Col-0; WT) and nai1 seedlings at 8 days after germination (8 DAG; A), and in wild-type (ecotype WS; WT) and lec1/lec2 embryos at 7 days after flowering (7 DAF; B), respectively. Error bars represent the SD of biological replicates (n = 3). (C–F) Expression of PYK10-EFGP in lec1 nai1 double (C, D) or lec1 single (E, F) mutant embryos at 7 DAF. The top and bottom of each paired panel show a Z-stack image consisting of several EGFP optical sections from the surface to the approximate median of the embryo and the corresponding DIC image overlaid with the EGFP image, respectively. The image in (D) is a graphically enhanced image of (C). Images in (E) and (F) are of the same embryo. However, the image in (E) was acquired with an excitation laser intensity reduced by 2 orders of magnitude relative to that applied to the image acquisition in (C) and (F). All other conditions for image acquisition were identical. Bar = 20 µm. The lec1 nai1 PYK10-EGFP embryos were obtained by crossing lec1 PYK10-EGFP into nai1. Thus the embryo of each genotype carried the same allele of the transgenic reporter. RNA preparation, real-time RT-PCR, and laser confocal microscopic observation were performed as described previously.¹² The mutant alleles used in the experiments were nai1-1, lec1-1 and lec2-1.^12,19 For the PYK10-EGFP reporter gene experiments, the lec1-1 allele in the Col-0 background was used.¹²

To see whether NAI1 could be a trans-acting factor responsible for the onset of PYK10 expression in lec embryos, we examined the effect of the nai1 mutation on PYK10-EGFP expression in lec1 embryos. Although the fluorescent signals in lec1 embryos were substantially reduced, the cell-by-cell onset of expression of the same PYK10-EGFP transgene was still recognized in a nai1 background (Figs. 1C–F). Thus, NAI1 was unlikely to be a factor that conditions or triggers the conversion of PYK10 chromatin from a repressed to a derepressed state. Instead, it appeared to be required for quantitative regulation. In accordance with this, the H3H27me3 level on the PYK10 promoter region chromatin in the nai1 seedlings was similarly lowered as in the wild type seedlings (Fig. 2A). Thus, the repressed state of PYK10 appeared to be converted to a derepressed state in the absence of NAI1 without a high level of transcriptional activation. We also examined levels of H3K4me3, an active histone mark, in the PYK10 chromatin. As expected from the expression state, an increased level of H3K4me3 on the PYK10 chromatin of wild type seedlings was detected compared to that in developing seeds (Fig. 2B). However, in nai1 seedlings, such an increase in H3K4me3 level was significantly reduced (Fig. 2B). Thus, 3 PYK10 chromatin states were distinguished, high H3K27me3/low H3K4me3, low H3K27me3/lowH3K4me3 and low H3K27me3/high H3K4me3. These results together indicate that NAI1 is required for converting the PYK10 chromatin to an active state, but not to a derepressed state, and/or maintaining the active state.

Figure 2.

Effect of the nai1 mutation and developmental phases on the histone methylation states of PYK10 chromatin. Chromatin immunoprecipitation analyses of H3K27me3 (A) and H3K4me3 (B) on PYK10 chromatin (promoter region) from developing seeds at 10 DAF, seedlings at 3 DAG and rosette leaves obtained immediately after bolting were performed as described previously.¹² Results from the immunoprecipitation with the specific antibody and the mock precipitation are indicated by (+) and (−), respectively. Data are expressed as a percentage of input DNA, with error bars representing the SD of biological replicates (n = 2–5). Asterisks indicate values that are significantly different between the treatments (p <0.05, Student's t test).

Considering the tight repression of NAI1 expression during the embryonic phase under the control of LAFL, NAI1 may also be repressed at the chromatin level in a similar way to that of its target genes such as PYK10, and derepressed in a cell-by-cell manner. This assumption can explain the discrete fluorescent intensity levels of expression among the PYK10-EGFP-expressing cells or cell clusters. Other quantitative regulators that are regulated in a similar manner to that of NAI1 may be involved in such multistable expression of PYK10.

In conclusion, we have shown that the switching of PYK10 expression upon the embryonic-to-postgerminative developmental phase transition is controlled by at least 2 steps of histone modification: removal of H3K27me3 and stimulation of H3K4me3. We have also shown that the bHLH transcription factor NAI1 is required for the latter but not for the former event. Although the 2 processes are experimentally uncoupled, they may not take place independently during the actual development because repression of NAI1 during the embryonic phase may be controlled in a similar manner to that of PYK10. Detailed analyses of NAI1 regulation as done for PYK10 will apparently be important for clarifying this point. Currently the molecular links between LAFL and the maintenance of the repressive state of postgerminative genes are totally missing. Possible mechanisms may range from direct interaction of LAFL with chromatins of postgerminative genes to transcriptional regulation of genes for chromatin metabolism. Biochemical and genetic examination of such possibilities regarding PYK10 and NAI1 regulation will eventually shed light on more general understanding of the molecular mechanisms underlying the robust switching of gene expression program upon developmental phase transitions.

Disclosure of Potential Conflicts of Interest

No potential conflicts of interest were disclosed.

Acknowledgments

The authors thank Dr. Hirokazu Katoh, Ms. Yuka Tsukiji and Ms. Rena Endo for excellent technical assistance, and Dr. Ikuko Hara-Nishimura for nai1 seeds.

Funding

This work was funded in part by the Ministry of Education, Culture, Sports, Science and Technology, Japan [Grants-in-Aid for Scientific Research on Priority Areas (grants 18075007 and 23012019)] and Research Program for Agricultural Science and Technology Development (PJ010015022014) of the Rural Development Administration, Republic of Korea.