Research ArticleEpigenetics

Arabidopsis ATXR2 deposits H3K36me3 at the promoters of LBD genes to facilitate cellular dedifferentiation

See allHide authors and affiliations

Sci. Signal.  28 Nov 2017:
Vol. 10, Issue 507, eaan0316
DOI: 10.1126/scisignal.aan0316

Epigenetic control of dedifferentiation

Some plant cells can dedifferentiate to form a mass of pluripotent cells called callus. This not only occurs at wound sites but can also be induced by specific laboratory culture conditions. Lee et al. found that the histone lysine methyltransferase ATXR2 promoted cellular dedifferentiation during callus formation in Arabidopsis thaliana by stimulating the expression of LBD genes, which encode transcription factors that promote cell cycle progression. ATXR2 localized to LBD promoters, stimulated the accumulation of lysine-methylated histones at these promoters, and was recruited to the promoters by the transcription factors ARF7 and ARF19. Epigenetic regulation is a key mechanism controlling cell potency and differentiation in both plants and animals, and these findings contribute to understanding the remarkable developmental plasticity of plant cells.


Cellular dedifferentiation, the transition of differentiated somatic cells to pluripotent stem cells, ensures developmental plasticity and contributes to wound healing in plants. Wounding induces cells to form a mass of unorganized pluripotent cells called callus at the wound site. Explanted cells can also form callus tissues in vitro. Reversible cellular differentiation-dedifferentiation processes in higher eukaryotes are controlled mainly by chromatin modifications. We demonstrate that ARABIDOPSIS TRITHORAX-RELATED 2 (ATXR2), a histone lysine methyltransferase that promotes the accumulation of histone H3 proteins that are trimethylated on lysine 36 (H3K36me3) during callus formation, promotes early stages of cellular dedifferentiation through activation of LATERAL ORGAN BOUNDARIES DOMAIN (LBD) genes. The LBD genes of Arabidopsis thaliana are activated during cellular dedifferentiation to enhance the formation of callus. Leaf explants from Arabidopsis atxr2 mutants exhibited a reduced ability to form callus and a substantial reduction in LBD gene expression. ATXR2 bound to the promoters of LBD genes and was required for the deposition of H3K36me3 at these promoters. ATXR2 was recruited to LBD promoters by the transcription factors AUXIN RESPONSE FACTOR 7 (ARF7) and ARF19. Leaf explants from arf7-1arf19-2 double mutants were defective in callus formation and showed reduced H3K36me3 accumulation at LBD promoters. Genetic analysis provided further support that ARF7 and ARF19 were required for the ability of ATXR2 to promote the expression of LBD genes. These observations indicate that the ATXR2-ARF-LBD axis is key for the epigenetic regulation of callus formation in Arabidopsis.


Plant somatic cells have the remarkable capability of dedifferentiating to form unorganized masses of pluripotent cells called callus under specific environmental conditions, at wound sites, and when explanted into culture (1). This plasticity of cellular differentiation allows plants to optimize their growth and development for specific environmental conditions (2). Callus formation is an early event in cellular dedifferentiation, and intricate genetic programs underlie callus formation in Arabidopsis thaliana (3). For example, the hormone auxin and its downstream signaling components play key roles in callus formation (4). The transcription factors AUXIN RESPONSE FACTOR 7 (ARF7) and ARF19 activate several members of the LATERAL ORGAN BOUNDARIES DOMAIN (LBD) gene family to stimulate callus formation (5). The ARF proteins bind directly to the LBD16 and LBD29 promoters to stimulate their expression and also indirectly promote LBD17 and LBD18 expression by ARF-activated LBD16 and LBD29 (6). These four LBD zinc finger transcription factors facilitate cell cycle progression by stimulating the expression of E2 PROMOTER BINDING FACTOR a (E2Fa) genes that promote DNA replication and cell division (7).

Chromatin structure is dynamically regulated by covalent chemical modifications at nucleosomes, including histone methylation, acetylation, phosphorylation, and ubiquitination and DNA methylation (8). Chromatin context influences the accessibility of transcriptional regulators and thereby gene expression (9), facilitating stable but reversible patterns of gene expression. Because of the large number of histone methyltransferases in many plant genomes (10), the histone methylation process has been extensively investigated (11, 12). The effects of histone methylation are variable depending on the position of the modified residue and the number of methyl groups added (13). In general, methylation of histone H3 lysine 4 (H3K4) and H3K36 is associated with transcriptional activation (14), whereas methylation of H3K9, H3K27, and H4K20 is characteristic of repressive epigenetic marks (15).

SET domain proteins—named after the three founding members SuVar (39), E(z), and Trx from Drosophila melanogaster—have potential lysine methyltransferase activity (16) and are grouped into four major classes: (i) enhancer of zeste [E(z)] homologs; (ii) absent, small, or homeotic disks 1 (Ash1) homologs and related proteins; (iii) trithorax (Trx) homologs and related proteins; and (iv) suppressor of variegation [Su(var)] homologs and related proteins (17). H3K4 methylation is mainly catalyzed in yeast (Saccharomyces cerevisiae) by the Set1 class and in D. melanogaster by the Trx class (1820). In Arabidopsis, 10 proteins have been identified as putative methyltransferases that not only deposit H3K4me3 but also deposit H3K36me3 at a lesser frequency (17, 2123), based on the structural similarity to SET domains of Set1 and Trx and the analysis of biochemical activities (21). These 10 proteins are ARABIDOPSIS TRITHORAX 1-5 (ATX1-5), ARABIDOPSIS TRITHORAX-RELATED 1-4 (ATXR1-4), and ATXR7 (24).

The trithorax group (TrxG) proteins, which have the ability to catalyze active mark establishment and play roles in the maintenance of gene transcription, are associated with various physiological processes in Arabidopsis. ATX1 (also known as SDG27), ATX2 (also known as SDG30), and ATXR7 (also known as SDG25) are involved in the activation of FLOWERING LOCUS C (FLC) expression, which suppresses the floral transition (11). Mutations in these ATX genes suppress in part FRIGIDA-induced flowering delay with decreased FLC expression and reduced H3K4me3 deposition (2426). Furthermore, ATXR7 further contributes to FLC activation by depositing both H3K4me3 and H3K36me3 marks at the locus (24, 27). In addition, ATX1 is involved in dehydration stress responses (28). ATX1-deficient mutants displayed enlarged stomatal openings, increased water loss, and decreased dehydration tolerance with decreased abscisic acid (ABA) accumulation (28). Notably, ATX1 binds to the promoter of the rate-limiting ABA biosynthetic gene 9-CIS-EPOXYCAROTENOID DIOXYGENASE 3 (NCED3) and activates its expression with a substantial increase of H3K4me3 accumulation during dehydration (28). Moreover, ATX1 also activates ABA-independent genes, such as COLD-REGULATED 15A (COR15A), ALCOHOL DEHYDROGENASE 1 (ADH1), and ABSCISIC ACID RESPONSIVE ELEMENTS-BINDING FACTOR 2 (ABF2), integrating ABA-dependent and ABA-independent pathways (28). ATXR3 is required for global H3K4me3 deposition (29). Consistent with this, atxr3 mutations affect the expression of many genes and result in pleiotropic phenotypes including dwarfism, curly leaves, early flowering, terminal flowers, and sterility (29, 30). These observations demonstrate that diverse aspects of plant development are under the control of TrxG proteins.

Accumulating evidence indicates that epigenetic regulation further ensures a proper cellular dedifferentiation process by allowing massive reprogramming of gene expression (3133). Here, we report that, in addition to its global impact on genome-wide H3K36me3 and H3K4me3 accumulation, ATXR2 binds to LBD promoters and increases H3K36me3 accumulation at these promoters to stimulate callus induction. ARF7 and ARF19 recruit ATXR2 to the LBD promoters and facilitate proper histone modification at these loci. These observations indicate that coordination of multiple auxin-related factors underlies callus formation and contributes to the massive transcriptional changes required for a robust transition of cell fate.


Callus formation is reduced in atxr2 mutants

Genome-wide accumulation of H3K4me3 and H3K36me3, epigenetic marks that are catalyzed by ATXs and ATXRs and characteristic of open chromatin, increases substantially during the leaf-to-callus transition in Arabidopsis (33). Because open chromatin formation is necessary during callus formation (34), we investigated epigenetic modifications during callus formation. To identify the enzyme(s) regulating cell fate change through active mark deposition, we analyzed the callus formation capabilities of several atxr mutants using leaf explants on callus-inducing medium (CIM) (fig. S1).

Among the mutants examined, the transferred DNA (T-DNA) insertional mutant atxr2-1 exhibited reduced callus formation (Fig. 1A and fig. S2). Fresh weight measurements revealed that callus formation was suppressed by 25% in leaf explants from atxr2-1 mutants compared with leaf explants from wild-type plants (Fig. 1B). To validate these observations, we obtained the additional T-DNA insertional mutant allele atxr2-3 (fig. S2) and measured callus formation rate (Fig. 1, A and B). The atxr2-3 mutant leaf explants also showed reduced callus formation upon callus induction, similar to atxr2-1 (Fig. 1, A and B). Only atxr2-1 was used in subsequent analyses based on these phenotypic similarities.

Fig. 1 Callus formation in leaf explants from atxr2 mutant and ATXR2-overexpressing plants.

(A) Callus formation in leaves explanted from young wild-type (Col-0), atxr2-1, atxr2-3, and 35S:ATXR2-MYC plants. Scale bar, 5 mm. n > 30 plants of each genotype. (B) Fresh weight of calli from wild-type (Col-0), atxr2-1, atxr2-3, and 35S:ATXR2-MYC plants. Bars indicate the SE of the mean. *P < 0.05 (Student’s t test). n = 30 calli of each genotype.

To provide further support for the role of ATXR2 in callus formation, we generated transgenic plants overexpressing ATXR2 (35S:ATXR2-MYC) and analyzed their callus formation capability (Fig. 1A and fig. S3). Leaf explants from 35S:ATXR2-MYC transgenic plants exhibited more robust callus formation than wild-type leaf explants (Fig. 1A and fig. S3). Fresh weight measurements further supported the rapid progression of cellular dedifferentiation and proliferation in 35S:ATXR2-MYC transgenic leaf explants (Fig. 1B and fig. S3). Thus, ATXR2 promoted callus formation.

ATXR2 activates LBD gene expression

To obtain clues about the signaling network involving ATXR2, we analyzed wild-type, atxr2-1, and 35S:ATXR2-MYC calli for the expression of genes involved in callus formation, including ARABIDOPSIS RESPONSE REGULATOR 1 (ARR1), ARR21, BABY BOOM (BBM), ENHANCER OF SHOOT REGENERATION 1 (ESR1), ESR2, LBD16, LBD17, LBD18, LBD29, LEAFY COTYLEDON 1 (LEC1), LEC2, WOUND-INDUCED DEDIFFERENTIATION 1 (WIND1), WIND2, and WIND3. Reverse transcription quantitative real-time polymerase chain reaction (RT-qPCR) analysis showed that accumulation of LBD transcripts was repressed in atxr2-1 mutant calli and greatly increased in 35S:ATXR2-MYC calli relative to wild-type calli (Fig. 2A). Reduced LBD expression was also observed in atxr2-3 mutant calli (fig. S4). Because LBDs were expressed mainly in the callus tissues compared to differentiated leaf tissues (fig. S5), it was possible that reduced LBD expression in atxr2-1 mutant calli may be attributable to reduced callus size. However, we harvested the plant materials at 7 DAC (days after explant onto CIM), when callus formation is starting, and thus, no clear difference in callus size was observed between mutants and wild-type plants (fig. S6). Furthermore, expression of other genes involved in callus formation was not altered in ATXR2-overexpressing (35S:ATXR2-MYC) plants (Fig. 2A). In particular, expression of the callus-specific marker WUSCHEL-RELATED HOMEOBOX 5 (WOX5) was also unchanged in atxr2-1 (fig. S7), indicating specific regulation of LBDs by ATXR2.

Fig. 2 Transcriptional activation of LBD genes by ATXR2.

(A) Expression profiling of the indicated genes involved in callus formation in 35S:ATXR2-MYC and atxr2-1 calli by reverse transcription quantitative real-time polymerase chain reaction (RT-qPCR) and normalized to each gene’s expression in wild-type (Col-0) plants. Bars indicate the SE of the mean. *P < 0.05 (Student’s t test). n = 3 biological replicates. (B) Kinetics of LBD expression in Col-0 and atxr2-1 leaf explants during callus formation at 2, 4, and 7 days after culturing on callus-inducing medium (DAC). Expression of each gene was normalized to the expression of that gene at time 2 in Col-0. *P < 0.05 (Student’s t test). n = 3 biological replicates. (C) Promoter analysis of LBD genes. The regions labeled A to M were identified as putative binding sites for ATXR2. Black lines above the labels indicate the regions that were amplified by PCR after chromatin immunoprecipitation (ChIP). Black boxes indicate exons. (D) Enrichment of ATXR2 on promoter regions as determined by ChIP-qPCR. Values for qPCR analysis were normalized to eIF4a in wild-type empty vector (EV) control plants. Bars indicate the SE of the mean. *P < 0.05 (Student’s t test). n = 3 biological replicates.

In addition, we also analyzed the expression kinetics of LBD genes during callus formation. The LBD genes were rapidly induced upon placing leaves on CIM and showed the greatest expression during the period of 4 to 7 DAC (Fig. 2B). However, LBD gene induction was reduced in atxr2 mutant calli (Fig. 2B and fig. S4), suggesting that ATXR2 promotes LBD expression during the process of callus formation.

On the basis of its similarity to other known histone-modifying enzymes, ATXR2 is predicted to bind to specific target promoters and modify the histones at those loci (24, 25). To determine whether ATXR2 binds to LBD gene promoters, we performed chromatin immunoprecipitation (ChIP) using 35S:ATXR2-MYC transgenic calli. Total protein extracts from control and 35S:ATXR2-MYC transgenic calli were immunoprecipitated with an antibody specific for MYC, and the DNA bound to MYC-tagged ATXR2 proteins was analyzed by qPCR assays using primers specific for several regions of LBD promoters (Fig. 2C). ATXR2 bound mainly to the LBD16 and LBD29 promoters (Fig. 2D). Control ChIP in the absence of the MYC-specific antibody did not enrich LBD16 and LBD29 fragments (fig. S8). To confirm the functionality of the ATXR2-MYC fusion for LBD activation, we conducted transient expression assays using Arabidopsis protoplasts, which are cells isolated from leaves and treated to remove the cell wall. The core cis-elements of the LBD promoters were fused in the recombinant reporter plasmid, and the construct was expressed with effector constructs expressing ATXR2 in Arabidopsis protoplasts isolated from wild-type and atxr2-1 leaves. ATXR2 bound to and transcriptionally activated the LBD genes in protoplasts from both wild-type plants and atxr2-1 mutants (fig. S9). In addition, the ATXR2-MYC fusion complemented the atxr2-1 mutation (fig. S9).

Two other LBD genes, LBD17 and LBD18, were not targeted by ATXR2 (Fig. 2D). However, expression of these genes may be influenced by LBD16 and LBD29 (6), suggesting that ATXR2 could influence the expression of LBD17 and LBD18 indirectly through its action on LBD16 and LBD29. These observations account for the reduced expression of all examined LBDs in atxr2-1 mutant calli. Another LBD gene examined as a negative control (LBD1) was not targeted by ATXR2 (fig. S10). These results suggest that ATXR2 specifically activates a subset of LBD genes by binding directly to the promoters and possibly modifying their chromatin context.

ATXR2 establishes H3K36me3 at LBD promoters

Sequence analysis places ATXR2 in the Trx family of proteins that preferentially and directly confer H3K4me3 accumulation (17). To examine this possibility, we examined the global accumulation of H3K4me3 in wild-type and atxr2-1 calli. We subjected proteins isolated from calli to immunoblot analysis with an antibody recognizing H3K4me3. Contrary to our expectations, H3K4me3 accumulation was only slightly altered in atxr2-1 mutant calli (fig. S11). We noted that ATXR2, like ATXR1 and ATXR4, has an interrupted SET domain (17), and thus hypothesized that ATXR2 might have biochemical activities that differ from other SET domain proteins in establishing epigenetic marks. To examine this hypothesis, we examined the accumulation of H3K9me3, H3K27me3, and H3K36me3 in wild-type and atxr2-1 calli. Notably, global H3K36me3 accumulation was reduced in atxr2-1 mutant calli at 4 DAC (Fig. 3A), whereas H3K9me3 and H3K27me3 accumulation was unchanged or increased (fig. S11). The biochemical function of ATXR2 in H3K36me3 deposition was further supported by our observation of increased H3K36me3 accumulation in 35S:ATXR2-MYC transgenic calli (fig. S12). These results suggest that ATXR2 primarily promotes global H3K36me3 deposition, and to a lesser degree H3K4me3, during callus formation.

Fig. 3 ATXR2 mediates H3K36me3 accumulation at LBD loci during callus formation.

(A) Global accumulation of H3K36me3 in wild-type (Col-0) and atxr2-1 leaf explants during callus formation. Bands from three independent blots were quantified (right). n =3 biological replicates. (B) H3K36me3 accumulation at the LBD loci indicated in Fig. 2C. Bars indicate the SE of the mean. *P < 0.05 (Student’s t test). n = 3 biological replicates. (C) Kinetics of H3K36me3 accumulation at the LBD16 and LBD29 promoters during callus formation in Col-0 and atxr2-1 plants. Bars indicate the SE of the mean. *P < 0.05 (Student’s t test). n = 3 biological replicates.

To assess the importance of ATXR2 in H3K36me3 deposition at LBD loci, we performed ChIP analysis with an antibody specific for H3K36me3 using wild-type, atxr2-1, and 35S:ATXR2-MYC calli. qPCR analysis showed that H3K36me3 accumulation at the LBD16 and LBD29 promoters was impaired in atxr2-1 mutant calli compared to wild-type calli (Fig. 3B) at 2 to 4 DAC (Fig. 3C). In contrast, this histone modification was enhanced in 35S:ATXR2-MYC transgenic calli (Fig. 3B). H3K36me3 accumulation at the LBD17 and LBD18 promoters was unchanged in atxr2-1 and 35S:ATXR2-MYC transgenic calli (Fig. 3B), which is consistent with the binding specificity of ATXR2 (Fig. 2D). In addition, H3K4me3 accumulation at the LBD promoters was unchanged in atxr2-1 mutant calli (fig. S13). These results indicate that ATXR2 promotes H3K36me3 deposition at the LBD loci to ensure efficient callus formation.

ATXR2 interacts with ARFs

We have shown that ATXR2 is recruited to the LBD promoters and directly catalyzes histone modification at cognate regions. One question this raises is how ATXR2 recognizes its target promoters. We hypothesized that ATXR2 may be guided to the LBD promoters by transcription factors that directly bind to nearby regions. ARF7 and ARF19 were strong candidates, because they bind to the promoters of the LBD16 and LBD29 genes (6) and because arf7 arf19 double mutants exhibit abnormal callus formation (5). We therefore decided to test the potential physical interactions between ATXR2 and ARF7 or ARF19.

We carried out yeast two-hybrid (Y2H) assays by coexpressing ATXR2 fused in-frame to the 3′ end of the GAL4 DNA binding domain (DBD) in yeast cells along with the GAL4 activation domain (AD) alone or fused to ARF7 or ARF19. Cell growth on selective medium showed that ATXR2 interacted with full-length ARF19 but not full-length ARF7 (Fig. 4A). We also constructed a series of deletion forms of ATXR2, ARF7, and ARF19 (Fig. 4B). Y2H analysis revealed that the N-terminal region of ARF19 containing the B3 domain was responsible for the interaction with ATXR2 (Fig. 4C). In addition, the middle region of ATXR2, which has the SET domain, associated with both ARF7 and ARF19 (Fig. 4D), suggesting that ATXR2 interacts with both these ARFs depending on structural compatibility in yeast cells.

Fig. 4 ATXR2 interacts with ARFs.

(A) Yeast two-hybrid assays using ATXR2 fused to the Gal4 DNA binding domain (DBD) and ARFs fused to the Gal4 transcriptional activation domain (AD). Full-length GAL4 was used as a positive control. LW, dropout medium without leucine (L) or tryptophan (W); LWAH, dropout medium without L, W, adenine (A), or histidine (H). n = 3 biological replicates. (B) Deletion constructs of ATXR2 and ARFs. Numbers indicate residue positions; black boxes indicate the ARF B3 domain; white boxes indicate the low-complexity region; gray boxes indicate the coiled-coil region of ARF7; hatched boxes indicate the SET domain of ATXR2. (C) Interactions of ATXR2 with ARF fragments. Bars indicate the SE of the mean. *P < 0.05 (Student’s t test). n = 3 biological replicates. (D) ARF interactions with ATXR2 fragments. Bars indicate the SE of the mean. *P < 0.05 (Student’s t test). n = 3 biological replicates. (E) Bimolecular fluorescence complementation (BiFC) assays in Arabidopsis protoplasts transiently expressing the indicated combinations of ATXR2 or ARFs YFP (yellow fluorescent protein) fusion constructs. Scale bar, 20 μm. n = 3 biological replicates. DIC, differential interference contrast. (F) Split-luciferase (LUC) assays. Partial fragments of Luciferase (nLuc and cLuc) were fused with ATXR2 or ARFs. Bars indicate the SE of the mean. *P < 0.05 (Student’s t test). n = 3 biological replicates.

To provide evidence for the interaction of ATXR2 and ARFs in vivo, we performed bimolecular fluorescence complementation using Arabidopsis protoplasts. The ATXR2 cDNA sequence was fused in-frame to the 5′ end of a gene sequence encoding the N-terminal half of yellow fluorescent protein (nYFP), and each ARF gene was fused in-frame to the 5′ end of a sequence encoding the C-terminal half of YFP (cYFP). The fusion constructs were then transiently coexpressed in Arabidopsis protoplasts. Yellow fluorescence was visualized exclusively in the nucleus in all ATXR2-ARF combinations (Fig. 4E and figs. S14 and S15). Binding of ATXR2 was specific to ARF7 and ARF19, because other ARFs did not associate with ATXR2 (fig. S16). To quantify the physical interaction between ATXR2 and these two ARFs, we used a split Luciferase (Luc) assay. ATXR2 was fused to the amino portion of Luc (nLuc), and ARFs were fused with carboxy portion of Luc (cLuc). Coexpression of ATXR2-nLuc and ARF-cLuc constructs in Arabidopsis protoplasts resulted in enhanced Luc activity (Fig. 4F). These results indicate that ATXR2 forms a complex with ARF7 and ARF19 in plant cells.

ARFs are required for ATXR2 function

Consistent with the physical association of ARFs and ATXR2, the arf7-1arf19-2 double mutant was defective in callus formation (Fig. 5, A and B) and showed a substantial reduction in LBD expression compared to wild-type plants (Fig. 5C), similar to the atxr2-1 mutant. ARF7 and ARF19 are known to bind directly to the LBD16 and LBD29 promoters (6) and stimulate the expression of these genes as well as other LBD genes including LBD17 and LBD18 (6). The specific binding of ATXR2 to the LBD16 and LBD29 promoters (Fig. 2D) and the interaction between ATXR2 and ARF7 and ARF19 support these observations (Fig. 4). The ATXR2-binding sites overlap with ARF-binding regions in the LBD16 and LBD29 promoters (Fig. 2D) (6). To verify a role for ARFs in epigenetic activation of LBDs, we analyzed the accumulation of H3K36me3 at the LBD16 and LBD29 promoters in arf7-1arf19-2 mutants. Notably, the LBD16 and LBD29 promoters showed reduced H3K36me3 deposition in arf7-1arf19-2 mutant calli (Fig. 5D) compared to wild-type calli, whereas H3K4me3 accumulation was not altered (fig. S17).

Fig. 5 Functional coordination of ATXR2 with ARF transcription factors.

(A) Callus formation in leaf explants from wild-type (Col-0) and arf7-1arf19-2 double-mutant plants. Scale bar, 5 mm. (B) Quantification (fresh weight) of callus formation in leaf explants from Col-0 and arf7-1arf19-2 plants. Bars indicate the SE of the mean. *P < 0.05 (Student’s t test). n = 30 leaf explants of each genotype. (C) LBD transcript accumulation in Col-0 and arf7-1arf19-2 mutants during callus formation. Expression of each gene was normalized to the expression of that gene at time 1 in Col-0. Bars indicate the SE of the mean. *P < 0.05 (Student’s t test). n = 3 biological replicates. (D) H3K36me3 accumulation at LBD loci in Col-0 and arf7-1arf19-2 calli. H3K36me3 abundance was normalized to the abundance at time 0 in Col-0. Bars indicate the SE of the mean. *P < 0.05 (Student’s t test). n = 3 biological replicates.

To confirm that ATXR2 binding to LBD promoters was mediated by ARFs, we generated 35S:ATXR2-MYC/arf7-1arf19-2 plants, in which protein accumulation of ATXR2 was similar to 35S:ATXR2-MYC/Col-0, by genetic crosses (fig. S18). ChIP analysis with an antibody specific for MYC revealed that the ATXR2 protein was recruited to the LBD promoters in a wild-type (Col-0) background, but ATXR2 binding to the LBD promoters was compromised by introducing the arf7-1 and arf19-2 mutations (Fig. 6A). Accordingly, increased LBD expression in 35S:ATXR2-MYC transgenic calli was also reduced in the arf7-1arf19-2 mutant background (Fig. 6B). Furthermore, the callus formation capability of 35S:ATXR2-MYC was largely dependent on ARFs (Fig. 6, C and D). These results indicate that the ATXR2 and ARF proteins are functionally interconnected in promoting LBD-dependent callus formation.

Fig. 6 Requirement of ARFs for ATXR2 function.

(A) Quantification of ATXR2 binding to LBD promoters in arf7-1arf19-2 double-mutant plants by ChIP analysis. Enrichment was quantified relative to the amount at each gene promoter in control (EV) plants. Bars indicate the SE of the mean. *P < 0.05 (Student’s t test). n = 3 biological replicates. (B) Expression of LBD16 and LBD29 in calli from arf7-1arf19-2 double-mutant plants overexpressing ATXR2 (35S:ATXR2-MYC/arf7-1arf19-2). Different letters represent a significant difference at P < 0.05 [one-way analysis of variance (ANOVA) with Fisher’s post hoc test). n = 3 biological replicates. (C) Callus formation in 35S:ATXR2-MYC/arf7-1arf19-2 leaf explants. n > 30 plants of each genotype. (D) Quantification (fresh weight) of callus formation in leaf explants from plants of the indicated genotypes. Different letters represent a significant difference at P < 0.05 (one-way ANOVA with Fisher’s post hoc test). Bars indicate the SE of the mean. n = 30 calli of each genotype. (E) Proposed role of ATXR2 in callus formation. ATXR2 interacts with ARF7 and ARF19 to bind to the promoters of LBD16 and LBD29. ATXR2 promotes LBD gene expression by catalyzing the deposition of H3K36me3 at these promoters. ATXR2 indirectly affects the expression of LBD17 and LBD18 through its direct effects on LBD16 and LBD29 expression. These activities facilitate callus formation. CIM, callus-inducing medium.

Finally, we examined whether ATXR2 depends on LBDs for the control of callus formation. 35S:ATXR2-MYC transgenic plants were crossed with 35S:LBD16-SRDX transgenic plants. Leaf explants of 35S:ATXR2-MYC plants exhibited enhanced callus formation (Fig. 1A and fig. S19), whereas reduced callus formation was observed in 35S:LBD16-SRDX leaf explants (fig. S19) (5). Callus formation of 35S:ATXR2-MYC; 35S:LBD16-SRDX was similar to that of 35S:LBD16-SRDX (fig. S19), indicating that LBDs are epistatic to ATXR2 in the regulation of callus formation (Fig. 6E).


Reversible transitions in cellular differentiation states require massive gene expression reprogramming (35). Chromatin modification is the most plausible regulatory mechanism that facilitates global gene expression changes, and consistent with this, epigenetic regulation is considered as a key molecular scheme underlying cellular reprogramming in eukaryotes (36, 37). Several lines of evidence further support that chromatin status is closely associated with cell identity. Differentiated cells have a closed chromatin status with accumulation of repressive marks, whereas embryonic and pluripotent dedifferentiated cells have a relatively open chromatin status with active mark deposition (38, 39).

Plant dedifferentiation is a multistep process that begins with the formation of partially dedifferentiated founder cells, followed by callus formation and the establishment of pluripotency in callus cells. Widespread epigenetic changes, including modifications of both DNA and histones, occur during callus formation. Despite the importance of chromatin modification during callus formation, few molecular components responsible for this process have been demonstrated in Arabidopsis. DNA methylation is likely essential for callus formation (40), and the global cytosine methylation landscape changes during callus formation (40, 41). Arabidopsis METHYLTRANSFERASE 1 (MET1) has been implicated in cell fate changes and is an ortholog of mammalian DNA METHYLTRANSFERASE 1 (DNMT1), which maintains CG methylation globally in the genome (40). MET1-deficient mutants exhibit defective callus formation with hypomethylation of some genic loci, including GLUTATHIONE S-TRANSFERASE TAU 10 (GSTU10), MITOGEN-ACTIVATED PROTEIN KINASE 12 (MAPK12), BETA-XYLOSIDASE 1 (BXL1), and WUSCHEL (WUS) (40, 42), as well as in many nongenic regions (40). Furthermore, chromomethylase 3 (cmt3) and domains rearranged methylase (drm) mutants, which have altered patterns of both CHG and CHH methylation (40), also display impaired callus formation (40), underscoring the importance of DNA methylation in cell fate changes.

Histone modification also plays a key role in callus formation in Arabidopsis. Upon callus induction, marks of active chromatin such as H3ac, H3K4me3, and H3K36me3 accumulate globally in the genome (32, 33). In addition to these global effects, local epigenetic modifications at genic regions are also important for robust cellular dedifferentiation. Some histone modifiers that act as negative regulators of gene transcription have been implicated as key drivers of callus formation. For example, transcripts of several genes encoding histone deacetylases accumulate during callus formation, and genetic mutations in HISTONE DEACETYLASE 9 (HDA9) or HD-TUINS PROTEIN 1 (HDT1) impede callus formation (32). The polycomb repressive complex 2 (PRC2) also enhances callus formation by establishing the repressive H3K27me3 mark. PRC2 promotes leaf-to-callus transition through repression of leaf identity genes such as SAWTOOTH 1 (SAW1), SAW2, and TEOSINTE BRANCHED1-CYCLOIDEA-PCF 10 (TCP10) during callus formation (31).

Here, we report that ATXR2 is a key component that not only promotes global deposition of H3K36me3 but also contributes to epigenetic control of the auxin signaling pathway during callus formation. The ATXR2-ARF-LBD circuitry is a key player in cell fate change. Ectopic expression of these components results in enhanced callus formation on CIM, whereas the genetic mutants are defective in callus induction. ATXR2 activates LBD expression by catalyzing H3K36me3 deposition at the LBD promoter regions. Notably, ARFs recruit ATXR2 to target promoters, and the ARF-ATXR2 complex ensures precise epigenetic modification at the LBD promoters, adding complexity to callus formation. It remains to be determined whether additional targets of ATXR2, other than the LBDs, play a role in promoting callus formation. Likewise, ATXR2 may have additional roles in callus formation because it can also catalyze the formation of H3K4me3 marks in addition to H3K36me3 mark. Thus, ATXR2 may make other contributions to the remarkable capability of cellular dedifferentiation in Arabidopsis.

Note that the roles of TrxG proteins in cellular reprogramming are well conserved in higher eukaryotes. For example, WD REPEAT DOMAIN 5 (WDR5) is the core subunit of the TrxG protein–containing complex in humans, and its expression correlates with the dedifferentiation state of cells (43). This protein interacts with the pluripotency-promoting transcription factor OCTAMER-BINDING TRANSCRIPTION FACTOR 4 (OCT4) to promote H3K4me3 accumulation and thus stimulate downstream genes, such as POU DOMAIN, CLASS 5, TRANSCRIPTION FACTOR 1 (POU5F1), NANOG, and SEX DETERMINING REGION Y-BOX 2 (SOX2) (43). Furthermore, a mouse TrxG protein, Absent, small or homeotic 2-like protein (Ash2l), is also implicated in establishing pluripotency (44). The Ash2l protein, which primarily confers the H3K4me3 mark, works together with chromatin remodelers and H3K9 demethylases to globally maintain an open chromatin landscape in mouse embryonic stem cells (44). The ATXR2 gene is an example of a TrxG protein possibly performing a similar function in a plant, contributing to the robust callus formation capability of Arabidopsis. On the basis of this initial finding regarding TrxG-dependent cellular dedifferentiation, conserved mechanisms underlying cellular dedifferentiation in eukaryotes can be further investigated.


Plant materials and growth conditions

A. thaliana (Columbia-0 ecotype) was used for all experiments unless otherwise specified. Arabidopsis seeds were surface sterilized and sown on 0.7% agar plates containing half-strengthened Murashige and Skoog media. Plants were grown under long-day conditions (16-hour light/8-hour dark cycles) with white fluorescent light (120 μmol photons m−2 s−1) at 22° to 23°C. The arf7-1arf19-2 mutant was described previously (6). The atxr2-1 (SAIL-600-E07) and atxr2-3 (SALK-095652) mutants were isolated from a T-DNA insertional mutant pool deposited in the Arabidopsis Biological Resource Center (ABRC;

To produce transgenic plants overexpressing the ATXR2 gene, we subcloned a full-length cDNA into the modified binary pBA002 vector under the control of the CaMV 35S promoter. Agrobacterium tumefaciens–mediated Arabidopsis transformation was then performed.

For callus induction, leaf explants of third leaf from 2-week-old plants were placed on CIM [B5 medium supplemented with 2,4-dichlorophenoxyacetic acid (0.5 μg/ml) and kinetin (0.05 μg/ml)], followed by incubation at 22°C in the dark for additional 2 weeks. Thirty calli of each genotype were collected to measure fresh weight. Three independent measurements were averaged. Statistically significant differences between wild-type and transgenic or mutant calli are determined by Student’s t test.

RT-qPCR analysis

Total RNA was extracted using TRI reagent (Takara Bio) according to the manufacturer’s recommendations. RT was performed using Moloney Murine Leukemia Virus reverse transcriptase (Dr. Protein) with oligo(dT20) to synthesize first-strand cDNA from 2 μg of total RNA. The cDNAs were diluted to 100 μl with Tris-EDTA (TE) buffer, and 1 μl of diluted cDNA was used for PCR amplification.

RT-qPCR reactions were performed in 96-well blocks using the StepOnePlus Real-Time PCR System (Applied Biosystems). The PCR primers used are listed in table S1. The values for each set of primers were normalized relative to the EUKARYOTIC TRANSLATION INITIATION FACTOR 4A1 (eIF4A) gene (At3g13920). All RT-qPCR reactions were performed in biological triplicates using total RNA samples extracted from three independent replicate samples. The comparative ΔΔCt method was used to evaluate the relative quantities of each amplified product in the samples. The threshold cycle (Ct) was automatically determined for each reaction with the analysis software set using default parameters. Specificity of the RT-qPCR reactions was determined by melt curve analysis of the amplified products.

Chromatin immunoprecipitation

Putative ATXR2 binding sites were predicted on the basis of the Web-based promoter analysis. Promoter regions containing multiple putative cis-elements were chosen as candidate binding sites. The epitope-tagged transgenic plant samples were cross-linked with 1% formaldehyde, ground to powder in liquid nitrogen, and then sonicated. The sonicated chromatin complexes were precipitated with salmon sperm DNA/protein A agarose beads (Millipore) and antibodies recognizing MYC, H3K4me3, H3K27me3, and H3K36me3 (Millipore). Precipitated DNA was purified using phenol/chloroform/isoamyl alcohol and sodium acetate (pH 5.2). The abundance of specific precipitated DNA fragments was quantified by qPCR using the primer pairs listed in table S2. Values were normalized according to input DNA abundance. Values for control plants were set to 1 after normalization against eIF4a for qPCR analysis.

Yeast two-hybrid assays

Y2H assays were performed using the BD Matchmaker system (Clontech). Full-length or truncated cDNAs of ARF7 and ARF19 were cloned into the pGADT7 vector for GAL4 AD fusion. Full-length or truncated cDNAs of ATXR2 were cloned into the pGBKT7 vector for GAL4 DBD fusion. Full-sized GAL4 transcription factor was expressed as a positive control (Clontech). The yeast strain AH109 harboring the LacZ and His reporter genes was used. The expression constructs were cotransformed into yeast AH109 cells, and transformed cells were selected by growth on SD/-Leu/-Trp medium and SD/-Leu/-Trp/-His/-Ade. Interactions between proteins were analyzed by measuring β-galactosidase activity using o-nitrophenyl-β-d-galactopyranoside as substrate.

Preparation of Arabidopsis protoplasts

Leaves from 4-week-old plants were cut into 0.5-mm pieces using a fresh razor blade. Twenty leaves were digested in 15 ml of enzyme solution [0.8% cellulase (Yakult), 0.2% macerozyme (Yakult), 0.4 M mannitol, 10 mM CaCl2, 20 mM KCl, 0.1% bovine serum albumin, and 20 mM MES (pH 5.7)], vacuumed for 20 min, and incubated in the dark for 5 hours at 22° to 23°C. Protoplasts were then passed through 40-μm stainless mesh and collected after a gentle wash with W5 media (154 mM NaCl, 125 mM CaCl2, 5 mM KCl, 2 mM MES, 5 mM glucose adjusted to pH 5.7 with KOH).

Bimolecular fluorescence complementation assays

The ATXR2 gene was fused in-frame to the 5′ end of a gene sequence encoding the C-terminal half of enhanced YFP (EYFP) in the pSATN-cEYFP-C1 vector (E3082). The ARF cDNA sequences were fused in-frame to the 5′ end of a gene sequence encoding the N-terminal half of EYFP in the pSATN-nEYFP-C1 vector (E3081). Using the polyethylene glycol (PEG) method of transformation, the expression constructs were transfected into Arabidopsis protoplasts that had been generated by standard methods (45). Expression of the fusion constructs was monitored by fluorescence microscopy using the Zeiss LSM510 confocal microscope (Carl Zeiss).

Immunoblot analysis

Harvested plant materials were ground in liquid nitrogen, and total cellular extracts were suspended in SDS–polyacrylamide gel electrophoresis (PAGE) sample loading buffer. The protein samples were then analyzed by SDS-PAGE (10% gels) and blotted onto Hybond-P+ membranes (Amersham Pharmacia). Epitope-tagged proteins were immunologically detected using antibodies specific for H3K4me3, H3K9me3, H3K27me3, or H3K36me3 (Millipore).

Transient gene expression assays

For transient expression assays using Arabidopsis protoplasts, reporter and effector plasmids were constructed. The reporter plasmid contains a minimal 35S promoter sequence and the GUS gene. The core elements on the LBD promoters were inserted into the reporter plasmid. To construct effector plasmids, cDNAs were inserted into the effector vector containing the CaMV 35S promoter. Recombinant reporter and effector plasmids were cotransformed into Arabidopsis protoplasts by PEG-mediated transformation. The GUS activities were measured by a fluorometric method. A CaMV 35S promoter–Luc construct was also cotransformed as an internal control. The Luc assay was performed using the Luciferase Assay System kit (Promega).


Fig. S1. Callus formation in leaf explants from atxr1 and atxr4 mutants.

Fig. S2. ATXR2 expression in atxr2 mutants.

Fig. S3. ATXR2 expression in 35S:ATXR2-MYC transgenic plants.

Fig. S4. Transcript accumulation of LBDs in atxr2-3 mutant calli.

Fig. S5. Spatial expression of LBD16 in leaf explants and calli.

Fig. S6. Phenotype of leaf explant–derived callus at 7 DAC.

Fig. S7. Transcript accumulation of WOX5 in atxr2-1 calli.

Fig. S8. ChIP assays using antibody-free resin.

Fig. S9. Transient expression assays.

Fig. S10. Binding of ATXR2 to the LBD1 promoter.

Fig. S11. Accumulation of H3K4me3, H3K9me3, and H3K27me3 in atxr2-1 mutants during callus induction.

Fig. S12. Accumulation of H3K4me3 and H3K36me3 in 35S:ATXR2-MYC calli.

Fig. S13. H3K4me3 accumulation at the LBD promoters in atxr2-1.

Fig. S14. Interactions of ATXR2 with ARF7 and ARF19.

Fig. S15. Interactions of ATXR2 with deletion constructs of ARF7 and ARF19.

Fig. S16. Interactions of ATXR2 with other ARFs.

Fig. S17. Accumulation of H3K4me3 at LBD promoters in arf7-1arf19-2 mutant calli.

Fig. S18. Protein accumulation of ATXR2 in 35S:ATXR2-MYC/Col-0 and 35S:ATXR2-MYC/arf7-1arf19-2.

Fig. S19. Callus formation of leaf explants from 35S:ATXR2-MYC x 35S:LBD16-SRDX plants.

Table S1. Primers used for PCR.

Table S2. Primers used for ChIP assays.


Acknowledgments: We thank the Nottingham Arabidopsis Stock Centre (NASC) and the Arabidopsis Biological Resource Center (ABRC) for Arabidopsis mutant seeds used in this work and M. S. Choi for critical reading of the manuscript. Funding: This work was supported by the Basic Research Laboratory (2017R1A4A1015620) and Basic Science Research (NRF-2016R1D1A1B03931139) programs provided by the National Research Foundation of Korea and by the Next-Generation BioGreen 21 Program (PJ01119204) provided by the Rural Development Administration. This paper was also supported by SEOK CHUN Research Fund, Sungkyunkwan University, 2016. Author contributions: P.J.S. conceived and designed the experiments. P.J.S. wrote the paper with the help of K.L. K.L. and O.-S.P. conducted experiments and contributed to the study design. K.L. analyzed the data. Competing interests: The authors declare that they have no competing interests.
View Abstract

Navigate This Article