Interplay between SCARECROW, GA and LIKE HETEROCHROMATIN PROTEIN 1 in ground tissue patterning in the Arabidopsis root

Abstract

Regulated cell division is critical for the development of multi-cellular organisms. In the Arabidopsis root, SCARECROW (SCR) is required for the first cell division, but represses the subsequent, longitudinal asymmetric cell divisions that generate the two cell types of the ground tissue - cortex and endodermis. To elucidate the molecular basis of the role of SCR in ground tissue patterning, we screened for SCR-interacting proteins using the yeast two-hybrid method. A number of putative SCR-interacting proteins were identified, among them LIKE HETEROCHROMATIN PROTEIN 1 (LHP1). In lhp1 mutants, a second longitudinal asymmetric cell division occurs in the ground tissue earlier than in wild-type plants. Similar to the scr mutant, this premature middle cortex phenotype is suppressed by the phytohormone gibberellin (GA). We provide evidence that the N-terminal domain of SCR is required for the interaction between SCR and LHP1 as well as with other interacting partners, and that this domain is essential for repression of asymmetric cell divisions. Consistent with a role for GA in cortex proliferation, mutants of key GA signaling components produce a middle cortex precociously. Intriguingly, we found that the spindly (spy) mutant has a similar middle cortex phenotype. As SPY homologs in animals physically interact with histone deacetylase, we examined the role of histone deacetylation in middle cortex formation. We show that inhibition of histone deacetylase activity causes premature middle cortex formation in wild-type roots. Together, these results suggest that epigenetic regulation is probably the common basis for SCR and GA activity in cortex cell proliferation.

Figure 1. Functional analysis of SCR domains

(a) Schematic of the radial pattern in the Arabidopsis root, showing longitudinal (left) and transverse (right) sections of a primary root and the asymmetric cell divisions (framed area) that give rise to the endodermis and cortex. The outermost cell layer at the root tip is the lateral root cap, which protects the inner meristem. (b) Confocal microscopy images showing the middle cortex in wild-type (WT) root after 14 days of germination (DAG) and sporadic cortex formation in the scr-1 mutant at 7 DAG. (c) Molecular structure of the SCR protein, and the various fragments used in the protein-protein interaction assay. (d) GFP fluorescence detection showing subcellular localization of the fusion proteins between GFP and SCR domains after transient expression in onion epidermal cells. (e) Confocal microscopy images of transgenic Arabidopsis roots expressing GFP fusion proteins with SCR domains or ND-truncated SCR (pSCR::GFP-SCRΔND). The insets show parts of the images at higher magnification (3.5× higher). QC, quiescent center; E, endodermis; C, cortex; S, stele; Col, columella; CEI/CEID, cortex/endodermis initial or daughter cell; MC, middle cortex; M, mutant cell layer; ND, N-terminal variable domain; CD, the central domain, which spans the VHIID motif and the leucine heptad repeats (LHR); PS, the C-terminal domain containing the conserved PFYRE and SAW motifs. Scale bar = 20 μm.

Figure 2. LHP1 acts together with SCR to regulate the timing of middle cortex formation

(a) Confocal microscope image of 1-week-old lhp1 mutant root showing the presence of a middle cortex layer. (b) GFP fluorescence detection in the gLHP1::GFP transgenic plants, showing the LHP1 expression pattern in primary root. The meristematic region is shown in the inset at higher magnification (2.5× higher). (c) ChIP-quantitative PCR assay for SCR and LHP1 binding to the MGP promoter. (d) Quantitative RT-PCR assay of SCR target genes in the lhp1–4 mutant and wild-type root. The error bars in (c) and (d) represent standard deviations between duplicate measurements. C, cortex; E, endodermis; MC, middle cortex. Scale bar = 20 μm.

Figure 3. Confocal microscope images of primary roots of GA signaling mutants 1 week after germination

(a–c) gid1, rgaΔ 17 and spy-3 mutants, respectively. (d–f) SCR and SHR expression and subcellular localization in the spy-3 mutant, as visualized by expression of the pSCR::GFP, pSCR::GFP-SCR and pSHR::SHR-GFP constructs. The insets show the framed areas at higher magnification (2× higher). C, cortex; E, endodermis; MC, middle cortex. Scale bar = 20 μm.

Figure 4. TSA causes premature middle cortex formation

(a,b) Confocal microscope images of wild-type roots 6 days after germination on MS medium containing 1 μg ml⁻¹ TSA. (c–j) Time-course study of middle cortex formation in wild-type roots that express the cortex marker pCO2:HYFP. Three-day-old seedlings were transferred onto MS medium containing 1 μg ml⁻¹ TSA, and confocal microscope images were taken 8 h (c,d), 24 h (e,f), 48 h (g,h) and 72 h (i,j) after transfer. The framed areas in (a), (c), (e) and (f) are also shown at higher magnification (1.5× higher). C, cortex; E, endodermis; MC, middle cortex; QC, quiescent center. Scale bar = 20 μm.

Figure 5. Model depicting the interplay between SCR, GA and the epigenetic machinery in cortex proliferation in the Arabidopsis root

Upon entry into the CEID, SHR together with SCR initially activates cell division genes, but the transcription is rapidly quenched by the interaction between SCR and LHP1, which may be preceded by changes in histone modifications due to other mechanisms. In parallel, GA also suppresses the cell division genes through SPY, which potentially interacts with and modulates the activity of associated HDACs.