The Histone Chaperone NRP1 Interacts with WEREWOLF to Activate GLABRA2 in Arabidopsis Root Hair Development

Abstract

NUCLEOSOME ASSEMBLY PROTEIN1 (NAP1) defines an evolutionarily conserved family of histone chaperones and loss of function of the Arabidopsis thaliana NAP1 family genes NAP1-RELATED PROTEIN1 (NRP1) and NRP2 causes abnormal root hair formation. Yet, the underlying molecular mechanisms remain unclear. Here, we show that NRP1 interacts with the transcription factor WEREWOLF (WER) in vitro and in vivo and enriches at the GLABRA2 (GL2) promoter in a WER-dependent manner. Crystallographic analysis indicates that NRP1 forms a dimer via its N-terminal α-helix. Mutants of NRP1 that either disrupt the α-helix dimerization or remove the C-terminal acidic tail, impair its binding to histones and WER and concomitantly lead to failure to activate GL2 transcription and to rescue the nrp1-1 nrp2-1 mutant phenotype. Our results further demonstrate that WER-dependent enrichment of NRP1 at the GL2 promoter is involved in local histone eviction and nucleosome loss in vivo. Biochemical competition assays imply that the association between NRP1 and histones may counteract the inhibitory effect of histones on the WER-DNA interaction. Collectively, our study provides important insight into the molecular mechanisms by which histone chaperones are recruited to target chromatin via interaction with a gene-specific transcription factor to moderate chromatin structure for proper root hair development.

Figure 1.

NRPs Are Required for Full Expression of GL2. (A) Relative expression levels of root hair-related genes in nrp1-1 nrp2-1 double mutant. Roots of the wild type and nrp1-1 nrp2-1 were collected at 12 DAG for RNA isolation and quantitative RT-PCR. Values are normalized to ACTIN2. Error bars show sd from three biological replicates. Mean values of relative gene expression levels in nrp1-1 nrp2-1 compared with that in the wild type (set as 1) are shown with error bars. Asterisk indicates statistically significant difference (P < 0.01). (B) Expression patterns of GL2:GUS reporter in root tips of the wild type (WT) and nrp1-1 nrp2-1 double mutant. Bar = 50 μm. (C) Transverse sections were prepared with the root tips of the wild type and nrp1-1 nrp2-1 harboring reporter GL2:GUS. Bars = 20 μm. (D) Time course of NRP1 induction. Gene expression in wild-type and transgenic plants harboring ES:YFP-NRP1 in nrp1-1 nrp2-1 background at 12 DAG was induced by 4 μM estradiol. Roots were collected at the indicated times for RNA isolation and quantitative RT-PCR examination. Values are normalized to ACTIN2. Error bars show sd from three biological replicates. Mean values of relative NRP1 expression levels in transgenic plants compared with that in the wild type (set to 1) are shown with error bars. (E) Time course of relative changes in GL2 and TTG1 transcript levels, determined using the same RNA as that used in (D). Mean values of relative GL2 and TTG1 expression levels in induced transgenic plants compared with those in uninduced ones (set to 1) are shown with error bars.

Figure 2.

NRPs Act Upstream of Ethylene/RHD6 Pathways and in Conjunction with WER-Containing Transcription Factor Complex. The wild type (WT) and nrp1-1 nrp2-1 were vertically cultured in normal culture medium (A) and medium containing 25 μM AVG (B), respectively. Mutants of rhd6 (C) and wer-1 (D) were introgressed into nrp1-1 nrp2-1 to obtain rhd6 nrp1-1 nrp2-1 and wer-1 nrp1-1 nrp2-1 triple mutants, respectively. Plants were grown vertically in normal culture medium. Images were taken at 12 DAG. Bars = 0.5 mm.

Figure 3.

Enrichment of NRP1 at GL2 Promoter Is Dependent on WER. (A) Schematic representation of structures of GL2 promoter and first two exon/introns. Black boxes represent exons, white box represents the untranscribed region, lines represent promoter and introns. Number-labeled bars (P1–P7 and T1–T3) represent regions amplified by primer pairs corresponding to numbers on x axis of graph below. The HindIII site represents start of GL2 functional promoter and WBS indicates the location of the WER binding sites. (B) Enrichment of YFP-NRP1 at the GL2 promoter. Roots of plants vertically grown and collected at 12 DAG were fixed with formaldehyde for ChIP analysis using polyclonal anti-GFP antibody. Error bars show sd from three biological replicates. Asterisks indicate statistically significant difference of YFP-NRP1 enrichment at GL2, FLC, and ACTIN2 (ACT2) between the wild type and wer-1 (P < 0.05).

Figure 4.

NRP1 Directly Interacts with Transcription Factor WER in Vitro and in Vivo. (A) Pull-down assay. Protein extracts from transgenic plants expressing YFP-NRP1 were incubated with beads coated with GST and GST-WER. Quantity/purity of GST and GST-WER beads were analyzed by SDS-PAGE gel stained by Coomassie blue (CBB) (left panel). Two percent of the input and pull-down fractions were analyzed by immunoblot using polyclonal anti-GFP antibody (indicated by the arrowhead, right panel). (B) BiFC analysis of interaction between NRP1 and WER in tobacco leaf cells. Bars = 50 μm. (C) SEC profiles (Superdex 200 10/300 GL) of NRP1 (blue) and WER (red), as well as their mixture at a molar ratio of 1:1 (black).

Figure 5.

Crystal Structure Shows That NRP1 Forms a Dimer through Its N-Terminal α-Helices. (A) Overall structure of NRP1 (19–225 amino acids) dimer shown as ribbon diagram. Dimerization domain (19–78 amino acids) is shown in blue. Structures of α2-α4 (79–106 amino acids), β1–β4 (107–156 amino acids), and α5–α7 (157–225 amino acids) in earmuff domain are shown in green, yellow, and red, respectively. (B) Structure of dimerization domain of NRP1. Hydrophobic amino acids within two antiparallel N-terminal long α-helices are labeled in black (α1) and red (α1’). (C) Labeled structure of earmuff domain. Dashed lines show peptides (163–184 amino acids) and (192–203 amino acids) between β4 and α6.

Figure 6.

Disruption of Dimerization or Acidic C Terminus of NRP1 Diminishes Its Interaction with Histones and with WER. (A) Diagrammatic representations of NRP1-mN and NRP1-ΔC. In NRP1-mN, three hydrophobic amino acids (Ile-31, Leu-34, and Ile-37) in dimerization domain were replaced by hydrophilic Ser residue. In NRP1-ΔC, acidic C terminus (226–256 amino acids) was deleted. (B) Glutaraldehyde coupling of NRP1 and its mutants. Open arrowheads indicate monomer of each uncoupled protein. Black arrowheads indicate dimer of coupled proteins formed by intermolecular cross-linking. Black arrow indicates coupled form of NRP1-mN, likely formed by intramolecular cross-linking. Bands of cross-linked proteins broaden in the SDS-PAGE, so 3-fold glutaraldehyde-treated proteins were loaded in the gel compared with the controls to provide a clear image. (C) SEC profiles (Superdex 200 10/300 GL) of NRP1, NRP1-mN, and NRP1-ΔC (each in blue) and H2A/H2B dimer (red), as well as their mixtures (molar ratio of 1:1 in black). Buffer was 20 mM Tris-HCl, pH 8.0, and 150 mM NaCl. Note that NRP1 and H2A/H2B dimer complex eluted as a single peak at 1:1 molar ratio, indicating that a NRP1 dimer could stably bind two molecules of H2A/H2B dimer. Neither NRP1-mN nor NRP1-ΔC mixed with H2A/H2B dimer displayed a single peak in SEC profiles. (D) Pull-down assay. Protein extracts of transgenic plants overexpressing YFP-NAP1;3, YFP-NRP1, YFP-NRP1-mN, and YFP-NRP1-ΔC were incubated with beads coated with GST-WER. Input and pull-down fractions were analyzed by immunoblot using polyclonal anti-GFP antibody.

Figure 7.

NRP1-mN and NRP1-ΔC Could Not Promote Full GL2 Expression, Leading to Loss of NRP1 Function in Planta. (A) Root hair phenotypes of wild-type (WT), nrp1-1 nrp2-1, and transgenic plants harboring ES:YFP-NRP1, ES:YFP-NRP1-mN, and ES:YFP-NRP1-ΔC in nrp1-1 nrp2-1 background, respectively. Plants were grown vertically in culture medium supplemented with 4 μM estradiol. Images were taken at 12 DAG. Bar = 0.5 mm. (B) Relative transcription of GL2 in roots of plants in (A). Values are normalized to ACTIN2. Error bars show sd from three biological replicates. Mean values of the relative GL2 levels compared with that in the wild type (set to 100%) are shown with error bars. Asterisks indicate statistically significant difference of relative GL2 transcription between the wild type and nrp1-1 nrp2-1 (P < 0.01). (C) Enrichment of YFP-NRP1, YFP-NRP1-mN, or YFP-NRP1-ΔC at GL2 promoter. Roots from transgenic plants in (A) were fixed with formaldehyde for ChIP analysis using polyclonal anti-GFP antibody. Error bars show sd from three biological replicates. Asterisks indicate statistically significant difference of enrichment at GL2, FLC, and ACT2 between YFP-NRP1 and YFP-NRP1-mN/YFP-NRP1-ΔC (P < 0.01).

Figure 8.

NRPs Promote Histone Release and Decrease Nucleosome Density at GL2 Promoter. (A) Schematic representation of structures of GL2 promoter and first two exons. (B) and (C) Histone occupancy at promoter regions of GL2. Roots of vertically grown wild type and nrp1-1 nrp2-1 at 12 DAG were fixed with formaldehyde for ChIP analysis using commercial antibodies against H2B (B) and H3 (C). Error bars show sd from three biological replicates. Asterisks indicate statistically significant difference of histone enrichment at GL2, FLC, and ACT2 between the wild type and nrp1-1 nrp2-1 (P < 0.05). (D) Nucleosome density at the GL2 promoter in nrp1-1 nrp2-1. Nuclei were isolated from roots of the wild type and nrp1-1 nrp2-1 at 12 DAG and were treated with MNase. The digested DNA fragments were purified for quantitative PCR analysis, and the intact genomic DNA without digestion were used as the input. Error bars show sd from three biological replicates. Asterisks indicate statistically significant difference of DNA level between the wild type and nrp1-1 nrp2-1 (P < 0.05).

Figure 9.

WER Forms a Stable Complex with Its Target DNA When NRP1 Associates with H2A/H2B Dimer. (A) SEC profiles (Superdex 200 10/300 GL) of WER (A₂₈₀ in cyan) and its DNA substrate (A₂₅₄ in black), as well as their complex at 1:1 molar ratio (A₂₅₄ in red; A₂₈₀ in blue). (B) EMSA of NRP1, H2A/H2B dimer, WER, and its substrate DNA. Lanes 1 to 14, 2 μM DNA (black arrow, set as relative molar ratio at 1). Lanes 2 to 4, WER titrated against DNA at molar ratios of 2:1, 4:1, and 6:1. Lanes 5 to 9, H2A/H2B dimer titrated against DNA at molar ratios of 1:1, 2:1, 4:1, 5:1, and 6:1, while amount of WER was held at 6:1 molar ratio. Lanes 10 to 14, NRP1 titrated against DNA at molar ratios of 1:1, 2:1, 4:1, 5:1, and 6:1, while WER and H2A/H2B dimer were both held at molar ratio of 6:1 against DNA. Lane 15, H2A/H2B dimer and NRP1 as in lane 14 without WER and its substrate DNA. All samples were separated on 6% native PAGE gel, stained with GelRed, and visualized by UV (upper panel). The GelRed-stained gel shows bands of DNA bound to H2A/H2B (black arrowhead) and DNA bound to WER (open arrowhead). Gel was further stained by CBB to show bands of NRP1₂-(H2A/H2B)₄ (green arrowhead), NRP1₂-(H2A/H2B)₂ (red arrowhead), and WER bound to DNA (purple arrowhead).