(A) Confocal projection of the differentiation zone of an ADF8p:NTF/ACT2p:BirA transgenic root showing expression of the NTF in hair cells. GFP signal is shown in green and propidium iodide staining of cell walls is shown in red. (B) Confocal projection of the differentiation zone of an GL2p:NTF/ACT2p:BirA transgenic root showing expression of the NTF in non-hair cells. (C) Confocal section of the post-meristematic region of a GL2p:NTF/ACT2p:BirA transgenic root. (D) Fluorescence micrograph of nuclei (one is shown in inset) isolated from ADF8p:NTF/ACT2p:BirA transgenic roots and incubated with streptavidin Dynabeads. GFP and beads are shown in green and DAPI staining of DNA is shown in blue. (E) Streptavidin western blot of whole cell extracts (input) and anti-GFP immunoprecipitates (IP) from roots of ACT2p:BirA, ADF8p:NTF/ACT2p:BirA, and GL2p:NTF/ACT2p:BirA plants. Top and bottom bands in each lane are endogenous biotinylated proteins and the middle band is the 42 kD NTF. (F) Diagram of the apparatus used for purification of biotinylated nuclei. See also .

(A) RT-PCR analysis of selected INTACT hair (H) cell-enriched genes in wild type and gl2-8 roots, where all epidermal cells are H cells. Expression of true H cell-specific genes is expected to be higher in gl2-8 given that the H cell type has a higher relative abundance in this genotype. Data represent the average of two biological replicates +/- SD. Asterisks indicate p values < 0.05. P values higher than 0.05 are indicated on the graph. See also and . (B) Observed versus expected percentage of genes in each Gene Ontology (GO) annotation category for hair cell-enriched genes. Chi squared tests were performed for each comparison. P values are indicated as *** < 0.001, ** < 0.01, and * < 0.03. (C) Same as in (B) but for non-hair cell-enriched genes. See also .

(A) Representative euchromatic chromatin landscapes of H3K4me3 and H3K27me3 in hair (H) and non-hair (NH) cells. Chromosome 1 genes are shown above the chromatin tracks. Genes on the top strand are shown above the black line and those below the line are on the bottom strand. Asterisks indicate genes where H3K4me3 and H3K27me3 overlap in both cell types. Each chromatin landscape is the average of two biological replicates, displayed on the same log-ratio scale. (B) Heat map of the H3K4me3 chromatin landscape relative to gene ends in H cells (-1 kb to +1 kb relative to transcription start and end sites). Genes are ranked from highest to lowest expression level in H cells. Yellow indicates positive, black indicates zero, and blue represents negative log₂ ratios. (C) Same as in (B) but for the H3K27me3 chromatin landscape. Heatmaps made the same way for each modification in non-hair cells showed a nearly identical pattern to those in hair cells, as shown in . See also .

(A) Heat map showing H3K4me3 and H3K27me3 differences between hair (H) and non-hair (NH) cell types (H cell profile minus NH cell profile) around the 5′ end of genes (-1 kb to +1 kb from transcription start site). The 946 H cell-enriched genes are ranked from highest to lowest fold difference in expression level between H and NH cells. Yellow represents higher modification levels in H cells while blue indicates lower levels in H cells, black represents no difference, and gray indicates no data where analysis was stopped when another genomic feature was encountered. (B) Same as in (A) except that the 118 NH cell-enriched genes were used, and these were ranked from highest to lowest fold difference between H and NH cells. Since the subtraction of profiles was done in the same direction as in (A), blue represents higher modification levels in NH cells while yellow indicates lower levels in NH cells. (C) Same heat map rows in (A) clustered into 3 groups (k_means = 3) over -1 kb to +1 kb. White bars delineate the three clusters. (D) Same heat map rows as in (B) clustered into 3 groups. All heat maps are shown at the same contrast level. (E and F) Comparison of H3K4me3 and H3K27me3 profiles in H and NH cells on two H cell-enriched genes, At5g70450 and At3g49960, respectively. Genes above the x-axis are on the top strand while those below the line are on the bottom strand. Dotted boxes surround the gene of interest in each case. (G and H) Comparison of H3K4me3 and H3K27me3 profiles in H and NH cells on two NH cell-enriched genes, At1g66800 and At5g42591, respectively. See also .