* Your assessment is very important for improving the workof artificial intelligence, which forms the content of this project
Download SABRE is required for stabilization of root hair
Survey
Document related concepts
Transcript
Physiologia Plantarum 153: 440–453. 2015 ISSN 0031-9317 SABRE is required for stabilization of root hair patterning in Arabidopsis thaliana Stefano Pietraa,† , Patricia Langa,†,‡ and Markus Grebea,b,* a b Umeå Plant Science Centre, Department of Plant Physiology, Umeå University, Umeå SE-90187, Sweden Institute of Biochemistry and Biology, Plant Physiology, University of Potsdam, Potsdam-Golm DE-14476, Germany Correspondence *Corresponding author, e-mail: [email protected] Received 18 March 2014; revised 18 June 2014 doi:10.1111/ppl.12257 Patterned differentiation of distinct cell types is essential for the development of multicellular organisms. The root epidermis of Arabidopsis thaliana is composed of alternating files of root hair and non-hair cells and represents a model system for studying the control of cell-fate acquisition. Epidermal cell fate is regulated by a network of genes that translate positional information from the underlying cortical cell layer into a specific pattern of differentiated cells. While much is known about the genes of this network, new players continue to be discovered. Here we show that the SABRE (SAB) gene, known to mediate microtubule organization, anisotropic cell growth and planar polarity, has an effect on root epidermal hair cell patterning. Loss of SAB function results in ectopic root hair formation and destabilizes the expression of cell fate and differentiation markers in the root epidermis, including expression of the WEREWOLF (WER) and GLABRA2 (GL2) genes. Double mutant analysis reveal that wer and caprice (cpc) mutants, defective in core components of the epidermal patterning pathway, genetically interact with sab. This suggests that SAB may act on epidermal patterning upstream of WER and CPC. Hence, we provide evidence for a role of SAB in root epidermal patterning by affecting cell-fate stabilization. Our work opens the door for future studies addressing SAB-dependent functions of the cytoskeleton during root epidermal patterning. Introduction Multicellular organisms comprise a variety of cells with diverse shapes and functions. In order for coordinated development of multicellular organisms to occur, it is therefore essential to differentiate sets of distinct cell types based on spatial and developmental cues. The establishment of different fates in a seemingly homogeneous population of cells based on their relative positions is referred to as pattern formation (Jürgens et al. 1991). Patterned cell-fate specification has been extensively studied in plants by analyzing the controlled differentiation of root and leaf epidermal cells into root hairs and trichomes, respectively (Marks 1997, † These authors contributed equally to this work. Present address: Department of Molecular Biology, Max Planck Institute for Developmental Biology, Spemannstrasse 37-39, Tübingen DE-72076, Germany. ‡ Abbreviations – COBL9, COBRA-LIKE 9; Col, Columbia; CPC, CAPRICE; ERH2, ECTOPIC ROOT HAIR 2; ERH3, ECTOPIC ROOT HAIR 3; GFP, green fluorescent protein; GL2, GLABRA2; GUS, 𝛽-glucuronidase; JKD, JACKDAW; kuq, kreuz und quer; MS, Murashige & Skoog; NTF, nuclear targeting fusion; RHD6, ROOT HAIR DEFECTIVE 6; SAB, SABRE; SCM, SCRAMBLED; SCZ, SCHIZORIZA; SEM, scanning electron microscopy; WER, WEREWOLF. 440 © 2014 The Authors. Physiologia Plantarum published by John Wiley & Sons Ltd on behalf of Scandinavian Plant Physiology Society. This is an open access article under the terms of the Creative Commons Attribution-NonCommercial License, which permits use, distribution and reproduction in any medium, provided the original work is properly cited and is not used for commercial purposes. Hülskamp 2004, Dolan 2006, Schiefelbein et al. 2009). These model systems provide a clear morphological read-out to dissect the regulatory mechanisms behind their spatial distribution (Dolan et al. 1994, Larkin et al. 1996). The root epidermis of Arabidopsis thaliana is composed of parallel files of cells that multiply by transversal cell divisions (Dolan et al. 1993). Each file acquires a specific fate based on positional information from the layer of cortical cells underneath: cells overlying the cleft between two cortical cells are considered to be in hair (H) position and generally produce a root hair at maturation, thus being called trichoblasts; cells on top of one single cortical cell are referred to as non-hair (N) cells and usually do not produce any hairs, being designated as atrichoblasts (Dolan and Costa 2001). Clonal propagation of cell files throughout the root ensures that the striped pattern of alternating cell fates is maintained, although sporadic longitudinal divisions can give rise to file duplications, increasing the number of epidermal files (Berger et al. 1998). In the leaf epidermis, trichomes are distributed evenly at a specific distance from each other (Larkin et al. 1997). In this ‘de novo patterning’, regular spacing is not achieved through signaling from underlying cells, but relies on cell to cell interactions that promote trichome differentiation in one cell while simultaneously repressing it in its neighbors (Schnittger et al. 1999). Despite substantial differences, the pathways that govern patterning of root hairs and trichomes are based on similar gene networks and regulatory mechanisms, which have been extensively studied in Arabidopsis by molecular genetics and modeling (Schiefelbein et al. 2009, Balkunde et al. 2010, Benítez et al. 2011). Fate selection revolves around an activator protein complex that includes the MYB transcription factor WEREWOLF (WER) in roots and its homolog GLABRA1 in leaves (Ishida et al. 2008). The complex induces expression of the homeobox transcription factor GLABRA2 (GL2), which mediates formation of hairless cells in the root whereas it has the opposite effect of initiating trichome outgrowth in leaves (Rerie et al. 1994, Di Cristina et al. 1996, Masucci et al. 1996). Functionality of the activator complex is negatively regulated by the small MYB-like DNA-binding protein CAPRICE (CPC) and its homologues through a mechanism ensuring lateral inhibition; these inhibitors are synthesized specifically in non-hair cells and trichomes but then relocalize to neighboring cells where they determine root-hair or trichome-less fate (Wada et al. 2002, Wester et al. 2009). While in trichomes other factors like the control of endoreplication participate in the regulation of epidermal patterning (Bramsiepe et al. 2010), an important Physiol. Plant. 153, 2015 contribution to breaking cell homogeneity in the root epidermis comes from positional cues originating from the cortex. Cortical signals, whose molecular nature is still unknown, are likely perceived by the receptor-like kinase SCRAMBLED (SCM), which reduces levels of the WER transcription factor specifically in H cells (Kwak and Schiefelbein 2007). Upstream of the patterning gene network, the JACKDAW (JKD) zinc finger protein operates from the cortex and through SCM to induce asymmetric distribution of fate determinants in epidermal cells, thereby directing cell-fate specification in a non-cell-autonomous way (Hassan et al. 2010). Another factor acting non-cell-autonomously on specification of epidermal fate is the heat shock transcription factor SCHIZORIZA (SCZ), which is expressed in ground tissue stem cells, endodermis and cortex and controls cell-fate segregation (Mylona et al. 2002, ten Hove et al. 2010). In mutants defective in either SCM or JKD function the differentiation of cells in the epidermis is more irregular with respect to their position over the cortex, while roots of scz mutants develop more hairs and lack the alternation of hair and non-hair files (Mylona et al. 2002, ten Hove et al. 2010). In addition to the core genes of the epidermal patterning network, several other genes are involved in the specification and initiation of root hairs. Some of them belong to pathways of hormone biosynthesis and signal transduction (Schiefelbein 2000, Grierson et al. 2001, Guimil and Dunand 2006). Ethylene and auxin have long been known to affect the development of root hairs (Masucci and Schiefelbein 1994, Masucci and Schiefelbein 1996) and, so far, all studies have placed their action just before hair outgrowth and downstream of the patterning mechanism (Dolan 2001). Interestingly, auxin and ethylene, together with other genes required for hair initiation such as ROOT HAIR DEFECTIVE 6 (RHD6; Masucci and Schiefelbein 1994), are also main players in the establishment of planar polarity of root hairs (Masucci and Schiefelbein 1994, Fischer et al. 2006). This process controls coordinated hair emergence in a specific position within trichoblasts over the whole root epidermal layer (Fischer et al. 2006). In this study, we analyze epidermal patterning of mutants defective in SABRE (SAB), a large gene implicated in planar polarity of root hair positioning that encodes for a 290 kDa protein with conserved domains of unknown function (Pietra et al. 2013). SAB has originally been isolated and characterized as a gene required for correct root morphogenesis and cell expansion (Benfey et al. 1993, Aeschbacher et al. 1995). Recently, SAB was shown to genetically interact with the gene encoding the microtubule-associated protein cytoplasmic linker protein 170 (CLIP170)-associated protein (CLASP). The two genes interact to stabilize orientation of 441 cell division planes and mitotic microtubule structures, organization of cortical microtubules, and establishment of planar polarity on both the root and leaf surface (Pietra et al. 2013). However, a role of SAB in cell-fate specification has not been addressed. Here, we uncover and characterize a defect in epidermal patterning of sab mutants that leads to the formation of ectopic root hairs but does not alter the density of leaf trichomes. Unstable expression of cell fate and differentiation markers in sab mutants as well as genetic interaction of SAB with WER and CPC, core components of the patterning pathway, suggest that SAB function acts on the regulatory network controlling root epidermal cell fate. Materials and methods Plant material and growth conditions Arabidopsis thaliana ecotype Columbia (Col) was used as wild-type control. The following mutant and transgenic lines were obtained from the Nottingham Arabidopsis Stock Centre (NASC) or the corresponding authors of the cited publications: sab-5 and sabkuq (Pietra et al. 2013), pGL2:NTF (Deal and Henikoff 2010), p35S:mCherry-TUA5 (Gutierrez et al. 2009, Sampathkumar et al. 2011), pGL2:GUS (Szymanski et al. 1998), pWER:ER-GFP and wer-1 (Lee and Schiefelbein 1999), pCOBL9:GFP (Brady et al. 2007), and cpc-1 (Wada et al. 1997). Seeds were surface sterilized and stratified at 4∘ C for 3 days prior to plating on Murashige & Skoog (MS) medium (1× MS salts, 1% sucrose, 1 M morpholinoethanesulphonic acid, 0.8% plant agar, pH 5.7) (Fischer et al. 2006). Plates were grown vertically under long-day conditions (16 h of 150 μmol m−2 s−1 light at 22∘ C and 8 h of dark at 18∘ C) with 60% humidity, and seedlings were analyzed after 5 days. Genotyping and generation of double mutants The sab-5 and sabkuq mutants were genotyped as previously described (Pietra et al. 2013), and homozygous mutant seedlings for analyses were phenotypically identified from the progeny of heterozygous plants. F2 seedlings from the crosses between sab-5 and either cpc-1 or wer-1 were selected for displaying the cpc-1 or wer-1 phenotype. They were genotyped for sab-5 heterozygosity and then tested to confirm cpc-1 or wer-1 homozygosity using the gene-specific primers CPC_F (TCT TGT CTT GTG AAT TAA GGA GAG G) and CPC_R (CAG GTT AGA GAC TCT TT TCT CTC G) or WER_F (CTT CAT TTG CTC AAA ATT AGT TTC TC) and WER_R (TGA ACC CAA AGT GAA CTC AAG TAG). Polymerase chain reaction (PCR) products of 442 WER_F and WER_R were digested with MboI allowing genotyping with a cleaved amplified polymorphism (CAPS) marker. Progeny of the selected plants showed 166/674 (24.6%) and 161/704 (22.9%) seedlings with a phenotype clearly different from cpc-1 and wer-1, respectively, and these were considered double mutants and used in the following analyses. Light microscopy Bright field microscopy images of 5-day-old seedlings and root close-ups were acquired with a Leica MZ 9.5 stereomicroscope (Leica Microsystems, Wetzlar, Germany) equipped with a Leica DC 300 digital camera. 𝛽-Glucuronidase (GUS)-stained roots and roots for hair patterning quantification were visualized with the 2.5× or 40× objectives of an Axioplan 2 microscope (Zeiss, Göttingen, Germany) equipped with Nomarski optics. Images were acquired with an Axiocam digital camera (Zeiss) using the AXIOVS40 V 4.8.2.0 software (Zeiss), and processed with IMAGEJ (http://imagej.nih.gov/ij/) and ADOBE PHOTOSHOP CS4 (Adobe Systems, San Jose, CA). Quantitative analysis of root hair patterning Seedlings were grown for 5 days and then mounted on microscope slides in clearing solution (80 g chloral hydrate, 30 ml glycerol and 10 ml dH2 O). Analysis was performed on three independent experiments per genotype, using at least 10 seedlings per experiment. Differentiated non-consecutive epidermal cells were randomly selected over the whole root, avoiding the region of transition between root and hypocotyl. Cells were scored for whether they carried a root hair (or root-hair bulge) or not, and their position with respect to the underlying cortical cell layer was determined by focusing through the entire epidermal layer: cells on top of a single cortical cell file were considered in N position, while cells on the cleft between two cortical cell files were considered in H position (Berger et al. 1998). Significance of the differences between genotypes was then analyzed by Fisher’s exact test. Scanning electron microscopy of leaves The third true leaf of 21-day-old Col and sab-5 plants was processed for scanning electron microscopy (SEM) as previously described (Pietra et al. 2013) and examined with a Stereoscan 360iXP microscope (Cambridge Instruments, Cambridge, UK). For each sample, three micrographs were recorded over the adaxial leaf surface: one overview image for counting the total number of trichomes and measuring the leaf area, and two images Physiol. Plant. 153, 2015 at higher magnification for estimating the number of cells per leaf. The latter were close-ups of 300 × 400 μm regions located at one quarter and three quarters of the distance between the leaf tip and the petiole, halfway between central vein and leaf margin. Micrographs from at least 10 samples per genotype were analyzed; number of trichomes and cells were counted manually and leaf outlines were drawn using the polygon selection tool of IMAGEJ to calculate leaf area. Data were processed with Microsoft Excel. Analysis of pGL2:GUS activity GUS activity was histochemically detected as previously described (Men et al. 2008). In brief, 5-day-old whole seedlings were stained for 8 h at 37∘ C, fixed for 1 h in 4% paraformaldehyde, cleared with 8:3:1 chloral hydrate:glycerol:distilled water and mounted on microscope slides for observation. To determine the frequency of gain or loss of pGL2:GUS activity, roots of sab-5 mutants were analyzed from the first rapidly elongating cells to the transition zone between root and hypocotyl. In the wild-type counting was done on a 4.6 mm long section of the root starting from the meristem, a region 1.25 times the average length of a sab-5 mutant root. As epidermal cells are on average about 25% shorter in sab-5 mutants compared with wild-type (Pietra et al. 2013), this allowed analyzing an equivalent number of cells, compensating for the differences in root length and cell elongation between wild-type and sab-5. Ten seedlings per genotype for each of two experiments were analyzed. A loss or gain of pGL2:GUS activity was scored as a single event regardless of how many consecutive cells in one file it involved. Re-establishment of the original status of pGL2:GUS activity after a fate switch was not considered as an additional switching event, while independent alternations in the same file were counted individually. Confocal laser scanning microscopy Live imaging of lines expressing fluorescent fusions was performed with a Zeiss LSM780 spectral confocal laser scanning microscopy (CLSM) system mounted on a Zeiss Axio Observer Z1 inverted microscope. Five-day-old seedlings were mounted on cover slips in MS medium and observed with a water-corrected C-Apochromat 40× 1.2 numerical aperture (NA) objective (M27, Zeiss). Laser excitation lines were 488 nm for green fluorescent protein (GFP) [and nuclear targeting fusion (NTF)] and 561 nm for mCherry, emission filters were 493–556 nm for GFP (and NTF) and 586–647 nm for mCherry. To Physiol. Plant. 153, 2015 determine the frequency of fate switches in lines expressing the pGL2:NTF marker, 10 roots of each genotype were analyzed from the meristem to the transition zone between root and hypocotyl, and scoring was done following the same criteria as for pGL2:GUS. For top views and cross-sections of lines expressing fate or differentiation markers and mCherry-TUA5 (TUBULIN ALPHA-5), Z stacks of planes at 0.45 μm distance were acquired over the periclinal face of roots. Top views were maximum intensity projections of the stacks, while for cross-sections slices of 3–50 μm were cut transversally out of the stacks and rotated three-dimensionally. Images were processed and analyzed using IMAGEJ and ADOBE PHOTOSHOP CS4, and assembled in Adobe Illustrator CS3 (Adobe Systems). Statistical analysis Significance of the differences between the number of root hair- and non-hair cells found in H position and N position in different genotypes were assessed by Fisher’s exact test with populations of >150 cells for each position from at least 30 roots and three independent experiments, and significance level P < 0.05 (http:// quantitativeskills.com/sisa/statistics/fisher.htm). Analysis of the number of cell fate switches detected with the pGL2:GUS and pGL2:NTF reporters in wild-type and sab-5 roots was performed by two-tailed, unpaired t-test with equal variance and significance level P < 0.05 (computed with Microsoft Excel), with n = 3 independent experiments of 10 seedlings per genotype. Results sab mutants have defects in root hair patterning The Arabidopsis root epidermis, with its regular pattern of hair-forming trichoblast files alternating with one or two files of hairless atrichoblasts, provides an ideal model to study epidermal cell-fate determination (Dolan et al. 1994; Fig. 1A). Here, we investigated whether the SAB gene is involved in patterning of epidermal cells. Benfey et al. (1993) reported that sab mutations do not cause abnormalities in radial symmetry and number of epidermal cells in the roots. Intriguingly, observation of sab mutant roots revealed an increased amount of hairs ectopically growing in non-hair files right next to other hair-bearing cells (Fig. 1B, C, arrowheads). In order to quantify the sab epidermal patterning defects, we calculated the proportion of cells that displayed a hair in either H position (over the anticlinal walls of two cortical cells) or N position (over the outer periclinal wall of a single cortical cell) for Col wild-type and two sab mutant 443 A D Col E Col B Col F sab-5 C sab-5 G sabkuq sab-5 Fig. 1. sab mutants are defective in root hair patterning, but not in leaf trichome spacing. (A–C) Bright field microscopy images of roots of 5-day-old (A) Col, (B) sab-5 and (C) sabkuq seedlings. Regular alternation of hair- and non-hair cell files in (A) Col is compromised in (B–C) sab mutants, where root hairs can emerge from adjacent files (arrowheads). (D–G) Scanning electron microscopy images of the third true leaf of 21-day-old (D–E) Col and (F–G) sab-5 plants. (D, F) Overall views of the adaxial leaf surface and (E, G) close-ups of epidermal cells located at one quarter of the distance between the leaf tip and the petiole, halfway between the central vein and leaf margin. Note, different total leaf area and cell density between Col and sab-5 (see Table 2 for quantification). Scale bars = (A–C, E, G) 100 μm, (D, F) 1 mm. alleles, sab-5 and sabkreuz und quer (sabkuq ; Table 1). In Col roots about 73% of the cells in H position carried a hair, and this percentage was similar in mutant roots (65.2% for sab-5 and 66.5% for sabkuq ). However, sab-5 (20.5%) and sabkuq (19.2%) mutants had more than four times as many hair cells in N position than Col (4.6%), indicating that in sab-5 and sabkuq a considerable number of cells in the N position did not follow their developmental program, resulting in a higher amount of ectopic hairs than in the wild-type. This strongly suggested that sab mutants have a patterning defect in the root epidermis, and that SAB might be required for either the specification or execution of epidermal cell fate. 444 Leaf trichome density is not affected in sab-5 mutants Several determinants of root epidermal cell fate such as GLABRA 2, GLABRA 3, TRANSPARENT TESTA GLABRA 1, TRIPTYCHON and CAPRICE are also involved in controlling patterning of trichomes on the leaf epidermis (Koornneef 1981, Hülskamp et al. 1994, Wada et al. 1997). The presence of a defect in root hair patterning of sab mutants raised the question as to whether the SAB gene may as well be involved in specification and spacing of leaf trichomes. To this end, we acquired SEM images of the third true leaf of 21-day-old Col wild-type and sab-5 mutant plants (Fig. 1D, F) and calculated the average number of trichomes per leaf, revealing that sab-5 leaves had almost half as many trichomes as wild-type (Table 2). In order to compensate for the smaller size of mutant leaves (Aeschbacher et al. 1995, Pietra et al. 2013) and a potential difference in cell density between sab-5 and wild-type, we measured total leaf areas and counted the number of epidermal cells in 0.12 mm2 regions on the adaxial surface of wild-type and sab-5 leaves (Fig. 1E, G; Table 2). We confirmed that wild-type leaves are over four times larger than those of sab-5, and uncovered a much higher density of cells in the sab-5 mutant leaf epidermis. This allowed us to estimate the total number of epidermal cells per leaf in wild-type and sab-5 and determine the average number of cells per trichome in each of the two genotypes (Table 2). Strikingly, this ratio was very similar between the two, suggesting that in sab mutants the relative amount of leaf epidermal cells that develop into trichomes does not differ from the wild-type. Thus, SAB does not likely have a role in regulating the density of trichome placement in the leaf epidermis suggesting that its effect on hair patterning may be root-specific. Cell file divisions and subsequent cell-fate duplications do not account for increased ectopic hair formation in sab-5 Root epidermal cells typically divide transversely to the direction of root growth, generating clonal files of cells that all originate from the same initial. However, as already observed by Berger et al. (1998), in rare cases one longitudinal division can determine a file duplication, initiating two daughter clones of cells. As the SAB gene has been shown to have an important role in cell division plane orientation (Pietra et al. 2013), we tested if a higher number of aberrant longitudinal divisions and a preferential duplication of the H cell fate could explain the increased number of hair cells observed in sab mutants. We counted the number Physiol. Plant. 153, 2015 Table 1. sab-mutant roots display hair-patterning defects. Percentage of root hair and non-hair epidermal cells in H position (on top of the border between two cortical cell files) and N position (on top of one single cortical cell file) in 5-day-old seedlings of wild-type Col and two sab mutant alleles. Average value and standard deviation between three independent experiments are displayed. In H position: P = 0.0693 Col vs sab-5, P = 0.1025 Col vs sabkuq . In N position: P = 2.3 × 10−6 Col vs sab-5, P = 2.3 × 10−5 Col vs sabkuq by Fisher’s exact test, n indicated in parenthesis. Cells in the H position Col sab-5 sabkuq Cells in the N position Hair cells (%) Non-hair cells (%) Hair cells (%) Non-hair cells (%) 73.2 ± 1.8 (249) 65.2 ± 5.1 (118) 66.5 ± 2.6 (141) 26.8 ± 1.8 (91) 34.8 ± 5.1 (63) 33.5 ± 2.6 (71) 4.6 ± 2.8 (9) 20.5 ± 7.1 (40) 19.2 ± 8.2 (30) 95.4 ± 2.8 (185) 79.5 ± 7.1 (155) 80.8 ± 8.2 (126) Table 2. Trichome patterning is unaffected in sab-5 mutants. Quantification of the number of trichomes and epidermal cells on the adaxial face of the third true leaf of 21-day-old wild-type Col and sab-5 plants through analysis of SEM images, and subsequent calculation of the cells/trichome ratio. Number of trichomes per leaf represents average ± standard deviation from 21 and 27 leaves for Col and sab-5, respectively. Total leaf area and number of epidermal cells represent average ± standard deviation from 10 leaves. Col Trichomes per leaf Total leaf area Epidermal cells per 0.12 mm2 leaf area Estimated epidermal cells per leaf Average epidermal cells per trichome 90 ± 20.68 50.66 ± 12.72 mm2 22 ± 15.01 sab-5 49 ± 12.52 11.41 ± 1.70 mm2 53 ± 14.60 9057 5049 101 104 of file duplications in wild-type and sab-5, and used the fluorescent pGL2:NTF marker (Deal and Henikoff 2010) and the pGL2:GUS reporter construct (Szymanski et al. 1998), which are specifically expressed in atrichoblast cells (Masucci et al. 1996), to assess the identity of epidermal cells in files that had undergone duplication. In 30 wild-type seedlings that expressed pGL2:GUS we observed 17 file duplication events, while in the same number of seedlings with the sab-5 phenotype we only found 15. These results indicated that mutations in SAB do not induce a higher number of longitudinal divisions in the root meristem. Moreover, when considering the identity of files undergoing division, 16 of 17 duplications in the wild-type took place in H files and in most cases (n = 14/16) gave rise to one H and one N file (Fig. 2A, B), with the exception of two events in which both daughter files acquired H fate (Fig. 2C). On the other hand, only 9 of 15 duplications in sab-5-looking seedlings occurred in H files, and also in this case they mainly produced one H and one N file (n = 8/9; Fig. 2D, E), except for one event that resulted in two N files. In wild-type roots, we observed longitudinal divisions Physiol. Plant. 153, 2015 in N files only once (n = 1/17); in contrast, longitudinal divisions in N files represented 40% of the events in sab-5 (n = 6/15) and in all cases led to a duplication of the N fate (Fig. 2F). These findings clearly indicated that sab-5 mutants did not have a bias toward divisions that would increase the number of H cells but on the contrary showed more duplications of N fate than wild-type seedlings. Taken together, file duplications are unlikely to account for the higher occurrence of ectopic hairs in sab mutants, and the effect of SAB on epidermal pattering may not directly depend on SAB function during the orientation of cell division. sab-5 affects stable expression of fate and differentiation markers in elongated root epidermal cells Root epidermal cell files acquire their hair or non-hair fate early in the elongation zone (Dolan et al. 1994). This fate is then generally maintained in the whole clonal progeny of the early cells, and translates in the homogeneous expression of fate markers throughout an entire cell file (Masucci et al. 1996, Lee and Schiefelbein 1999). We could confirm this homogeneous expression within files when observing the N-fate markers pGL2:NTF and pGL2:GUS in the epidermis of wild-type roots: here files of cells that consistently expressed the marker alternated with files of cells that did not show any marker expression (Fig. 3A, B; Fig. S1A, Supporting information). However, in rare instances we noticed one or more cells of different fate compared with the other cells of the file they belonged to, causing two consecutive cells of the same file to display opposite marker expression (Fig. S1B). When studying expression of the markers in sab-5-looking roots we repeatedly observed similar fate alternations, leading to cells expressing N-fate markers in trichoblast files and cells not expressing them in atrichoblast files (Figs 3C–E and S1C, D). To assess whether fate alterations were more frequent in sab-5 than in wild-type, we quantified their number analyzing 10 seedlings expressing the pGL2:NTF marker 445 A B C wt D wt E wt F sab-5 sab-5 sab-5 Fig. 2. Duplication of root epidermal cell files can lead to differential cell fates in daughter files. (A, D) Live imaging and (B, C, E, F) GUS staining of (A–C) wild-type (wt) and (D, F) sab-5 seedlings expressing N-fate markers. (A, D) Confocal laser scanning microscopy images of meristematic root epidermal cells where N fate is indicated by expression of the nuclear pGL2:NTF marker gene (green) and cell outlines are labeled with the tubulin marker mCherry-TUBULIN ALPHA-5 (mCherry-TUA5; magenta). Note the duplication of H files resulting in one H file and one N file, which in D is composed of only three cells. (B, C, E, F) Representative differential interference contrast (DIC) images from a total of 30 wild-type and 30 sab-5-looking seedlings expressing the pGL2:GUS marker, where blue color indicates N fate. (B, E) H file dividing into one H file and one N file; representative images of 14 and 8 similar divisions observed in wild-type and sab-5, respectively. (C) Duplication of an H file into two H daughter files, representative of two images from wild-type seedlings. Note that the two daughter files return one file after only four cells. (F) Division of an N file resulting in the duplication of the N fate, representative of events observed once in wild-type and six times in sab-5. One additional event (not shown) involved a sab-5 H file dividing into two N files. Arrowheads indicate file divisions and return of duplicated files to one file. Scale bars = 10 μm. and 20 seedlings carrying the pGL2:GUS reporter for each genotype. Strikingly, in all experiments the number of alternations in seedlings with the sab-5 phenotype was almost double than in wild-type, adding up to significantly different totals of 124 and 73 events, 446 respectively (P = 0.00135 by t-test; n = 3 experiments of 10 seedlings each). Hence, SAB was required for stabilizing the expression of the pGL2 promoter in the root epidermis, although it was still unclear whether this role extended to epidermal fate specification in general. In order to test if the transcriptional fusion of another positive regulator of non-hair cell fate was affected in a similar way, we analyzed expression of the pWER:GFP marker (Lee and Schiefelbein 1999) in wild-type and sab-5 roots. Strikingly, also in this case the number of fate alternations was substantially higher in the sab-5 mutant compared with wild-type (Fig. 3F–J), supporting the hypothesis that overall stabilization of N fate was impaired in sab-5. Finally, to test whether cells with H fate were also influenced by the lack of SAB activity, we examined expression of the pCOBRA-LIKE 9:GFP marker, which is specifically active in hair cells [pCOBL9 (COBRA-LIKE 9):GFP; Brady et al. 2007]. Here, we similarly observed a much more irregular expression pattern in epidermal cell files of sab-5-mutant roots, when compared with the uniform and stable activity of the promoter in wild-type roots (Fig. 3K–O). Taken together, these results indicated that lack of SAB function destabilizes the file-specific expression of cell fate and differentiation markers in N and H cell files. Our findings suggest that SAB function may either contribute to determine or to reinforce cell-fate specification in the root epidermis. Genetic interactions between SAB and WER or CPC, two core regulators of root epidermal cell fate To further test SAB function in root epidermal cell-fate determination, we investigated whether sab mutants genetically interact with mutants of genes known to regulate epidermal cell patterning. The WER MYB protein is one of the core components of the complex that induces N fate in root epidermal cells, and loss-of-function wer-1 mutants display strongly increased formation of hairs in N position, resulting in characteristic ‘hairy’ seedlings (Lee and Schiefelbein 1999; Fig. 4A–B, G–H). As activity of the pWER promoter was affected by mutation of the SAB gene (Fig. 3F–J), we generated sab-5;wer-1 double mutants and analyzed their patterning phenotype to test for potential functional genetic interaction. sab-5 seedlings showed reduced growth and shorter, swollen roots compared with Col wild-type (Pietra et al. 2013; Fig. 4A, D), and the overall phenotype of sab-5;wer-1 double mutants was very similar to that of sab-5 mutants (Fig. 4D, E). However, closer observation of double-mutant roots revealed a clearly higher amount of ectopic hairs (Fig. 4J, K). Accordingly, quantitative analysis confirmed a significant difference in the number of ectopic root hairs in N position between sab-5 and Physiol. Plant. 153, 2015 pCOBL9:GFP;mCherry-TUA5 pWER:GFP;mCherry-TUA5 pGL2:NTF;mCherry-TUA5 A C D before * sab-5 E after wt B * wt sab-5 sab-5 F H I before * * sab-5 J after * wt * G wt sab-5 sab-5 K M N before * * * sab-5 O * wt * after * L wt sab-5 sab-5 Fig. 3. SAB stabilizes expression of cell fate and differentiation markers along root epidermal cell files. (A–E) Live imaging of (A, B) wild-type and (C–E) sab-5 mutant seedlings expressing the N-fate nuclear marker pGL2:NTF (green). (F–J) Visualization of the endoplasmic reticulum (ER)-localized N-fate marker pWER:GFP (green) in epidermal cells of (F, G) wild-type and (H–J) sab-5 seedlings. (K–O) Imaging of the hair-cell differentiation marker pCOBL9:GFP (green) in roots of (K, L) wild-type and (M–O) sab-5 seedlings. All seedlings also express the tubulin marker mCherry-TUA5 (magenta), indicating cell outlines. (A, C, F, H, K, M) Top views showing the stable expression of fate and differentiation markers along meristematic cell files of wild-type roots and the disturbed patterning of the markers in cell files of sab-5 mutants. (B, D–E, G, I–J, L, N–O) Cross-sections displaying regular patterning of wild-type epidermal cells with respect to the underlying cortical layer, as opposed to differences in fate between cells of the same file displayed in consecutive sections of sab-5 seedlings. Upper images (before) show marker expression before an observed difference in fate, and lower images (after) display a section of the same seedling taken closer to the root tip. Differential fate of the cells marked by an asterisk, which belong to the same cell file. (I, J) Dotted lines with arrowheads indicate the slightly twisted orientation of cells in sab-5 mutant roots. Scale bars = 20 μm. both wer-1 and sab-5;wer-1 mutants (Table 3), while no significant difference was observed between wer-1 and the double mutant (Table 3). The number of hairs in H position did not differ between sab-5 and wer-1, and was thus not indicative for evaluating interaction between the two mutants. Together, these findings suggested that the Physiol. Plant. 153, 2015 wer-1 defect in non-hair epidermal cell-fate specification is epistatic to the phenotype of sab-5 mutants. In order to also address potential genetic interaction between SAB and a gene promoting H fate of root epidermal cells, we crossed sab-5 with cpc-1, a knock-out mutant defective in the positive regulator of hair fate 447 A Col B C wer-1 cpc-1 D G J sab-5 Col sab-5 E H K sab-5; wer-1 wer-1 sab-5; wer-1 F I L sab-5; cpc-1 cpc-1 sab-5; cpc-1 Fig. 4. Double mutant analysis reveals genetic interaction of WER and CPC with SAB in root hair patterning. Seedling and root hair patterning phenotypes of sab-5;wer-1 and sab-5;cpc-1 double mutants. (A–F) Five-day-old seedlings of (A) Col wild-type, of (B) wer-1, (C) cpc-1 and (D) sab-5 single mutants, and of (E) sab-5;wer-1 and (F) sab-5;cpc-1 double mutants; small seedlings with curved cotyledons and short, thick root in sab-5;wer-1 and sab-5;cpc-1, similar to sab-5. (G–L) Bright field microscopy images of the root epidermis of 5-day-old (G) Col, (H) wer-1, (I) cpc-1, (J) sab-5, (K) sab-5;wer-1 and (L) sab-5;cpc-1 seedlings. wer-1 and sab-5 mutants display an increased number of ectopic root hairs, although to different extents, while cpc-1 mutants almost completely lack hairs throughout the whole root (see Table 3 for quantification). Note how, with respect to root hair patterning, sab-5;wer-1 and sab-5;cpc-1 double mutants closely resemble wer-1 and cpc-1 single mutants, respectively, rather than sab-5. Scale bars = (A–C) 1 mm, (D–F) 500 μm, (G–L) 100 μm. Table 3. Genetic relationship of wer-1 and cpc-1 with sab-5 in root-hair patterning. Percentage of root hair and non-hair epidermal cells in H and N positions in 5-day-old sab-5, wer-1, cpc-1 single-mutant and sab-5;wer-1, sab-5;cpc-1 double-mutant seedlings. Average value and standard deviation among three independent experiments are displayed. In H position: P = 0.0522 sab-5 vs wer-1, P = 5.9 × 10−24 sab-5 vs cpc-1, P = 0.6380 sab-5 vs sab-5;wer-1, P = 2.0 × 10−30 sab-5 vs sab-5;cpc-1, P = 0.0114 wer-1 vs sab-5;wer-1, P = 0.3734 cpc-1 vs sab-5;cpc-1. In N position: P = 4.9 × 10−10 sab-5 vs wer-1, P = 0 sab-5 vs cpc-1, P = 2.5 × 10−7 sab-5 vs sab-5;wer-1, P = 0 sab-5 vs sab-5;cpc-1, P = 0.2960 wer-1 vs sab-5;wer-1, P = 1 cpc-1 vs sab-5;cpc-1 by Fisher’s exact test, n indicated in parenthesis. Cells in the H position sab-5 wer-1 cpc-1 sab-5;wer-1 sab-5;cpc-1 Hair cells (%) Non-hair cells (%) Hair cells (%) Non-hair cells (%) 75.7 ± 3.2 (143) 84.2 ± 1.7 (155) 23.1 ± 4.1 (39) 73.2 ± 8.3 (139) 19.2 ± 2.7 (39) 24.3 ± 3.2 (46) 15.8 ± 1.7 (29) 76.9 ± 4.1 (130) 26.8 ± 8.3 (51) 80.8 ± 2.7 (164) 24.0 ± 4.7 (46) 55.6 ± 1.9 (100) 1.7 ± 1.8 (3) 49.7 ± 2.7 (93) 1.4 ± 1.0 (3) 76.0 ± 4.7 (146) 44.4 ± 1.9 (80) 98.3 ± 1.8 (175) 50.3 ± 2.7 (94) 98.6 ± 1.0 (214) CPC (Wada et al. 1997). The cpc-1 mutation causes the formation of only a few and irregularly distributed hairs (Wada et al. 1997; Fig. 4C, I), which was strikingly and significantly different compared with the sab-5 phenotype when analyzing the proportion of both H cells and N cells bearing root hairs (Table 3). Intriguingly, as in the case of sab-5;wer-1 double mutants, sab-5;cpc-1 seedlings had an overall phenotype resembling that of 448 Cells in the N position sab-5 single mutants, i.e. impaired growth and short bloated roots (Fig. 4F). However, all sab-5;cpc-1 mutants displayed extremely few root hairs (Fig. 4L), in a manner highly similar to the cpc-1 mutant. Quantitative analysis demonstrated that while the percentage of hair cells in H and N position was significantly different between sab-5;cpc-1 and sab-5, no difference could be detected between sab-5;cpc-1 and cpc-1 (Table 3). Physiol. Plant. 153, 2015 Hence, we concluded that the cpc-1 mutation is able to completely suppress the sab-5 root hair patterning phenotype, indicating an epistatic relationship similar to the one discovered for wer-1. Thus, our findings suggest that SAB genetically interacts with WER and CPC to determine root epidermal hair cell fate and its stable, ordered patterning in the root epidermis. Discussion Epidermal cell differentiation in Arabidopsis is controlled by a well-studied network of genes that directs emergence of trichomes and root hairs in a regular pattern. Some of the components that induce root hair initiation are also involved in determining planar polarity of hair positioning, however, a few of those have a role in epidermal cell-fate determination. In this study, we discover that the SAB gene, recently shown to be a player in planar polarity establishment, does also affect root hair patterning. SAB interacts genetically with core genes of the patterning pathway and lack of SAB function affects the expression of cell fate markers, suggesting that SAB function is required for stable cell-fate patterning in the root epidermis. SAB affects patterning in the root epidermis Our analyses of root hair-forming epidermal cells clearly revealed that sab mutants have an increased amount of ectopic hairs in the N position. This is similar to what has been shown for a number of mutants in genes involved in N fate specification, including GL2 and WER (Di Cristina et al. 1996, Masucci et al. 1996, Lee and Schiefelbein 1999), and indicates that SAB has an effect on patterning of root epidermal cells. The excessive number of ectopic hairs in sab mutants cannot be justified by an increased amount of H fate duplications in epidermal cell files of sab roots, as the frequency of file duplications in sab mutants was even lower than in wild-type. When considering the fate of files affected by duplications, in control seedlings 94% of the files were in H position; this is in line with the findings of Berger et al. (1998), who reported that these events are up to 20 times more frequent in H files than in N files. Surprisingly, however, in sab mutants only 60% of the duplications occurred in H files, suggesting that SAB might also function in selection of the fate of cells undergoing aberrant longitudinal divisions. A similar defect has been observed in mutants defective in the core pathway of epidermal cell-fate specification (Berger et al. 1998), and is consistent with an involvement of SAB in epidermal patterning. Physiol. Plant. 153, 2015 Intriguingly, ectopic hairs are frequently found in mutants of the ECTOPIC ROOT HAIR 3 (ERH3) gene, which encodes a p60 katanin involved in microtubule severing and re-organization (Webb et al. 2002). Strikingly, SAB also directs microtubule organization, and similarly to ERH3 it is required for proper arrangement of mitotic microtubule structures, division plane orientation and branching of leaf trichomes. In addition, katanin mutants have defects in anisotropic growth of cells in the shoot apical meristem, resulting in abnormal tissue morphogenesis (Uyttewaal et al. 2012). This suggests analogies in the way ERH3/KATANIN and SAB affect epidermal patterning, consistent with the view that SAB function in cell-fate determination may be based on its effects on microtubules. Furthermore, the impaired regulation of anisotropic cell growth is equally obvious in erh3 roots, which are radially swollen compared with wild-type (Schneider et al. 1997). This is true as well for the ectopic root hair 2/pom-pom 1 mutant, which also shows an increase of ectopic root hairs (erh2/pom1; Schneider et al. 1997), indicating a possible link between root hair initiation and radial cell expansion. This hypothesis is interesting considering that the increased radial dimension of erh2/pom1 is mainly due to abnormal expansion of the cortical cell layer (Hauser et al. 1995), which is also a morphological feature of sab mutants (Benfey et al. 1993). Therefore, SAB function during anisotropic cell expansion, particularly of cortical cells, may have an influence on epidermal patterning, although this would require future testing. Furthermore, enlargement of cortical cells is caused by iron deficiency in plants (Schmidt 1999) and Fe-deficient conditions are known to increase root hair formation in N position, an adaptation that is also typical of reduced phosphorus bioavailability (Schmidt and Schikora 2001). Interestingly, sab mutants exhibit enhanced sensitivity to P starvation (Yu et al. 2012). Thus, a higher amount of ectopic hairs could theoretically be explained by a hyperactive physiological response inducing hair outgrowth even under normal P availability. Such a scenario, however, could neither explain the unstable expression of fate markers in sab mutants nor the epistasis of core patterning pathway components over SAB, because P and other environmental factors have been found to act on epidermal cell differentiation downstream of the patterning pathway (Müller and Schmidt 2004). Similarly, it is unlikely that induction of ectopic root hairs in sab is the consequence of an effect on ethylene signaling. SAB has been proposed to act antagonistically to ethylene (Aeschbacher et al. 1995, Yu et al. 2012), and ethylene is required for promoting root hair outgrowth (Tanimoto et al. 1995), so it may be argued that loss of 449 SAB function could induce ethylene signaling and emergence of hairs in N position. However, again, ethylene affects root hair differentiation downstream of GL2 and of the patterning gene network (Masucci and Schiefelbein 1996). Therefore, the effect of sab on fate markers and its genetic interaction with wer and cpc are not consistent with such a scenario. Taken together, our data show that SAB is a novel gene that plays a role in establishment of root epidermal cell fate likely acting upstream of WER and CPC function, and future work can reveal the exact mechanisms involved. The effect of SAB on epidermal patterning is restricted to the root The gene network that determines cell fate and regular spacing of Arabidopsis leaf trichomes closely resembles the one directing patterning of epidermal non-hair fate in roots. Accordingly, many mutants that display aberrant root hair patterning display opposite defects in trichome density; mutants with fewer root hairs have an increased number of trichomes and vice versa. We therefore speculated that the newly discovered effect of SAB on root epidermal patterning could be extended to leaf trichome patterning. This hypothesis was appealing, as SAB has a role in branching of leaf trichomes and establishment of planar polarity in leaves as well as in roots (Pietra et al. 2013). However, quantitative analysis revealed a very similar number of epidermal cells per trichome in sab mutant leaves compared with the wild-type, indicating that SAB does not contribute to leaf trichome density or patterning. SAB may act upstream of WER and CPC in patterning of root epidermal cell fate Valuable information on the relationship between SAB and the core network of epidermal cell differentiation came from studying the expression of fate makers in sab background and the genetic interaction of sab with mutants in components of the patterning pathway. Expression of reporters marking either H or N cells in the root became more randomized upon loss of SAB function. The pCOBL9:GFP reporter is active in tip growth of root hairs (Brady et al. 2007), and destabilization of its expression suggested not only that SAB affects H cells as well as N cells, but also that SAB function has an effect on selection of cell fate, and not on hair initiation or outgrowth downstream of COBL9. The analysis of promoter GL2 reporter lines further supported this idea, considering that GL2 transcription is a target of the central activator complex that regulates epidermal cell fate (Schiefelbein et al. 2009); as lack of SAB activity induced 450 changes in GL2 expression, SAB is likely to act upstream of GL2. Finally, we observed that similar expression of pWER:GFP was less stable in sab mutants, which could either imply that SAB functions earlier than WER or that it feeds on WER transcription. Of the three genes analyzed, WER is the one acting earliest in fate determination, and destabilization of its expression could explain the GL2 and COBL9 marker phenotypes, further supporting an effect of SAB upstream of these latter components. However, when analyzing the effect of sab on WER expression it is hard to explain the root hair patterning defect of sab mutants; the increased number of fate alternations in sab was not consistently affecting one or the other cell fate, and thus did not account for a higher amount of ectopic hairs in N position. Hence, SAB seems to affect maintenance of the same fate within cell files rather than the specification of one or the other cell fate. Nevertheless, additional clues on the effect of SAB on cell-fate determination came from studying its genetic interaction with the WER and CPC genes. Overall seedling phenotypes of the sab;wer and sab;cpc double mutants closely resembled those of sab single mutants, indicating that with respect to its roles in seedling morphology SAB is epistatic to WER and CPC (Fig. 4). Strikingly, however, root hair patterning defects of wer and cpc mutants were epistatic to those of sab in non-hair and both hair and non-hair cells, respectively (Table 3). This suggested that SAB may function upstream of WER and CPC, the two components of the patterning gene network. Furthermore, the observation that mutants lacking both SAB and CPC activity produce very few hairs favors the possibility that the activator complex is functional even in the absence of SAB, consistent with the idea that SAB is not part of the core patterning pathway. When considering the effect of the wer and cpc mutations on sab seedlings, there are some intriguing similarities to what has been described for the receptor-like kinase SCM: scm;wer and scm;cpc double mutants have phenotypes undistinguishable from wer and cpc, respectively (Kwak and Schiefelbein 2007). Loss-of-function scm mutants display a higher number of non-hair cells in the H position, in addition to increased frequency of hair cells in the N position (Kwak et al. 2005), a feature they share with erh3 mutants and mutants defective in the zinc-finger protein gene JKD (Webb et al. 2002, Hassan et al. 2010). Moreover, expression of promoter GL2 and of a pWER reporter constructs is perturbed in scm and jkd, resembling the unstable patterns observed in sab mutants (Kwak et al. 2005, Hassan et al. 2010; Fig. 3). The SAB protein is expressed in all root cell types (Pietra et al. 2013) and could therefore cell-autonomously affect stable hair patterning in epidermal cells, although Physiol. Plant. 153, 2015 an additional non-cell-autonomous function from cortex cells cannot be excluded. For example, JKD acts non-cell-autonomously on epidermal patterning from the cortical layer (Hassan et al. 2010). However, these considerations support the idea that, similar to SCM and JKD, SAB functions in stabilizing fate acquisition operating upstream of the core patterning pathway. Taken together, SAB likely acts upstream of the core epidermal cell-fate pathway to stabilize patterning of the root epidermis. Defects of sab mutants in anisotropic cell expansion and microtubule organization suggest that SAB could influence epidermal patterning by affecting microtubule-mediated anisotropic cell growth. Roles of SAB have been described in both root epidermal and cortical cells, and although SAB might act cell-autonomously on hair and non-hair fate specification, future analyses are needed to assess whether its functions in the cortex contribute to epidermal patterning. Our findings reveal SAB as a prime example of a gene required for stabilization of epidermal cell fate and planar polarity of root hair positioning. They thus stimulate further studies on the interactions between these two processes and on the role of the microtubule cytoskeleton during epidermal patterning. Author contributions M. G. conceived the project; M. G., S. P. and P. L. designed experiments and interpreted experimental results; P. L. and S. P. prepared experimental material and performed the experiments; P. L. performed the majority of experiments under supervision of S. P. and M. G.; S. P. and M. G. wrote the manuscript; all authors commented on and edited the manuscript. Acknowledgements – We thank P. Benfey, D. W. Ehrhardt, S. Henikoff, S. Persson and J. Schiefelbein for sharing published materials, and P. Hörstedt at Umeå Core Facility Electron Microscopy for providing excellent SEM services including SEM micrographs. We acknowledge the Nottingham Arabidopsis Stock Centre for providing seed stocks. We are grateful to A. Gustavsson, C. Kiefer and M. Nakamura for critical reading of the manuscript and for useful comments. The work on SAB was supported by grants 2009-4846 and 2013–4405 from the Swedish Research Council (Vetenskapsrådet) to M. G. References Aeschbacher RA, Hauser M-T, Feldmann KA, Benfey PN (1995) The SABRE gene is required for normal cell expansion in Arabidopsis. Genes Dev 9: 330–340 Physiol. Plant. 153, 2015 Balkunde R, Pesch M, Hülskamp M (2010) Trichome patterning in Arabidopsis thaliana from genetic to molecular models. Curr Top Dev Biol 91: 299–321 Benfey PN, Linstead PJ, Roberts K, Schiefelbein JW, Hauser M-T, Aeschbacher RA (1993) Root development in Arabidopsis: four mutants with dramatically altered root morphogenesis. Development 119: 57–70 Benítez M, Monk NAM, Alvarez-Buylla ER (2011) Epidermal patterning in Arabidopsis: models make a difference. J Exp Zool 316: 241–253 Berger F, Hung C-Y, Dolan L, Schiefelbein JW (1998) Control of cell division in the root epidermis of Arabidopsis thaliana. Dev Biol 194: 235–245 Brady SM, Song S, Dhugga KS, Rafalski JA, Benfey PN (2007) Combining expression and comparative evolutionary analysis. The COBRA gene family. Plant Physiol 143: 172–187 Bramsiepe J, Wester K, Weinl C, Roodbarkelari F, Kasili R, Larkin JC, Hülskamp M, Schnittger A (2010) Endoreplication controls cell fate maintenance. PLoS Genet 6: e1000996 Deal RB, Henikoff S (2010) A simple method for gene expression and chromatin profiling of individual cell types within a tissue. Dev Cell 18: 1030–1040 Di Cristina M, Sessa G, Dolan L, Linstead PJ, Baima S, Ruberti I, Morelli G (1996) The Arabidopsis Athb-10 (GLABRA2) is an HD-Zip protein required for regulation of root hair development. Plant J 10: 393–402 Dolan L (2001) The role of ethylene in root hair growth in Arabidopsis. J Plant Nutr Soil Sci 164: 141–145 Dolan L (2006) Positional information and mobile transcriptional regulators determine cell pattern in the Arabidopsis root epidermis. J Exp Bot 57: 51–54 Dolan L, Costa S (2001) Evolution and genetics of root hair stripes in the root epidermis. J Exp Bot 52: 413–417 Dolan L, Janmaat K, Willemsen V, Linstead PJ, Poethig S, Roberts K, Scheres B (1993) Cellular organisation of the Arabidopsis thaliana root. Development 119: 71–84 Dolan L, Duckett CM, Grierson CS, Linstead PJ, Schneider K, Lawson E, Dean C, Poethig S, Roberts K (1994) Clonal relationships and cell patterning in the root epidermis of Arabidopsis. Development 120: 2465–2474 Fischer U, Ikeda Y, Ljung K, Serralbo O, Singh M, Heidstra R, Palme K, Scheres B, Grebe M (2006) Vectorial information for Arabidopsis planar polarity is mediated by combined AUX1, EIN2, and GNOM activity. Curr Biol 16: 2143–2149 Grierson CS, Parker JS, Kemp AC (2001) Arabidopsis genes with roles in root hair development. J Plant Nutr Soil Sci 164: 131–140 Guimil S, Dunand C (2006) Patterning of Arabidopsis epidermal cells: epigenetic factors regulate the complex epidermal cell fate pathway. Trends Plant Sci 11: 601–609 451 Gutierrez R, Lindeboom JJ, Paredez AR, Emons AMC, Ehrhardt DW (2009) Arabidopsis cortical microtubules position cellulose synthase delivery to the plasma membrane and interact with cellulose synthase trafficking compartments. Nat Cell Biol 11: 797–806 Hassan H, Scheres B, Blilou I (2010) JACKDAW controls epidermal patterning in the Arabidopsis root meristem through a non-cell-autonomous mechanism. Development 137: 1523–1529 Hauser M-T, Morikami A, Benfey PN (1995) Conditional root expansion mutants of Arabidopsis. Development 121: 1237–1252 ten Hove CA, Willemsen V, de Vries WJ, van Dijken A, Scheres B, Heidstra R (2010) SCHIZORIZA encodes a nuclear factor regulating asymmetry of stem cell divisions in the Arabidopsis root. Curr Biol 20: 452–457 Hülskamp M (2004) Plant trichomes: a model for cell differentiation. Nat Rev Mol Cell Biol 5: 471–480 Hülskamp M, Miséra S, Jürgens G (1994) Genetic dissection of trichome cell development in Arabidopsis. Cell 76: 555–566 Ishida T, Kurata T, Okada K, Wada T (2008) A genetic regulatory network in the development of trichomes and root hairs. Annu Rev Plant Biol 59: 365–386 Jürgens G, Mayer U, Torres-Ruiz RA, Berleth T, Miséra S (1991) Genetic analysis of pattern formation in the Arabidopsis embryo. Development 113(suppl 1): 27–38 Koornneef M (1981) The complex syndrome of ttg mutants. Arab Inf Serv 18: 45–51 Kwak S-H, Schiefelbein JW (2007) The role of the SCRAMBLED receptor-like kinase in patterning the Arabidopsis root epidermis. Dev Biol 302: 118–131 Kwak S-H, Shen R, Schiefelbein JW (2005) Positional signaling mediated by a receptor-like kinase in Arabidopsis. Science 307: 1111–1113 Larkin JC, Young N, Prigge M, Marks MD (1996) The control of trichome spacing and number in Arabidopsis. Development 122: 997–1005 Larkin JC, Marks MD, Nadeau J, Sack F (1997) Epidermal cell fate and patterning in leaves. Plant Cell 9: 1109–1120 Lee MM, Schiefelbein JW (1999) WEREWOLF, a MYB-related protein in Arabidopsis, is a position-dependent regulator of epidermal cell patterning. Cell 99: 473–483 Marks MD (1997) Molecular genetic analysis of trichome development in Arabidopsis. Annu Rev Plant Physiol Plant Mol Biol 48: 137–163 Masucci JD, Schiefelbein JW (1994) The rhd6 mutation of Arabidopsis thaliana alters root-hair initiation through an auxin- and ethylene-associated process. Plant Physiol 106: 1335–1346 Masucci JD, Schiefelbein JW (1996) Hormones act downstream of TTG and GL2 to promote root hair 452 outgrowth during epidermis development in the Arabidopsis root. Plant Cell 8: 1505–1517 Masucci JD, Rerie WG, Foreman DR, Zhang M, Galway ME, Marks MD, Schiefelbein JW (1996) The homeobox gene GLABRA 2 is required for position-dependent cell differentiation in the root epidermis of Arabidopsis thaliana. Development 122: 1253–1260 Men S, Boutté Y, Ikeda Y, Li X, Palme K, Stierhof Y-D, Hartmann M-A, Moritz T, Grebe M (2008) Sterol-dependent endocytosis mediates post-cytokinetic acquisition of PIN2 auxin efflux carrier polarity. Nat Cell Biol 10: 237–244 Müller M, Schmidt W (2004) Environmentally induced plasticity of root hair development in Arabidopsis. Plant Physiol 134: 409–419 Mylona P, Linstead P, Martienssen R, Dolan L (2002) SCHIZORIZA controls an asymmetric cell division and restricts epidermal identity in the Arabidopsis root. Development 129: 4327–4334 Pietra S, Gustavsson A, Kiefer C, Kalmbach L, Hörstedt P, Ikeda Y, Stepanova AN, Alonso JM, Grebe M (2013) Arabidopsis SABRE and CLASP interact to stabilize cell division plane orientation and planar polarity. Nat Commun 4: 2779 Rerie WG, Feldmann KA, Marks MD (1994) The GLABRA2 gene encodes a homeo domain protein required for normal trichome development in Arabidopsis. Genes Dev 8: 1388–1399 Sampathkumar A, Lindeboom JJ, Debolt S, Gutierrez R, Ehrhardt DW, Ketelaar T, Persson S (2011) Live cell imaging reveals structural associations between the actin and microtubule cytoskeleton in Arabidopsis. Plant Cell 23: 2302–2313 Schiefelbein JW (2000) Constructing a plant cell. The genetic control of root hair development. Plant Physiol 124: 1525–1531 Schiefelbein JW, Kwak S-H, Wieckowski Y, Barron C, Bruex A (2009) The gene regulatory network for root epidermal cell-type pattern formation in Arabidopsis. J Exp Bot 60: 1515–1521 Schmidt W (1999) Mechanisms and regulation of reduction-based iron uptake in plants. New Phytol 141: 1–26 Schmidt W, Schikora A (2001) Different pathways are involved in phosphate and iron stress-induced alterations of root epidermal cell development. Plant Physiol 125: 2078–2084 Schneider K, Wells B, Dolan L, Roberts K (1997) Structural and genetic analysis of epidermal cell differentiation in Arabidopsis primary roots. Development 124: 1789–1798 Schnittger A, Folkers U, Schwab B, Jürgens G, Hülskamp M (1999) Generation of a spacing pattern: the role of TRIPTYCHON in trichome patterning in Arabidopsis. Plant Cell 11: 1105–1116 Physiol. Plant. 153, 2015 Szymanski DB, Jilk RA, Pollock SM, Marks MD (1998) Control of GL2 expression in Arabidopsis leaves and trichomes. Development 125: 1161–1171 Tanimoto M, Roberts K, Dolan L (1995) Ethylene is a positive regulator of root hair development in Arabidopsis thaliana. Plant J 8: 943–948 Uyttewaal M, Burian A, Alim K, Landrein B, Borowska-Wykrȩt D, Dedieu A, Peaucelle A, Ludynia M, Traas J, Boudaoud A, Kwiatkowska D, Hamant O (2012) Mechanical stress acts via katanin to amplify differences in growth rate between adjacent cells in Arabidopsis. Cell 149: 439–451 Wada T, Tachibana T, Shimura Y, Okada K (1997) Epidermal cell differentiation in Arabidopsis determined by a Myb homolog, CPC. Science 277: 1113–1116 Wada T, Kurata T, Tominaga R, Koshino-Kimura Y, Tachibana T, Goto K, Marks MD, Shimura Y, Okada K (2002) Role of a positive regulator of root hair development, CAPRICE, in Arabidopsis root epidermal cell differentiation. Development 129: 5409–5419 Webb M, Jouannic S, Foreman J, Linstead PJ, Dolan L (2002) Cell specification in the Arabidopsis root epidermis requires the activity of ECTOPIC ROOT HAIR 3 – a katanin-p60 protein. Development 129: 123–131 Wester K, Digiuni S, Geier F, Timmer J, Fleck C, Hülskamp M (2009) Functional diversity of R3 single-repeat genes in trichome development. Development 136: 1487–1496 Yu H, Luo N, Sun L, Liu D (2012) HPS4/SABRE regulates plant responses to phosphate starvation through antagonistic interaction with ethylene signalling. J Exp Bot 63: 4527–4538 Supporting Information Additional Supporting Information may be found in the online version of this article: Fig S1. SAB function contributes to stable expression of the N-fate marker pGL2:GUS in the root epidermis. Edited by Y. Helariutta Physiol. Plant. 153, 2015 453