Phytosulfokine (PSK) is a secreted disulfated pentapeptide that controls root and shoot growth. The ubiquitous expression of PSK precursor and of the LRR receptor kinase genes in Arabidopsis raised the question of whether PSK acts as an autocrine growth factor in planta. Expression of PSKR1 under the control of tissue- and cell type-specific promoters in a receptor null background strongly suggests that PSK is a non-cell autonomous signal that controls growth through localized activity in the epidermis. pskr1–3 pskr2–1 seedlings had shorter roots and hypocotyls than the wild type, whereas 35S: PSKR1 or 35S: PSKR2 seedlings were larger, indicating that receptor abundance limits growth in planta. The preferential expression of PSKR1 in the epidermis of CER6: PSKR1 pskr1–3 pskr2–1 seedlings was sufficient to promote wild-type growth. Moreover, in GL2:PSKR1 pskr1–3 pskr2–1 seedlings that express PSKR1 in atrichoblasts of the root epidermis, root growth was restored to wild-type levels. In pskr1–3 pskr2–1 seedlings, trichoblasts and atrichoblasts were shorter than in the wild type. Trichoblasts of GL2:PSKR1 pskr1–3 pskr2–1 seedlings, which are unable to sense PSK, nonetheless had acquired wild-type length, suggesting that PSK acts as a non-cell autonomous signal. Inhibition of brassinosteroid (BR) biosynthesis with brassinazole (BZ) caused a loss of responsiveness to PSK in wild-type, tpst–1 (tyrosylprotein sulfotransferase–1), PSKR1ox12 and CER6:PSKR1–3–1 seedlings, as did the genetic knock-out of BR synthesis in det2–1 and of BR perception in bri1–9, suggesting that BR mediates PSK-dependent growth. Quantitative PCR analysis of BR-related genes in wild-type, pskr1–3 pskr2–1, PSKR1ox and tpst–1 seedlings showed largely unchanged transcript levels of BR biosynthesis genes.
Multicellular plants grow and develop as an integrated unit with individual plant cells being glued together through their walls, rendering them immotile. Growth and development are therefore dependent on the coordinated proliferation and expansion of individual cells, and rely on cell–cell communication mediated by both chemical and mechanical signals. The autocrine growth factor phytosulfokine (PSK) is a 5–amino acid disulfated peptide of the sequence Tyr(SO3H)-Ile-Tyr(SO3H)-Thr-Gln. PSK is produced from preproproteins encoded by five genes in Arabidopsis thaliana that are expressed throughout the plant life cycle, suggesting that PSK plays a ubiquitous role in plant growth and development (Matsubayashi et al., 1997; Yang et al., 2000; Yang et al., 2001; Lorbiecke and Sauter, 2002; Stührwohldt et al., 2011). Post-translational sulfation of the PSK precursor proteins is required for peptide activity, and is catalyzed by the enzyme tyrosylprotein sulfotransferase (TPST) in the cis-Golgi (Komori et al., 2009). Post-translational proteolytic processing of the PSK4 proprotein from Arabidopsis was shown to be catalyzed by a subtilisin serine protease in the apoplast, but other proteases are predicted to exist (Srivastava et al., 2008). The active PSK peptide binds to plasma membrane-localized leucine-rich repeat (LRR) receptor kinases that are conserved in Daucus carota ssp. sativus (carrot; Matsubayashi et al., 2002) and Arabidopsis (Matsubayashi et al., 2006). Two PSK receptor genes, PSKR1 and PSKR2, were identified in Arabidopsis.
PSK was first identified as a signal that promotes the proliferation of cells kept at a low density in in vitro cell cultures of both dicot and monocot plants (Matsubayashi et al., 1997; Matsubayashi et al., 1997; Hanai et al., 2000; Yang et al., 2000). Carrot cells overexpressing the receptor kinase showed enhanced proliferation after application of PSK (Matsubayashi et al., 2002). In planta, PSK signaling participates in the control of root and shoot growth. Root growth and hypocotyl elongation are promoted by PSK, mainly through the signaling of cell elongation rather than cell division (Matsubayashi et al., 2006; Kutschmar et al., 2009; Stührwohldt et al., 2011). Seedlings of the T–DNA insertion knock-out lines pskr1–2 and pskr1–3 had shorter roots and shorter hypocotyls, whereas pskr2–1 seedlings had slightly shorter roots but wild-type (wt) hypocotyl lengths, indicating that root elongation was predominantly, and hypocotyl elongation was solely, controlled through PSKR1, with PSKR2 contributing to seedling growth control only to a minor degree. PSKR1 was shown to be expressed in roots, hypocotyl, leaves, stem and flowers, and expression of PSKR2 in hypocotyl was also reported (Kutschmar et al., 2009; Stührwohldt et al., 2011). PSKR1:GUS analysis indicated that the PSKR1 promoter was active in all cell layers in the growth region of the root. Expression of PSKR1 in the hypocotyl was very weak, as indicated by PSKR1:GUS analysis, and was not specific to defined cell layers. Hence expression analysis did not clarify whether PSK is active in defined tissues.
In higher plants, the shoot meristem gives rise to all above-ground organs and has an organized pattern: the shoot epidermis is derived from the L1 layer, the photosynthesizing cells of the sub-epidermis are produced from the L2 layer and cells comprising the ground tissues descend from L3 (Stewart and Burk, 1970; Szymkowiak and Sussex, 1996; Marcotrigiano, 2001). As the cells of the different layers are immotile because of the connecting cell walls, they are forced to grow together in a coordinated manner. When the inner layers of a growing stem are separated from the outer epidermis they rapidly extend, whereas the epidermis contracts because the different layers are under differential tension (Peters and Tomos, 1996). It was hypothesized that the dividing, expanding or turgid thin-walled cells of the inner layers are the driving force for stem and hypocotyl growth, whereas the expansion-limiting thick-walled epidermal cell layer determines the rate of growth (Kutschera, 1992, 2008; Niklas and Paolillo, 1997; Kutschera and Niklas, 2007; Savaldi-Goldstein et al., 2007). Furthermore, it was shown that the epidermis can compensate for defects in cell division through excessive epidermal cell expansion (Serralbo et al., 2006; Bemis and Torii, 2007). Cell division in the inner layers was not controlled by the epidermis (Bemis and Torii, 2007), but suppressed epidermal cell division through L1-specific expression of the CDK inhibitor ICK1 was partially rescued by the inner layers, presumably via intercellular movement of cell cycle regulators (Serralbo et al., 2006). Nonetheless, reduced epidermal cell division put mechanical constraints on the cell expansion of the inner layers, which was compensated for in the L2 by generating smaller cells. Taken together, current findings support the idea that the epidermis is the growth-limiting cell layer, and that both biochemical signals as well as mechanical constraints contribute to interlayer communication to ensure coordinated growth.
The structure of a root is generated from a single layer of stem cells that surround the quiescent center, which produce the central cylinder including the pericycle, a surrounding endodermis, cortex and an epidermis that separates the root from its environment (Stahl and Simon, 2010). Likewise, root elongation depends upon the coordinated growth of cells in all layers, and hence on interlayer communication. Cell–cell communication can take place via the symplastic movement of signaling molecules or via extracellular diffusible signals. PSK is a soluble extracellular peptide proposed to participate in cell–cell communication. The expression patterns of PSK and PSKR genes suggest that receptor and ligand are ubiquitously present without specifying a specific function. In this study we asked to what extent PSK signaling controlled elongation growth by comparing signaling knock-out and overexpression mutants. We further investigated if PSK signaling of cell expansion was required in all tissues or if signaling in the epidermal cell layer, or even in defined epidermal cells, was sufficient to promote elongation growth. Our data support the view that root and shoot growth is dependent on PSK receptor abundance in the epidermis, and that PSK perception in a subset of epidermal cells is sufficient to drive PSK-dependent elongation growth.
The abundances of PSKR1 and PSKR2 limit seedling growth
To analyze if seedling growth in Arabidopsis was controlled at the level of PSK perception, we generated plants that overexpressed either PSKR1 or PSKR2. To elevate the abundance of PSK receptors, the coding regions of the receptor genes PSKR1 and PSKR2 were put under the control of the constitutively overexpressing 35S cauliflower mosaic virus promoter. RT-PCR analysis of PSKR1 and PSKR2 transcripts in two independent transgenic lines each revealed enhanced gene expression (Figures 1a and S1a). In order to study the effect of PSKR1 overexpression on growth, wt seedlings and seedlings of PSKR1ox12 and PSKR1ox2 lines were grown on plates with half-strength MS medium in the presence or absence of 1 μm PSK (Figures 1 and 2) for comparison, wt seedlings and pskr1–3 seedlings with a T–DNA insertion in the PSKR1 gene that eliminates expression were raised on the same plates (Figure 1b). Hypocotyls of etiolated PSKR1ox12 and PSKR1ox2 seedlings were significantly longer than wt hypocotyls, indicating that the abundance of PSKR1 was limiting hypocotyl elongation (Figure 1b,c). This conclusion was supported by the observation that pskr1–3 hypocotyls were on average 29% shorter than wt (Stührwohldt et al., 2011; Figure 1b).
The root elongation of light-grown seedling roots was also significantly enhanced by 35.9% in PSKR1ox12 (Figure 2a,c) and by 23.3% in PSKR1ox2 (Figure 2d), as compared with wt, and resulted in nearly twofold longer roots in PSKR1ox12 as compared with pskr1–3 seedlings (Figure 2a). The addition of 1 μm PSK resulted in even longer roots in wt and PSKR1ox12 (Figure 2c), indicating that PSK-dependent root growth was limited both by the concentration of the ligand and at the level of receptor abundance. Growth promotion at the post-seedling stage by PSKR1 overexpression was observed in four-week-old PSKR1ox12 plants, which had larger rosettes than the wt (Figure 2b).
PSKR2 was shown previously to have only a minor impact on root growth. pskr2–1 seedling roots were slightly shorter than wt roots, compared with pskr1–3 roots (Kutschmar et al., 2009; Figure S1b). So far no functional differences were reported for the two receptor proteins, and differences in knock-out phenotypes may be a result of the differential abundance of the two receptors. In order to analyze whether PSKR2 could promote root growth similarly to PSKR1 when expressed at high levels, two independent 35S:PSKR2 lines were generated and analyzed (Figure S1a–f). Roots of light-grown PSKR2ox1 and PSKR2ox6 seedlings were significantly longer by 25 and 32.8%, respectively, as compared with the wt (Figure S1b,c), indicating that PSKR2 can enhance root growth to a similar degree as PSKR1 when expressed at high levels. Likewise, overexpression of PSKR2 promoted hypocotyl elongation, as analyzed in etiolated seedlings (Figure S1e,f). The addition of PSK did not further promote root or hypocotyl growth in PSKR2-overexpressing lines, indicating that PSK was present at saturating levels, as described previously (Figure S1f; Stührwohldt et al., 2011). PSKR2-overexpressing plants also had larger rosettes, indicating that PSK receptors control seedling and vegetative shoot growth (Figure S1d).
Hypocotyl and root growth are controlled by PSK signaling in the epidermis
The epidermis is considered as the growth-limiting tissue layer, and was shown previously to control brassinosteroid-dependent elongation growth in Arabidopsis seedlings (Savaldi-Goldstein et al., 2007; Hacham et al., 2010). We therefore asked whether PSK signaling of elongation growth in the epidermis was sufficient to promote growth. To answer this question, we expressed PSKR1 under the control of the CER6 promoter in the pskr1–3 pskr2–1, double-receptor knock-out mutant background. Based on expression data obtained from the Arabidopsis eFP browser database, CER6 is active in the stem and leaf epidermis, but not in leaf mesophyll cells (http://bar.utoronto.ca/efp/cgi-bin/efpWeb.cgi; Winter et al., 2007). The pskr1–3 pskr2–1 mutant was generated by crossing pskr1–3 and pskr2–1 plants, and the homozygous double knock-out mutant was previously shown to be devoid of PSKR1 and PSKR2 transcripts (Kutschmar et al., 2009). A 1.2–kb promoter fragment upstream of the ATG start codon of the epidermis-specific CER6 gene was cloned in front of the PSKR1 coding region, and the resulting expression vector was transformed into pskr1–3 pskr2–1 plants. Three independent CER6:PSKR1 pskr1–3 pskr2–1 transformants were analyzed and shown to express PSKR1 to levels comparable with or slightly higher than those found in the wt (Figure 3a). Analysis of CER6:GUS plants showed GUS activity in cotyledons, in the hypocotyl and in the primary root tip of etiolated (Figure 3b,c) and de-etiolated seedlings (Figure 4a–e). Cross-sections confirmed that GUS activity was localized preferentially in the epidermal cell layer of the hypocotyl, as reported previously (Millar et al., 1999; Hooker et al., 2002; Figures 3c and 4c).
Hypocotyl lengths were determined in 5–day-old etiolated seedlings of three independent CER6:PSKR1 pskr1–3 pskr2–1 lines termed 4–2, 3–1 and 14, and compared with hypocotyls of wt, pskr1–3, pskr2–1 and pskr1–3 pskr2–1 seedlings (Figure 3d,e). Hypocotyls of pskr1–3 and pskr1–3 pskr2–1 seedlings were shorter than wt hypocotyls. Hypocotyls of the three independent CER6:PSKR1 pskr1–3 pskr2–1 lines 4–2, 3–1 and 14 displayed partially or fully restored hypocotyl elongation (Figure 3e). Hypocotyls reached wt length in lines 3–1 and 14, indicating that PSK perception in the epidermis was sufficient to mediate PSK-dependent hypocotyl growth. Analysis of 4–week-old, soil-grown plants showed that the rosette size of CER6:PSKR1 plants was similar to wt, whereas shoots of pskr1–3 and pskr1–3 pskr2–1 knock-out mutants were smaller than wt or CER6:PSKR1 pskr1–3 pskr2–1 plants (Figure 3f). Hence, shoot growth retardation in the PSK receptor knock-out background was restored by epidermal expression of PSKR1. Furthermore, epidermal perception of PSK restored wt growth not only in seedlings but also during plant maturation (Figure 3f).
When analyzing de-etiolated CER6:GUS seedlings we observed GUS activity in cotyledons, hypocotyl (Figure 4a–c) and in the growth region of the root tip (Figure 4d,e). As in the shoot, GUS expression in the root was detected in the epidermal cell layer (Figure 4e). In order to see whether epidermal PSKR1 expression in the root was sufficient to restore the short-root phenotype of PSK receptor knock-out seedlings, we compared root lengths of pskr1–3 and pskr2–1 single knock-out and pskr1–3 pskr2–1 double knock-out seedlings with CER6:PSKR1 pskr1–3 pskr2–1 lines 4–2, 3–1 and 14 (Figure 4f,g). The wt root length was fully restored in lines 4–2 and 3–1, and was partially restored in line 14. These results support the view that perception of PSK in the epidermis is sufficient to promote hypocotyl and root growth. It is also possible that residual CER6 promoter activity in tissues other than the epidermis contributed to growth promotion by PSKR1.
PSK promotes cell elongation in a non-cell-autonomous manner
The perception of PSK in the epidermis of either shoot or root appeared to be sufficient to drive elongation growth. In order to verify that PSKR1 activity in the epidermis was responsible for growth promotion, we next expressed PSKR1 under the control of the GLABRA2 (GL2) promoter in the pskr1–3 pskr2–1 receptor knock-out background (Figure 5). The GL2 promoter drives gene expression in atrichoblasts of the root epidermis and in cells of the lateral cap, but not in trichoblast cells or in other root cell types of the root, as shown previously by in situ RNA hybridization and a sensitive GFP reporter (Hung et al., 1998; Lin and Schiefelbein, 2001; Hacham et al., 2010). This highly preferential expression is visualized in GL2:GUS seedlings, which show GUS staining in defined cell files of the growth region at the root tip (Figure 5a,b). RT-PCR analysis of wt, pskr1–3 pskr2–1 and GL2:PSKR1 pskr1–3 pskr2–1 lines confirmed PSKR1 expression in wt and GL2:PSKR1 pskr3–1 pskr2–1 lines (Figure 5c). Root growth was subsequently analyzed in 5–day-old seedlings (Figure 5d,e). GL2:PSKR1–1 roots reached wt lengths and seedling roots of GL2:PSKR1–2 were significantly longer than wt roots, indicative of a PSKR overexpressor phenotype. The short-root phenotype of pskr1–3 pskr2–1 was hence fully restored when PSK perception occurred in single cell files of the root epidermis. Atrichoblast-specific expression of PSKR1 also partially restored wt growth in 4–week-old GL2:PSKR1–1 plants (Figure 5f).
In order to analyze whether PSK signaling promoted elongation in the trichoblasts, which do not express a PSK receptor, we determined cell lengths in the growth region of wt, pskr1–3 pskr2–1, GL2:PSKR1–1 pskr1–3 pskr2–1 and CER6:PSKR1–3–1 pskr1–3 pskr2–1 roots (Figure 6). On average, mature atrichoblasts were longer than trichoblasts in all genotypes (Figure 6a). Both mature atrichoblast and trichoblast cells were significantly shorter in the PSK receptor double knock-out line pskr1–3 pskr2–1 than in wt, and were comparable in length in wt, GL2:PSKR1–1 pskr1–3 pskr2–1 and CER6:PSKR1–3–1 pskr1–3 pskr2–1 roots, indicating that expression of PSKR1 in atrichoblasts was sufficient to drive cell elongation of trichoblast cells. A more detailed analysis revealed shorter trichoblast and atrichoblast cells in pskr1–3 pskr2–1 roots not only at maturity but also during the elongation phase, and comparable cell length profiles in wt, GL2:PSKR1–1 pskr1–3 pskr2–1 and CER6:PSKR1–3–1 pskr1–3 pskr2–1 (Figure 6b). The cell length analyses thus support the view that a signal is transmitted in response to PSK from PSK perceiving atrichoblasts to promote elongation growth in PSK receptor-less trichoblasts.
PSK signaling of root growth requires brassinosteroid synthesis
Based on previous observations that brassinosteroid-dependent growth was driven by brassinosteroid (BR) perception or synthesis in the epidermis (Savaldi-Goldstein et al., 2007), we hypothesized that PSK signaling of cell growth in the epidermis might be mediated by BR. To test this hypothesis, wt and tpst–1 (tyrosylprotein sulfotransferase–1) seedlings were treated with 1 μm brassinazole (BZ), an inhibitor of BR biosynthesis (Asami et al., 2001), and/or with 1 nm 24–epibrassinolide (BL) and root growth in response to PSK was analyzed (Figure 7). The tpst–1 mutant lacks the ability for sulfation of the PSK precursor peptides to generate active ligands. tpst–1 seedlings have smaller shoots and shorter roots than the wt, a phenotype that is partially restored by treatment with 1 μm PSK (Stührwohldt et al., 2011; Figure 7b,c). Moreover, tpst–1 seedlings have shorter epidermal cells, and treatment of tpst–1 seedlings with 1 μm PSK fully restored the mature wt cell length in roots (Figure 7a). This finding is in accord with previous reports showing that PSK controls elongation growth (Kutschmar et al., 2009), whereas other sulfated peptides of the RGF growth factor family control meristematic activity in roots (Matsuzaki et al., 2010).
Treatment of wt or tpst–1 seedlings with 1 nm BL did not significantly alter root length (Figure 7b,c). Treatment with 1 μm BZ on the other hand resulted in shorter roots in both genotypes, indicating that growth inhibition in tpst–1 was further aggravated in the absence of BR. The addition of 1 nm BL to BZ-treated tpst–1 seedlings partially restored responsiveness to PSK in the tpst–1 background (Figure 7b,c). The data thus support the view that PSK signaling of root elongation is mediated through BR signaling. To see whether BR was required for PSK signaling in the epidermis, we analyzed root growth in CER6:PSKR1–3–1 pskr1–3 pskr2–1 and in the overexpression line PSKR1ox12 (Figure 7d). Both lines showed comparable root growth that was enhanced in the presence of 1 μm PSK. Treatment with 1 μm BZ strongly inhibited root growth and resulted in unresponsiveness to treatment with 1 μm PSK, suggesting that PSK signaling in the epidermis is dependent on the presence of BR signaling.
As an independent means to verify the involvement of BRs in the PSK signaling of growth we employed mutants that are defective in either BR synthesis or signaling. det2–1 (deetiolated2–1) is a BR biosynthesis mutant (Chory et al., 1991, 1994). DET2 is a steroid 5α–reductase responsible for the conversion of (24R)–24–methylcholest-4-En-3-1 to (24R)–24–methyl-5α-cholestan-3-1 in the early BR synthetic pathway (Noguchi et al., 1999). The BR receptor mutant bri1-200A has a T–DNA insertion 1245 bp downstream of the ATG in the LRR coding region of the BR receptor gene BRI1 (BRASSINOSTEROID INSENSITIVE 1), which leads to the knock-down of BRI1 gene expression. BZR1 (BRASSINAZOLE-RESISTANT 1) and BES1 (BRI1-EMS SUPPRESSOR 1) are transcription factors that mediate BR responses (Clouse, 2011). det2–1 seedlings displayed a dwarf phenotype with very short roots similar to tpst–1 seedlings (Figure 8a,b). Roots of bzr1–1D, bes1–1 and bri1–200A seedlings were only slightly shorter than wt roots (Figure 8b). BZR1 and BES1 can act in a redundant fashion such that the knock-out of one gene may not result in a growth phenotype. Treatment with 1 μm PSK promoted root elongation in wt and tpst–1, included as controls, in bes1–1, bzr1–1D and to a lesser, but significant, extent in bri1-200A, which is in accord with the weak phenotype of these BR signaling mutants. By contrast, root growth of bri1–9 seedlings of the A. thaliana ecotype Wassilweskija with a loss-of-function mutation in the BRI1 receptor is not promoted by PSK (Figure 8c). Similarly, knock-out of BR synthesis in det2–1 abolished responsiveness to PSK (Figure 8b), indicating that BR synthesis and perception are a prerequisite for the PSK signaling of cell elongation.
In order to test whether PSK signaling affects BR synthesis or signaling at the transcriptional level, we analyzed expression of the biosynthetic genes DET1, DWF4 (DWARF4; Asami et al., 2001), CYP85A1 (Cytochrome P450 85A1; BR6ox, BRASSINOSTEROID-6-OXIDASE 1; Bancos et al., 2002), SMT2 (STEROL METHYLTRANSFERASE 2; Bouvier-Navé et al., 1998) and CPD (CONSTITUTIVE PHOTOMORPHOGENESIS AND DWARFISM; Bancos et al., 2002), and of the signaling genes BRI1 (BRASSINOSTEROID INSENSITIVE 1; Clouse et al., 1996) and BAK1 (BRI1-ASSOCIATED RECEPTOR KINASE 1; Nam and Li, 2002) in wt, tpst–1, pskr1–3 pskr2–1 and PSKR1ox seedlings by qPCR (Figure 9). Expression of all genes was normalized to Actin2 and GAPC (GLYCERALDEHYDE-3-PHOSPHATE DEHYDROGENASE C). The results indicated that transcript levels of genes encoding for enzymes of BR synthesis were not significantly changed in PSK signaling and synthesis mutants. BAK1 transcript levels were unchanged as well. BRI1 transcript levels were slightly reduced to 0.60–fold and 0.62–fold in PSKR1ox and tpst–1 roots, respectively (Figure 9). Overall, the genetic alteration of PSK signaling did not significantly alter the expression of key BR synthesis genes, indicating that interaction may occur at the protein level, possibly in the BR signaling cascade. Taken together, genetic and pharmacological data indicate that PSK perception in atrichoblasts generates a BR-dependent signal that travels to neighboring trichoblast cells.
PSK signaling was shown to control root and shoot growth in Arabidopsis (Matsubayashi et al., 2006; Kutschmar et al., 2009; Stührwohldt et al., 2011). Root elongation was enhanced in Arabidopsis seedlings supplemented with PSK, indicating that growth was limited by ligand availability (Kutschmar et al., 2009). Signaling can likewise be limited by the availability of the respective receptor, which is regulated by synthesis, modification and degradation. In Arabidopsis cells overexpressing PSKR1, (3H)PSK binding activity in the microsomal fractions was enhanced (Matsubayashi et al., 2006). Arabidopsis plants overexpressing PSKR1 showed enlarged leaves and delayed senescence (Matsubayashi et al., 2006). In addition, the overexpression of PSKR1 or PSKR2 was shown in this study to promote hypocotyl and root elongation growth, indicating that PSK receptor abundance was limiting seedling and plant growth, and that transcriptional upregulation can overcome this limitation. Taken together, PSK signaling results in growth-promoting activity, whereby the degree of growth promotion can be influenced by regulating the abundance of either ligand or receptor. The abundance of the plasma membrane-bound BR receptor BRI1 was shown to be regulated at the protein level through dephosphorylation catalyzed by PP2A phosphatase. Dephosphorylated BRI1 is degraded, thus terminating BR signaling (Wu et al., 2011). PSK receptor proteins show great structural similarity to BRI1, and it is conceivable that the regulation of PSK receptor abundance likewise takes place at the protein level to modulate the PSK response.
Plant growth is regulated by phytohormones that interact in a signaling network, which includes communication between cell layers (Savaldi-Goldstein and Chory, 2008). Hormones act in a cell or tissue-type specific manner. The epidermis is considered as the cell layer that restricts growth. Stems cut in half curl outwards, indicating that the inner tissues can expand when relieved from strain exerted by the outer tissue layer (Kutschera and Niklas, 2007). Promoters used for tissue-specific expression may be active at very low levels, not detected by common methods employing fluorescent protein or in situ hybridization. Nonetheless, preferential expression of PSKR1 in the epidermis in a PSK receptor null mutant background was sufficient to promote hypocotyl as well as root growth, and to restore the wt growth phenotype.
Brassinosteroids (BRs) are known to regulate cell elongation (Savaldi-Goldstein and Chory, 2008). Subnanomolar levels of brassinolide were shown to promote root growth, whereas concentrations of 1 nm and higher inhibited root growth, as did genetic suppression of BR synthesis, indicating that BRs promote root growth in a dose-dependent manner (Müssig et al., 2003). BR perception specifically in the epidermis was found to be crucial for BR-dependent root growth (Savaldi-Goldstein and Chory, 2008; Hacham et al., 2010). Similarly, BR-dependent shoot growth was shown to be driven by BR perception or brassinolide synthesis in the epidermis (Savaldi-Goldstein et al., 2007). Epidermis-specific expression of the BR receptor BRI1 or of the key biosynthesis CPD gene resulted in a full BR-stimulated growth response, whereas expression in other tissues did not.
The coincident activity of PSK and BR in the epidermis and the fact that both PSK and BR promote cell elongation prompted us to ask if PSK signaling of elongation growth may be mediated by BR signaling. BZ is a specific inhibitor that inhibits the cytochrome P450 monoxygenase DWF4 that acts as steroid 22–hydroxylase within the BR biosynthesis pathway (Asami et al., 2001). Treatment of wt seedlings with BZ caused a typical BR-deficient phenotype with short hypocotyls and short roots. PSK-deficient tpst–1 seedlings display severe root growth inhibition, which was further aggravated in the presence of BZ. The addition of the highly active BR 24–epibrassinolide to BZ-treated wt or tpst–1 seedlings partially reverted root growth inhibition. In tpst–1 seedlings treated with BZ, responsiveness to PSK was lost but was restored when BR was supplied in addition to BZ, indicating that BR signaling was required for PSK-dependent root growth promotion. The requirement for BR could be direct or indirect. Signaling of growth through the PSK and BR pathways likely occurs in epidermal cells, as PSK responsiveness was also lost in the presence of BZ in CER6:PSKR1 seedlings that express PSKR1 preferentially in the epidermis, even though residual activity in other cell layers cannot be excluded. BRs are perceived by the leucine-rich repeat receptor kinase BRI1 at the plasma membrane. Signal transduction occurs through phosphorylations and dephosphorylations, resulting in the activation of the transcription factors BES1 and BZR1 (Clouse, 2011). Genetic disruption of either BR synthesis in det2–1 or of BR signaling in bri1–9 rendered seedlings unresponsive to PSK, providing additional evidence that the PSK response is mediated by the BR signaling pathway. Knock-out of either BES1 or BZR1 did not compromise PSK-dependent growth, which may be owed to the fact that these transcription factors can act in a redundant fashion (Clouse, 2011).
PSK was described as an autocrine growth factor based on the observation that it promotes the proliferation of suspension-cultured cells kept at a low density (Matsubayashi et al., 1997). Data on the mobility of the PSK peptide or PSK signal in planta have so far not been provided. Using atrichoblast-specific GL2 promoter-driven PSK receptor expression, we showed that localized PSK perception in atrichoblasts caused not only this subpopulation of epidermal cells to elongate, but also trichoblasts that most likely were unable to perceive PSK. PSK signaling of growth thus appears to occur non-cell autonomously. Hence, despite the ubiquitous expression of PSK precursor genes, the PSK peptide does not act as an autocrine growth factor in planta. PSK activity is transmitted to neighboring cells and to distant cell layers, such as the cortex and central cylinder in the root, via mobile or possibly mechanical signals. The epidermis limits growth as a result of the constraints imposed on cell expansion by the stiff cell wall. Cells of the inner tissues expand once the epidermis is removed, indicating that they are compressed (Kutschera, 2008). When epidermal cells grow the compression imposed by the epidermis on inner tissue layers is reduced. This altered tissue tension may be perceived as a mechanical signal either at the plasma membrane or in the cell wall of cells in inner tissue layers. Signaling to the protoplast may involve cortical microtubules, which are also known to determine the direction of cell growth through directional deposition of newly synthesized cellulose microfibrils (Paredez et al., 2006). In the context of organogenesis at the shoot apical meristem, Uyttewaal et al. (2012) recently described a crucial role for the microtubule-severing katanin in harmonizing the growth rates of adjacent cells. Depending on the degree of mechanical stress, the presence of katanin led to more homogeneous or to more heterogeneous growth within the epidermal tissue. As all cells are directly or indirectly linked to each other via their cell walls, it is easily conceivable that mechanical stress is transmitted not only between cells of one tissue but also between tissue layers to coordinate whole-organ growth.
Although the nature of this coordinating signal remains elusive, the regulation of elongation growth by PSK was shown here to require BR, thus linking PSK peptide signaling with hormone regulation of growth.
Plant material and growth conditions
All experiments were performed with A. thaliana ecotype Columbia–0, with the exception of bri1–9 analysis. bi1–9 is a loss-of-function mutant of BRI1 of A. thaliana ecotype Wassilewskija (Jin et al., 2007). The pskr1–3, pskr2–1 and tpst–1 insertion lines used in this study have been described previously (Matsubayashi et al., 2006; Amano et al., 2008; Kutschmar et al., 2009; Stührwohldt et al., 2011). In pskr1–3, a T–DNA is inserted in the kinase domain of PSK receptor 1 and pskr2–1 has a T–DNA in the 11th LRR domain of PSK receptor 2. tpst–1 has a T–DNA inserted in the fifth exon. Two CER6:GUS lines were obtained from Dr Shauna Somerville (Department of Plant and Microbial Biology, Berkeley, CA, USA). We further used homozygous det2–1 (N6159), bzr1–1D (N65987), bes1–1 (N651123) and bri1–200A (N678032) mutants obtained from the Nottingham Arabidopsis Stock Centre (NASC, University of Nottingham, Nottingham, UK). det2–1 and bzr1–1D have been previously described (Chory et al., 1991, 1994). bes1–1 has a T–DNA insertion very close to the stop codon and bri1-200A in the LRR coding region of the BR receptor BRI1. The bri1–9 mutant of Arabidopsis ecotype Wassilewskija has a point mutation in the BR binding site, causing a Ser662 → Phe exchange and BR insensitivity (Jin et al., 2007). Seeds of bri1–9 and wt Arabidopsis ecotype Wassilewskija were kindly provided by Birgit Kemmerling (University of Tübingen, Germany).
Plants were grown in a 2:3 sand:humus mixture that was frozen at –80°C for 2 days to avoid insect contamination and then watered regularly with tap water. Seeds were stratified at 4°C in the dark for 2 days and then transferred to a growth chamber, where they were grown under long-day conditions with 16 h of light (70 μm ol photons m−2 s−1) at 22°C. For growth experiments on sterile plates, Arabidopsis seeds were surface sterilized for 20 min in 1 ml 2% (w/v) sodium hypochlorite, washed five times with autoclaved water and laid out under sterile conditions on square plates containing half-concentrated MS media (Murashige and Skoog, 1962) and 1.5% (w/v) sucrose solidified with 0.38% (w/v) gelrite (Duchefa, http://www.duchefabiochemie.nl). The media were supplemented with 1 μm PSK (NeoMPS, http://www.neomps.com), 1 nm 24–epibrassinolide (Duchefa) or 1 μm BZ (TCI Europe, http://www.tcichemicals.com), as indicated.
Plasmid constructs, transformation and histochemical GUS detection
For the CER6 promoter-driven epidermis-specific PSKR1 (At2g02220) expression, a 1.2–kb genomic fragment upstream of the CER6 (At1g68530) coding region was amplified by PCR using the forward primer 5′–TTTATTCGAGCTCCTTCGATATCGGTTGTTGAC–3′ and the reverse primer 5′–AAATTTGGACTAGTCGTCGGAGAGTTTTAATGT–3′. For the GLABRA2 (GL2) promoter-driven atrichoblast-specific PSKR1 expression, a 2.1–kb genomic fragment upstream of the GL2 (At1g79840) coding region was amplified by PCR using the forward primer 5′–TTTATTCGAGCTCGTTTCCTTCACTATACGTC–3′ and the reverse primer 5′–CCACTAGTCTGTCCCTAGCTAGCTTC–3′. The fragments were cloned into the vector pB7WG2.0 (VIB, http://www.vib.be) by replacing the CaMV 35S promoter. Point mutations were introduced within the SpeI site of the CER6 and of the GL2 promoter PCR fragments by overlap extension PCR (Horton et al., 1989), using primers 5′–CCAACAAAATCAAGTTTTTGCTAAAAATAGTTT–3′ and 5′–ACCCGCAAATAAACTATTTTTAGCA–3′ for CER6, and primers 5′–TACTGCTACGTACATACCCCTACTATAATAGTCAG–3′ and 5′–CCCAATCGAATCTAATACACTGACTATTATAGT–3′ for GL2. The 3.1–kb coding sequences of the intronless PSKR1 and PSKR2 genes were amplified by PCR from genomic DNA using primers 5′–ACGCGTCGACATGCGTGTTCATCGTTTTTGTGTGATCG–3′ and 5′–TACCGGATATCCTAGACATCATCAAGCCAAGAGAC–3′ for PSKR1 and primers 5′–AAATTTGTCGACATGGTGATCATTCTCC–3′ and 5′–ATAGTTTAGCGGCCGCTCATTGTTGTTGAACAGAC–3′ for PSKR2. The PSKR1 and PSKR2 PCR fragments were cloned into the vector pB7WG2.0 downstream of the CER6, GL2 or CaMV 35S promoter using the Gateway cloning system (Invitrogen, http://www.invitrogen.com). For CER6 and GL2 promoter-driven PSKR1 expression, the PSKR1 coding sequence was cloned into the vector pB7WG2.0 downstream of the CER6 or GL2 promoter, respectively.
Arabidopsis plants were transformed with Agrobacterium tumefaciens (GV3101) using the floral-dip method (Clough and Bent, 1998). Selection of transformed plants was performed by spraying with 200 μm BASTA (AgrEvo, http://www.agrevo.de). Homozygous plants were identified by BASTA selection. β–Glucuronidase (GUS) assays were performed as described by Kutschmar et al. (2009). Overview pictures were taken with an Olympus BX 41 microscope (with a Color-View II camera and cell a software) or with an Olympus zoom stereo microscope SZ 61 (with a CMOS color SC 100 camera and analySIS®getIT! software; Olympus, http://www.olympus.com). For cell type-specific analysis of GUS expression, roots or hypocotyls were embedded in Technovit 7100 according to the manufacturers' instructions (Heraeus Kulzer, http://heraeus-dental.com). Sections of 15–25 μm thick were cut with a Leica RM 2255 microtome and analyzed using an Olympus BX 41 microscope. Images were taken with a Color-View II camera and cell a software (Olympus).
RT-PCR and qPCR analysis
RT-PCR expression analysis of PSKR1 and PSKR2 was performed on 5–day-old seedlings and on 14–day-old plants using total RNA reverse transcribed with an oligo dT primer. The cDNA was amplified with the forward primer 5′–CAAAGACCAGCTCTTCCATCG–3′ and the reverse primer 5′–CTGTGAACGATTCCTGGACCT–3′ for Actin2, which was used as a control, with primers 5′–GAGCGTTGCAATACAATCAG–3′ and 5′–CAGTACTTACATGCGTCTCGT–3′ for PSKR1, and with primers 5′–CTCTTCAAGGCTACTGCAAGCATG–3′ and 5′–CATTGT TGTTGAACAGACTCCATAG–3′ for PSKR2 cDNA. PCR amplifications were performed for 29 cycles for Actin2 and 37 cycles for PSKR1 and PSKR2, as described by Kutschmar et al. (2009).
Quantitative reverse transcription polymerase chain reaction (qRT-PCR) was performed with the QuantiTect SYBR Green RT-PCR Kit (Qiagen, http://www.qiagen.com), according to the manufacturer's instructions. The reverse transcription products were amplified using gene-specific primers 5′–TTTGGAGAGGCGATTGAGTG–3′ and 5′–CTCTTCCTTGAACTTGGCAATG–3′ for DET2, 5′–CAGAGGATGAAGCAGAGATGAG–3′ and 5′–TGAGATCGAGAATTTGCTCCG–3′ for DWF4, 5′–CCAAAAGAAATCGAAACCGCC–3′ and 5′–CCAAAAGAAATCGAAACCGCC–3′ for SMT2, 5′–ACAGAGCAGAAAACAGAGTGAG–3′ and 5′–GAAGGAGAGCGGAACAGAG–3′ for CYP85A1, 5′–CCCAAACCACTTCAAAGATGCT–3′ and 5′–GGGCCTGTCGTTACCGAGTT–3′ for CPD, 5′–CCAAAGTTTCAGGTGTTCAAC–3′ and 5′–CTCCAAAATCCGGTGAATCCG–3′ for BRI1, 5′–GAAGAAGTGGAGCAGCTAATC–3′ and 5′–CAGCTAAACCATCTCCTTCAAG–3′ for BAK1, and 5′–ATCAAGGAGGAATCCGAAGG–3′ and 5′–AAGTCGACCACACGGGAAT–3′ for GAPC and 5′–CAAAGACCAGCTCTTCCATCG–3′ and 5′–GTTGTCTCGTGGATTCCAGCA–3′ for Actin2, which were used as controls for standardization. Reactions were performed with a 7300 real-time PCR-system (Life Technologies, http://www.lifetechnologies.com). All data were normalized to GAPC and Actin2 and analyzed using the 2−ΔΔct method (Livak and Schmittgen, 2001). Three independent biological repeats were performed and statistical analysis was performed as described below.
Growth measurements and statistical analysis
Hypocotyl and root lengths were determined from photographs using ImageJ (National Institutes of Health, http://rsb.info.nih.gov/ij). Epidermal cell lengths were determined on intact roots using an Olympus microscope BX41 and an Olympus Color-View II camera, with a 20-fold magnification and image software cell a (Olympus). Statistical analysis was performed with minitab (Minitab Inc., http://www.minitab.com). An anova (Tukey's test) or a two-sample Stiden'ts t–test was performed to test the statistical significance of means. Before statistical analysis, constant variance and normal distribution of data were verified and the P value was set to P < 0.001 if one of these conditions was not achieved. P values for the Pearson product moment correlation are indicated in the figure legends.
We gratefully thank Dr Shauna Somerville (Department of Plant and Microbial Biology, Berkeley, CA, USA) for supplying homozygous CER6:GUS seeds and Dr Birgit Kemmerling (University of Tübingen, Germany) for providing seeds of wt Arabidopsis ecotype Wassilewskija and bri1–9.