SEARCH

SEARCH BY CITATION

Keywords:

  • Arabidopsis;
  • asymmetric division;
  • cell fate;
  • Fackel;
  • sterol;
  • stomata

Summary

  1. Top of page
  2. Summary
  3. Introduction
  4. Results
  5. Discussion
  6. Experimental Procedures
  7. Acknowledgements
  8. References
  9. Supporting Information

Asymmetric cell division is important for regulating cell proliferation and fate determination during stomatal development in plants. Although genes that control asymmetric division and cell differentiation in stomatal development have been reported, regulators controlling the process from asymmetric division to cell differentiation remain poorly understood. Here, we report a weak allele (fk–J3158) of the Arabidopsis sterol C14 reductase gene FACKEL (FK) that shows clusters of small cells and stomata in leaf epidermis, a common phenomenon that is often seen in mutants defective in stomatal asymmetric division. Interestingly, the physical asymmetry of these divisions appeared to be intact in fk mutants, but the cell-fate asymmetry was greatly disturbed, suggesting that the FK pathway links these two crucial events in the process of asymmetric division. Sterol profile analysis revealed that the fk–J3158 mutation blocked downstream sterol production. Further investigation indicated that cyclopropylsterol isomerase1 (cpi1), sterol 14α–demethylase (cyp51A2) and hydra1 (hyd1) mutants, corresponding to enzymes in the same branch of the sterol biosynthetic pathway, displayed defective stomatal development phenotypes, similar to those observed for fk. Fenpropimorph, an inhibitor of the FK sterol C14 reductase in Arabidopsis, also caused these abnormal small-cell and stomata phenotypes in wild-type leaves. Genetic experiments demonstrated that sterol biosynthesis is required for correct stomatal patterning, probably through an additional signaling pathway that has yet to be defined. Detailed analyses of time-lapse cell division patterns, stomatal precursor cell division markers and DNA ploidy suggest that sterols are required to properly restrict cell proliferation, asymmetric fate specification, cell-fate commitment and maintenance in the stomatal lineage cells. These events occur after physical asymmetric division of stomatal precursor cells.


Introduction

  1. Top of page
  2. Summary
  3. Introduction
  4. Results
  5. Discussion
  6. Experimental Procedures
  7. Acknowledgements
  8. References
  9. Supporting Information

Stomatal patterning in plants is an excellent model system for studying molecular mechanisms of cell self-renewal, asymmetric cell division and cell-fate determination. During stomatal development, a precursor cell such as a meristemoid mother cell or a meristemoid undergoes an asymmetric division to produce a new meristemoid and a large sister cell. The new meristemoid may either continue to undergo asymmetric division one to three times, or be directly converted into a guard mother cell, which produces two guard cells by a symmetric division. The large sister cell may either differentiate into a mature pavement cell or regain cell division activity to produce a new meristemoid and a large sister cell (Bergmann and Sack, 2007; Pillitteri and Torii, 2012). In this process, the one-celled spacing pattern rule must be followed (Geisler et al., 2000). This process is regulated by both intercellular and intracellular signaling pathways to make appropriate cell-fate decisions. For example, as the primary receptors of epidermal patterning factors (EPFs), the members of ERECTA (ER) family and their co-receptor TOO MANY MOUTHS (TMM) regulate the initiation and spacing divisions of the stomatal lineage (Nadeau and Sack, 2002; Shpak et al., 2005; Hara et al., 2007; Hunt and Gray, 2009; Abrash and Bergmann, 2010; Sugano et al., 2010; Abrash et al., 2012; Lee et al., 2012). A novel protein, BREAKING OF ASYMMETRY IN THE STOMATAL LINEAGE (BASL), accumulates in a polarized crescent at the cell periphery to control asymmetric division during stomatal formation (Dong et al., 2009). Several bHLH and MYB transcription factors [FAMA, SPEECHLESS (SPCH), MUTE, ICE1/SCREAM, SCREAM2, FOUR LIPS (FLP) and MYB88] control three consecutive steps of cell specification and differentiation in stomatal development (Lai et al., 2005; Ohashi-Ito and Bergmann, 2006; MacAlister et al., 2007; Pillitteri et al., 2007; Kanaoka et al., 2008). Some cell-cycle regulators such as cyclin B1 (CYCB1), cyclin-dependent kinase B1;1 (CDKB1;1) and cyclin A2 (CYCA2) are implicated in the self-renewal and differentiation of stomatal lineage cells (Boudolf et al., 2004; Xie et al., 2010; Vanneste et al., 2011). In addition, other proteins such as STOMATAL DENSITY AND DISTRIBUTION1 (SDD1), GLUCAN SYNTHASE-LIKE 8 (GSL8) and photoreceptors are involved in stomatal patterning (Berger and Altmann, 2000; Kang et al., 2009; Guseman et al., 2010). More recently, brassinosteroid (BR) was found to regulate stomatal development via BR-INSENSITIVE2 (BIN2)-mediated phosphorylation of YODA (YDA) and SPCH (Gudesblat et al., 2012; Kim et al., 2012). Many of these stomatal regulators play important roles in regulating cell self-renewal, orientation of asymmetric division, physical asymmetric division, and differentiation. However, the regulatory events that connect asymmetric division and cell differentiation are not well understood.

Sterols play significant roles in maintaining cell integration and cell-to-cell communication. They are components of the cell membrane (Simon-Plas et al., 2011) and biosynthetic precursors of steroid hormones, including mammalian androgens (Attard et al., 2009), insect ecdysteroids (Gilbert et al., 2002) and plant BRs (Clouse, 2000). Sterol biosynthesis in plants has been thoroughly investigated within the last decade. The sterol precursor cycloartenol is methylated either once or twice to produce a mixture of sterols, including sitosterol, stigmasterol and campesterol. Campesterol is the precursor of BRs (Clouse, 2000) (Figure 1). Recently, it was demonstrated that the lanosterol pathway of yeast and animals also exists in plants as a minor pathway (Ohyama et al., 2009). Identification of plant sterol mutants has helped to uncover the biological functions of sterols in mediating morphogenesis, cell differentiation, cell polarity and cell patterning (Lindsey et al., 2003; Boutté and Grebe, 2009). Enzymes regulating early sterol biosynthesis, such as squalene epoxidase 1 (SQE1), cycloartenol synthase 1 (CAS1), cyclopropylsterol isomerase 1 (CPI1), sterol 14αdemethylase (CYP51A2), FACKEL (FK), HYDRA 1 (HYD1) and sterol methyltransferases SMT1, SMT2 and SMT3, play additional roles in plant development in addition to their roles in BR biosynthesis (Jang et al., 2000; Schrick et al., 2000, 2002, 2004a; Carland et al., 2002, 2010; Souter et al., 2002; Kim et al., 2005, 2010; Babiychuk et al., 2008; Men et al., 2008; Pose et al., 2009), whereas enzymes catalyzing late stages of sterol biosynthesis, such as DWARF 7 (DWF7), DWF5 and DWF1, mainly control BR biosynthesis (Klahre et al., 1998; Choe et al., 1999, 2000). Although the BR signaling cascade has been well documented, BR-independent roles of sterols mediating plant development are largely unknown.

image

Figure 1. The sterol biosynthetic pathway in Arabidopsis.

A branch in the pathway results from either one or two methylation steps, producing two major end products, stigmasterol and campesterol. Campesterol is the precursor of brassinosteroids. Enzymes catalyzing these steps are shown in blue, and their corresponding mutants are shown in red (Jang et al., 2000; Schrick et al., 2000, 2002, 2004a; Men et al., 2008).

Download figure to PowerPoint

Here we describe a mutant exhibiting severe stomatal development and patterning defects from an ethyl methane sulfonate Arabidopsis mutant pool. Map-based cloning revealed that a point mutation in FK is responsible for the stomatal defective phenotypes. We provide genetic and developmental data to demonstrate that early steps of sterol biosynthesis play fundamental roles in stomatal lineage cell-fate determination after physical asymmetric division.

Results

  1. Top of page
  2. Summary
  3. Introduction
  4. Results
  5. Discussion
  6. Experimental Procedures
  7. Acknowledgements
  8. References
  9. Supporting Information

Identification of a stomatal mutant, fk–J3158

To identify additional components regulating stomatal development, we performed ethyl methane sulfonate mutagenesis in an E361 line, which displays a normal phenotype with regard to plant growth and stomatal patterning. A GFP has been engineered in the mature stomata of E361, which allows easy stomatal mutant screening under a fluorescence microscope (Gardner et al., 2009) (Figure 2a). A stomatal mutant, J3158, was identified from approximately 6000 M2 plants. Unlike the wild-type plant (Col-0), J3158 displayed severe defects in stomatal cell division and patterning, which resulted in clusters of small cells and stomata (Figure 2b, brackets, and Table S1). Map-based cloning indicated an adenine-to-guanine substitution 2170 bp from the translation initiation codon of At3 g52940 in J3158 (Figure 2c,d), changing amino acid 291 from a serine to a leucine in the FK C14 reductase of the sterol biosynthetic pathway (Figure 1). Ser291 is conserved in the signature motif LLXSGWWGXXRH of the FK protein (Jang et al., 2000). The defective stomatal development in J3158 was complemented by over-expression of FK cDNA (all five transgenic lines showed complemented phenotypes) (Figure S1ac,e,f), confirming that the J3158 phenotypes are indeed caused by the mutation in FK. Over-expression of FK in Col-0 (35S::FK) resulted in normal stomatal development (n = 7 lines) (Figure S1df). However, unlike other FK mutant alleles such as fk–J79 and fk–X224 (Jang et al., 2000; Schrick et al., 2000, 2004a), J3158 seedlings are fully fertile in soil and display much weaker defects, including multiple cotyledons, dark green rosettes, dwarfed growth, prolonged lifespan, opposite siliques, shriveled seeds, and defects in vascular patterning, root growth and embryo development (Figures S2 and S3). J3158 was also crossed with fk–J79 (heterozygote). The resulting F1 plants with the genotype J3158- fk–J79- did not show complemented stomatal development and plant growth defect phenotypes (= 16 lines), further indicating that J3158 and fk–J79 are different alleles of FK. We therefore renamed this FK mutant allele fk–J3158. In addition, we re-examined the leaf epidermis of fk–J79 and fk–X224, and found that both had stomatal patterning defects similar to fk–J3158 with even more severe defective phenotypes (determined by the percentage of small cells: 44.4% of fk–J3158, 50.9% of fk–J79, and 69.9% of fk–X224) (Figure 2e,f and Table S1). These data indicate that fk–J3158 is a weak allele of FK.

image

Figure 2. Mutants of the sterol C–14 reductase FK display stomatal patterning defects

(a,b) Stomatal phenotypes in 5 dpg rosette leaves of Col-0 (a) and fk–J3158 (b) in the E361 background. Brackets indicate clustered small cells and stomata in fkJ3158. Scale bars = 50 μm.

(c) A single G[RIGHTWARDS ARROW]A substitution at position 2170 of At3g52940 was identified by map-based cloning.

(d) Schematic representation of the T–DNA insertion into the promoter of FK (fkJ79), the genomic rearrangement in exon 10 with a breakpoint induced by X–rays (fkX224), and the single base mutation (S291L) in exon 11 induced by ethyl methane sulfonate (fkJ3158). Black boxes represent exons.

(e,f) Stomatal phenotypes in 5 dpg rosette leaves of fkJ79 (e) and fkX224 (f). Scale bars = 50 μm.

Download figure to PowerPoint

Sterol profile analysis showed an accumulation of Δ8,14-sterols in the fk–J3158 mutant

The levels of sterols are expected to be altered in fk–J3158 if it is a true weak allele of FK. The sterol profile and quantification of greenhouse-grown rosettes from Col-0 and fk–J3158 revealed a replacement of 80% of the pathway end-products (campesterol and sitosterol) by Δ8,14sterols (Δ8,14-campesterol, -sitosterol and -isofucosterol) (Table 1). This clearly indicates that the FK S291L mutation in fk–J3158 very poorly catalyzes C14 reduction. In fact, the substrate of the enzyme, 4αmethylergostatrienol, was most likely converted to Δ8,14sterols by enzymes downstream of FK in the sterol biosynthetic pathway (Figure S4). In addition, this sterol profile was very similar to that of Arabidopsis seedlings treated with 15azasterol, an inhibitor of FK (Schaller, 2003). These data suggest that the mutation in fk–J3158 blocks downstream sterol production and accumulates abnormal sterol intermediates.

Table 1. Major sterols in Col-0 and fkJ3158 leaves
SterolCol-0fkJ3158
  1. a

    Shown in Figure S4.

  2. b

    mg/g dry matter.

Campesterol (1)a0.27 ± 0.02bTrace
Δ8,14–ergostadienol (2)Not detected0.10 ± 0.02
Sitosterol (3)1.62 ± 0.100.37 ± 0.03
Δ8,14–stigmastadienol (4)Not detected1.16 ± 0.13
Isofucosterol (5)TraceNot detected
Δ8,14,24–stigmastatrienol (6)Not detected0.21 ± 0.07

Early steps of sterol biosynthesis are involved in Arabidopsis stomatal patterning

To test whether other enzymes of the sterol biosynthetic pathway are required for proper stomatal development, we first examined stomatal phenotypes in sqe1–5, smt1/cph–G213, cpi1–1, cyp51A2–3 and hyd1 mutants. Over-production of small cells (approximately four times the wild-type level) was observed in the leaf epidermis of sqe1–5 (Figure 3b and Table S1). However, these small cells appeared to maintain three normal sequential asymmetric divisions (almost no more than four small cells adjacent to each other, unlike the small-cell cluster of fk mutants), and no stomatal clusters were observed. Although smt1/cph–G213 mutants showed abnormal embryogenesis and post-embryonic development, they displayed wild-type stomatal patterning (Figure 3c and Table S1). Interestingly, the mutants of the three individual steps closest to FK in sterol biosynthesis, cpi1–1, cyp51A2–3 and hyd1, exhibited clustered small cells and stomata, resembling that of the fk mutants (Figure 3d–f and Table S1). Next, we investigated the stomatal development of mutants in the late steps of sterol biosynthesis (e.g. smt2/cvp1–3 and dwf5), the BR biosynthetic pathway [e.g. de-etiolated 2–28 (det2–28 and constitutive photomorphogenesis and dwarfism (cpd)], three receptors of BR (bri1–701 brl1 brl3 triple mutants), and the GSK3/SHAGGY-like kinase of the BR signaling pathway (bin2–1). Although the stomatal densities of some mutants were increased, probably by the reduced end sterols and BRs, no clustered small cells and stomata phenotypes similar to fk were observed in any of these mutants (Figure S5a,c,e,g,i,k,mr). Moreover, the fk–J3158 stomatal development and patterning defects were not rescued by exogenous application of 24epibrassinolide (BL), stigmasterol, campesterol or a mixture of the three substances. We conclude that the absence of final sterols and BRs partially affected the stomatal development of fk–J3158, but appeared unable to coordinate the defective stomatal patterning in a regulated manner (Table S2). Hence, these results suggest that the early steps of the sterol biosynthetic pathway are required for stomatal development, which may be independent of final sterol levels and the BR pathway.

image

Figure 3. Stomatal patterning defects in some mutants of sterol biosynthesis.

(a–f) Stomatal phenotypes in 5 dpg rosette leaves of Col-0 (a), sqe15 (b), smt1/cphG213 (c), cpi11 (d), cyp51A23 (e) and hyd1 (f). Scale bars = 50 μm.

Download figure to PowerPoint

Fenpropimorph, an inhibitor of FK, phenocopies stomatal development and patterning defects of fk

Fenpropimorph (FEN) is an inhibitor of FK sterol C14 reductase in Arabidopsis (He et al., 2003). It is also reported to inhibit at least three other enzymes (cycloeucalenolobtusifoliol isomerase, C8,7 isomerase and Δ7reductase, in various organisms (Mercer, 1993). In previous studies of sterol biosynthesis, FEN was used as an effective tool to verify and extend genetic data (He et al., 2003; Schrick et al., 2004a). In the present study, we treated Col-0 with 0.01, 0.1 and 1.0 μm FEN. As expected, FEN-treated Col-0 displayed clustered small cells and stomata, similar to the fk mutants (Figure 4a,b and Table S1). Col-0 treated with 0.1 or 1.0 μm FEN showed similar levels of stomatal development defects, but the phenotypes were more severe than for those treated with 0.01 μm FEN [as determined by the percentage of small cells in rosette leaves: 23.7% for Col-0 versus 31.6, 60.5 and 67.8% for 0.01, 0.1 and 1.0 μm FEN-treated Col-0 at 5 days post-germination (dpg), respectively]. We therefore selected 0.1 μm FEN for subsequent studies. To further examine whether the defective phenotypes of FEN-treated Col-0 and fk mutants were caused by the same molecular mechanisms, RNASeq analysis was performed. The expression patterns of most genes in fk–J3158 were analogous to those of FEN-treated Col-0 (Figure 4c and Data S1). These results indicate that FEN inhibits sterol biosynthesis and effectively mimics the stomatal development and patterning defects of fk mutants, further confirming the effect of the sterol biosynthetic pathway on stomatal development in Arabidopsis.

image

Figure 4. FEN phenocopies the stomatal defects of fkJ3158.

(a,b) Stomatal phenotypes in 5 dpg rosette leaves of Col-0 (a) and 0.1 μM FEN-treated Col-0 (b) in the E361 background. Scale bars = 50 μm.

(c) Cluster analysis of the genes differentially expressed in fkJ3158 and FEN-treated Col-0 (see Data S1).,

Download figure to PowerPoint

Sterols mediate an additional signaling pathway required for stomatal development

To reveal the relationship between sterol biosynthesis and known genetic pathways of stomatal development, we generated double mutants by using fk–J3158 and several stomatal mutants, including tmm–1, sdd1–1, flp–1 and ice1–J3106. Compared with the single tmm and sdd1 mutants, which produced stomata clusters and showed enhanced stomata density, fk–J3158 tmm–1 and fk–J3158 sdd1–1 double mutants generated small-cell clusters similar to fk–J3158, and showed a greatly increased number of small cells and stomatal density (Figure 5a,b,d,e,m,n, brackets). The fk–J3158 flp–1 double mutant showed small-cell and stomata clusters similar to fk–J3158, and four-lip or caterpillar-like guard cells similar to flp–1 (Figure 5g,h, brackets). The double mutant fk–J3158 ice1–J3106 also displayed dual phenotypes of fk–J3158 and ice1–J3106, exhibiting small-cell clusters, inward-spiral meristemoid division, caterpillar-like guard cells and reduced stomatal density (Figure 5j,k,o, brackets). Furthermore, these double mutant phenotypes were completely recapitulated when each of the four single mutants was treated with 0.1 μm FEN, respectively (Figure 5c,f,i,l,m–o, brackets). These results indicate that it is feasible to mimic the double mutant phenotypes by treating the known stomatal mutants with FEN. To this end, the smt2/cvp1–3 mutant, which shows normal stomatal development, was treated with FEN, and displayed a stomatal phenotype similar to FEN-treated Col-0 (Figure S5a,b,m). The FEN-treated dwf5, det2–28, cpd, bri1–701 brl1 brl3 and bin2–1 mutants displayed additive effects with regard to small-cell and stomata density (Figure S5cl,nr). Likewise, the FEN-treated epf1–1, epf2–1, yda–Y295 and basl–2 mutants also showed dual phenotypes (Figure S6). These results suggest that sterols function independently of the TMM-, SDD1-, FLP-, BASL- and BR-mediated signaling pathways in stomatal development.

image

Figure 5. The sterol biosynthetic pathway acts independently of TMM, SDD1, FLP and ICE1 in stomatal development.

(a–l) Stomatal phenotypes in 5 dpg rosette leaves of tmm1 (a), fkJ3158 tmm1 (b), FEN-treated tmm1 (c), sdd11 (d), fkJ3158 sdd11 (e), FEN-treated sdd11 (f), flp1 (g), fkJ3158 flp1 (h), FEN-treated flp1 (i), ice1J3106 (j), fkJ3158 ice1J3106 (k) and FEN-treated ice1J3106 (l). Scale bars = 50 μm. Yellow brackets indicate stomata clusters or caterpillar-like guard cells. Red brackets indicate small-cell clusters.

(m–o) Statistical analysis of stomata and small cells in fkJ3158 tmm1 and FEN-treated tmm1 (m), fkJ3158 sdd11 and FEN-treated sdd11 (n), and fkJ3158 ice1J3106 and FEN-treated ice1J3106 (o). Values are mean ± standard deviation (SD). Asterisks indicate a statistically significant difference compared with Col-0, tmm1, sdd11 and ice1J3106 in small cells; hash symbols (#) indicate a statistically significant difference compared with Col-0, fkJ3158 and FEN-treated Col-0 in stomata (Student's t-test, < 0.01, = 30).

Download figure to PowerPoint

SPCH, MUTE and FAMA are epistatic to the sterol biosynthetic pathway in certain stomatal development contexts

Three bHLHs (SPCH, MUTE and FAMA) are the final switches that determine three specific types of cell differentiation: protodermal cells to meristemoids, meristemoids to guard mother cells, and guard mother cells to guard cells (Ohashi-Ito and Bergmann, 2006; MacAlister et al., 2007; Pillitteri et al., 2007). In mute, all meristemoids arrested and failed to differentiate into guard mother cells (Figure 6a). Likewise, fama–1 displayed immature caterpillar-like guard cells (Figure 6c). When treated with FEN, mute showed clustered small cells and no stomata, and fama–1 demonstrated clustered small cells and caterpillar-like guard cells (Figure 6b,d). These dual phenotypes suggest that MUTE and FAMA additively interact with the sterol biosynthetic pathway, but are epistatic to this pathway in some developmental contexts. In spch–1, all protodermal cells failed to establish the stomatal cell lineage and became pavement cells (Figure 6e). FEN-treated spch–1 displayed a phenotype similar to that of the single mutants without small cells and stomata (Figure 6f), suggesting that the sterol biosynthetic pathway functions after the first entry division of protodermal cells in stomatal development. However, it is difficult to conclude that this pathway directly feeds into SPCH, because SPCH is epistatic to all mutations that affect division late in the stomatal lineage (MacAlister et al., 2007). Moreover, ICE1 interacts with SPCH directly (Kanaoka et al., 2008), and FK and ICE1 function independently, as the phenotype of each single mutant was present in the double mutant (Figure 5j–l,o). These results suggest that the sterol biosynthetic pathway and bHLHs function independently during stomatal development, but that bHLHs are epistatic to the sterol pathway in the differentiation processes of meristemoids, guard mother cells and guard cells.

image

Figure 6. SPCH, MUTE and FAMA are epistatic to sterol biosynthesis in certain stomatal developmental contexts.

(a–f) Stomatal phenotypes in 5 dpg rosette leaves of mute (a), FEN-treated mute (b), fama1 (c), FEN-treated fama1 (d), spch1 (e) and FEN-treated spch1 (f). Scale bars = 50 μm.

Download figure to PowerPoint

Abnormal stomatal lineage cell divisions of fk–J3158 occur after the first asymmetric entry division

To further determine the function of sterol biosynthesis in Arabidopsis stomatal development, we examined successive phases of leaf epidermal development from various leaves of fk–J3158 (Figure 7a–h). Protodermal cells of fk–J3158 showed no difference from the wild-type 12 h after germination (Figure 7a,e). Once asymmetric cell divisions appeared, over-produced and bunched small cells were observed in fk–J3158 (Figure 7b–d,f–h). High mitotic activity was maintained not only in young leaf epidermis but also in expanded adult epidermis (Figure 7i–k). However, some of these bunched small cells ultimately developed into clustered small pavement cells instead of stomata, as demonstrated by their pavement cell-like lobes and necks (Figure 7i,j, brackets). These observations indicate that the stomatal development and patterning defects in fk–J3158 are caused by abnormal division of stomatal precursor cells after the first asymmetric entry division, consistent with the result for FEN-treated spch–1 (Figure 6f).

image

Figure 7. Stomatal development in various growth processes of fkJ3158 leaves.

(a–h) Stomatal development on rosette leaves of fkJ3158 (e–h) compared with Col-0 (a–d) at 12 h, 1, 2 and 4 dpg, respectively. Scale bars = 50 μm.

(i–j) Scanning electron microscopy images of stomatal phenotypes in 24 dpg adult rosette leaves of Col-0 (i) and fkJ3158 (j). Bracket indicates small pavement cell clusters. Scale bars = 50 μm.

(k) Quantification of various epidermal cells in 24 dpg rosette leaves. Value are means ± SD. Asterisks indicate a statistically significant difference compared with Col-0.

Download figure to PowerPoint

The sterol biosynthetic pathway functions in cell-fate commitment and maintenance of the stomatal lineage after physical asymmetric division

To balance the pavement cell and guard cell population in wild-type leaves, the large daughter cell, after asymmetric division, is committed to develop into a pavement cell, and the small daughter cell maintains mitotic activity to ultimately differentiate into guard cells (Bergmann and Sack, 2007). We discovered that abnormal stomatal lineage cell divisions in fk–J3158 occur after the first entry division of protodermal cells. To determine how these clustered small cells, stomata and small pavement cells are formed in fk–J3158, we performed time-lapse observations of the same fields on living rosette leaves and some stomatal precursor markers. We found that some stomatal precursors in fk–J3158 underwent typical amplifying divisions like Col-0, whereas others showed abnormal cell division (Figure 8). In abnormal cell division, both daughter cells maintained mitotic activity and synchronously divided, finally resulting in clustered small cells and stomata (Figure 8a–g, colored outlines, and Figure S7). At the same time, the mitotic spindles in dividing small-cell clusters of fk–J3158 were clearly visible by the GFP–TUBULIN6 (TUB6) marker (Nakamura et al., 2004) (Figure 9a,b, Figure S8ac and Movies S1S3), further demonstrating the synchronous division of two daughter cells. Similar phenomena were also observed in FEN-treated Col-0 (Figure 9c). We know that MUTE is expressed in some meristemoids that are about to differentiate into guard mother cells (Pillitteri et al., 2007) (Figure 9d). In fk–J3158 and FEN-treated Col-0, pMUTE::GFP signals were observed in some small-cell clusters (Figure 9e,f), suggesting that the asymmetric expression pattern of MUTE in the stomatal lineage is lost in the fk mutants, and that the clustered small cells expressing MUTE will most likely undergo terminal differentiation and form clustered stomata. Furthermore, fk–J3158 and FEN-treated Col-0 were labeled by pTMM::TMM-GFP and pCYCB1;1::CYCB1;1-GUS, which are expressed in early stomatal lineage cells, including meristemoid mother cells, meristemoids and guard mother cells (Donnelly et al., 1999; Nadeau and Sack, 2002) (Figure 9g,j). We found that GFP and GUS signals were detected in some clustered small cells, but not in those that likely withdrew from stomatal lineage cell division and finally converted to small pavement cell clusters (Figure 9h,i,k,l and Figure S8df). In addition, we found another case in which only the big daughter cell from asymmetric division developed into guard cells in fk–J3158 (Figure 8c,d, yellow asterisks, and Figure S7c). Furthermore, in some instances, two adjacent small cells from different mother cells synchronously differentiated into guard cells (Figure S7 h,i).

image

Figure 8. fkJ3158 stomatal lineage histories.

(a–g) fkJ3158 stomatal lineage histories on the same field of 3, 4, 5 and 6 dpg rosette leaves (a–d), and 3, 4 and 5 dpg (e–g), respectively. Normal stomatal development (white outlines) and abnormal small-cell division and differentiation (colored outlines) are shown. The yellow asterisks in (c) indicate typical asymmetric divisions from four neighboring small cells. The yellow asterisk in (d) indicates the big daughter cell but the small one developed into guard cells. Scale bars = 50 μm.

(h) Serial stomatal development of Col-0 and fkJ3158, drawn on the basis of the results in (a–g).

Download figure to PowerPoint

image

Figure 9. Expression patterns of TUB6, MUTE, TMM, CYCB1;1 and BASL in fkJ3158 and FEN-treated Col-0.

(a–c) GFPTUB6 expression in 5 dpg rosette leaves of Col-0 (a), fkJ3158 (b) and FEN-treated Col-0 (c). Brackets indicate synchronously dividing stomatal lineage cells. Scale bars = 50 μm. (d–f) pMUTE::GFP expression in 5 dpg rosette leaves of Col-0 (d), fkJ3158 (e) and FEN-treated Col-0 (f). Brackets indicate some small-cell clusters with pMUTE::GFP expression. Scale bars = 50 μm.

(g–l) Expression of pTMM::TMM-GFP (g–i) and pCYCB1;1::CYCB1;1-GUS (j–l) in 5 dpg rosette leaves of Col-0, fkJ3158 and FEN-treated Col-0, respectively. Brackets indicate some small-cell clusters without division activity. Scale bars = 50 μm.

(m–o) pBASL::GFP-BASL expression in 5 dpg rosette leaves of Col-0 (m) and FEN-treated Col-0 (n,o). Arrows indicate BASL peripheral crescents in the small-cell clusters. Scale bars = 50 μm.

(p–s) Real-time imaging was performed using FEN-treated Col-0 rosette leaves expressing pBASL::GFP-BASL. Arrows indicate BASL peripheral crescents in the small-cell clusters. Asterisks indicate the nucleus in small-cell clusters.

Download figure to PowerPoint

We next addressed the question of whether, if the FK mutation leads to abnormal cell proliferation and cell-fate commitment in stomatal development because of the synchronous division of two daughter cells, it affects physical asymmetric division. We quantified two daughter cell sizes from asymmetric division, and observed expression of the asymmetric cell division (ACD) marker (Dong et al., 2009). In Col-0 leaves, 70% of meristemoid mother cell and meristemoid divisions resulted in the smaller cells being <40% of the combined area of the two daughter cells. Similarly, in fk–J3158, 72% of meristemoid mother cell and meristemoid divisions that produced daughter cells undergoing incorrect cell division or cell fate (detected by time-lapse agarose impressions) exhibited the same size asymmetry as in Col-0 (= 180 cell pairs) (Figure 8 and Figure S7). Additionally, measurement of the division asymmetry using TUB6-labeled preprophase bands of microtubules, cytokinesis and phragmoplasts, in Col-0 and fk–J3158 leaves also showed similar results (= 200 cells) (Figure 9a–c and Figure S8c). Moreover, expression of the asymmetric cell division marker pBASL::GFP-BASL was strongly polarized in the clustered small cells of FEN-treated Col-0 (Figure 9m–o and Figure S8gi). Time-lapse imaging revealed that, during the mitotic division of small-cell clusters, pBASL::GFP-BASL exhibited polarized localization in every cell membrane (Figure 9p–s and Figure S8j–l). These results clearly indicate that the physical asymmetry of cell division is normal during stomatal development of the fk mutants.

Thus, a model for the stomatal development of fk mutants is shown in Figure S8(m). After physical asymmetric division of meristemoid mother cells or meristemoids, two daughter cells can freely change their cell-fate commitments to develop into stomata or pavement cells or maintain stomatal stem cell activity. Overall, these data indicate that the sterol biosynthetic pathway is required for specification of the asymmetric daughter cell fates and maintenance of the stomatal lineages during development.

FK affects cell-cycle gene expression and cell ploidy in Arabidopsis

Because stomatal precursor cells in adult leaves of fk–J3158 maintained high mitotic activity, we examined whether the cell cycle was affected. We investigated the expression of a set of cell-cycle genes, such as CYCB1, CYCD3, INHIBITOR OF CDK PROTEIN 1 (ICK1), ICK2, S PHASE KINASE-ASSOCIATED PROTEIN 1 (SKP1) and E2FC, in adult leaves by real-time RTPCR. The results show that ICK1, ICK2, SKP1 and E2FC, genes that repress cell division (Gutierrez, 2009), were down-regulated by approximately one-third (Figure 10a). At the same time, expression of CYCB1 and CYCD3 in fk–J3158 was significantly increased by approximately one-half and three times, respectively, compared to wild-type (Figure 10a). A previous study showed that over-expression of CYCD3 inhibits cell differentiation and dramatically increases cell numbers in leaves (Dewitte et al., 2003). This may be explained by the FK mutation indirectly enhancing CYCD3 expression, resulting in over-production of small cells. The high CYCB1 transcription level in fk–J3158 suggests that more cells entered the G2/M phase (Gutierrez, 2009). Moreover, through RNASeq analysis, we found that more than 36 cell cycle-related genes such as CDK, cyclin, ICK and E2F genes were also affected in fk–J3158 and FEN-treated Col-0 (Figure S9), suggesting that sterol biosynthesis is involved in cell-cycle regulation during stomatal development. Recently, Vanneste et al. (2011)showed that CYCA2 genes are required for proper guard mother cell division and differentiation. In fk–J3158 and FEN-treated Col-0, CYCA2;1 and CYCA2;2 were specifically down-regulated (Figure S9). This finding is consistent with the role of sterols in the proper restriction of cell proliferation in stomatal development.

image

Figure 10. Cell-cycle gene expression and cell ploidy in fkJ3158.

(a) Cell-cycle gene expression determined by real-time RT–PCR in the 3rd and 4th rosette leaves of 24-day-old Col-0 and fkJ3158. Asterisks indicate statistically significant differences compared with Col-0 (Student's t-test, < 0.01, = 3).

(b) Frequency distribution of epidermal cells in the 4th rosette leaves of 24-day-old Col-0 and fkJ3158.

(c,d) Ploidy distribution of the 4th rosette leaf nuclei from 24-day-old Col-0 and fkJ3158, obtained using a flow cytometer (c) and by analysis of cell ploidy from (c) (d). The asterisk indicates a statistically significant difference compared with Col-0 (Student's t-test, < 0.01, = 5).

Download figure to PowerPoint

It has been reported that epidermal pavement cell sizes reflect the cell nuclear DNA levels to a certain extent (Melaragno et al., 1993). Analysis at the cellular level by morphometry showed that fk–J3158 adult leaves contain more, but smaller, epidermal cells than Col-0 adult leaves (Figure 10b), implying that fk–J3158 exhibits more low-level DNA ploidy. Flow cytometric analyses indicated that fk–J3158 mature rosette leaves have higher 4C levels of DNA ploidy but lower 8C and 16C levels of DNA ploidy than those of Col-0 (Figure 10c,d), suggesting that more cells enter into the mitotic cycle, but not the endocycle, in fk–J3158 than in the wild-type. This finding is also consistent with the high expression of CYCB1 and CYCD3 in fk–J3158. The altered gene expression and cell ploidy partially reflect the abnormal cell proliferation and differentiation in fk stomatal development.

Discussion

  1. Top of page
  2. Summary
  3. Introduction
  4. Results
  5. Discussion
  6. Experimental Procedures
  7. Acknowledgements
  8. References
  9. Supporting Information

Some unidentified sterols may function as signaling molecules that affect stomatal development in Arabidopsis

Previous studies indicated that several phytohormones regulate stomatal density and guard cell movement (Saibo et al., 2003; Hirayama and Shinozaki, 2007; Acharya and Assmann, 2009). Recently, kinase BIN2 of the BR signaling pathway was found to regulate stomatal development via its interaction with YDA and SPCH (Gudesblat et al., 2012; Kim et al., 2012). In this paper, we show that the early steps of sterol biosynthesis are involved in stomatal patterning, which is independent of the final products of sterol biosynthesis and BRs. Why are stomatal defects only seen in the cpi1, cyp51A2, fk and hyd1 mutants, but not in other sterol biosynthetic mutants? Are the stomatal defects caused by the lack of end products or by altered sterol products? The fact that the fk, cpi1, cyp51A2 and hyd1 mutants exhibit the same stomatal defects implies that they share a common regulatory mechanism in stomatal development. This suggests that it is not the protein target of a regulatory pathway but rather the product of a biosynthetic pathway that affects stomatal development. Sterol profiles in mutants of the sterol biosynthetic pathway have been analyzed in detail. In sqe1 and smt1 mutants, the sterol biosynthetic pathway is normal, because SQE1 and SMT1 have several homologous genes that may compensate for their functions (Schrick et al., 2002; Pose et al., 2009). As a result, these two mutants did not show any stomatal patterning defects. However, in the fk, cpi1, cyp51A2 and hyd1 mutants, the levels of all of the downstream sterols are dramatically reduced and those of the upstream sterols are partially accumulated. In addition, some abnormal sterol derivatives that are not detected in wild-type leaves also accumulated in these mutants (Jang et al., 2000; Schrick et al., 2000, 2004a; Kim et al., 2005; Men et al., 2008) (Table 1). It seems impossible that the accumulated upstream sterols, before the four steps, and the absence of the sterols, after the steps of DWF5, may result in the stomatal development defects. This is because the dwf5 mutant in the late step of sterol biosynthesis, which shows increased upstream sterols and decreased downstream products when the step is blocked, did not display abnormal stomatal patterning (Figure S5c) (Choe et al., 2000). Moreover, as different mutants produce different sterol derivatives (e.g. 5αcholesta-8,14dien-3βol in fk mutants, 24dihydrocycloeucalenol in cpi1 mutants, 14amethyl-24(241) dihydrofecosterol in cyp51A2 mutants, and stigmasta-monoen-3βol in hyd1 mutants), these sterol derivatives may not cause stomatal development defects. However, the possibility cannot be completely excluded that although the derivative structures are different, their functions may be similar. Based on all these results, we propose that mutations in the four early steps of the sterol biosynthetic pathway cause depletion of some unidentified intermediate sterols between the steps of HYD1 and DWF5, which possibly mediate a signaling pathway that affects stomatal development in Arabidopsis.

It has been postulated that specific sterols required for appropriate signaling involved in embryo and vascular tissue development are absent or reduced in some mutants (Clouse, 2000; Schrick et al., 2002). To further understand the function of sterol biosynthesis in stomatal development, we carefully examined various defective mutant phenotypes of this pathway (Table S3). The smt1, fk and hyd1 mutants displayed embryonic defects, and the smt1, cpi1, cyp51A2, fk, hyd1 and smt2/cvp1 mutants showed abnormal vascular development. All mutants showed arrested growth of roots and hypocotyls, except the smt2/cvp1 mutants. However, stomatal defects were only found in the cpi1, cyp51A2, fk and hyd1 mutants, suggesting that sterols are required for stomatal patterning via a mechanism that is distinct from their roles in embryogenesis and vascular development. In addition, fk–J3158 is a weak allele of FK, displaying weaker phenotypes not only in stomatal development but also in whole-plant development in comparison with other previously reported fk alleles. This mutant is completely fertile, which makes it an ideal candidate for future genetic analysis of the sterol pathway.

The class III and IV homeodomain leucine zipper (HDZIP) transcription factors contain steroidogenic acute regulatory protein-related lipid transfer (START) domains that have been hypothesized to be regulated by sterols in the meristem and vascular development of Arabidopsis (Schrick et al., 2004b; Carland et al., 2010; Pullen et al., 2010). Two members of the class IV HDZIP family, PROTODERMAL FACTOR 2 (PDF2) and MERISTEM LAYER 1 (ATML1), were specifically expressed in stomatal lineage cells, and their double mutant displayed stomatal defects (Abe et al., 2003; Nakamura et al., 2006). The results of these studies suggest the possibility that sterols bind START domain proteins to affect stomatal development.

Sterol biosynthesis mediates an additional stomatal signaling pathway to affect daughter cell fates after asymmetric division

Asymmetric cell division in the development of the stomatal lineage is an important strategy for balancing self-renewal and differentiation of dividing cells (Bergmann and Sack, 2007). Mechanisms determining the different fates of the two daughter cells from an asymmetric division in stomatal development are still poorly understood. In this study, we proposed a model in which sterol biosynthesis is involved in the cell-fate commitment of a daughter cell from asymmetric division through an additional genetic pathway (Figure 11), which was supported by two independent lines of evidence. First, genetic analyses indicate that, during stomatal development, sterol biosynthesis functions independently of the TMM-, SDD1-, BASL- and BR-mediated pathways (Figure 5, and Figures S5 and S6), which are mainly involved in selection of the cell division site, physical asymmetric division and self-renewal (Berger and Altmann, 2000; Geisler et al., 2000; Shpak et al., 2005; Dong et al., 2009; Kim et al., 2012). However, this pathway still requires three bHLHs as final switches to determine the three cell-type differentiations (Figure 6). Second, analyses of the stomatal development process and expression patterns of cell division markers clearly suggest that the main functions of the sterol pathway components are in cell-fate commitment and maintenance of the stomatal lineage after physical asymmetric division (Figures 8 and 9, and Figures S7 and S8). However, this pathway does not appear to be required for regulation of division orientation or physical asymmetry.

image

Figure 11. Model of the early steps of sterol biosynthesis involved in stomatal development.

The EPF–TMM/ER–MAPK cascade, SDD1 and BASL help bHLH and MYB transcription factors to determine three cell-type differentiations (Casson and Gray, 2008; Lampard et al., 2009; Pillitteri and Torii, 2012). The sterol biosynthetic pathway restricts the cell cycle to affect cell-fate commitment and maintenance of the stomatal lineage after asymmetric division.

Download figure to PowerPoint

In Drosophila, studies of neuroblast and ovarian stem cell development have suggested that either intrinsic mitotic spindle orientation (physical asymmetric division) and cell-fate determinant segregation, or extrinsic cell–cell communication are responsible for the difference in the two daughter cells during asymmetric division. These mechanisms appear to be combined in only a very small number of mammalian stem cell systems (Horvitz and Herskowitz, 1992; Hawkins and Garriga, 1998; Knoblich, 2008). Because cell–cell communication-controlled stem cell division offers a high degree of flexibility, this mechanism is more common in adult stem cells (Knoblich, 2008). In the present study, we sought to determine whether FK is important for stomatal cell-fate determinants or for cell–cell communication of the stomatal lineage. If FK functions in segregation of stomatal cell-fate determinants, its mutation should result in only two abnormal cases, i.e. where the big daughter cell from asymmetric division develops into guard cells in fk–J3158 (Figure 8c,d, yellow asterisk, and Figure S7c), or both daughter cells synchronously develop into guard cells (Figure S7dg). However, there are two exceptions in fk–J3158: either both the daughter cells maintain meristemoid activity (Figure 9e,f,h,i,k,l, and Figure S8d–f) or they withdraw from the stomata lineage cell division and differentiate into pavement cells (Figures 7j and 9e,f,h,i,k,l, and Figure S8d–f). As the adjacent small cells from different mother cells synchronously differentiated into guard cells (Figure S7 h,i), we tentatively presume that the sterol biosynthetic pathway coordinates the daughter cell fates by cell–cell communication after asymmetric division, rather than through intrinsic determinant segregation.

The smt2/frl1 mutation causes an enhanced level of endo-reduplication in petal tips and rosette leaves (Hase et al., 2005; Carland et al., 2010). In contrast, we found that, in fk–J3158 leaves, more cells entered into mitosis (Figure 10 and Figure S9), suggesting that sterols restrict the cell cycle to a certain extent in order to affect cell proliferation in stomatal development. The close relationship between sterols and the cell cycle has also been extensively reported in other systems, such as Drosophila, humans and mice (Duman-Scheel et al., 2002; Frietze et al., 2008; Shamma et al., 2009). Our findings suggest that sterol-mediated cell-cycle regulation generally exists not only in animals but also in plants, and is involved in cell proliferation and differentiation. However, understanding the detailed molecular mechanism by which sterols are involved in the cell cycle to affect stomatal development remains a challenge.

Experimental Procedures

  1. Top of page
  2. Summary
  3. Introduction
  4. Results
  5. Discussion
  6. Experimental Procedures
  7. Acknowledgements
  8. References
  9. Supporting Information

Plant materials and growth conditions

Arabidopsis thaliana Columbia0 (Col-0) was used as the wild-type. The mutants and transgenic plants used in the present study were as follows: E361, tmm–1, pTMM::TMM-GFP, flp–1, sdd1–1, mute, epf1–1, pMUTE::GFP, ice1–J3106, fama–1, spch–1, basl–2, pBASL::GFP-BASL, yda–Y295, epf2–1, pCYCB1;1::CYCB1;1-GUS, GFP–TUB6, sqe1–5/dry2, smt1/cph–G213, fk–X224, fk–J79, cpi1–1, cyp51A2–3, hyd1, smt2/cvp1–3, dwf5, det2–28, cpd, bri1–701 brl1 brl3 and bin2–1. Their source details are provided in Methods S1. Seedlings were germinated on half-strength MS medium (Murashige and Skoog, 1962) for 810 days, and transferred to soil for growth under 16 h light/8 h dark cycles at 2022°C.

Mapping and isolation of FK

The fk–J3158 mutation was isolated from a population of ethyl methane sulfonate-mutagenized M2 E361 seeds using dental resin impressions (Geisler et al., 2000), and was mapped to a 130 kb genomic region on chromosome three near the 20 Mb position. Thirty-five candidate open reading frames in the region were sequenced. Sequencing of At3 g52940 identified a G/A mutation. FK cDNA was fused with the 35S promoter in the pCAMBIA1300 vector and transformed into fk–J3158 and Col-0 for complemented and 35S::FK plants, respectively (Xiang et al., 2007). FK expression was detected in all transgenic plants, and the G/A mutation locus in the fk–J3158 background was sequenced. Primer sequences are shown in Table S4.

Imaging and microscopy analysis

Differential interference contrast images of the leaf epidermis were obtained using a Nikon ECLIPSE 80i microscope (http://www.nikon.com/). Fluorescence of GFP and propidium iodide signals was simultaneously captured using an Olympus FV1000MPE2 confocal microscope (http://www.olympus-global.com/en/). Scanning electron microscopy images were obtained using a Hitachi S3400N scanning electron microscope (http://www.hitachi.com/gateway.html). Time-lapse imaging was performed using a Leica DM6000CS confocal microscope (http://www.leica-camera.com/) and a Zeiss Observer Z1 AXIO fluorescence microscope (http://www.zeiss.com/). Detailed protocols, sample preparations, microscopy, and data analyses are available in Methods S2.

Sterol analysis

Sterol isolation, measurement and identification of fk–J3158 and Col-0 plants were performed as previously described (Babiychuk et al., 2008). Freeze-dried material was subjected to a saponification step. Non-saponifiable lipids were extracted using n–hexane, and the dried extract was acetylated. Steryl acetates were quantified by gas chromatography coupled to flame ionization detection using lupenyl-3,28diacetate as an internal standard. Structures of identified compounds were confirmed by mass spectrometry.

Treatment with sterols, BL and FEN

Col-0 and fk–J3158 seeds were germinated on sterol-, BL- or FEN-containing half-strength MS medium. Stigmasterol, campesterol and BL (Sigma, http://www.sigmaaldrich.com/) were dissolved in absolute ethanol, and added to the medium at 1, 5 or 10 μm for stigmasterol and campesterol, and 1, 5, 10 or 100 nm for BL. Final concentrations of 0.01, 0.1 or 1.0 μm FEN (Sigma) were prepared as described by Schrick et al. (2004a).

RNA–Seq analysis

Total RNA of Col-0, fk–J3158 and FEN-treated Col-0 rosette leaves were isolated from 15-day-old seedlings on half-strength MS plates. RNA integrity and concentration were evaluated using an Agilent 2100 Bioanalyzer (http://www.home.agilent.com/). Each RNA sample was submitted to Illumina Inc (http://www.cynergene.com/). for sequencing, and all data were analyzed at the Beijing Genomics Institute, China. Genes were considered significantly differently expressed if they had a P value <0.005, a false discovery rate (FDR) <0.01 and an estimated absolute log2 fold change >1 in sequence counts. We performed cluster analysis of gene expression patterns using Cluster software (http://bonsai.hgc.jp/~mdehoon/software/cluster/software.htm#ctv) and Java TreeView software (http://jtreeview.sourceforge.net/).

Construction of double mutants

The double mutants in the present study [fk–J3158 tmm–1, fk–J3158 sdd1–1, fk–J3158 flp–1, fk–J3158 ice1–J3106, fk–J3158 pTMM::TMM-GFP, fk–J3158 pCYCB1;1::CYCB1;1-GUS and fk–J3158 GFP-TUB6] were checked by DNA sequencing (for primer sequences, see Table S4). Single stomatal mutants were treated with 0.1 μm FEN on half-strength MS medium to mimic the double mutant phenotypes.

GUS staining assays

Five- to fifteen-day-old pCYCB1;1::CYCB1;1-GUS seedlings were placed directly into GUS reaction buffer (10 mm EDTA, 0.1% Triton X100 and 2 mm or 10 mm potassium ferri/ferrocyanide in 50 mm phosphate buffer, pH 7.0) plus 0.5 mg mL−1 Xglucuronic acid (Sigma). After vacuum treatment for 20 min, they were incubated overnight at 37°C. Stained tissues were cleared in 8:3:1 w/v/v chloral hydrate/glycerol/water, and detected using differential interference contrast microscopy.

RNA extraction and real-time RT–PCR

Total RNA was isolated from the leaves of 15-day-old seedlings or the 3rd or 4th rosette leaves of 24-day-old seedlings using a plant RNA kit (Omega, http://www.omegabiotek.com/) according to the manufacturer's instructions. One microgram of total RNA was used for reverse transcription using the PrimeScript RT reagent kit (TaKaRa, http://www.takara.com.cn/) according to the manufacturer's instructions. Real-time RTPCR was performed using SYBR Premix Ex Taq II (TaKaRa) and a Stratagene Mx3000P PCR instrument (http://www.stratagene.com) (for primer sequences, see Table S4).

Flow cytometric analyses

Mature rosette leaves of Col-0 and fk–J3158 seedlings growing on half-strength MS medium for 24 days were chopped with a razor blade, and 0.5 mL cold Galbraith buffer was added (Galbraith et al., 1983). The extracts were filtered through 60 and 40 mm filters, and the isolated nuclei were stained using 50 mg mL−1 propidium iodide for 45 min. Approximately 10 000 nuclei were measured, and DNA histograms were generated using a Beckman-Coulter Epics XL flow cytometer (http://www.bclifesciences.com).

Acknowledgements

  1. Top of page
  2. Summary
  3. Introduction
  4. Results
  5. Discussion
  6. Experimental Procedures
  7. Acknowledgements
  8. References
  9. Supporting Information

This work is supported by the National Natural Science Foundation of China (grant numbers 31070247, 91017002 and 31271460), the National Basic Research Program of China (grant number 2009CB941500), the Ministry of Agriculture of the People's Republic of China (grant numbers 2011ZX08009-003-002 and 2009ZX08009-029B), and Doctor Newcomer Award of the Ministry of Education of the People's Republic of China in 2010. We thank Alex Webb (University of Cambridge), Fred D. Sack (UBC), Thomas Altmann (MPI-MP), Keiko U. Torii ( University of Washington), Tatsuo Kakimoto (Osaka University), Dominique C. Bergmann (Stanford University), Peter Doerner (University of Edinburgh), Miguel A. Botella (University of Malaga), Kathrin Schrick (Kansas State University), Jyan-Chyun Jang (Ohio State University), Markus Grebe (Umeå University), Ho Bang Kim (Myongji University), Keith Lindsey (Durham University), Timothy Nelson (Yale University), Xuelu Wang (Fudan University) for providing seeds of mutants and transgenic plants. We are indebted to Xiaoping Gou, Liping Guan, Qingxiang Gao and Longfeng Yan for helpful discussion and technical support.

References

  1. Top of page
  2. Summary
  3. Introduction
  4. Results
  5. Discussion
  6. Experimental Procedures
  7. Acknowledgements
  8. References
  9. Supporting Information

Supporting Information

  1. Top of page
  2. Summary
  3. Introduction
  4. Results
  5. Discussion
  6. Experimental Procedures
  7. Acknowledgements
  8. References
  9. Supporting Information
FilenameFormatSizeDescription
tpj12190-sup-0001-SupportingInformation.docWord document10447K

Figure S1. Complementation of the fk–J3158 mutant phenotype by the wild-type gene.

Figure S2. Seedling phenotypes of fk–J3158, fk–J79 and fk–X224.

Figure S3. Phenotypes of leaf, vascular tissues, root and embryo of fk–J3158.

Figure S4. The sterol C14 reduction step, and sterols included in the present study.

Figure S5. The post-cycloeucalenol branch of the sterol biosynthetic pathway acts independently of the downstream SMT2/CVP1, DWF5 and BR pathways in stomatal development.

Figure S6. The sterol biosynthetic pathway acts independently of the stomatal development regulators EPF1, EPF2, YDA and BASL.

Figure S7. fk–J3158 stomatal lineage histories in various cases of cell division and differentiation.

Figure S8. Expression patterns of TUB6, TMM and BASL in fk–J3158 and FEN-treated Col-0.

Figure S9. RNASeq analysis of cell cycle-related gene expression in 15-day-old fk–J3158 and FEN-treated Col-0.

Table S1. Numbers of differentiated cells identified in fk allele mutants, FEN-treated Col-0, and other mutants of the early steps of the sterol biosynthetic pathway.

Table S2. Numbers of differentiated cells identified in Col-0 and fk–J3158 after treatment with 24epibrassinolide, stigmasterol or campesterol.

Table S3. Summary of defect phenotypes in mutants of the early steps of the sterol biosynthetic pathway in Arabidopsis.

Table S4. Primers used in this study.

Methods S1. Plant materials.

Methods S2. Imaging and microscopy analysis.

tpj12190-sup-0002-DataS1.xlsapplication/msexcel124KData S1. Clustering analysis of the union of differentially expressed genes in fk–J3158 and FEN-treated Col-0.
tpj12190-sup-0003-MovieS1.AVIvideo/avi6397KMovies S1S3. Time-lapse imaging of TUB6GFP in dividing small-cell clusters of fk–J3158.
tpj12190-sup-0004-MovieS2.AVIvideo/avi1952K 
tpj12190-sup-0005-MovieS3.AVIvideo/avi929K 

Please note: Wiley Blackwell is not responsible for the content or functionality of any supporting information supplied by the authors. Any queries (other than missing content) should be directed to the corresponding author for the article.