• sterols;
  • cellulose;
  • cell wall;
  • embryogenesis;
  • GC–MS;
  • Arabidopsis


  1. Top of page
  2. Summary
  3. Introduction
  4. Results
  5. Discussion
  6. Experimental procedures
  7. Acknowledgements
  8. References

A crucial role for sterols in plant growth and development is underscored by the identification of three Arabidopsis sterol biosynthesis mutants that exhibit embryonic defects: fackel (fk), hydra1 (hyd1), and sterol methyltransferase 1/cephalopod (smt1/cph). We have taken a dual approach of sterol profiling and ultrastructural analysis to investigate the primary defects underlying the mutant phenotypes. Comprehensive gas chromatography (GC)–MS analysis of hyd1 in comparison to fk reveals an abnormal accumulation of unique sterol intermediates in each case. Sterol profiling of the fk hyd1 double mutant provides genetic evidence that FK C-14 reductase acts upstream of HYD1 C-8,7 isomerase. Despite distinct differences in sterol profiles, fk and hyd1 as well as smt1/cph share ultrastructural features such as incomplete cell walls and aberrant cell wall thickenings in embryonic and post-embryonic tissues. The common defects are coupled with ectopic callose and lignin deposits. We show that all three mutants exhibit a deficiency in cellulose, but are not reduced in pectin and sugars of the cell wall and cytosol. The sterol biosynthesis inhibitors 15-azasterol and fenpropimorph also cause cell wall gaps in dividing root cells and a reduction in bulk cellulose, corroborating that the cell wall abnormalities are due to altered sterol composition. Our results demonstrate that sterols are crucial for cellulose synthesis in the building of the plant cell wall.


  1. Top of page
  2. Summary
  3. Introduction
  4. Results
  5. Discussion
  6. Experimental procedures
  7. Acknowledgements
  8. References

Sterols and their derivatives are essential for all eukaryotic organisms, whether synthesized de novo or taken up as nutrients, and they play multifaceted roles in cell biology and development (Edwards and Ericsson, 1999). As extremely stable components of biological membranes, their structures are highly conserved, characterized by a hydrophobic tetracyclic ring system, a monohydroxy modification at C-3 and a side chain at C-17 (Bloch, 1983). The plant sterol biosynthesis pathway generates a mixture of sterols unique to plants collectively called phytosterols (Benveniste, 1986; Ourisson, 1994). As many as 61 sterols and pentacyclic triterpenes are reported to occur in maize seedlings (Guo et al., 1995). In Arabidopsis, the major constituent is sitosterol, followed by campesterol, stigmasterol, and a variety of biosynthetic intermediates and other minor sterols (Hartmann, 1998; Patterson et al., 1993).

In contrast to animals, which contain a wide variety of steroid hormones but only one major sterol, cholesterol, a Δ5-sterol lacking alkylation of C-24, plants synthesize both methyl- and ethyl-modified sterols. Campesterol, a plant methylsterol, is the precursor to the C28-brassinosteroids (BRs), which function in post-embryonic growth and development (Bishop and Yokota, 2001; Fujioka and Yokota, 2003). The plant ethylsterols stigmasterol and sitosterol are thought to play important roles in plant membrane systems. Stigmasterol has been shown to modulate the activity of plasma membrane H(+)-ATPase (Grandmougin-Ferjani et al., 1997). Sitosterol, the predominant ethylsterol, seems to regulate the integrity of the plasma membrane in a more general way (Schuler et al., 1991), although a direct role as molecular primer for the synthesis of cellulose has recently been postulated by Peng et al. (2002). The function of other plant sterols is largely uncharacterized, and it is intriguing that sterol content varies according to developmental context. This is exemplified by cholesterol, which generally comprises a small fraction of the sterol content in plants, but has been found in extraordinary amounts (>50% sterol content) in meristematic apices (Hobbs et al., 1996) and leaf surface lipids (Noda et al., 1988).

Genetic evidence for the role of sterol molecules in embryonic development came from studying a distinct group of Arabidopsis pattern formation mutants (Mayer et al., 1991; reviewed by Schrick and Laux, 2001): fackel (fk) and two other non-allelic mutants (sterol methyltransferase 1/cephalopod (smt1/cph) and hydra1 (hyd1)) exhibit cell division and expansion defects, as well as patterning defects such as multiple shoot meristems in embryogenesis (Schrick et al., 2000, 2002). The finding that these mutants represent enzymatic lesions in sterol biosynthesis upstream of BRs, coupled with the inability to rescue their defects by exogenous application of BRs, led to the hypothesis that novel steroid molecules are crucial in development (reviewed by Clouse, 2000, 2002; Fujioka and Yokota, 2003; Schaller, 2003).

Figure 1 illustrates the enzymatic steps mediated by FK (C-14 reductase), HYD1 (C-8,7 isomerase), and SMT1/CPH (C-24 methyltransferase). The only sterol biosynthesis mutants identified thus far that exhibit embryonic defects act upstream of the SMT2 step: smt2/cotyledon vascular pattern 1 (cvp1) mutants have been shown to have post-embryonic, but not embryonic vascular patterning defects (Carland et al., 2002). Altered sterol composition in sterol biosynthesis mutants seems to affect various other biological processes relevant to growth and patterning. For example, auxin and ethylene signaling appear perturbed in fk (also called hyd2) and hyd1 mutants (Souter et al., 2002). Moreover, smt1/cph mutants display defects in cell polarity and localization of pin-formed (PIN) proteins, which are postulated to be auxin efflux carriers (Willemsen et al., 2003). cvp1, fk, and hyd1 mutants also exhibit changes that have been interpreted as defects in cell polarity, such as aberrant cell division plane orientation (Carland et al., 2002; Schrick et al., 2000) and altered epidermal morphology (Souter et al., 2002). FK gene transcription and thus sterol biosynthesis appear to be upregulated by several hormones, including auxin, brassinolide, gibberellin, cytokinin, and ethylene (He et al., 2003). However, the underlying mechanisms by which sterols influence cell polarity and other signaling pathways are not known. A current challenge toward elucidating the roles of sterols in plant cells is to characterize the molecular and ultrastructural defects caused by mutations in sterol biosynthesis genes.


Figure 1. Sterol biosynthesis in Arabidopsis.

The pathway from cycloartenol, the first cyclic sterol intermediate, is shown. Enzymes are depicted in bold and steps for which mutants have been identified are shown in italics. The SMT1/CPH (C-24 sterol methyltransferase; Diener et al., 2000; Schrick et al., 2002), FK (C-14 reductase; Jang et al., 2000; Schrick et al., 2000), and HYD1 (C-8,7 isomerase; Grebenok et al., 1998; Schrick et al., 2002; Souter et al., 2002) steps are highlighted by boxes. A branch in the pathway at the SMT2/CVP1 (C-28 methyltransferase) step (Carland et al., 2002; Schaeffer et al., 2001) leads to the major sterol end-products, sitosterol, stigmasterol (left), and campesterol, the precursor to the BRs (right). The end-product cholesterol is produced from cycloartenol via an independent pathway. Although the sterol biosynthesis pathway is depicted here as a simple linear pathway, genetic evidence and sterol profiles of the corresponding mutants suggest that a complex grid of enzymatic activities is realistic (Schrick et al., 2002).

Download figure to PowerPoint

A potentially fruitful approach toward understanding the primary defects in sterol biosynthesis mutants is to characterize the sterol composition in these mutants by gas chromatography (GC)–MS analysis. Both fk (Jang et al., 2000; Schrick et al., 2000) and smt1/cph (Diener et al., 2000; Schrick et al., 2002) sterol profiles have been analyzed: fk mutants accumulate Δ8,14 sterols and exhibit a reduction in both sitosterol and campesterol branch sterols, as expected from a block at the C-14 reductase step. In contrast, smt1/cph mutants accumulate both cycloartenol and cholesterol, but surprisingly are only reduced in sitosterol branch sterols. Thus, the sterol profiles of fk and smt1/cph mutants differ considerably despite their similar embryo phenotypes. Sterol extracts from hyd1 and fk plants appear to have similarly reduced levels of sitosterol and campesterol (Souter et al., 2002), but a comprehensive sterol profile for hyd1 has not been documented.

Here, we report complete sterol profiles for hyd1 in comparison to fk, which reveal the accumulation of a unique set of abnormal sterols in each mutant. Our analysis demonstrates that FK acts upstream of HYD1 in a linear portion of the sterol biosynthesis pathway. To characterize the fk, hyd1, and smt1/cph sterol biosynthesis mutants at the subcellular level, we used transmission electron microscopy (TEM) of cryofixed embryos and post-embryonic roots. These ultrastructural studies reveal a common defect in cell wall formation in all three sterol biosynthesis mutants. Consistent with a defect in cellulose synthesis as the cause of cell wall defects, the sterol biosynthesis mutants exhibit reduced levels of cellulose coupled with ectopic deposits of both callose and lignin. Arguing against a general defect in cell wall biogenesis, the mutants are not altered in other cell wall components such as pectin and neutral sugars. Treatment of wild-type seedlings with sterol biosynthesis inhibitors mimics the fk, hyd1, and smt1/cph mutant phenotypes. Our results provide evidence that sterol production is critical for both cell elongation and cell wall expansion during the building of the primary cell wall and raise the possibility that there is direct link between plasma membrane sterols and cellulose synthesis.


  1. Top of page
  2. Summary
  3. Introduction
  4. Results
  5. Discussion
  6. Experimental procedures
  7. Acknowledgements
  8. References

hyd1 and fk sterol biosynthesis mutants display distinct sterol profiles

A comprehensive GC–MS sterol analysis of mutants having specific lesions in sterol biosynthesis is necessary to elucidate the primary defects associated with their phenotypes. Souter et al. (2002) reported that hyd1 and fk mutants have similarly reduced levels of sitosterol and campesterol. However, their analysis was limited to the three predominant sterols: sitosterol, campesterol, and stigmasterol. In this study, our goal was to obtain a complete profile of sterol biosynthesis intermediates and end-products that accumulate or are lacking in hyd1 and fk in comparison to the wild-type.

We examined dedifferentiated hyd1, fk, and wild-type callus cells derived from mutant seedling material, propagated as previously described by Schrick et al. (2002). Callus cell cultures were chosen because they are expected to contain actively dividing cells. Consistent with a block at the C-14 reductase and C-8,7 isomerase steps, the levels of upstream sterols such as cycloartenol and 24-methylenecycloartanol were increased in both mutants relative to the wild-type (Table 1). An increase in the level of cholesterol in both mutants was anticipated under the assumption that cholesterol can be synthesized in a separate pathway from cycloartenol independently of C-14 reductase and C-8,7 isomerase (Figure 1). The levels of upstream cycloeucalenol and obtusifoliol were increased in fk, but not in hyd1.

Table 1.  Mass spectral analysis of sterols from fk and hyd1 mutants display distinct sterol profiles
  • Sterols determined by GC–MS are listed according to branches of the sterol biosynthesis pathway. Callus cell cultures were used for sterol extraction. The wild-type control was Ler and the mutant alleles were fk-X224 and hyd1-R216. nd, not detected.

  • a Values are given in μg g−1 FW.

  • b

    Percentage of the total sterol content is shown in parentheses.

Upstream branch
 Cycloartenol5.9a (5.3)b46 (19)16 (11)
 24-Methylenecycloartanol18 (16)40 (17)61 (40)
 Cycloeucalenol0.8 (0.7)2.3 (1.0)0.6 (0.4)
 Obtusifoliol0.1 (0.09)0.2 (0.08)0.1 (0.07)
 4α-Methyl-5α-ergosta-8,14,24(28)-trien-3β-ol0.01 (0.009)2.0 (0.8)0.03 (0.02)
Sitosterol branch
 Isofucosterol0.6 (0.5)ndnd
 Sitosterol51 (46)ndnd
 Sitostanol2.0 (1.8)0.01 (0.004)(0.03) (0.02)
 6-Oxositostanol0.4 (0.4)0.01 (0.004)(0.05) (0.03)
 Stigmasterol11 (9.9)ndnd
Campesterol branch
 Episterol0.01 (0.009)ndnd
 24-Methylenecholesterol0.4 (0.3)ndnd
 Campesterol19 (17)0.07 (0.03)0.08 (0.05)
 Campestanol0.8 (0.8)0.008 (0.003)0.01 (0.007)
 6-Oxocampestanol0.1 (0.1)0.004 (0.002)0.009 (0.006)
 Brassicasterol0.05 (0.05)ndnd
Cholesterol branch
 Cholesterol0.5 (0.5)6.8 (2.8)8.8 (5.8)
 Cholestanol0.04 (0.04)0.09 (0.04)0.09 (0.06)
 6-Oxocholestanol0.008 (0.007)0.004 (0.002)0.008 (0.005)
Abnormal sterols
 5α-Cholesta-8,14-dien-3β-olnd2.3 (1.0)0.02 (0.01)
 5α-Ergosta-8,14-dien-3β-olnd15 (6.2)0.2 (0.1)
 5α-Stigmasta-8,14-dien-3β-olnd89 (37)1.9 (1.3)
 4α-Methyl-5α-stigmasta-8-14,24(28)-trien-3β-olnd3.5 (1.5)nd
 Unknown ergosta-trien-3β-olnd2.2 (0.9)nd
 Unknown stigmasta-trien-3β-olnd31 (13)0.5 (0.3)
 Unknown ergosta-monoen-3β-olndnd4.5 (3.0)
 Unknown stigmasta-monoen-3β-olnd0.5 (0.2)58 (38)

In both mutants, appreciable levels of abnormal biosynthetic intermediates were detected. Characteristic Δ8,14 sterols (5α-cholesta-8,14-dien-3β-ol, 5α-ergosta-8,14-dien-3β-ol, and 5α-stigmasta-8,14-dien-3β-ol) accumulated in hyd1 and fk, but not in wild-type samples. However, the level of the Δ8,14 sterols was much lower in hyd1 mutants (1.4% of total sterols) in comparison to fk mutants (44.2% of total sterols), and fk mutants accumulated an additional biosynthetic intermediate, 4α-methyl-5α-stigmasta-8,14,24(28)-trien-3β-ol. Both mutants showed the accumulation of different abnormal sterol molecules of unresolved structure (Table 1, last four rows). In hyd1 cells, an unknown stigmasta-monoen-3β-ol accumulated to levels representing 38% of the total sterol content, but was only detected at a level of 0.2% of total sterols in fk. Conversely, an unknown stigmasta-trien-3β-ol accumulated at a level of 13% of the total sterols in fk, but accumulated only weakly in hyd1 cells (0.3% of total sterols).

Strikingly, in both mutants, the levels of downstream sterols of the sitosterol and campesterol branches were dramatically reduced. Members of the sitosterol branch that were not detected in the mutants included isofucosterol, sitosterol, and stigmasterol, while sitostanol and 6-oxositostanol showed a marked reduction relative to the wild-type. Among campesterol branch sterols that showed decreased levels were episterol, 24-methylenecholesterol, campestanol, 6-oxocampestanol, as well as campesterol, which was decreased from 17% in the wild-type to 0.05 and 0.03% in hyd1 and fk, respectively, suggesting that downstream BRs are either reduced or absent. Consistent with this presumption, Jang et al. (2000) found BR levels in fk-J79 plants to be significantly decreased, while the most potent BR brassinolide was not detected at all. Thus, in contrast to smt1/cph mutants in which campesterol levels are not reduced (Diener et al., 2000; Schrick et al., 2002), both hyd1 and fk can be viewed as BR biosynthesis mutants in addition to being sterol biosynthesis mutants.

Sterol analysis of hyd1 fk double mutants reveals that FK acts upstream of HYD1

The FK C-14 reductase step is thought to immediately precede the HYD1 C-8,7 isomerase step, although genetic evidence for this is not reported (Bach and Benveniste, 1997; Figure 1). It is striking that fk hyd1 double mutants are viable and exhibit seedling phenotypes that are indistinguishable from either single mutant, in contrast to double mutants between smt1/cph and hyd1 or fk, which display embryonic lethality (Schrick et al., 2002). As hyd1 and fk mutants have distinct sterol profiles (Table 1), we sought to determine whether the double mutant has a more fk- or hyd1-like sterol profile. Seedling progeny from a doubly heterozygous parent were grown on rich agar medium as previously described by Schrick et al. (2002). Whereas wild-type siblings had flowered within this time period, homozygous mutant progeny showed a dwarfed cabbage-like appearance devoid of inflorescences. Individual hyd1, fk, and fk hyd1 plants were genotyped by allele-specific PCR polymorphisms (Schrick et al., 2002), pooled according to genotype, and subjected to sterol analysis.

In Table 2, we present the sterol profile of fk hyd1 double mutant plants. The double mutant, like hyd1 and fk single mutants, accumulated the upstream sterol 24-methylenecycloartanol, while the levels of other upstream sterols such as cycloartenol or obtusifoliol appear unaltered. Like hyd1 and fk, the double mutant accumulated Δ8,14 sterols to appreciable levels (Tables 1 and 2). fk and fk hyd1 accumulated very high levels of Δ8,14 sterols (71 and 60% of total sterols, respectively), while hyd1 mutants accumulated lower levels (18% of total sterols). Moreover, fk and fk hyd1 accumulated a similar amount of 24-methylenecycloartanol (4 and 7% of total sterols, respectively) in comparison to hyd1 (17% of total sterols). In addition, sterol molecules of unresolved structure accumulated in all three mutants. An unknown ergosta-trien-3β-ol accumulated in fk and fk hyd1, but not in hyd1 mutants. Similarly, an unknown stigmasta-monoen-3β-ol accumulated in the double mutant at about the same level (23% of total sterols) as that in fk (14% of total sterols), whereas it accumulated to a higher level in hyd1 (39% of total sterols). Thus, the level of abnormal sterol accumulations in the double mutant more closely resembles that of fk.

Table 2.  Mass spectral analysis indicates similar sterol profile in fk single and fk hyd1 double mutants
SterolWild-typefkhyd1fk hyd1
  • Sterols determined by GC–MS are listed according to branches of the sterol biosynthesis pathway. Mature plants grown on rich growth medium (Schrick et al., 2002) were used for sterol extraction. The wild-type control was Columbia and the mutant alleles were fk-5D8 and hyd1-E508. nd, not detected.

  • a Values are given in μg g−1 FW.

  • b

    Percentage of the total sterol content is shown in parentheses.

Upstream branch
 Cycloartenol1.2a (0.7)b0.05 (0.2)0.03 (0.3)0.3 (0.6)
 24-Methylenecycloartanol0.7 (0.4)1.0 (3.7)1.6 (17)3.4 (7.0)
 Cycloeucalenol0.3 (0.2)0.3 (1.1)0.1 (1.0)0.3 (0.6)
 Obtusifoliol0.1 (0.06)0.05 (0.2)0.02 (0.2)0.05 (0.1)
 4α-Methyl-5α-ergosta-8,14,24(28)-trien-3β-ol0.02 (0.01)0.06 (0.2)0.01 (0.1)0.3 (0.6)
Sitosterol branch
 24-Ethylidenelophenol0.1 (0.06)ndndnd
 Avenasterol0.2 (0.1)ndndnd
 Isofucosterol4.8 (2.9)ndndnd
 Sitosterol105 (64)0.1 (0.4)0.1 (1.0)0.1 (0.2)
 Sitostanol5.4 (3.3)0.02 (0.08)0.05 (0.5)0.03 (0.06)
 6-Oxositostanol0.2 (0.2)0.04 (0.2)0.04 (0.4)0.04 (0.08)
 Stigmasterol7.8 (4.8)ndndnd
Campesterol branch
 24-Methylenelophenol0.1 (0.6)ndndnd
 Episterol0.1 (0.09)ndndnd
 24-Methylenecholesterol1.4 (0.9)0.03 (0.1)0.01 (0.1)0.02 (0.04)
 Campesterol28 (17)0.1 (0.4)0.1 (1.0)0.1 (0.2)
 Campestanol0.6 (0.4)0.01 (0.04)0.01 (0.1)0.01 (0.02)
 6-Oxocampestanol0.03 (0.02)0.01 (0.04)0.01 (0.1)0.01 (0.02)
 Brassicasterol2.4 (1.5)ndndnd
Cholesterol branch
 Cholesterol4.3 (2.6)1.0 (3.7)1.6 (17)1.4 (2.9)
 Cholestanol0.2 (0.1)0.01 (0.04)0.01 (0.1)0.01 (0.02)
 6-Oxocholestanol0.02 (0.01)0.01 (0.04)0.01 (0.1)0.01 (0.02)
Abnormal sterols
 5α-Cholesta-8,14-dien-3β-olnd0.08 (0.3)0.01 (0.1)0.2 (0.4)
 5α-Ergosta-8,14-dien-3β-olnd1.8 (6.7)0.2 (2.1)4.2 (8.6)
 5α-Stigmasta-8,14-dien-3β-olnd17 (64)1.5 (16)25 (51)
 Unknown ergosta-trien-3β-olnd0.02 (0.08)nd0.2 (0.4)
 Unknown stigmasta-trien-3β-olnd1.0 (3.7)0.03 (0.3)1.1 (2.3)
 Unknown ergosta-monoen-3β-olnd0.2 (0.8)0.4 (4.1)0.9 (1.8)
 Unknown stimasta-monoen-3β-olnd3.8 (14)3.7 (39)11 (23)

Sterols downstream of the sitosterol branch were reduced in the fk hyd1 double mutant as they were in both single mutants. Whereas the wild-type contained 64% sitosterol, the hyd1, fk, and fk hyd1 mutants contained 1, 0.4, and 0.2% sitosterol, respectively. Moreover, sitosterol branch biosynthetic intermediates 24-ethylidenelophenol, avenasterol, isofucosterol as well as the end-product stigmasterol were detected in wild-type, but not in the mutants. Similarly, downstream sterols from the campesterol branch of the biosynthesis pathway appeared equally reduced in all three mutants. The biosynthetic intermediates 24-methylenelophenol and episterol were detected in the wild-type, but not in any of the mutants, consistent with a deficiency in BRs.

The total amounts of cholesterol, cholestanol, and 6-oxocholestanol appeared similarly reduced in fk and the double mutant. For example, the percentage of cholesterol was increased in hyd1 (17% of total sterols) relative to the wild-type (about 3% of total sterols), but not in fk or the double mutant (4 and 3% of total sterols, respectively). Taken together, our data show that the sterol profile of the fk hyd1 double mutant closely approximates that of fk, consistent with the model of the sterol biosynthesis pathway placing FK upstream of HYD1 (Figure 1).

Growth and patterning defects of sterol biosynthesis mutants are phenocopied by 15-azasterol and fenpropimorph

Chemical inhibitors are powerful tools that can be used to verify and extend genetic data. In this study, we examined the degree to which two inhibitors of sterol biosynthesis are able mimic mutations in FK, HYD1, and SMT1/CPH. The antimycotic agent 15-aza-24-methylene-Δ-homocholesta-8,14-dien-3β-ol (15-azasterol) has been shown to be a strong specific inhibitor of sterol C-14 reductase in plant cell cultures (Schmitt et al., 1980). fk mutants are insensitive to the application of 15-azasterol, consistent with the idea that the FK sterol C-14 reductase is a specific target of the drug (Schrick et al., 2002). In contrast to 15-azasterol, the fungicide fenpropimorph represents a broad-range sterol biosynthesis inhibitor that is reported to inhibit at least four enzymes: cycloeulenol-obtusfoliol isomerase, C-14 reductase, C-8,7 isomerase, and Δ7-reductase (see Figure 1; reviewed by Mercer, 1993). He et al. (2003) recently reported sterol profiles for fenpropimorph-treated seedlings, which demonstrate that fenpropimorph is a potent dose-dependent inhibitor of Arabidopsis C-14 reductase.

Our data indicate that 15-azasterol and fenpropimorph similarly inhibit hypocotyl and root growth of germinating Arabidopsis seedlings at a concentration of 1 µm (Figure 2a). Compared to the untreated control (Figure 2c), inhibitor-treated seedlings (Figure 2d,e) display an overall dwarfed appearance. Like fk, hyd1, and smt1/cph mutant seedlings (Figure 2b; for fk see Figure 2i), inhibitor-treated seedlings display malformed leaf tissue (Figure 2f–h) and patterning defects, as exemplified by abnormal trichome nests or multibranched trichomes in the epidermis (Figure 2j,k).


Figure 2. Comparison of sterol biosynthesis mutants with fenpropimorph- and 15-azasterol-treated seedlings.

(a) Wild-type seedlings germinated for 14 days in the presence of 0 µm, 1.0 µm 15-azasterol (AZA), or 1.0 µm fenpropimorph (FEN) were measured for the total hypocotyl plus root length.

(b–k) Typical seedling phenotypes of fk, hyd1, and smt1/cph sterol biosynthesis mutants grown on rich media for 14 days are shown. (b) fk and hyd1 exhibit reduced stature and a cabbage-like appearance, whereas smt1/cph shows a comparatively weaker phenotype. In comparison to (c) untreated seedlings, both (d) AZA- and (e) FEN-treated seedlings display an overall dwarfed appearance, a short root and hypocotyl, as well as reduced outgrowth of apical structures. (f,g) AZA- and (h) FEN-treated seedlings exhibit deformed leaves at the seedling stage similar to that observed in (b) fk, hyd1, and smt1/cph. The mis-specification of trichome cells, such as cluster or nest formation in fk mutants (i) (arrow) was also observed in sterol biosynthesis inhibitor-treated seedlings. Examples of (j) trichome nests and (k) multibranched trichomes are shown from FEN-treated seedlings (arrows). Bars = 0.5 mm.

Download figure to PowerPoint

Sterol biosynthesis mutants display distinctive cell wall defects in the embryo

All three sterol biosynthesis mutants, fk, hyd1, and smt1/cph, display obvious cell expansion and division defects visible by light microscopy after the globular stage of embryogenesis (Schrick et al., 2000, 2002). To examine the nature of these defects in more detail, we used cryofixation (high-pressure freezing) and freeze-substitution for efficient ultrastructural preservation of embryonic plant cells. Ovule sections from plants heterozygous for fk, hyd1, and smt1/cph were examined by TEM. Mutant embryos were identified by their aberrant cell morphologies. While globular stage embryos showed relatively few aberrations (Figure 3a), later stages typically showed more pronounced deviations (Figure 3b) from the ordered cell morphologies in wild-type embryos (Figure 3c). The most obvious defect observed in embryos from all three mutants was the appearance of incomplete cell walls (Figure 3b,d–i). Cell wall stubs examined in serial sections appeared well-rounded at their boundary (Figure 3g), arguing against the possibility that the cell wall stubs were an artifact of mechanical stress. A wide spectrum of atypical cell wall phenotypes was visualized in the mutants although the frequency of cell wall gaps was variable. Abnormal cell wall thickening and multinucleate cells were also observed (Figure 3i). The multinucleate phenotype appears to be an indirect consequence of the inability to complete cell wall formation. In addition, we observed abnormally high numbers and variable sizes of vacuoles in the embryonic mutant cells, as well as rare dead cells.


Figure 3. Embryonic cell wall defects in sterol biosynthesis mutants.

TEM of ultrathin sections from cryofixed and freeze-substituted wild-type and sterol biosynthesis mutants. Overviews of embryos from (a) hyd1 and (b) smt1/cph exhibit disordered cell morphologies and incomplete cell walls (arrows) compared to the normal cell morphologies of the wild-type at (c) the heart stage. Details from (d) fk, (e,f) hyd1, and (g,h) smt1/cph mutant embryos show cell wall thickenings (asterisk) and incomplete cell walls (arrows). Detail of (i) a multinucleate cell with three nuclei (N) from a smt1/cph embryo. Bars = 20 µm in (a–c) and 5 µm in (d–i).

Download figure to PowerPoint

While the embryonic cell wall defects were easily visualized, we did not observe any obvious defects in nuclear, endoplasmic reticulum (ER), or Golgi membranes or compartments. It is reported that the plasma membrane, in comparison to other plant membrane systems, contains the greatest sterol content in proportion to protein amount coupled with the highest sterol to phospholipid molar ratio (Hartmann and Benveniste, 1987). The finding that sterol biosynthesis mutants display specific cell wall defects support the idea that sterol composition of the plasma membrane is integral to cell wall synthesis and function.

Post-embryonic cell wall defects in the roots of sterol biosynthesis mutants and inhibitor-treated seedlings

To examine the effect of the two different sterol biosynthesis inhibitors on cell wall formation, we germinated wild-type seedlings on media containing either 1 µm 15-azasterol or 1 µm fenpropimorph, and subjected their roots to subcellular analysis by TEM. No apparent difference between 15-azasterol- and fenpropimorph-treated roots were observed at the cellular level. The morphologies of inhibitor-treated roots were compared to roots from the untreated wild-type and from fk, hyd1, and smt1/cph sterol biosynthesis mutants. In contrast to the elongated differentiated cells in wild-type root tip (Figure 4a), the mutants (Figure 4b–d) and inhibitor-treated seedlings (Figure 4e,f) exhibited cell elongation defects, abnormal cell morphologies of variable sizes and highly disordered rows of cells. The root epidermal cells typically formed an irregular periphery in comparison to the smooth epidermal layer in the wild-type. The most obvious cellular defect in the mutants and inhibitor-treated seedlings was the abnormal appearance of incomplete cell walls (Figure 4g,i,j) and cell wall thickenings (Figure 4g,h), similar to that observed in fk, hyd1, and smt1/cph embryos. Thus, the sterol biosynthesis inhibitors 15-azasterol and fenpropimorph, which both seem to target FK C-14 reductase (He et al., 2003; Schrick et al., 2002), phenocopy the cell wall abnormalities found in fk, hyd1, and smt1/cph mutants. The results indicate that sterols are crucial for cell wall formation in dividing cells during both embryonic and post-embryonic development.


Figure 4. Cell wall defects in roots from sterol biosynthesis mutants and inhibitor-treated seedlings.

TEM of ultrathin sections of 14-day-old roots from wild-type and mutant seedlings in comparison to sterol biosynthesis inhibitor-treated seedlings. Root tips from (a) the wild-type display ordered rows of cells in contrast to those from the sterol biosynthesis mutants (b) smt1/cph, (c) hyd1, and (d) fk, which show altered morphologies among all cell types. Highly vacuolated cells were frequently observed in the mutants although the defects were variable. Wild-type seedlings germinated on (e) 1 µm 15-azasterol or (f) 1 µm fenpropimorph exhibit altered morphologies in root tip cells similar to those of fk, smt1/cph, and hyd1. Details of (g) fk and (h) hyd1 root cells showing incomplete cell walls (arrows) and aberrant cell wall thickenings (asterisk). (i) Higher magnification of smt1/cph root cells, with a multinucleate cell (N denotes nuclei) and an incomplete cell wall (arrow). Similar cellular defects were observed when wild-type seedlings were germinated on medium containing 1 µm 15-azasterol or fenpropimorph. A field of root cells from (j) 15-azasterol- and (k) fenpropimorph-treated seedlings showing incomplete cell walls (arrows) and binucleate cell. Detail of (l) a fenpropimorph-treated root with a nucleus confined between a cell wall gap. Bars = 20 µm in (a–f) and 5 µm in (g–l).

Download figure to PowerPoint

Sterol biosynthesis mutants and inhibitor-treated seedlings exhibit reduced levels of cellulose but show normal pectin and sugar profiles

Cellulose microfibrils constitute the principal scaffolding components of the cell wall. Incomplete cell walls and cell expansion defects in fk, hyd1, and smt1/cph mutants and inhibitor-treated seedlings could be explained by an impairment of cellulose synthesis. To investigate this possibility, we measured the cellulose fraction of cell wall material from seedlings using a quantitative colorimetric assay (Updegraff, 1969). As the sterol biosynthesis mutant seedlings typically exhibit very short roots, we examined wild-type seedlings both with and without roots as controls. Figure 5(a) shows that the levels of bulk cellulose in the wild-type with roots versus those without roots were not significantly different. In contrast, we observed that cellulose levels were consistently reduced from approximately 50% of the wild-type level in fk, to 60 and 70% in hyd1 and smt1/cph mutants, respectively (Figure 5a). Moreover, we confirmed that 15-azasterol- and fenpropimorph-treated seedlings exhibit significantly reduced levels of bulk cellulose in comparison to the wild-type (Figure 5a). The level of reduction is about 60% of the wild-type level, which is similar to that observed for the sterol biosynthesis mutants.


Figure 5. Sterol biosynthesis mutants and inhibitor-treated seedlings exhibit reduced levels of cellulose, but are not altered in galacturonic acids, cell wall sugars, or free sugars.

(a) Amount of cellulose (µg cellulose mg−1 DW) in wild-type seedlings without root (WT-(nrt)), with root (WT) versus sterol biosynthesis mutant seedlings fk, hyd1, smt1/cph, and WT seedlings treated with 1 µm 15-azasterol (AZA) and 1 µm fenpropimorph (FEN), respectively, and dwf1 mutants.

(b) Amount of pectin (µg uronic acids mg−1 DW) in WT seedlings (WT-(nrt) and WT) as compared to the sterol biosynthesis mutants (fk, hyd1, and smt1/cph), WT inhibitor-treated seedlings (AZA, FEN), and dwf1 mutant seedlings.

(c) Amount of neutral sugars (rhamnose (rha), fucose (fuc), arabinose (ara), xylose (xyl), mannose (man), and galactose (gal); µg sugar mg−1 DW) are shown in each case as in (b).

(d) Percentage glucose (glc), fructose (fru), and sucrose (suc) of the total free sugars is shown for WT seedlings (WT-(nrt) and WT) as compared to fk, hyd1, smt1/cph, and WT inhibitor-treated seedlings (AZA, FEN). SDs for (a) and (b–d) represent three independent seedling samples measured thrice each, and three to five independent measurements, respectively.

Seedlings were grown for 14 days on rich media.

Download figure to PowerPoint

All three sterol biosynthesis mutants analyzed are predicted to act in the same upstream branch of the sterol biosynthesis pathway (Figure 1). We examined whether a downstream lesion in the sterol C-24 reductase step (DWARF1 (DWF1)/DIMINURO (DIM)), which results in cell elongation defects (Choe et al., 1999; Klahre et al., 1998), results in a cellulose deficiency. Figure 5(a) shows that the corresponding BR-deficient dwarf mutant dwf1, in contrast to the sterol biosynthesis mutants fk, hyd1, and smt1/cph, displays wild-type levels of cellulose.

A decrease in cellulose levels may reflect a general reduction in cell wall biogenesis in the mutants and inhibitor-treated seedlings. To examine this possibility, we measured pectin and neutral sugar amounts in cell wall fractions from fk, hyd1, smt1/cph, and inhibitor-treated seedlings. Pectin is a major component of the primary cell wall that encompasses a range of galacturonic acid-rich polysaccharides. We quantified cell wall pectin using a colorimetric assay that directly measures uronic acid levels. Figure 5(b) shows that uronic acid levels are not significantly altered in the mutants and inhibitor-treated seedlings relative to wild-type controls. Using GC, we quantified the levels of an assortment of neutral sugars (rhamnose, fucose, arabinose, xylose, mannose, galactose) from cell wall material of the sterol biosynthesis mutants, and inhibitor-treated seedlings relative to the wild-type (Figure 5c). We detected a slight reduction in arabinose levels in the wild-type sample without roots in comparison to the wild-type control with roots. However, there was no significant reduction of any cell wall monosaccharide in the mutants or inhibitor-treated seedlings relative to the wild-type. In contrast, we observed what appears to be a slight increase in the level of arabinose in fk, hyd1, smt1/cph, or the inhibitor-treated seedlings relative to the wild-type. However, this increase is also present in the dwf1 mutant seedlings, which display normal levels of cellulose. Thus, the cellulose deficiency present in the sterol biosynthesis mutants and inhibitor-treated seedlings cannot be correlated with a change in other cell wall components.

As sterols may regulate membrane-associated metabolic processes, we explored the possibility that sterol composition affects cellulose synthesis indirectly by altering the levels of free glucose. We used HPLC analysis to examine the levels of soluble sugars in fk, hyd1, and smt1/cph, and 15-azasterol- or fenpropimorph-inhibitor-treated seedlings (Figure 5d). The sugar profile, comprised glucose, fructose, and sucrose of the wild-type without roots is indistinguishable from that of the mutants or inhibitor-treated seedlings. Thus, we were unable to correlate a cellulose synthesis defect with a change in sugar profiles. The data argue against the possibility that the deficiency in bulk cellulose arises simply from a deficiency in freely available glucose levels.

fk, hyd1, and smt1/cph embryos accumulate ectopic callose

In Arabidopsis, callose serves an intermediate in primary cell wall synthesis, and it is replaced by cellulose during cell wall maturation. Callose is not normally found in the primary cell wall, although it rapidly accumulates in response to wounding or mechanical stress (Aist, 1976). Abnormal callose deposits are reported for two cellulose-deficient mutants: cyt (cytokinesis-defective)1 (Nickle and Meinke, 1998) and kobito1 (kob1; Pagant et al., 2002). The cellulose synthase inhibitor dichlorobenil results in radial swelling, incomplete cell walls and the accumulation of callose (Nickle and Meinke, 1998). Furthermore, tunicamycin, an inhibitor of N-glycosylation, also causes ectopic accumulation of callose in drug-treated root tips, pointing to the importance on N-glycosylation for cellulose synthesis (Lukowitz et al., 2001). Thus, the appearance of ectopic callose appears to be correlated with cellulose deficiency.

We examined fk, hyd1, and smt1/cph mutant embryos for callose deposition. Embryos dissected from plants heterozygous for the sterol biosynthesis mutants were probed for callose using a simple aniline blue staining protocol. In contrast to their wild-type siblings, fk, hyd1, and smt1/cph mutant embryos were found to exhibit ectopic deposits of callose at the bent-cotyledon stage of embryogenesis (Figure 6). The callose deposits were typically localized to small patches representing finite groups of cells (Figure 6f,l), suggesting local spreading effects. As we did not observe any particular zone of the embryo with more patches than another, the abnormal callose deposits appear to be randomly distributed in different embryos. Perhaps cell wall thickenings that are visible by TEM contain callose deposits. Our data indicate that a reduction in bulk cellulose in the sterol biosynthesis mutants is coupled with an abnormal appearance of callose during embryonic development.


Figure 6. Sterol biosynthesis mutants display ectopic callose in the embryo.

Nomarski images of (a) an aniline blue-stained wild-type embryo at the bent-cotyledon stage and same stage embryos of (c,e) fk, (g,i,k) hyd1, and (m,o,q) smt1/cph mutants are shown to the left of and adjacent to (b,d,f,h,j,l,n,p,r) matched UV fluorescences. Patches of fluorescent staining (arrows) in the mutant versus those in the wild-type indicate the abnormal accumulation of callose. The staining appears randomly distributed in different mutant embryos. Localized patches of fluorescence, representing small clusters of cells, are frequently observed (arrows in (f,l)). Examples of (h,j,p,r) polar staining appear to be more prevalent than (d,n) uniformly distributed staining. Bars = 0.5 mm.

Download figure to PowerPoint

Ectopic lignification in sterol biosynthesis mutants and inhibitor-treated seedlings

The accumulation of ectopic lignin as well as callose occurs in wild-type seedlings treated with dichlorobenil or isoxaben, which are both potent cellulose synthesis blockers (Desprez et al., 2002). Abnormal lignin deposits are also reported in kob1 mutants, which are deficient in cellulose (Pagant et al., 2002). Using a phloroglucinol–HCl solution, we probed the accumulation of lignin in the wild-type and in fk, hyd1, and smt1/cph sterol biosynthesis mutant seedlings. A patchy red staining that was detected in the mutants (Figure 7c–h), but not in the wild-type (Figure 7a,b) indicated the presence of ectopic lignification. The lignification appeared to occur at random positions in both the root and in aerial tissues and was not observed in all parts of these tissues. No red staining was detected in dwf1 mutant seedlings (Figure 7l), consistent with the postulate that the defects in cell wall biogenesis are specific to the upstream sterol biosynthesis mutants. Similar to the fk, hyd1, and smt1/cph mutants, wild-type seedlings germinated in the presence of 15-azasterol (Figure 7i,j) or fenpropimorph (Figure 7k) displayed ectopic lignification. Taken together, our observations support the idea that inhibitors of sterol biosynthesis, like genetic lesions in the upstream steps of sterol biosynthesis, lead to ectopic lignification as a result of compromised cellulose synthesis.


Figure 7. Sterol biosynthesis mutants and inhibitor-treated seedlings accumulate ectopic lignin.

Phloroglucinol-stained (a) wild-type and (l) dwf1 seedlings exhibit weak or no red staining, indicating that very little or no lignin is present during normal seedling growth. In contrast, staining of (c) hyd1, (e,f) fk, and (g,h) smt1/cph mutant seedlings, as well as (i,j) 15-azasterol- and (k) fenpropimorph-treated seedlings resulted in distinct red patches, indicating the presence of ectopic lignin in the root, hypocotyl and/or apical tissues. Details of (b) a wild-type root in comparison to (d) a hyd1 root show a small patch of phloroglucinol-stained lignin (red) in the mutant, but not in the wild-type. Scale bars = 0.5 mm.

Download figure to PowerPoint


  1. Top of page
  2. Summary
  3. Introduction
  4. Results
  5. Discussion
  6. Experimental procedures
  7. Acknowledgements
  8. References

Steroid molecules are ubiquitous to eukaryotic organisms and play a fundamental role in cell and developmental biology. In plants, a special class of steroids, BRs, act as hormones to promote post-embryonic growth. The identification of BR dwarf mutants uncovered various BR-mediated developmental processes such as cell expansion and proliferation, differentiation of vascular tissues, and photomorphogenesis (reviewed by Altmann, 1999; Bishop and Yokota, 2001). Analogously, the characterization of a set of three sterol biosynthesis mutants that act upstream of the BRs, fk (Jang et al., 2000; Schrick et al., 2000), hyd1 (Schrick et al., 2002; Souter et al., 2002), and smt1/cph (Diener et al., 2000; Schrick et al., 2002), has revealed the importance of sterol composition for both embryonic and post-embryonic development. In the present study, we apply two separate approaches to further explore the role of sterols in plant cells: GC–MS sterol profiling and ultrastructural analysis of dividing cells from sterol biosynthesis mutants.

Differences and commonalities in fk, hyd1, and smt1/cph sterol profiles

We present detailed sterol profiles of fk and hyd1 mutants, which point to the importance of the corresponding enzymes, sterol C-14 reductase and C-8,7 isomerase, respectively, for sterol biosynthesis (Tables 1 and 2). A previous study (Souter et al., 2002) showed that hyd1 and fk mutants are similarly reduced in sitosterol and campesterol, consistent with our results. Extending these findings, our data show a clear distinction between fk and hyd1 sterol profiles. This is exemplified by the strong accumulation of an unknown stigmasta-monoen-3β-ol (38% of total sterols) in hyd1 versus a very weak accumulation (0.2% of total sterols) in fk callus cells (Table 1). The stigmasta-monoen-3β-ol is likely to be a modified form of 4α-methylfecosterol, the intermediate that is expected to accumulate by a block at the C-8,7 isomerase step. Conversely, the abnormal accumulation of Δ8,14 sterols is elevated in fk cells (44% of total sterols) relative to hyd1 (1.4% of total sterols). Remarkably, null mutations in FK or HYD1 do not completely block sterol biosynthesis, although both represent single genes in Arabidopsis. The finding that the fk hyd1 double mutant, like both fk and hyd1 single mutants, produces a residual amount of downstream products strongly supports the contention that a C-14 reductase- and C-8,7 isomerase-independent pathway can function to produce sterols when FK, HYD1, or both are absent. As the sterol profile of the fk hyd1 double mutant is more similar to that of fk than to that of hyd1 (Table 2), we conclude that FK sterol C-14 reductase indeed acts upstream of HYD1 C-8,7 isomerase. The two enzymes, which are predicted to be integral membrane proteins of the endoplasmic reticulum, are postulated to physically interact (Schrick et al., 2002).

Can the phenotypic defects in fk, hyd1, and smt1/cph be correlated with sterol composition? Sterol profiles reveal clear differences in each mutant. For example, whereas fk and hyd1 accumulate high levels of Δ8,14 sterols or an unknown stigmasta-monoen-3β-ol, respectively (Table 1), smt1/cph mutants show aberrantly high proportion of cycloartenol and cholesterol (Schrick et al., 2002). The observations argue against specific effects of altered sterol compositions. Perhaps the mutant phenotypes result from a non-specific toxic accumulation of abnormal sterol intermediates. Regarding sterol end-products, fk and hyd1 are reduced in both sitosterol and campesterol branch sterols, whereas smt1/cph appears to be reduced in only sitosterol branch sterols. Thus, all three mutants display a common reduction in sitosterol. In smt1/cph, the sitosterol deficiency is not as severe (30–50% of the wild-type; Schrick et al., 2002) as that in fk and hyd1, both of which had only trace amounts of sitosterol in our analysis. A potential role for sitosterol in cell wall formation, based on a postulate by Peng et al. (2002), is discussed below. Additionally, other steroid products could contribute to the mutant phenotypes. The GC–MS detection method may lack the sensitivity to monitor the absence or accumulation of important low-abundant sterols that could be affected in the mutants.

Sterol composition is critical for cell wall formation and cellulose synthesis

Our ultrastructural studies reveal a common set of cell wall defects in dividing embryonic and root cells of fk, hyd1, and smt1/cph mutants. These defects were linked to a specific reduction in cellulose coupled with ectopic callose and lignin deposits. Although the reduction in bulk cellulose is not dramatic (50–70% of the wild-type level), it is striking that a similar decrease is reported for other cellulose-deficient mutants. For example, cobra1 (cob1) mutants, which are defective in a predicted glucosylphosphatidyl (GPI)-anchored protein exhibit a reduction of cellulose to 67% of the wild-type level (Schindelman et al., 2001). Similarly, kob1 mutants, which are defective in a novel plasma membrane protein, have reduced cellulose levels that are also approximately 67% of the wild-type level (Pagant et al., 2002).

Cellulose synthesis is mediated by a plasma membrane-bound hexameric protein complex comprised of multiple cellulose synthase catalytic subunits (CESAs) encoded by 10 genes in Arabidopsis (Richmond and Somerville, 2000). Strong mutant alleles of CESA1 (Gillmor et al., 2002) and CESA6 (Fagard et al., 2000) exhibit cell expansion defects. The reported cellulose deficiencies for CESA1 lesions range from a reduction of cellulose content to 50% of the wild-type level (Arioli et al., 1998), to a reduction of about fourfold (Gillmor et al., 2002). A temperature-sensitive mutation in KORRIGAN (KOR), a gene encoding a putative endo-1,4-β-glucanase that is believed to interact with cellulose synthase, results in a similar reduction in cellulose (40% of the wild-type level; Sato et al., 2001).

It is noteworthy that the severity of the cellulose deficiency does not necessarily translate to a direct effect on the catalytic subunits of cellulose synthase. For instance, an extreme cellulose reduction of sevenfold from the wild-type level is reported for knopf (knf) embryos (Gillmor et al., 2002). KNF encodes an α-glucosidase I that is thought to indirectly affect cellulose synthesis by catalyzing the first step in N-glycan trimming on proteins that promote or regulate cellulose synthesis (Boisson et al., 2001). Similarly, the mannose-1-phosphate guanylyltransferase-deficient cyt1 mutant, which exhibits embryonic lethality, displays a dramatic reduction in cellulose content of about fivefold (Lukowitz et al., 2001; Nickle and Meinke, 1998). The CYT1 enzyme is required for N-glycosylation, a function that also appears to influence cellulose production in an indirect manner.

Other embryonic mutants that are defective in cytokinesis, such as knolle (kn) and keule (keu), which do not correspond sterol biosynthesis lesions, are not reduced in cellulose accumulation (W. Lukowitz, C. S. Gillmor, and C. Somerville, personal communication), and are reported to exhibit minimal callose accumulation in the embryo (Nickle and Meinke, 1998). In the present study, we demonstrate that BR dwf1 mutants, which show cell elongation defects (Choe et al., 1999; Klahre et al., 1998), do not exhibit a cellulose deficiency. Moreover, in contrast to fk, hyd1, and smt1/cph mutants and sterol biosynthesis inhibitor-treated seedlings, dwf1 mutants lack ectopic lignification. These observations further support the idea that lesions in upstream sterol biosynthesis genes, such as FK, HYD1, and SMT1/CPH, and not those affecting later steps in the pathway, display distinct phenotypes that cannot be explained simply by a reduction in BRs (Clouse, 2000, 2002; Fujioka and Yokota, 2003).

Sterols indirectly affect cellulose synthesis?

As major constituents of the plasma membrane, sterols may influence the stability and function of membrane-bound enzymes involved in cellulose synthesis. Lipid rafts act as highly dynamic microdomains of sphingolipids and cholesterol in animal cells (reviewed by Simons and Toomre, 2000), and in plants may contain specific types of sterols, some of which promote cellulose synthesis. It is striking that the level of cellulose deficiency in cob mutants (Schindelman et al., 2001) is quite similar to that of fk, hyd1, and smt1/cph sterol biosynthesis mutants. GPI-anchored proteins related to COB have been identified from tobacco membrane microdomains that have properties akin to mammalian lipid rafts (Peskan et al., 2000). The COB protein is asymmetrically localized in polarized root cells (Schindelman et al., 2001) and could promote oriented cell expansion via recruitment of enzyme machinery for cellulose deposition through interactions with local sterol environments. It would be interesting to clarify whether cellulose machinery components are properly expressed and functional in sterol biosynthesis mutants using immuno-precipitation and localization experiments.

In addition, it is possible that sterol composition affects the expression of cellulose synthesis machinery at the transcriptional level. However, He et al. (2003) showed that the endo-1,4-β-glucanase gene KOR, which is thought to play an intimate role in cellulose synthesis (Peng et al., 2002), exhibits normal transcript levels in fk seedlings. Yet, the transcript levels of CESA components in sterol biosynthesis mutants have not been reported. Whereas transcription or some other aspect of signal transduction may affect cellulose indirectly, below we discuss various scenarios for a direct role of sterols in cellulose synthesis (Figure 8).


Figure 8. Model for the requirement of sterols in the construction of the plant cell wall.

In this model, sterols promote the formation of cellulose microfibrils, which are essential building materials for both cell wall elongation and expansion. Sterols are found in lipid raft microdomains, in the enveloping plasma membrane or as free sterols in intracellular membrane systems. Lipid raft and/or plasma membrane sterols may positively influence the membrane environment, conformation, and activity of the cellulose synthesis machinery. Alternatively or in addition, glucoside-conjugated sterols (such as sitosterol-β-glucosides), may directly prime cellulose synthesis in the formation of sterol-cellodextrin conjugates, as proposed by Peng et al. (2002). Free sterols residing in other membrane compartments may also affect cellulose synthesis or the formation of cellulose microfibrils indirectly via signal transduction mechanisms.

Download figure to PowerPoint

A direct role for sitosterol in cellulose synthesis?

Sitosterol-β-glucoside, a glucose (glc)-conjugated form of sitosterol, is postulated to be a primer for the initiation of cellulose synthesis (Peng et al., 2002). Consistent with this idea, sitosterol-β-glucoside is synthesized on the inner face of the plasma membrane (Cantatore et al., 2000) where cellulose is made. Sitosterol–cellodextrin, which carries a chain of glc molecules, is hypothesized to serve as a critical intermediate in the assembly of cellulose microfibrils. Thus, mutants that are defective in sitosterol production, such as fk, hyd1, and smt1/cph, are predicted to have reduced levels of sitosterol-β-glucosides, sitosterol–cellodextrin, and cellulose microfibrils. However, there does not appear to be a direct correlation between sitosterol and cellulose levels among the mutants examined. While sitosterol production is affected more severely in fk and hyd1 than in smt1/cph, the cellulose deficiencies are comparable in all three mutants. It would be interesting to examine whether smt2/cvp1 mutants, which show a similar reduction in sitosterol as smt1/cph mutants (30–50% of wild-type; Carland et al., 2002), also exhibit a cellulose deficiency. Our results show that dwf1 mutants, which are clearly defective in sitosterol production (Klahre et al., 1998), are not deficient in cellulose. Perhaps other sterols produced in dwf1 mutants are able to substitute for the lack of sitosterol. Alternatively, sterols in addition to sitosterol may have in vivo roles in cellulose synthesis as described below.

Peng et al. (2002) reported that the herbicide 2,6-dichlorobenzonitrile (DCB), a cellulose synthesis inhibitor, blocks the synthesis of sitosterol-β-glucosides in cotton. A simple interpretation of this result is that DCB inhibits cellulose synthesis by disabling UDP-glc:sterol glucosyltransferase, the enzyme that transfers glc monomers to sitosterol. However, DCB may inhibit additional essential functions unrelated to the production of sitosterol-β-glucosides. UDP-glc:sterol glucosyltransferase is represented by two genes in Arabidopsis (Warnecke et al., 1997; M. Doblin, personal communication). Whereas the single mutants exhibit a threefold reduction of sitosterol-β-glucosides, double mutants display a 25-fold reduction in sitosterol-β-glucosides levels without altering cellulose content (W. Scheible, C. Somerville, and H. Schaller, personal communication). These results suggest that sitosterol-β-glucosides are not limiting factors for cellulose synthesis in Arabidopsis, and that very low levels may be sufficient to initiate cellulose synthesis. A glc incorporation experiment with cotton membranes showed that UDP-glc is linked to a mixture of sterols, which has more than 95% sitosterol (Peng et al., 2002). Yet, minor sterols besides sitosterol could play essential roles in cellulose synthesis and/or other mechanisms underlying cell wall elongation and expansion. Future studies will continue to focus on the judicious use of sterol biosynthesis mutants and functional assays as tools to uncover specific cellular and molecular defects, and ultimately to assign roles to the underlying sterols involved.

Experimental procedures

  1. Top of page
  2. Summary
  3. Introduction
  4. Results
  5. Discussion
  6. Experimental procedures
  7. Acknowledgements
  8. References

Plant materials and growth conditions

The Arabidopsis thaliana ecotypes Landsberg erecta (Ler) and Columbia (Col) were used as wild-type controls. fk, cph, and hyd1 mutant alleles were propagated as heterozygotes. The mutant alleles were fk-X224 (Schrick et al., 2000), cph-T357, and hyd1-R216 (Schrick et al., 2002) unless otherwise mentioned. The dwf1 allele used was dwf1-6 (Choe et al., 1999; Schrick et al., 2002). Plants were grown on soil as previously described by Schrick et al. (2002). To generate mutant seedling material, seeds from heterozygous plants were germinated on rich media containing 0.8% agar, 1% sucrose, and 1× MS (Murashige and Skoog, 1962). Preparation of callus cell cultures was as described by Schrick et al. (2000), except that the liquid-rich medium contained 3% sucrose.

Sterol biosynthesis inhibitors

Seeds were germinated on inhibitor-containing rich media having 0.8% agar, 1% sucrose and 1× MS (Murashige and Skoog, 1962). 15-Azasterol (Lilly Research Laboratories, Greenfield, IL, USA) and fenpropimorph (Riedel-deHaën, Seelze, Germany) 2000× solutions for a final concentration of 1.0 µm were made by dissolving 1.000 and 0.607 mg ml−1, respectively, in absolute ethanol.

Transmission electron microscopy (TEM)

For TEM analysis, ovules were cryofixed by high-pressure freezing (Bal-Tec HPM 010, Balzers) in hexadecene (Merck Sharp & Dohme, Haar, Germany) (Studer et al., 1989), freeze substituted in acetone containing 2% osmium tetroxide and 0.5% uranylacetate, and embedded in Spurr's resin (Serva, Heidelberg, Germany). Ultrathin sections of ovules were stained with ethanolic uranyl acetate and lead citrate. Root tips were conventionally fixed with 2.5% glutaraldehyde/2% formaldehyde, embedded in agarose and treated with 1% osmium tetroxide and 1% aqueous uranyl acetate. Dehydration was with a graded series of acetone followed by embedding in Spurr's resin. Ultrathin sections of roots were stained with aqueous uranyl acetate and lead citrate.

Sterol extraction and GC–MS analysis

Fresh weight of callus cells (0.5–9.0 g) and seedlings (0.4–0.9 g) was determined. Plant material was frozen in liquid nitrogen and freeze dried overnight. Sterol purification and quantification were carried out as described previously by Fujioka et al. (2002) and He et al. (2003).

Cellulose analysis

Plant material was extracted in 70% ethanol at 70°C, washed twice with 70% ethanol, once with 100% acetone, and air dried overnight. DW (2–8 mg) was measured with a microbalance (Sartorius, Göttingen, Germany). Cell wall material was hydrolyzed by boiling in acetic acid:nitric acid:water (8 : 1 : 2) for 30 min. The insoluble fraction was recovered by centrifugation and washed twice with water, twice with acetone, and air dried overnight. To hydrolyze the cellulose-containing pellet to glucose monomers, samples were incubated with sulfuric acid (67%) for 1 h at room temperature in a thermoshaker. Cellulose content was measured colorimetrically with anthrone (Sigma-Aldrich, München, Germany) reagent as described by Updegraff (1969).

Uronic acids and neutral sugars analysis

Samples were prepared as above for cellulose analysis. Two to nine milligrams of cleared and dried plant material was subjected to Saeman hydrolysis (Adams, 1965). Each sample was subsequently analyzed for both uronic acid content and neutral sugar composition. Uronic acids were measured by the hydroxybiphenyl method using galacturonic acid as a standard (Blumenkrantz and Asboe-Hansen, 1973; Taylor and Buchanan-Smith, 1992). Neutral sugars of the non-cellulosic cell wall fraction were determined by GC of alditol acetates with myoinositol as the internal standard (Blakeney et al., 1983; Reiter et al., 1997). Relative sugar composition values were calculated as percentages of the DW material.

HPLC analysis of free sugars

Seedlings were washed in deionized H2O, transferred briefly to Kimwipes, and placed in microfuge tubes. Samples were frozen in liquid nitrogen and freeze dried. DW (8–18 mg) was determined using a microbalance (Sartorius). Samples were re-frozen in liquid nitrogen and ground with an 8-mm steel ball in a Retsch mill. Sugars were extracted by boiling in 80% methanol for 2 min, followed by centrifugation (17 949 g, 5 min) to sediment the insoluble fraction. The remaining pellet was extracted again by boiling in 20% methanol for 2 min as in the previous step. The 80 and 20% methanol supernatants were unified and evaporated in a speed-vac at room temperature for 5 h. Samples were re-suspended in 0.5 ml HPLC-grade H2O (28–63 µl mg−1 DW), centrifuged (17 949 g, 3 min), and filtered (0.2 µm). HPLC separation was performed using a high pH anion exchange column (Dionex Carbo Pac PA1, Dionex GmbH, Idstein, Germany) with 30 mm NaOH as eluent and pulsed amperometric detection (Biometra PED-300, Biometra, Göttingen, Germany). The instrumentation used was a Modular HPLC System Kontron (Pump 420, Autosampler 460) with kromasystem 2000 software (Kontron Instruments, Milano, Italy).

Aniline blue staining for callose

Bent-cotyledon stage embryos were dissected from ovules using a thin-walled hypodermic needle (0.45 mm × 12 mm) and transferred directly to 70% ethanol. The embryos were left in 100% (v/v) ethanol at 4°C overnight. Samples were re-hydrated with 70, 50, 30, and 15% (v/v) ethanol, and two changes of deionized H2O, incubating each change for 10 min. The embryos were incubated for 30 min in freshly prepared 0.05% aniline blue (Schmid GmbH + Co., Köngen/N, Germany) dissolved in 0.07m K3O4P · 3H2O, pH 8.6, and filter sterilized. The staining solution was replaced with 20% glycerol, and the embryos were transferred to clean microscope slides and covered with 20 mm × 60 mm cover slips fixed by thin strips of double-stick tape on either side. Samples were destained by daylight (or artificial light at the lab bench) for 30 min prior to microscopy with a Zeiss Axiophot UV fluorescence microscope. Digital images were processed using a Zeiss AxioCam with axiovision 3.1 software.

Phloroglucinol staining for lignin

A 2% phloroglucinol–HCl solution was made by first dissolving phloroglucinol (1,3,5-trihydroxybenzene, Sigma-Aldrich) in 20% ethanol, and then adding 0.2× volume 12N HCl. The solution was filtered (0.45 µm) and used for lignin staining of seedlings directly without prior fixation. Lignin staining was observed as a red dye under standard dissecting microscope optics for seedlings. Digital images of seedlings were processed with a Nikon Coolscan using wintv software.


  1. Top of page
  2. Summary
  3. Introduction
  4. Results
  5. Discussion
  6. Experimental procedures
  7. Acknowledgements
  8. References

The authors thank Heather Youngs and Chris Somerville of the Department of Biology Carnegie Institution at Stanford University for help with analysis of uronic acids and neutral sugars, and Heinz Schwarz at the Max Planck Institute for Developmental Biology in Tübingen for help with high pressure freezing of embryos. We are grateful to Ariane Alvarez, Vivian Jeske, Makoto Kobayashi, Bettina Stadelhofer, and Hannah Steigele for technical assistance, Stewart Gillmor, Wolfgang Lukowitz, Titus Neumann, and Animesh Ray for critical reading of the manuscript, and Monika Doblin, Wolfgang Lukowitz and Chris Somerville for communicating pertinent data prior to publication. 15-Azasterol (A25822B azasterol) was a gift from Lilly Research Laboratories (Greenfield, IN, USA). The dwf1-6 seeds were provided by the Nottingham Arabidopsis Stock Centre. This work was funded through a research fellowship and a grant to K.S. from the Deutsche Forschungsgemeinschaft (SCHR 744/1-1,1-2,1-3).


  1. Top of page
  2. Summary
  3. Introduction
  4. Results
  5. Discussion
  6. Experimental procedures
  7. Acknowledgements
  8. References
  • Adams, G.A. (1965) Complete acid hydrolysis. Meth. Carbohydr. Chem. 5, 269276.
  • Aist, J.R. (1976) Papillae and related wound plugs of plant cells. Annu. Rev. Phytopathol. 14, 145163.
  • Altmann, T. (1999) Molecular physiology of brassinosteroids revealed by the analysis of mutants. Planta, 208, 111.
  • Arioli, T., Peng, L., Betzner, A.S. et al. (1998) Molecular analysis of cellulose biosynthesis in Arabidopsis. Science, 279, 717720.
  • Bach, T.J. and Benveniste, P. (1997) Cloning of cDNAs or genes encoding enzymes of sterol biosynthesis from plants and other eukaryotes: heterologous expression and complementation analysis of mutations for functional characterization. Prog. Lipid Res. 36, 197226.
  • Benveniste, P. (1986) Sterol biosynthesis. Annu. Rev. Plant Physiol. 37, 275308.
  • Bishop, G.J. and Yokota, T. (2001) Plants steroid hormones, brassinosteroids: current highlights of molecular aspects on their synthesis/metabolism, transport, perception and response. Plant Cell Physiol. 42, 114120.
  • Blakeney, A.B., Harris, P.J., Henry, R.J. and Stone, B.A. (1983) A simple and rapid preparation of alditol acetates for monosaccharide analysis. Carbohydr. Res. 113, 291299.
  • Bloch, K.E. (1983) Sterol structure and membrane function. Crit. Rev. Biochem. 14, 4791.
  • Blumenkrantz, N. and Asboe-Hansen, G. (1973) New method for quantitative determination of uronic acids. Anal. Biochem. 54, 484489.
  • Boisson, M., Gomord, V., Audran, C., Berger, N., Dubreucq, B., Granier, F., Lerouge, P., Faye, L., Caboche, M. and Lepiniec, L. (2001) Arabidopsis glucosidase I mutants reveal a critical role of N-glycan trimming in seed development. EMBO J. 20, 10101019.
  • Cantatore, J.L., Murphy, S.M. and Lynch, D.V. (2000) Compartmentation and topology of glucosylceramide synthesis. Biochem. Soc. Trans. 28, 748750.
  • Carland, F.M., Fujioka, S., Takatsuto, S., Yoshida, S. and Nelson, T. (2002) The identification of CVP1 reveals a role for sterols in vascular patterning. Plant Cell, 14, 20452058.
  • Choe, S., Dilkes, B.P., Gregory, B.D. et al. (1999) The Arabidopsis dwarf1 mutant is defective in the conversion of 24-methylenecholesterol to campesterol in brassinosteroid biosynthesis. Plant Physiol. 119, 897907.
  • Clouse, S.D. (2000) Plant development: a role for sterols in embryogenesis. Curr. Biol. 10, R601R604.
  • Clouse, S.D. (2002) Arabidopsis mutants reveal multiple roles for sterols in plant development. Plant Cell, 14, 19952000.
  • Desprez, T., Vernhettes, S., Fagard, M., Refregier, G., Desnos, T., Aletti, E., Py, N., Pelletier, S. and Höfte, H. (2002) Resistance against herbicide isoxaben and cellulose deficiency caused by distinct mutations in same cellulose synthase isoform CESA6. Plant Physiol. 128, 482490.
  • Diener, A.C., Li, H., Zhou, W., Whoriskey, W.J., Nes, W.D. and Fink, G.R. (2000) STEROL METHYLTRANSFERASE 1 controls the level of cholesterol in plants. Plant Cell, 12, 853870.
  • Edwards, P.A. and Ericsson, J. (1999) Sterols and isoprenoids: signaling molecules derived from the cholesterol biosynthetic pathway. Annu. Rev. Biochem. 68, 157185.
  • Fagard, M., Desnos, T., Desprez, T., Goubet, F., Refregier, G., Mouille, G., McCann, M., Rayon, C., Vernhettes, S. and Hofte, H. (2000) PROCUSTE1 encodes a cellulose synthase required for normal cell elongation specifically in roots and dark-grown hypocotyls of Arabidopsis. Plant Cell, 12, 24092424.
  • Fujioka, S. and Yokota, T. (2003) Biosynthesis and metabolism of brassinosteroids. Annu. Rev. Plant Biol. 54, 137164.
  • Fujioka, S., Takatsuto, S. and Yoshida, S. (2002) An early C-22 oxidation branch in the brassinosteroid biosynthetic pathway. Plant Physiol. 130, 930939.
  • Gillmor, C.S., Poindexter, P., Lorieau, J., Palcic, M.M. and Somerville, C. (2002) Alpha-glucosidase I is required for cellulose biosynthesis and morphogenesis in Arabidopsis. J. Cell Biol. 156, 10031013.
  • Grandmougin-Ferjani, A., Schuler-Muller, I. and Hartmann, M.A. (1997) Sterol modulation of plasma membrane H+-ATPase activity from corn roots reconstituted into soybean lipids. Plant Physiol. 113, 163174.
  • Grebenok, R.J., Ohnmeiss, T.E., Yamamoto, A., Huntley, E.D., Galbraith, D.W. and Della Penna, D. (1998) Isolation and characterization of an Arabidopsis thaliana C-8,7 sterol isomerase: functional and structural similarities to mammalian C-8,7 sterol isomerase/emopamil-binding protein. Plant Mol. Biol. 38, 807815.
  • Guo, D.A., Venkatramesh, M. and Nes, W.D. (1995) Developmental regulation of sterol biosynthesis in Zea mays. Lipids, 30, 203219.
  • Hartmann, M.-A. (1998) Plant sterols and the membrane environment. Trends Plant Sci. 3, 170175.
  • Hartmann, M.-A. and Benveniste, P. (1987) Plant membrane sterols: isolation, identification and biosynthesis. Meth. Enzymol. 148, 632650.
  • He, J.-X., Fujioka, S., Li, T.-C., Kang, S.G., Seto, H., Takatsuto, S., Yoshida, S. and Jang, J.-C. (2003) Sterols regulate development and gene expression in Arabidopsis. Plant Physiol. 131, 12581269.
  • Hobbs, D., Hume, J.H., Rolph, C.E. and Cooke, D.T. (1996) Changes in lipid composition during floral development of Brassica campestris. Phytochemistry, 42, 335339.
  • Jang, J.C., Fujioka, S., Tasaka, M., Seto, H., Takatsuto, S., Ishii, A., Aida, M., Yoshida, S. and Sheen, J. (2000) A critical role of sterols in embryonic patterning and meristem programming revealed by the fackel mutants of Arabidopsis thaliana. Genes Dev. 14, 14851497.
  • Klahre, U., Noguchi, T., Fujioka, S., Takatsuto, S., Yokota, T., Nomura, T., Yoshida, S. and Chua, N.H. (1998) The Arabidopsis DIMINUTO/DWARF1 gene encodes a protein involved in steroid synthesis. Plant Cell, 10, 16771690.
  • Lukowitz, W., Nickle, T.C., Meinke, D.W., Last, R.L., Conklin, P.L. and Somerville, C.R. (2001) Arabidopsis cyt1 mutants are deficient in a mannose-1-phosphate guanylyltransferase and point to a requirement of N-linked glycosylation for cellulose biosynthesis. Proc. Natl. Acad. Sci. USA, 98, 22622267.
  • Mayer, U., Torres-Ruiz, R.A., Berleth, T., Misera, S. and Jürgens, G. (1991) Mutations affecting the body organization in the Arabidopsis embryo. Nature, 353, 402407.
  • Mercer, I.E. (1993) Inhibitors of sterol biosynthesis and their applications. Prog. Lipid Res. 32, 357416.
  • Murashige, T. and Skoog, F. (1962) A revised medium for rapid growth and bioassays with tobacco tissue cultures. Physiol. Plant. 15, 473497.
  • Nickle, T.C. and Meinke, D.W. (1998) A cytokinesis-defective mutant of Arabidopsis (cyt1) characterized by embryonic lethality, incomplete cell walls, and excessive callose accumulation. Plant J. 15, 321332.
  • Noda, M., Tanaka, M., Seto, Y., Aiba, T. and Oku, C. (1988) Occurrence of cholesterol as a major sterol component in leaf surface lipids. Lipids, 23, 439444.
  • Ourisson, G. (1994) Peculiarities of sterol biosynthesis in plants. J. Plant Physiol. 143, 434439.
  • Pagant, S., Bichet, A., Sugimoto, K., Lerouxel, O., Desprez, T., McCann, M., Lerouge, P., Vernhettes, S. and Höfte, H. (2002) KOBITO1 encodes a novel plasma membrane protein necessary for normal synthesis of cellulose during cell expansion in Arabidopsis. Plant Cell, 14, 20012013.
  • Patterson, G.W., Hugly, S. and Harrison, D. (1993) Sterols and phytyl esters of Arabidopsis thaliana under normal and chilling temperatures. Phytochemistry, 33, 13811383.
  • Peng, L., Kawagoe, Y., Hogan, P. and Delmer, D. (2002) Sitosterol-β-glucoside as a primer for cellulose synthesis in plants. Science, 295, 147150.
  • Peskan, T., Westermann, M. and Oelmüller, R. (2000) Identification of low-density Triton X-100-insoluble plasma membrane microdomains in higher plants. Eur. J. Biochem. 267, 69896995.
  • Reiter, W.-D., Chapple, C. and Somerville, C.R. (1997) Mutants of Arabidopsis thaliana with altered cell wall polysaccharide composition. Plant J. 12, 335345.
  • Richmond, T.A. and Somerville, C.R. (2000) The cellulose synthase superfamily. Plant Physiol. 124, 495498.
  • Sato, S., Kato, T., Kakegawa, K. et al. (2001) Role of the putative membrane-bound endo-1,4-beta-glucanase KORRIGAN in cell elongation and cellulose synthesis in Arabidopsis thaliana. Plant Cell Physiol. 42, 251263.
  • Schaeffer, A., Bronner, R., Benveniste, P. and Schaller, H. (2001) The ratio of campesterol to sitosterol that modulates growth in Arabidopsis is controlled by STEROL METHYLTRANSERFASE 2;1. Plant J. 25, 605615.
  • Schaller, H. (2003) The role of sterols in plant growth and development. Prog. Lipid Res. 42, 163175.
  • Schindelman, G., Morikami, A., Jung, J., Baskin, T.I., Carpita, N.C., Derbyshire, P., McCann, M.C. and Benfey, P.N. (2001) COBRA encodes a putative GPI-anchored protein, which is polarly localized and necessary for oriented cell expansion in Arabidopsis. Genes Dev. 15, 11151127.
  • Schmitt, P., Scheid, F. and Benveniste, P. (1980) Accumulation of Δ8,14-sterols in suspension cultures of bramble cells cultured with an azasterol antimycotic agent (A25822B). Phytochemistry, 19, 525530.
  • Schrick, K. and Laux, T. (2001) Zygotic embryogenesis: The formation of an embryo from a fertilized egg. In Current Trends in the Embryology of Angiosperms (Bhojwani, S.S. and Soh, W.Y., eds). Dordrecht: Kluwer Academic Publishers, pp. 249277.
  • Schrick, K., Mayer, U., Horrichs, A., Kuhnt, C., Bellini, C., Dangl, J., Schmidt, J. and Jürgens, G. (2000) FACKEL is a sterol C-14 reductase required for organized cell division and expansion in Arabidopsis embryogenesis. Genes Dev. 14, 14711484.
  • Schrick, K., Mayer, U., Martin, G., Bellini, C., Kuhnt, C., Schmidt, J. and Jürgens, G. (2002) Interactions between sterol biosynthesis genes in embryonic development of Arabidopsis. Plant J. 31, 6173.
  • Schuler, I., Milon, A., Nakatani, Y., Ourisson, G., Albrecht, A.M., Benveniste, P. and Hartmann, M.A. (1991) Differential effects of plant sterols on water permeability and on acyl chain ordering of soybean phosphatidylcholine bilayers. Proc. Natl. Acad. Sci. USA, 88, 69266930.
  • Simons, K. and Toomre, D. (2000) Lipid rafts and signal transduction. Nat. Rev. Mol. Cell Biol. 1, 3139.
  • Souter, M., Topping, J., Pullen, M., Friml, J., Palme, K., Hackett, R., Grierson, D. and Lindsey, K. (2002) hydra mutants of Arabidopsis are defective in sterol profiles and auxin and ethylene signaling. Plant Cell, 14, 10171031.
  • Studer, D., Michel, M. and Müller, M. (1989) High pressure freezing comes of age. Scanning Microsc. Suppl. 3, 253269.
  • Taylor, K.A. and Buchanan-Smith, J.G. (1992) A colorimetric method for the quantitation of uronic acids and a specific assay for galacturonic acid. Anal. Biochem. 201, 190196.
  • Updegraff, D.M. (1969) Semimicro determination of cellulose in biological materials. Anal. Biochem. 32, 420424.
  • Warnecke, D.C., Baltrusch, M., Buck, F., Wolter, F.P. and Heinz, E. (1997) UDP-glucose: sterol glucosyltransferase: cloning and functional expression in Escherichia coli. Plant Mol. Biol. 5, 597603.
  • Willemsen, V., Friml, J., Grebe, M., Van Den Toorn, A., Palme, K. and Scheres, B. (2003) Cell polarity and PIN protein positioning in Arabidopsis require STEROL METHYLTRANSFERASE1 function. Plant Cell, 15, 116.