OPEN ACCESS ARTICLE

This Article Free via Open Access: OA OA Abstract Full Text (PDF) OA All Versions of this Article: 145/4/1444    most recent pp.107.102582v1 Alert me when this article is cited Alert me if a correction is posted Services Email this article to a friend Similar articles in this journal Similar articles in ISI Web of Science Similar articles in PubMed Alert me to new issues of the journal Download to citation manager Citing Articles Citing Articles via ISI Web of Science (1) Citing Articles via Google Scholar Google Scholar Articles by Sindhu, A. Articles by Johal, G. Search for Related Content PubMed PubMed Citation Articles by Sindhu, A. Articles by Johal, G. Agricola Articles by Sindhu, A. Articles by Johal, G.

DEVELOPMENT AND HORMONE ACTION

Maize Brittle stalk2 Encodes a COBRA-Like Protein Expressed in Early Organ Development But Required for Tissue Flexibility at Maturity^1,[C],[OA]

Department of Botany and Plant Pathology (A.S., A.O., N.C.C., G.J.) and Department of Biological Sciences (T.L., M.C.C.), Purdue University, West Lafayette, Indiana 47907–2054; Department of Agronomy, University of Missouri, Columbia, Missouri 65211 (D.S.M.); and University of Florida Genetics Institute and Agronomy Department, Gainesville, Florida 32610 (W.V.)

	ABSTRACT

TOP ABSTRACT RESULTS DISCUSSION MATERIALS AND METHODS LITERATURE CITED

 
The maize (Zea mays) brittle stalk2 (bk2) is a recessive mutant, the aerial parts of which are easily broken. The bk2 phenotype is developmentally regulated and appears 4 weeks after planting, at about the fifth-leaf stage. Before this time, mutants are indistinguishable from wild-type siblings. Afterward, all organs of the bk2 mutants turn brittle, even the preexisting ones, and they remain brittle throughout the life of the plant. Leaf tension assays and bend tests of the internodes show that the brittle phenotype does not result from loss of tensile strength but from loss in flexibility that causes the tissues to snap instead of bend. The Bk2 gene was cloned by a combination of transposon tagging and a candidate gene approach and found to encode a COBRA-like protein similar to rice (Oryza sativa) BC1 and Arabidopsis (Arabidopsis thaliana) COBRA-LIKE4. The outer periphery of the stalk has fewer vascular bundles, and the sclerids underlying the epidermis possess thinner secondary walls. Relative cellulose content is not strictly correlated with the brittle phenotype. Cellulose content in mature zones of bk2 mature stems is lowered by 40% but is about the same as wild type in developing stems. Although relative cellulose content is lowered in leaves after the onset of the brittle phenotype, total wall mass as a proportion of dry mass is either unchanged or slightly increased, indicating a compensatory increase in noncellulosic carbohydrate mass. Fourier transform infrared spectra indicated an increase in phenolic ester content in the walls of bk2 leaves and stems. Total content of lignin is unaffected in bk2 juvenile leaves before or after appearance of the brittle phenotype, but bk2 mature and developing stems are markedly enriched in lignin compared to wild-type stems. Despite increased lignin in bk2 stems, loss of staining with phloroglucinol and ultraviolet autofluorescence is observed in vascular bundles and sclerid layers. Consistent with the infrared analyses, levels of saponifiable hydroxycinnamates are elevated in bk2 leaves and stems. As Bk2 is highly expressed during early development, well before the onset of the brittle phenotype, we propose that Bk2 functions in a patterning of lignin-cellulosic interactions that maintain organ flexibility rather than having a direct role in cellulose biosynthesis.

Although mutants in cell wall-related genes are valuable materials to address the functions of cell wall polymers and their interactions, very few cell wall mutants have been characterized for the grasses (Yong et al., 2005). Some of the first mutations whose phenotypes suggested a cell wall defect were the brittle stalk (bk) mutants of maize (Zea mays), synonymous with the brittle culm (bc) mutants of rice (Oryza sativa) and barley (Hordeum vulgare; Langham, 1940; Kokubo et al., 1989; Li et al., 2003). In maize, bk2 is characterized by a brittleness of all aerial plant organs, where the tissues easily snap when bent (Langham, 1940). Curiously, the brittle phenotype only appears after about 4 weeks of growth, under field or greenhouse conditions, about the time of appearance of the fifth leaf. Before this stage, bk2 plants are indistinguishable from their wild-type siblings, as the leaves are flexible and can be bent easily without breaking (Langham, 1940).

To understand the molecular basis of the brittle phenotype, we cloned the Bk2 gene and characterized cell structure and cell wall composition in the bk2 mutant. The Bk2 gene is orthologous to the rice Bc1 gene, a member of the COBRA gene family that encodes glycosylphosphatidylinositol (GPI)-anchored proteins of unknown function (Schindelman et al., 2001; Li et al., 2003; Ching et al., 2006). A mutation in the founding member of the COBRA family in Arabidopsis (Arabidopsis thaliana) results in severe inhibition of root elongation and the progressive radial swelling of the cortical cells away from the root tip (Hauser et al., 1995). The cobra mutant is defective in cellulose synthesis and microfibril orientation in roots (Schindelman et al., 2001; Roudier et al., 2005). In contrast, rice Bc1 is expressed primarily in vascular regions of the leaves and culm, and the phenotype results in organ brittleness rather than any obvious cell deformities. As with the rice bc1 mutants, mutations in Arabidopsis COBRA-LIKE4 (COBL4), the most similar homolog of the rice Bc1 gene (Li et al., 2003), result in plants with normal morphology but weaker stems (Brown et al., 2005). In addition, cobl4 exhibits an irregular xylem phenotype, similar to mutants compromised in secondary wall cellulose synthesis or lignin deposition (Brown et al., 2005).

Because cobl mutants have decreased cellulose contents, COBL proteins are thought to be involved in cellulose synthesis or its regulation (Kokubo et al., 1991; Schindelman et al., 2001; Li, et al., 2003; Roudier et al., 2005). Also, the brittle phenotypes of bc1 and bk2 were determined to be a loss of mechanical strength associated with a reduction in cellulose content (Ching et al., 2006). However, this conclusion is inconsistent with the phenotype of other cellulose-deficient Arabidopsis mutants of reduced tensile strengths but which are not brittle (Turner et al., 2001). How the loss of function of COBL4, BK2, and BC1 proteins results in a brittle phenotype remains mysterious.

Consistent with the reports by Li et al. (2003) and Ching et al. (2006), we show here that decreases in the proportion of cellulose to total cell wall mass in leaves and mature stems of bk2. However, this was not true of the developing stem. Further, the bk2 mutant compensates for the loss of cellulose with an increase in wall mass, mostly xylans, per total dry mass. Additional structural features, most likely related to phenolic esters, are better correlated with the brittle phenotype. We show here that the Bk2 gene is strongly expressed during the first 2 weeks in development, but the expression is substantially lower by the time of the appearance of the brittle phenotype at the 4-week stage. In addition, after expression of the phenotype, Klason lignin content is enriched in bk2 stems but there is a loss of phloroglucinol-positive material and UV autofluorescence in the mature vascular bundles and sclerids underlying the epidermis. In contrast to the results of Ching et al. (2006), the mechanical or flexural strengths of wild-type and bk2 internodes, as measured by simple one-point bend tests, were indistinguishable. However, the wild-type internodes continue to bend under increasing stress, but the bk2 internodes bent slightly under the stress and then snapped upon continued applied stress. Our observations suggest that loss of cellulose synthesis is not tightly correlated with the brittle phenotype and that lack of proper formation of a lignocellulosic interaction during a change in lignin architecture may result in the drastic loss of flexibility exhibited in the mutant. Further, the high expression of Bk2 during early development before the onset of the brittle phenotype and its ability to establish the number and size of peripheral vascular bundles in the basal internodes indicate that Bk2 plays a broader fundamental role in developmental patterning.

	RESULTS

TOP ABSTRACT RESULTS DISCUSSION MATERIALS AND METHODS LITERATURE CITED

 

Phenotypic Characteristics of bk2

In the uniform backcross progeny that we generated, the mutant bk2 and wild-type siblings segregated in a 1:1 ratio, indicating that the recessive bk2 allele transmits normally through both the male and female gametophytes. In this progeny, the mutant (bk2/bk2 homozygotes) and wild-type siblings (bk2/+ heterozygotes) were indistinguishable in all aspects of growth, development, and vigor, except that the bk2 mutants were brittle and broke readily (Fig. 1A ). All of the plant organs are brittle, including leaves, the ear, the tassel, and brace roots, as well as the stems. Field-grown mutants have to be staked as even a slight wind gust causes bk2 to snap at the lower internodes (Fig. 1B).

View larger version (123K): [in this window] [in a new window] [as a PowerPoint slide]   Figure 1. Characteristics of the bk2 phenotype. A, Bending of a brittle mutant leaf causes it to snap and break cleanly, leaving razor-cut-like edges. B, Breaking of bk2 mutants in the field as a result of wind damage. [See online article for color version of this figure.]

bk2 Mutants Exhibit Some Changes in Stem Anatomy

To determine if the brittleness of bk2 mutants arises from changes in cell shape, size, or structure, we compared internode stem cells of bk2 with those of their wild-type siblings using bright-field and scanning electron microscopies. Sections derived from the first, third, and fourth internodes (counted from the bottom) of the greenhouse-grown mutant and wild-type siblings were used for these comparisons.

While the wild-type stems produce secondary vascular bundles in the rind region of their lower internodes at maturity, the mutant stems do not (Fig. 2, A and B ). As a result, the density of peripheral vascular bundles in bk2 stems is only 1.7 ± 0.2 per mm compared to 3.5 ± 0.4 per mm in wild-type internodes, even though there is little difference in stem circumference. The total number of peripheral bundles was almost halved, with an average of 37 for bk2 and 72 for wild type. The density of peripheral bundles was slightly reduced in the third internode of bk2, 2.4 ± 0.3 per mm versus 2.9 ± 0.2 per mm, with total peripheral bundles of 53 for bk2 and 84 for wild type. Although most cells in the rind region had slightly thinner walls in the mutant compared to the wild type, the difference was more conspicuous in the walls of sclerenchyma underlying the epidermis and surrounding the vascular bundles (Fig. 2, C–H). Whereas the thinner bk2 sclerenchyma walls are homogeneous in appearance, those of the wild type show separation of primary and secondary wall. There are relatively more cells in bk2 stem cortex between the epidermis and the outer ring of vascular bundles than in those of wild type. The lowermost internodes tend to form a continuous layer of sclerids between the outermost ring of vascular bundles in wild-type plants but not in bk2 mutants. Taken together, these observations indicate that the bk2 mutation results in developmental alterations in the structure and anatomy of the maize stalk.

View larger version (141K): [in this window] [in a new window] [as a PowerPoint slide]   Figure 2. Anatomy of a bk2 stalk. A and B are cross sections of the corresponding mutant and wild-type stems, respectively, imaged with bright-field microscopy. C to H, Scanning electron micrographs of mutant (C, E, and G) and wild-type (D, F, and H) stems taken from cross sections.

We tagged the Bk2 gene using a directed mutagenesis approach involving Mutator (Mu; Multani et al., 2003). Mu-active plants were used as females and fertilized with pollen collected from the bk2 tester homozygous for the original mutant allele, herein referred to as the reference allele (bk2-ref). About 60,000 of the resulting progeny were planted in the field and screened for bk2 mutants around week 7 after planting. Brittle mutants were screened by taking a leaf from each of the progeny plants between the index and the middle finger and then bending with the thumb. Two brittle mutants were identified, but only one could be propagated. Sixteen plants from the progeny of this mutant (named bk2-Mu1) with B73 were genotyped using two RFLP markers, BNL8.17, which we found to be located 4.0 cM proximal to the bk2 locus, and UMC95, which is located 2.5 cM distal to bk2. This analysis allowed us to identify plants that inherited the newly generated mutant allele, bk2-Mu1. These bk2-Mu1 heterozygotes were crossed again to the bk2 tester (bk2-ref) to generate progeny that segregated 1:1 for brittle and wild-type plants containing or lacking the bk2-Mu1 mutant allele, respectively.

Cloning of Bk2

As we were identifying the tagged bk2 gene by standard cosegregation analyses (Multani et al., 2003), a rice bc1 mutant, whose brittle plant phenotype matches closely with that of the maize bk2, was reported (Li et al., 2003). Because bc1 mapped to a region of the rice genome that exhibited a syntenic relationship with the maize region harboring bk2, we hypothesized that they may be related. Database searches for maize sequences with homology to the rice Bc1 gene revealed a full-length EST that exhibited 87% identity with the rice Bc1 cDNA (data not shown). A complete genomic clone corresponding to this EST was also found in The Institute for Genomic Research database and was named ZmBc1.

Primers were designed from the ZmBc1 sequence and used to clone a part of the gene by PCR. A radiolabeled probe derived from this amplicon was hybridized to the same blots used earlier to map bk2 using RFLP markers (Fig. 3A ). No recombinants were found between ZmBc1 and the bk2 brittle phenotype, suggesting that the two were closely linked. To address whether the bk2 phenotype results from a mutation in ZmBc1, we took advantage of the bk2-Mu1 mutant allele. DNA isolated from this mutant was subjected to a reverse-genetics PCR-amplification test to assess if it was caused by insertion of a Mu element. Using the genomic ZmBc1 sequence as a guide, several primers were designed from the gene (Fig. 3B). We especially targeted the 5' end of the gene for primers because Mu elements are known to insert preferentially in this gene region (May et al., 2003). Each of the ZmBc1-specific primers were used in combination with a primer from the Mu terminal inverted repeat (TIR) to run PCR reactions in which the template DNA was derived from bk2-Mu1. PCR reactions yielding amplification products revealed that a Mu element was inserted 119 nucleotides upstream of the start codon of ZmBc1 in the bk2-Mu1 mutant allele (Fig. 3B). Typical of Mu elements (May et al., 2003), a 9-bp duplication of the target site was found to flank this Mu insertion.

View larger version (52K): [in this window] [in a new window] [as a PowerPoint slide]   Figure 3. Cloning and confirmation of the bk2 gene. A, Southern gel blot showing linkage of bk2 with the maize homolog of the rice Bc1 gene (ZmBc1). B, The relative location of the two insertions (triangles) and of various primers (arrows) that were designed to amplify the bk2 gene. C, Diagrammatic representation of the bk2 gene showing three exons (gray rectangles), the relative location of the start codon (ATG), and the location of insertions (triangles) in the bk2-Mu1 and bk2-ref mutant alleles. Mu-1 is inserted 119 bp upstream of ATG, and the 5' UTR is 90 bp long. D, The TIR of the novel insertion in the bk2-ref allele. Out of 46 bases, only the terminal 7 bp exhibit perfect identity.

The coding region of the Bk2 gene (ZmBc1) is 1,353 bp long and its predicted product has 84% and 70% identity with the rice BC1 and Arabidopsis COBL4 proteins, respectively (Fig. 4 ). In addition, the maize Bk2 and rice Bc1 genes comprise three exons with identical exon/intron boundaries. The predicted BK2 protein is 450 amino acids long in contrast to the 468 amino acids of BC1. BK2 and BC1 differ from each other substantially in the N-terminal signal peptide and the hydrophobic C-terminal sequence, the latter of which is predicted to be clipped at the -site following attachment of the GPI anchor in the endoplasmic reticulum (Li et al., 2003). The 26 amino acids, corresponding to locations 243 to 268 of BC1, are missing in BK2. A gap of 32 amino acids is at the same location in Arabidopsis COBL4 in relation to rice BC1.

View larger version (88K): [in this window] [in a new window] [as a PowerPoint slide]   Figure 4. Protein sequence alignment of ZmBK2, OsBC1, and AtCOBL4. Numbers at the left are the positions of the amino acids in each protein, with gaps (dashes) included to maximize alignments. Identical and similar amino acids are shaded in black and gray, respectively. The predicted signal peptide is underlined, with cleavage site marked by an arrowhead. A consensus amino-acid sequence CCVS in all COBL proteins is boxed. The -amino acid at the predicted GPI cleavage site is denoted by an arrowhead.

The insertion in bk2-ref reveals a new family of transposons with at least two unusual characteristics. First, they have an imperfect TIR of 46 bp, of which only the outermost 7 bp exhibit a perfect match between the left and the right borders (Fig. 3D). Second, they cause a 10-bp duplication at the site of insertion, another novelty among the transposable elements identified thus far in plants. However, Ching et al. (2006), who also cloned this insertion and named it KITE, reported this duplication to be of 8 bp. About 50 copies of KITE have been detected in the maize databases. Southern blots have also revealed equivalent copies of this tranposon in most maize inbreds and in Teosinte (data not shown). Hundreds of copies of a modified version of KITE have also been detected in the rice genome (A. Sindhu and G. Johal, unpublished data).

The bk2 Brittle Phenotype Is Independent of Tensile Strength

Sections of wild-type and bk2 leaves were tested for break strength. Thin sections of leaves taken parallel to the vascular tissue broke at a force about twice that of leaf sections made perpendicularly, but the average breaking strengths of wild type and bk2 were indistinguishable (Fig. 5A ). Similarly, the strain profiles in a one-point bend test of internodes of wild type and bk2 were indistinguishable. Whereas the wild type only bent upon increasing stress, the bk2 internodes snapped at variable points beyond the maximum stress exerted (Fig. 5B). These results contrast with those of Ching et al. (2006), who reported a difference in the strain curves. We found that the strain was quite variable among samples from both wild type and bk2, but the average maximum force required to crease the tissue was about the same (Fig. 5, A and B, insets). Whereas bk2 snapped, wild type continued to bend with continually less resistance to the displacement. Whether interpreted as a consequence of a loss of mechanical strength (Ching et al., 2006) or flexibility, these results indicate an architectural defect in the cell walls of the bk2 mutant (Fig. 5B).

View larger version (41K): [in this window] [in a new window] [as a PowerPoint slide]   Figure 5. Mechanical strength and flexibility of leaves and internodes. A, Examples of stress-strain curves of force versus displacement for bk2 and wild-type leaf sections taken parallel (with vein) and perpendicular (against vein) to the long axis of the leaf blade. Inset, Average values and SD (n = 11–13 samples) for maximum force before breakage for the bk2 and wild-type leaves. B, Examples of stress-strain curves of force versus displacement for bk2 and wild-type stems in a one-point bend test. While both bk2 and wild-type stems reach a maximum force by about 4 mm, all bk2 stems snap at some point upon continued bending, observed as the sudden loss of force applied. Inset, Average values and SDs (n = 22 for wild type, n = 20 for bk2), where F is determined as the maximum force tolerated before severe bending.

We characterized cellulose and noncellulosic monosaccharide distributions in the leaf and internode walls of bk2 and wild-type plants at stages before and after the brittle phenotype developed. Leaf and internode tissues were pulverized in liquid nitrogen, portions were freeze-dried directly to give total dry mass of the tissue, and cell walls were isolated from the remaining material. Relative cellulose content could be quantified as a proportion of the total dry mass and total cell wall mass (Fig. 6, A and B ). Before the appearance of the brittle phenotype, only the juvenile leaves are present and little difference in proportions of wall mass or cellulose content were observed (data not shown). After the appearance of the phenotype, wild-type cellulose mass represents 13% to 17% of the total dry mass of the juvenile and adult leaves and 13% to 19% of the dry mass of the stem, and in each tissue the reduction in cellulose is more drastic with increased maturity of the tissues (Fig. 6A). In contrast, the proportions of the cell wall mass per total dry mass were greater in bk2 than in wild type (Fig. 6A). The proportion of cell wall cellulose varied from 31% to 39% in wild-type leaves and 23% to 33% in bk2 leaves (Fig. 6B). However, while bk2 exhibited a stark cellulose deficiency in the mature stem, the percentage cellulose in the cell walls of bk2 developing stems was slightly greater than that of wild type.

View larger version (30K): [in this window] [in a new window] [as a PowerPoint slide]   Figure 6. Determination of cell wall and cellulose content in juvenile leaves 3 and 4, adult leaves 6 and 7, and mature (S1, basal one-third) and developing internodes (S2, middle one-third, and S3, top one-third below the tassel stem) of wild type and bk2 demonstrating the brittle phenotype. Samples of total mass and isolated walls were weighed and digested with acetic-nitric reagent (Updegraff, 1969). The undigested cellulose was pelleted by centrifugation and washed several times with water. Glc equivalents were determined by phenol-sulfuric acid assay (DuBois et al., 1956) and converted to weight amounts for determination of weight%. Error bars are the mean ± SD of at least four samples. A, Cell wall and cellulose content is calculated as weight% of the total dry mass. B, Cellulose content is calculated as dry mass of isolated cell walls.

View larger version (14K): [in this window] [in a new window] [as a PowerPoint slide]   Figure 7. Noncellulosic cell wall monosaccharide distributions of wild-type and bk2 juvenile and adult leaves after appearance of the brittle phenotype. Samples are digested with 2 M TFA at 120°C for 90 min to hydrolyzed noncellulosic polysaccharides to monosaccharides. The TFA is evaporated, and alditol acetates derivatives are separated and quantified by gas-liquid chromatography. Mol% was calculated based on response-factor values of standards treated similarly during derivatization. A, Juvenile leaves (leaf 3). B, Adult leaves (leaf 6).

Using scanning electron microscopy, we observed specific decreases in secondary wall thickness in sclerids of the rind and surrounding vascular bundles in the bk2 stem (Fig. 2, G and H). We used Fourier transform infrared (FTIR) microspectroscopy to assess wall phenotypes in isolated cell walls from juvenile and adult leaves, and from sclerid-enriched fragments of the rind of mature and developing internodes. FTIR spectroscopy is a vibrational spectroscopic method that can quantitatively detect, without derivatization, the relative amounts of functional groups, including carboxylate esters, phenolic esters, amides and carboxylic acids, and a carbohydrate fingerprint region, with absorbances at certain wave number serving as diagnostic of cellulose, cross-linking glycans, and pectins in maize and other grasses (Tsuboi, 1957; Liang and Marchessault 1959; Séné et al., 1994; Carpita et al., 2001). Thirty to 40 spectra were obtained from wild-type and bk2 plants for each organ and developmental stage. The spectra of cell walls from internodes and leaves were similar, but there were differences between walls of bk2 and wild type for these tissues (Fig. 8 ). Consistent with chemical analyses of cellulose (Fig. 6B), the digital subtraction spectra, subtracting the average spectrum of bk2 walls from that of wild type, show wild-type mature stems are relatively enriched in absorbances characteristic of cellulose (wave numbers 1,157, 1,108, and 1,061 cm^–1) and that of phenolic esters (wave number 1,728 cm^–1) is relatively enriched in bk2 (Fig. 8B). A cellulose deficiency is also observed in bk2 adult leaves, but comparison of digital subtraction spectra of the immature stems and juvenile leaves do not indicate a substantial deficiency in the bk2 mutant (Fig. 8, D, F, and H).

View larger version (30K): [in this window] [in a new window] [as a PowerPoint slide]   Figure 8. Averaged FTIR spectra (A, C, E, and G) and digital subtraction spectra (B, D, F, and H) of isolated cell walls of wild-type and bk2 organs after the appearance of the brittle phenotype. In all spectra, wild-type spectra are in black, and bk2 are in gray scale. A and B, Mature basal internodes at tasseling. C and D, Developing internode taken just below the tassel internode. E and F, Juvenile leaves (leaf 3). G and H, Adult leaves (leaf 7).

View larger version (36K): [in this window] [in a new window] [as a PowerPoint slide]   Figure 9. Principal components analysis of FTIR spectra of cell walls of wild-type and bk2 leaves and stems after the appearance of the brittle phenotype. PC scores plots and PC loadings 1 and 3 of walls of bk2 and wild type from: A and B, mature basal internode; C and D, developing internode just below the tassel internode; E and F, juvenile leaves (leaf 3). G and H, Loadings 4 and 5 of walls of bk2 and wild-type adult leaves (leaf 7). Insets: Values reflect the classification percentages of populations of wild-type and bk2 samples from the first five PC scores. A value of 50% means the populations cannot be discriminated.

Whereas the relative proportion of Klason lignin in juvenile leaves is essentially unchanged before or after the appearance of the brittle phenotype, lignin was markedly enriched in the cell walls of bk2 mature and developing stems (Fig. 10 ). Despite this enrichment in total content, there is significantly less phloroglucinol-staining material in the mutant stem slices and sections compared to those derived from wild-type siblings (Fig. 11, A–C ). This staining difference is only detected in the mature stems, as developing stems of both bk2 and wild-type tissues do not stain. Consistent with these results, we found that cell walls of mature bk2 stems were less autofluorescent than wild type when viewed under UV light (Fig. 11, D and E). These results are consistent with enhanced levels of Klason lignin observed in rice bc1 but differ in the Wiesner (phloroglucinol)-staining behavior (Li et al., 2003). The total amounts of extractable UV-absorbing esterified hydroxycinnamates, primarily p-coumarate, ferulate, and other phenolic substances released upon saponification of isolated walls, were significantly higher in bk2 than in wild type in all tissues (Table I ). Whereas comparable amounts of p-coumaric and ferulic acid (FA) were found in leaves, the stems had substantially higher amounts of p-coumaric acid (pCA) than FA. These results are consistent with the finding of higher levels of phenolic esters by infrared spectroscopy (Figs. 8 and 9).

View larger version (19K): [in this window] [in a new window] [as a PowerPoint slide]   Figure 10. Klason lignin determination of cell walls of juvenile leaves before and after appearance of the brittle phenotype and of mature and developing stems. Klason lignin was determined gravimetrically after correction for ash, according to Theander and Westerlund (1986), as modified by Hatfield et al. (1994). Values are mean ± SD of four samples.

View larger version (89K): [in this window] [in a new window] [as a PowerPoint slide]   Figure 11. The quality of lignin is altered in bk2. A, Phloroglucinol staining of freshly-cut mutant (bk2) and wild-type (wt) stalk sections. B and C, A freshly-cut, phloroglucinol-stained transverse sections of bk2 and wild-type stems. D and E, UV autofluorescence of sections from bk2 and wild-type stems.

Reverse transcriptase-coupled PCR (RT-PCR) was used to examine the transcriptional pattern of Bk2 using gene-specific primers. Bk2 was expressed in most parts of the maize plant (B73 inbred) including roots, shoots, and stems. However, no RT-PCR product could be amplified from plants homozygous either for bk2-ref or bk2-Mu1, indicating that both of these mutant alleles are null.

To address how the expression of the Bk2 gene relates to the developmentally specified phenotype of the mutant, total RNA was also extracted from leaves of B73 at weeks 1, 2, 3, 4, and 8 and subjected to semiquantitative RT-PCR. Bk2 transcripts are more abundant at early stages of growth (weeks 1 and 2) than at later stages (weeks 3, 4, and 8; Fig. 12 ). Northern gel-blot analyses of polyA⁽⁺⁾-RNA confirmed a higher transcript level at weeks 1 and 2 with transcript abundance markedly decreasing in weeks 3 and 4 (Fig. 12).

View larger version (66K): [in this window] [in a new window] [as a PowerPoint slide]   Figure 12. The Bk2 gene is expressed much earlier than the appearance of the brittle phenotype. A, An RT-PCR assay showing that the expression of Bk2 is higher during earlier stages of maize development. In contrast, the expression of the Actin gene in these samples remained unchanged. B, Northern-based expression analysis of Bk2 confirming the RT-PCR data shown in A. PolyA(+)-enriched RNA was used for the northern analysis. Gel blot in the lower section shows the relative amount of polyA(+) RNA in different samples. Leaf materials were pulverized in liquid nitrogen and RNA extracted from aliquots of the frozen powder. For RT-PCR analyses, the topmost fully expanded leaf was used; for polyA(+)-enriched northern gel blots, the entire seedling or plants were used.

	DISCUSSION

TOP ABSTRACT RESULTS DISCUSSION MATERIALS AND METHODS LITERATURE CITED

Mutations in the COB gene of Arabidopsis were so named because of a change in the orientation of cell expansion in the root from the longitudinal axis to the radial axis, widening the root diameter progressively away from the tip and giving roots the appearance of a the head of a cobra (Hauser et al., 1995). Because this change in orientation was accompanied by a drastic lowering in the crystalline cellulose in the root cell walls, Schindelman et al. (2001) suggested that COB is either directly or indirectly involved in cellulose synthesis. The COB gene was identified by map-based cloning and predicted to contain an N-terminal signal sequence for secretion, a highly hydrophobic C-terminal sequence, and a GPI anchor sequence near the C terminus (Udenfriend and Kodukula, 1995a).

The cleavage site in known GPI-anchored proteins, termed , contains only Ser, Asn, Ala, Gly, Asp, and Cys, with only Ala, Gly, Thr, and Ser found at the +2 site (Udenfriend and Kodukula, 1995b). The +2 residue of known GPI-anchored proteins is usually followed by a spacer of five to seven amino acids rich in charged and/or Pro residues followed, in this instance, by a stretch of 20 to 25 hydrophobic residues. The ZmBK2 meets these criteria, with Asn at the site and Ala occupying the +2 position, and a Pro at the +3 position (Fig. 4). The 12-member Arabidopsis COBL gene families are all predicted to encode GPI-anchored proteins. With the exception of two members, COBL6 and COBL9, whose expression is floral specific, the rest of the COBL genes are expressed to different extents in all organs (Roudier et al., 2002). BK2 is most closely related to the rice BC1 and Arabidopsis COBL4, with 88% and 70% identity, respectively. The rice BC1 has a unique 25-amino acid insertion rich in Gly residues that is absent from both maize BK2 and Arabidopsis COBL4 proteins (Fig. 4).

Similar to other COBL proteins, BK2 possesses a conserved region between amino acids 53 to 217, characteristic of a metal-binding domain of plant phytochelatin synthases (E = 3.6 e^–99, with 83%–91% similarity and 72%–80% identity to four plant phytochelatins). Leuchter et al. (1998) complemented the Schizosaccharomyces pombe mutant defective in phytochelatin synthesis by expressing partial COB cDNA, and wheat (Triticum aestivum) COB1 possesses phytochelatin synthase-like activity that improved resistance to cadmium when expressed in Escherichia coli (Wang et al., 2006). A set of conserved Cys residues involved in metal ion binding was also observed (Roudier et al., 2002). While it is unlikely that the COBL proteins function as phytochelatin synthases, the metal-binding domains may be conserved for an alternative function (Wang et al., 2006).

Amino acids 37 to 76 of BK2 also show weak similarity (7%–13%) to a family II cellulose-binding module related to that of several bacterial type II cellulases (Simpson et al., 2000), but the Trp residues found in this region are conserved in BK2 and COBL1-5, and, with the presence of an adjacent conserved Gly, is predicted to confer binding specificity to cellulose over xylan (Linder and Teeri, 1997). COBRA is one of several plant proteins predicted to be GPI anchored that have been solubilized from plasma membranes by a phophatidylinositol-dependent phospholipase C (Roudier et al., 2005). Others include several arabinogalactan proteins (Youl et al., 1998), other cell expansion orientation factors SKU5 (Sedbrook et al., 2002), and SOS5 (Shi et al., 2003), and several disparate types of proteins, including hydrolases, proteases, and receptor-like kinases (Borner et al., 2003).

Although mutation of COBRA and other GPI-anchored proteins affect orientation of expansion, this phenotype is not observed in either bc1 or bk2 mutants. A reduction in secondary cell wall deposition indicates that the activities of these COBL proteins may be restricted to wall deposition in specialized cells after elongation. An independent expression-profiling study showed that the most closely related Arabidopsis COBL4 is coregulated with secondary wall cellulose synthases (Brown et al., 2005). Three independent insertional cobl4 mutants exhibit an irregular xylem phenotype and display a brittle phenotype (Brown et al., 2005).

There are no reports of irregular xylem phenotypes in grass mutants. In dicots, defects in secondary cell wall formation, whether due to defects in cellulose synthases or other components, lead to the collapsed or irregular xylem phenotype (Turner et al., 2001). However, none of the cellulose-deficient irx cell wall mutants having defects in CesA genes has a brittle phenotype. Only irx2, defective in KORRIGAN (KOR), a membrane-associated endo-(1 4)-β-D-glucanase gene (Lane et al., 2001), and cobl4, defective in an ortholog of the rice Bc1 and maize Bk2 genes (Brown et al., 2005), exhibit brittleness in addition to the irx phenotype and reduced cellulose content. These results suggest that COBRA has a function distinct from a role in cellulose biogenesis.

What Is Directly Responsible for the Brittle Phenotype?

The appearance of the brittle phenotype coincides with the transition from juvenile to adult leaves and a qualitative change in the deposition of lignin. This is apparent from various phase-change and brown midrib mutants (Moose and Sisco, 1994; Marita et al., 2003; Lauter et al., 2005). Whereas cellulose content is compromised in the bk2 mutant (Fig. 6), lignin deposition is enhanced (Fig. 10). Lack of a proper cellulosic matrix in the secondary wall may lead to an aberrant lignin polymerization, rendering it incapable of withstanding bending stress.

Despite the drastic increase in lignin content of the bk2 stems, acid phloroglucinol-stained material is reduced in the cells of these stems. This finding is in sharp contrast to that observed in rice (Li et al., 2003). These differences may be attributed to different lignin structures in different species with different stem architectures. Alternatively, the increases in lignin content may not be restricted to the vascular bundles and rind tissues in maize. Although the Wiesner stain reacts with cinnamaldehyde residues in lignin and the color intensity approximately reflects the total lignin content, our anomolous results indicate that there may be other changes in cell wall phenolics that contribute to cell wall flexibility. Our infrared analyses indicate that phenolic esters are increased in bk2 relative to wild type. We measured increased levels of hydroxycinnamic acids in bk2, consistent with the increased amounts of (arabino) xylans. Ferulate esters attached to the arabinosyl units of the xylan become cross-linked into the lignin structure by diverse kinds of linkages (Iiyama et al., 1994). Coniferaldehyde is the precursor of FA (Nair et al., 2004) and enhanced ferulate ester production may drain the aldehyde pools, resulting in decreased phloroglucinol staining.

Bk2 Expression Prestages the Appearance of Phenotype

Consistent with a secondary wall function, rice Bc1 demonstrates expression associated primarily with vascular bundles and the rind (Li et al., 2003). However, maize Bk2 expression is highest very early in leaf development but greatly reduced when the brittle phenotype presents itself (Fig. 12). The decrease in cellulose but not wall mass occurs in leaves before the onset of the brittle phenotype, and while it may predispose the organs to the brittle nature, it cannot be the direct cause. Such an expression pattern was observed not only when whole seedlings/plants were examined, but also when different plant parts such as leaf blades, leaf sheaths, or tiny stems were excised and examined separately (data not shown). Given the anatomical alterations in vascular bundle distribution as well as secondary wall defects, early expression predisposes the tissues to brittleness long before the onset of the actual phenotype. In line with our findings, Bc1 also expresses most strongly in tiller buds, long before they develop sclerenchyma cells (Li et al., 2003).

Does BK2 function similarly to other COBL proteins that function during primary wall growth and cellulose deposition? Alterations in cell walls can be an indirect consequence of mutations in more global developmental programs and expression of cell wall-related genes is modified in many different signal transduction pathways. Therefore, a cell wall mutant phenotype is not necessarily a result of a mutation in a gene directly involved in cell wall synthesis or assembly. For example, cellulose deficiency is a consequence of several kinds of mutations other than a defect in cellulose synthase catalytic subunits. Mutations in another plasma membrane-associated protein, kobito (Pagant et al., 2002), an endoglucanase, kor (Nicol et al., 1998), and a katanin-like microtubule-severing protein fra2 (Burk and Ye, 2002), all result in cellulose deficiencies. Further, plants also compensate cellulose deficiencies by altered wall metabolism. For example, tomato and tobacco (Nicotiana tabacum) cells with type I walls habituated to grow in cellulose synthesis inhibitors increased pectin cross-linking, whereas barley cells with type II walls do so by increased phenolic cross-linking (Shedletzky et al., 1992). Thus, one should not infer that all of these genes have a primary function in the biosynthesis of cellulose but that several pathways may regulate cellulose biosynthesis as a downstream target. The observation that total wall mass actually increases as cellulose is lowered in bk2 (Fig. 6A) indicates that COBLs function in global integration of wall extension and deposition rather than directly in cellulose synthesis. The location of COBRA proteins at the interface of plasma membrane and cell wall and their association with lipid microdomains (rafts) suggest a potential integrative function in regulation or integration of secretion and wall assembly (Martin et al., 2005).

	MATERIALS AND METHODS

TOP ABSTRACT RESULTS DISCUSSION MATERIALS AND METHODS LITERATURE CITED

 

Plant Material

The maize (Zea mays) mutant bk2 was obtained from the Maize Genetics Cooperative as a stock (no. 916C) that had been maintained over many generations as a backcross progeny. To have a uniform background, the material was propagated by backcrossing bk2 mutants with their wild-type heterozygotes an additional five times. These backcrosses resulted in a uniform background where the mutant (bk2/bk2 homozygotes) and wild-type plants (bk2/+ heterozygotes) were indistinguishable from each other except for the appearance of a brittle phenotype at 4 weeks in the bk2 mutants. In this backcross progeny, the mutant and wild-type plants segregated in a 1:1 ratio of homozygotes to heterozygotes. Upon harvest of plant material, it was frozen immediately with liquid nitrogen before being placed at –80°C until all samples had been collected.

Identification of RFLP Markers Flanking bk2

We generated another backcross population of bk2 with B73. One hundred and fifty plants from this population were assessed for the brittle phenotype, as well as the genotype of a number of RFLP markers known to map in the vicinity of bk2 on the proximal part of the long arm of chromosome 9. The data obtained placed bk2 between UMC95 and BNL8.17, with BNL8.17 being 4.0 cM proximal and UMC95 2.5 cM distal to the bk2 locus.

Southern and Northern Gel-Blot Analyses and PCR

DNA extraction and Southern gel-blot analyses were done as described before (Multani et al., 2003). For RNA, leaf materials were collected in bags and immediately frozen in liquid nitrogen and stored at –80°C. For RT-PCR analyses, the topmost fully expanded leaves at each age were used, whereas for polyA-enriched northern gel blots, the entire seedling or plants were used. Before extraction, the leaf materials were pulverized in liquid nitrogen and RNA extracted from aliquots of the frozen powder. Total RNA was extracted using LiCl₂ precipitation, as described by Lagrimini et al. (1987). PolyA⁽⁺⁾ mRNA fraction was isolated from total RNA using the PolyATract mRNA Isolation System II kit from Promega. For northern gel-blot analysis, the PolyA⁽⁺⁾ enriched RNA was fractionated on 1.2% (w/v) agarose gels. Samples of uniform quantity were prepared and mixed with 2x loading dye (547 µL of formamide, 182.5 µL of formaldehyde, 200 µL of 10x MOPS buffer {200 mM MOPS[Na], pH 7.0, containing 80 mM sodium acetate}, 66 µL of glycerol, 4 µL of 0.5 M EDTA, 5 µL of 10 mg mL^–1 EtBr), and heated for 65°C to 70°C for 10 min and then immediately placed in ice before loading. Transfer, hybridization, and subsequent exposure of blots to x-ray films were as described in Multani et al. (2003).

The PCR reaction conditions and the Mu-TIR primer to seek the insertion of a Mu element in bk2-Mu1 were essentially as described before (Multani et al., 2003), except we added 5% glycerol to the reaction mixture. Various bk2-specific primers used in this study were: 1F, 5'-GCGATGGTGACGATGAGCAAC-3'; 3F, 5'-ACTGGCACGTCAAGCTCAAC-3'; 4F, 5'-TGCGTCTCCTTCTCCTCCTTC-3'; http://www.idtdna.com/analyzer/Applications/Instructions/Default.aspx?AnalyzerDefinitions=true - MeltTemp; 5F, 5'-GGACTCCTACCCCTACCTGC-3'; 6F, 5'-TGCACACACACTGGCACTTG-3'; http://www.idtdna.com/analyzer/Applications/Instructions/Default.aspx?AnalyzerDefinitions=true - MeltTemp; 7F, 5'-TTTTATTCGCCCTGCCTCTAG-3'; 8F, 5'-ATCTAGCTGTGCCTGTGCCTG-3'; 1R, 5'-CGATGGTGCTGTTGTAGAAGGAG-3'; http://www.idtdna.com/analyzer/Applications/Instructions/Default.aspx?AnalyzerDefinitions=true - MeltTemp; 3R, 5'-ACCAGGAGCAGCAACAGTAGG-3'; 4R, 5'-TTGAGCCCGTAGAACATGCC-3'; 5R, 5'-GGGTTAGCAAATGTTCAGGCAAC-3'; 7R, 5'-ATGGACCAGATCACCTCCTTC-3'; 8R, 5'-ATGGTAACCGAACCTGCAACG-3'.

To elucidate the molecular nature of the structural polymorphism in bk2-ref, the Expand High Fidelity PCR system of Roche was used. The reaction conditions were set up as recommended by the manufacturer, except for the addition of 5% (v/v) glycerol. The cycle was repeated 30 times and the final elongation time was 15 min at 72°C. All amplified fragments were cloned in the pGEM-T vector from Promega.

For RT-PCR, total RNA was isolated from greenhouse-grown B73 plants and 1 µg of total RNA was used for each reaction for all samples. The first strand of cDNA was synthesized using the RETROscript kit from Ambion, essentially as described by the manufacturer. Next, 5 µL from each primary reaction was used for the second-step PCR reaction using bk2-specific primers 1F and 1R. RT-PCR controls were derived from the actin gene that was amplified using the following primers: forward primer 5'-TGTTTCGCCTGAAGATCACCCTGTG-3'; reverse primer 5'-TGAACCTTTCTGACCCAATGGTGATGA-3'.

Light and Scanning Electron Microscopies

The third stalk internode of greenhouse-grown plants at anthesis was used for microscopy. For scanning electron microscopy, stalk slices (approximately 1 mm in height) were prepared with a sharp razor and observed directly with a JFM-840 SEM (JEOL). For light microscopy, stalk slices were cut with a razor blade and fixed in 3% glutaraldehyde in 0.1 M potassium phosphate buffer, pH 6.8, under low vacuum in the desiccators overnight. Following dehydration, the samples were embedded in JB-4 resin following manufacturer's instructions (Electron Microscopy Sciences), and then sectioned (5-µm thickness) using a SORVALL type JB-4 Porter-Blum microtome (DuPont Company). The sections were dropped into a water bath (approximately 60°C) to expand them and four slides containing three to four sections each were prepared for each sample. They were either viewed under bright field or UV for autofluorescence (Nikon E800 Fluorescence Microscope; Fryer Company). For lignin staining, stalk slices were cut freehand and then stained with Wiesner's solution (2%, w/v) phloroglucinol in ethanol:50% HCl (v/v) in water (95:5, v/v). Samples of material were stained immediately or were first fixed in formalin-acetic acid-ethanol overnight before staining, and imaged with either Olympus Vanox-S brightfield or Nikon SMZ-U stereo microscope.

Leaf and Internode Tensile Strength Measurements

The tensile strength of bk2 and wild-type leaves was tested using a tensile grip probe with a TA.XTplus texture analyzer (Texture Technologies) on 2-month-old sections from the sixth leaf (counted from the bottom). Leaf sections (2 mm x 20 mm), avoiding the midrib and major veins, were cut with razor blades either parallel or perpendicular to the minor veins. The thickness of the leaf was measured with a Nikon SMZ 1500 microscope. The leaf section ends were firmly attached to tensile grip plates with an inelastic adhesive (Advanced Formula Instant Krazy glue), leaving a 5-mm gap between the plates. The tensile grip plates were moved apart at 0.1 mm per second. The Texture Exponent 32 version 4 software displayed fracture breakage of the leaf as distance (mm) by force (g). Maximum stress tolerated before breakage was calculated from force measured at breakage and the total tissue load-bearing volume.

A simple one-point bend test was performed to determine brittleness of bk2 and wild-type internodes. Individual fresh stem sections 20-cm long were cut, and the diameters of the cut section ends were measured with a Nikon SMZ 1500 microscope. The one-point bend rig was attached to the TA.XTplus texture analyzer and the probe was raised to a defined height and calibrated. The stem section was placed across two supports of the rig 40 mm apart, with 20 mm on either side of the supports. The probe for the one-point bend test moved toward the sample at a speed of 0.5 mm per second. Stress at fracture was calculated according to the equation:

where is the stress at fracture, F is the maximum force before deformation, L is the distance between the supports, and d is the stem diameter (Timoshenko and Young, 1962).

Cell Wall Isolation

The cell walls of various maize tissue samples were pulverized in liquid nitrogen with a mortar and pestle and divided into two portions. One was freeze dried to yield total dry mass, and for the second, cell walls were isolated by homogenization in 1% SDS in 50 mM Tris[HCl], pH 7.2, with a glass-glass motorized grinder (Kontes-Duall, Thomas Scientific). The samples were then heated at 80°C in 1% SDS for 15 min. All samples were washed five times with water through centrifugation and pelleting before the cells were broken again with the glass-glass motorized grinder. Samples were washed three times with water, two times with warm 50% ethanol, and three times with water. The remaining pellet was resuspended in water and freeze dried.

Determination of Cellulose, Lignin, and Saponifiable Phenolic Substances

Samples of total dry mass, ethanol-extracted material, and isolated cell walls were assayed for cellulose by acetic-nitric assay (Updegraff, 1969). Briefly, a known weight of isolated cell walls was hydrolyzed in 3 mL of acetic acid-nitric acid in water (8:2:1, v/v/v) at 100°C for 90 min in sealed 5-mL conical Reacti-vials (Pierce Chemical). The insoluble materials (mostly cellulose) were then pelleted by centrifugation and the pellet was washed five times with water. Cellulose in the pellet was assayed for Glc equivalents by phenol-sulfuric colorimetric assay (Dubois et al., 1956). Klason lignin determinations were performed as described by Theander and Westerlund (1986) as modified (Hatfield et al., 1994). To determine total saponifiable hydroxycinnamates and other phenolic substances, isolated walls were suspended in 1 M NaOH for 30 min at 37°C. After cooling, HCl was added to 1 M in excess of NaCl and the phenolic substances were partitioned into ethyl acetate. The ethyl acetate was evaporated and the residues resuspended in 1 mL of 50% (v/v) methanol in water. Absorbance was measured between 200 and 400 nm to estimate total phenolics per mg cell wall. The phenolic substances were separated by reverse-phase HPLC in a gradient of 1.5% (v/v) acetic acid in water to 35% acetonitrile (v/v) in 1.5% acetic acid in water (v/v), and FA and pCA were quantified by UV absorbance based on standards treated similarly (Hemm et al., 2003).

Pyrolysis-mass spectrometry is a rapid analytical technique to determine plant cell wall composition (Ralph and Hatfield, 1991; Vermerris and Boon, 2001). Pyrolysis-mass spectrometry analyses of stem and leaf tissue from wild-type and bk2 cell wall preparations were performed on a Varian 1200 mass spectrometer (Varian) equipped with a PC-3 direct exposure probe (SIS). Between 5 and 10 µg of cell wall material in methanol was applied to a platinum filament and air dried. The sample was pyrolyzed inside the ion source of the mass spectrometer by resistive heating of the filament with a current of 0.5 amp at about 500°C. The pyrolysate was ionized using electron impact ionization at 70 eV, and the mass spectrometer was operated at 1,200 V. The mass range was scanned from m/z 50 to 350 with a cycle time of 0.199 s. Samples were analyzed in triplicate. The data were collected and analyzed using Varian WorkStation software.

Determination of Noncellulosic Sugars

Samples of isolated cell walls (approximately 0.5–1 mg) were hydrolyzed with 2 M TFA containing 400 nmol of myoinositol as internal standard for 90 min at 120°C in 1-mL conical Reacti-vials (Pierce Chemical). The cellulose and other undigested polymers were pelleted by centrifugation. This material was suspended in water and assayed for Glc equivalents by phenol-sulfuric colorimetric assay (Dubois et al., 1956). The TFA containing the soluble sugars was evaporated under a stream of air and the sugars were reduced with NaBH₄ and converted to alditol acetates as described previously (Gibeaut and Carpita, 1991). The alditol acetates were separated by gas-liquid chromatography on a 0.25-mm x 30-m vitreous silica capillary column of SP-2330 (Supelco). Temperature was held at 80°C for 1 min upon injection then programmed from 80°C to 170°C at 25°C per min, then to 240°C at 5°C per min, with a 6-min hold at the upper temperature. Monosaccharide identity, as pentose, hexose, and deoxysugar, was confirmed by electron-impact mass spectrometry (Carpita and Shea, 1989). Equimolar standards were also converted to alditol acetates to calculate response factors for quantitation of mol% relative to the myoinositol standard.

FTIR Microspectroscopy

For FTIR microscopy, approximately 5 µL of a slurry of maize cell walls from wild-type and bk2 maize leaves and stems were spotted in individual wells of IR-reflective, gold-plated microscope slides (Thermo-Electron). The slides with cell wall samples were supported on the stage of a Continuum series microscope accessory to a 670 IR spectrophotometer with a liquid nitrogen-cooled mercury-cadmium telluride detector (Thermo-Electron). Spectra were taken in the transflectance mode, where a beam is transmitted through the wall sample, reflected off the gold-plated slide, and then is transmitted through the sample a second time before detection. One hundred twenty-eight scans were coadded at 8 cm^–1 resolution. Each set of 10 independent samples was spotted at least five times, with a minimum of five spectra taken from isolated primary cell walls in each spot from different areas of the sample. The resulting spectra were then analyzed with WIN-DAS software (Kemsley, 1998).

Sequence data from this article can be found in the GenBank/EMBL data libraries under accession numbers AY328909 (rice BC1 cDNA), DQ139874 (maize Bk2 cDNA), and AL391144 (Arabidopsis COBL4).

	ACKNOWLEDGMENTS

 
We thank Javier Campos for his assistance with the Klason lignin determinations, and Jason Stout, in the laboratory of Clint Chapple, for determinations of the hydroxycinnamic acids by HPLC. This is journal article number 2007–18194 of the Purdue Agricultural Experiment Station.

Received May 19, 2007; accepted October 1, 2007; published October 11, 2007.

	FOOTNOTES

 
¹ This work was supported by a grant from the National Science Foundation Plant Genome Research Program (to N.C.C., M.C.C., and W.V.), Purdue University start-up funds (to G.J.), and University of Florida funds to purchase a pyrolysis mass spectrometer (to W.V.).

² Present address: Department of Plant Pathology, Iowa State University, Ames, IA 50011.

³ Present address: Department of Biological Sciences, University of Missouri, Columbia, MO 65211.

⁴ Present address: Pioneer Hi-Bred International, P.O. Box 1000, Johnston, IA 50131.

The author responsible for distribution of materials integral to the findings presented in this article in accordance with the policy described in the Instructions for Authors (www.plantphysiol.org) is: Gurmukh Johal (gjohal@purdue.edu).

^[C] Some figures in this article are displayed in color online but in black and white in the print edition.

^[OA] Open Access articles can be viewed online without a subscription.

www.plantphysiol.org/cgi/doi/10.1104/pp.107.102582

^* Corresponding author; e-mail gjohal{at}purdue.edu.

	LITERATURE CITED

TOP