•The evolution of compositional polymers and their complex arrangement and deposition in the cell walls of terrestrial plants included the acquisition of key protein functions.
•A membrane-bound endoglucanase, termed Korrigan (KOR), has been shown to be required for proper cellulose synthesis. To date, no extensive characterization of the gymnosperm KOR has been undertaken.
•Characterization of the white spruce (Picea glauca) gene encoding KOR (PgKOR) shows conserved protein features such as polarized targeting signals and residues predicted to be essential for catalytic activity. The rescue of the Arabidopsis thaliana kor1-1 mutant by the expression of PgKOR suggests gene conservation, providing evidence for functional equivalence. Analyses of endogenous KOR expression in white spruce revealed the highest expression in young developing tissues, which corresponds with primary cell wall development. Additionally, RNA interference of the endogenous gymnosperm gene substantially reduced growth and structural glucose content, but had no effect on cellulose ultrastructure.
•Partial functional conservation of KOR in gymnosperms suggests that its role in cell wall synthesis dates back to 300 million yr ago (Mya), predating angiosperms, which arose 130 Mya, and shows that proteins contributing to proper cellulose deposition are important conserved features of vascular plants.
Plant evolution has been characterized by distinctly defined adaptive events, including the development of a stiff, upright stem that permitted the expansion of the growth habit of plants migrating from an aquatic to a terrestrial ecosystem. What were the key factors in the evolution of the cell wall of terrestrial plants? It is possible that the evolution of the proteins necessary for the formation of the critical cellulose–hemicellulose network of the cell wall was a key element in allowing plants to move from aquatic to terrestrial environments.
The arrangement of polysaccharides in both primary and secondary cell walls is an important determinant for cell shape, growth and cell wall properties in vascular plants (Estevez et al., 2009; Siddhanta et al., 2009). This unique arrangement could be a key factor in cell wall evolution. The cellulose of nonvascular plants has a higher level of crystallinity and a larger surface area than higher plant cellulose (Ek et al., 1998; Stromme et al., 2002). In addition, highly specialized proteins such as endotransglycosylases and expansins are present in higher plants that modify the complex cross-linking of cellulose with hemicelluloses during growth and morphogenesis (Cosgrove, 2005). It is this cross-linking that permits the cell wall to be strong yet flexible. Some of these proteins act on the hemicellulose–cellulose network in the cell wall, while other proteins act during cellulose deposition.
One particular group of enzymes, the endoglucanases (E.C.188.8.131.52), have the capacity to hydrolyse the β-1,4 linkage of the cellulose chain, and have been shown to be more active on amorphous cellulose than on crystalline cellulose (Carrard et al., 2000). A membrane-bound endoglucanase called Korrigan (KOR) has been shown to be required for synthesis of the ordered, load-bearing cellulose–hemicellulose network (Nicol et al., 1998; Sato et al., 2001). KOR was originally isolated in a mutant Arabidopsis thaliana plant (kor1-1) that showed pronounced architectural alterations in the primary cell wall when grown in the absence of light (Nicol et al., 1998). It was further shown that KOR possesses a single N-terminal membrane-spanning domain and therefore presumably acts at the plasma membrane–cell wall interface (Nicol et al., 1998). Additional KOR mutations (kor1-2) have been isolated and shown to cause the formation of aberrant cell plates, incomplete cell walls, and multinucleated cells, leading to abnormal seedling morphology (Zuo et al., 2000). In addition, the identification of irregular xylem mutants of KOR, such as irregular xylem 2 (Szyjanowicz et al., 2004), indicates that KOR is required for normal xylem vessel development. Related membrane-bound endoglucanases have been identified in tomato (Lycopersicon esculentum; Brummell et al., 1997a,b), oilseed rape (Brasica napus; Molhoj et al., 2001a,b), rice (Oryza; Zhou et al., 2006), and loblolly pine (Pinus taeda; Nairn et al., 2008).
Given the irregular xylem phenotype in Arabidopsis thaliana, it is interesting that expression of sequences related to the A. thaliana KOR (AtKOR) was correlated with cellulose deposition in secondary xylem of Populus spp., suggesting a role for KOR in wood formation (Sterky et al., 1998; Hertzberg et al., 2001; Bhandari et al., 2006). Hybrid poplar endoglucanase activity was experimentally demonstrated (Rudsander et al., 2003; Master et al., 2004). Within the angiosperms, KOR from hybrid aspen (Populus tremula L. × tremuloides Michx.; PttKOR) complemented the A. thaliana mutants and overexpression of the PttKOR in A. thaliana led to lower cellulose crystallinity (Takahashi et al., 2009). When KOR expression was suppressed by RNA interference (RNAi) in hybrid poplar (Populus alba × grandidentata; PaxgKOR), there was less cellulose present but it was more crystalline, while levels of the hemicellulose xylan increased (Maloney & Mansfield, 2010). In addition, the irregular xylem phenotype shown in A. thaliana kor mutants was also observed for poplar secondary xylem vessels (Maloney & Mansfield, 2010).
Although it has been speculated that KOR predates the split between angiosperms and gymnosperms (Molhoj et al., 2002; Nairn et al., 2008), to date no detailed characterization of a gymnosperm KOR has been undertaken. Should functional conservation of these membrane-bound endoglucanases in the gymnosperms exist, it would suggest that their role in vascular plant evolution dates back to 300 million yr ago (Mya), predating angiosperms, which arose 130 Mya. Here we report on the isolation and characterization of a membrane-bound endoglucanase gene (denoted PgKOR) from white spruce (Picea glauca). The rescue of the A. thaliana kor1-1 irregular xylem and dwarf phenotype by the expression of PgKOR provides evidence for functional equivalence of the gymnosperm gene in the angiosperm. Analyses of endogenous KOR expression in white spruce revealed the highest expression in young developing tissues, which corresponds with primary cell wall development. Additionally, decreased expression of PgKOR using RNAi-mediated suppression in white spruce trees resulted in substantially reduced growth and structural glucose content, but had no effect on cellulose ultrastructure. The lack of an effect on the cellulose ultrastructure suggests that, while PgKOR can partially replace the function of AtKOR, there still might be some functional divergence after the split of gymnosperms and angiosperms.
Materials and Methods
PgKOR isolation and construct development
RNA was extracted from the green leader portion of Picea glauca (Moench) Voss line 653 stem tissue. To identify and characterize the white spruce homologue of the Arabidopsis thaliana (L.) Heynh. KOR gene, the initial sequence was obtained by blasting against a spruce expressed sequence tag (EST) library (http://www.treenomix.ca), from which a set of contigs were selected based on their 73% sequence similarity to the AtKOR gene. A large portion of the gene was isolated with a forward primer (PtrKORFw) designed using the already characterized Populus tremuloides KOR (PtrKOR; GenBank AY535003), which has 83% similarity with the AtKOR sequence, and a reverse primer (PgKORESTRv) designed using the EST screen. In order for any amplification to occur, the forward primer had to be designed within the PtrKOR open reading frame (ORF), and we were therefore unable to amplify the entire PgKOR ORF. In order to obtain the remainder of the ORF, 5′ RLM RACE (RNA Ligase Mediated Rapid Amplification of cDNA Ends; Ambion, Austin, TX, USA) was employed to complete the 5′ sequence of the gene. Following sequence verification, oligonucleotides (PgKORgateFw and PgKORgateRv) were designed to generate a PCR product from cDNA for direct cloning into the pENTR/D-TOPO vector (Invitrogen, Carlsbad, CA, USA). The clone was then sequenced to verify the correct insertion of the PgKOR gene, and Gateway technology (Invitrogen) was used to insert the gene into the 35S::hRLUC::attR destination vector (Subramanian et al., 2004). The resulting binary plasmid, 35S::hRLUC::PgKOR, was transformed into Agrobacterium tumefaciens strain GV3101 and used for the complementation of the kor1-1 mutant.
The PgKOR genomic sequence was PCR-amplified using genomic DNA from the green leader portion of P. glauca 653 stem tissue and four sets of oligonucleotides designed within the PgKOR ORF (PgKORgen1FW&RV, PgKORgen2FW&RV, PgKOEgen3FW&RV and PgKORgen4FW&RV). The sequencing results for the four PCR fragments were then assembled to obtain the entire PgKOR genomic DNA sequence.
The PgKOR-RNAi construct was built using two oligonucleotides, PgRNAiFW and PgRNAiRV, with the addition of either 5′BamHI and 3′ClaI (sense) or 5′XhoI and 3′KpnI (antisense) restrictions sites. These oligonucleotides were used to amplify a 400-base pair fragment of the PgKOR coding region from cDNA. The fragments were then digested with the appropriate restriction enzymes and ligated into the pKANNIBAL (Helliwell & Waterhouse, 2003) cloning vector. Finally, the NotI fragment from pKANNIBAL containing the hairpin RNA (hpRNA) cassettes was subcloned into the binary vector pART27 (Gleave, 1992) and used for plant transformations.
Amino acid sequence alignment for PgKOR with all of the known A. thaliana glycosyl hydrolase family 9 proteins and all the other known plant membrane-bound endoglucanases was performed with ClustalW in the BioEdit program (Hall, 1999). The evolutionary history was inferred using the neighbour-joining method (Saitou & Nei, 1987). The optimal tree with the sum of branch length = 7.72127000 is shown in Fig. 2. The percentage of replicate trees in which the associated taxa clustered together in the bootstrap test (1000 replicates) is shown next to the branches (Felsenstein, 1985). The tree is drawn to scale, with branch lengths given in the same units as those of the evolutionary distances used to infer the phylogenetic tree. The evolutionary distances were computed using the JTT matrix-based method (Jones et al., 1992) and are given in units of the number of amino acid substitutions per site. The analysis involved 49 amino acid sequences and all positions containing gaps and missing data were eliminated. There were a total of 253 positions in the final data set. The final tree is rooted to an Oryza sativa endoglucanse (GenBank #BAG94612), which is not a class IX member. Evolutionary analyses were conducted in mega5 (Tamura et al., 2011).
Plant strains and growth conditions
The T-DNA insertion mutant A. thaliana line kor1-1 (Nicol et al., 1998) was acquired from the Arabidopsis Biological Resource Center (Columbus, OH, USA). The T-DNA insertion of kor1-1 carries a gene that results in resistance to kanamycin. Seeds were sterilized and germinated at room temperature in a 16 h light : 8 h dark cycle on half-strength Murashige and Skoog (MS) medium (Murashige & Skoog, 1962) containing 50 mg l−1 kanamycin sulfate (Sigma, St Louis, MO, USA) and no sugar. Plantlets were transferred to soil after the first four primary leaves had emerged, and growth until maturity was allowed to continue under the same conditions. The F1 progeny from the kor1-1 line were compared with the Wassilewskija (WS) ecotype, the background for kor1-1.
White spruce somatic embryo tissue (Pg653) was acquired from Dr K. Klimaszewska (CFS, Quebec, Canada). Embryos were matured according to Klimaszewska et al. (2001). Plantlets were grown for 18 months, and then destructively harvested.
kor1-1 plants were transformed by floral dip using A. tumefaciens carrying the PgKOR construct. The harvested seeds were selected on MS medium containing 50 mg ml−1 kanamycin and 85 mg ml−1 PESTANAL (glufosinate ammonium; Sigma). Plantlets resistant to both selection agents were transferred to soil and placed in a growth chamber and grown under the same conditions described previously. Pg653 somatic embryo tissue was transformed with the PgKOR-RNAi construct according to Klimaszewska et al. (2001).
Genomic DNA extraction and screening
Genomic DNA was extracted from either the green leader portion of P. glauca 653 stem tissue or single leaves taken from young soil-grown A. thaliana plants using a modified hexadecyltrimethylammonium bromide (CTAB) extraction method (Rogers & Bendich, 1994). Briefly, tissue was placed in microcentrifuge tubes and ground to a powder with liquid nitrogen. Then 1 ml of CTAB extraction buffer (2% (w/v) CTAB (Sigma), 100 mM Tris-HCl, pH 8.0, 1.4 M NaCl, 20 mM EDTA, 1% (w/v) polyvinylprylidone, and 0.2% (v/v) 2-mercaptoethanol) was added to each tube, and the tube was incubated at 65°C for 60 min. An equal volume of chloroform was added, and the tube was vortexed and then centrifuged in a microcentrifuge for 10 min. Genomic DNA was precipitated from the aqueous phase by the addition of 1 volume of isopropanol, incubation at −20°C for 10 min, and centrifugation for 5 min. Genomic DNA was re-suspended in RNase buffer (25 mM Tris-HCl, pH 7.5, 10 mM EDTA, and 100 mg ml−1 RNase A) and incubated at 37°C for 30 min. Two volumes of ethanol were added, and the genomic DNA was recovered by centrifugation. Finally, the DNA was re-suspended in 50 μl of EB buffer (10 mM Tris-HCl, pH 8.0, and 1 mM EDTA), quantified at A260, and stored at 4°C. PCR was performed on this DNA to determine the genotype (homozygous or heterozygous for the T-DNA insertion) of the plants with the genomic primers (RP and LP) specifically designed to amplify genomic regions flanking the T-DNA insertion, as well as the T-DNA specific TAG7 primer, according to Nicol et al. (1998). The presence or absence of the PgKOR construct was determined with the primers hLUC3′ Fw and PgKORgate Rv (for all primer sequences, see Supporting Information Table S1).
RNA extraction and real-time PCR
Total RNA was extracted from c. 500 mg of frozen ground 8-wk-old whole A. thaliana plants using TRIzol reagent (Sigma) according to the manufacturer’s instructions. Total RNA was isolated from white spruce tissues according to Kolosova et al. (2004). RNA yield was measured by absorption at 260 nm, and 10 μg was treated with DNAase (TURBO DNA-free; Ambion). Then 1 μg of the resulting DNA-free RNA was evaluated on a 1% Tris-acetate EDTA agarose gel in order to determine quality. Equal quantities of RNA (1 μg) were used for the synthesis of cDNA with SuperScript II reverse transcriptase (Invitrogen) and (dT)16 primers, according to the manufacturer’s instructions. Samples were run in triplicate with Platinum SYBR Green qPCR Master mix (Invitrogen) on an Mx3000p real-time PCR system (Stratagene, La Jolla, CA, USA). The real-time PCR analysis of the A. thaliana lines was performed using the primers PgKORRTFw and PgKORRTRv or AtKORRTFw and AtKORRT3′UTRRv, while the endogenous PgKOR expression in the wild type and the PgKOR-RNAi lines was performed using the primers PgKORRNAiRTFw and PgKORRNAiRTRv. After analysis of the dissociation curves to ensure single band amplification, critical threshold (Ct) values were quantified in triplicate and transcript abundances were determined based on changes in Ct values relative to elongation initiation factor 5A (Gutierrez et al., 2008; primers AtEIF5AFw and AtEIF5ARv) for A. thaliana and actin (primers PgActinFw and PgActinRv) for white spruce using the following equation: . Conditions for all PCR reactions were as follows: 95°C for 10 min, followed by 40 cycles of 95°C for 30 s, 55°C for 1 min, and 72°C for 30 s.
Structural carbohydrate analyses
Ten-wk-old A. thaliana stems and 18-month-old glasshouse-grown white spruce stems were ground in a Wiley mill to pass a 0.4-mm screen (40 mesh) and Soxhlet-extracted overnight in hot acetone to remove extractives. Lignin and carbohydrate contents were determined with a modified Klason (Coleman et al., 2009), in which extracted ground stem tissue (50 mg) was treated with 3 ml of 72% H2SO4 and stirred every 10 min for 2 h. Samples were then diluted with 112 ml of deionized water and autoclaved for 1 h at 121°C. The acid-insoluble lignin fraction was determined gravimetrically by filtration through a pre-weighed medium coarseness sintered-glass crucible, while the acid-soluble lignin component was determined spectrophotometrically by absorbance at 205 nm. Carbohydrate contents were determined by using anion exchange high-performance liquid chromatography (Dx-600; Dionex, Sunnyvale, CA, USA) equipped with an ion exchange PA1 (Dionex) column, a pulsed amperometric detector with a gold electrode, and a SpectraAS3500 auto injector (Spectra-Physics, Santa Clara, CA, USA).
Microfibril angle estimates were generated by X-ray diffraction. The 002 diffraction spectra from five individual 18-month-old white spruce trees from each of the transgenic lines and 10-wk-old A. thaliana stems were screened for T value distribution and symmetry on a Bruker D8 discover X-ray diffraction unit equipped with an area array detector (general area detector diffraction system (GADDS)). Wide-angle diffraction was used in the transmission mode, and the measurements were performed with CuKα1 radiation (λ = 1.54 Å). The X-ray source was fitted with a 0.5-mm collimator, and the scattered photon was collected by a GADDS detector. Both the X-ray source and detector were set to theta = 0°. The degree of cellulose crystallinity was determined for white spruce and A. thaliana stems as described previously (Mansfield et al., 1997; Coleman et al., 2009) using the same X-ray parameters as for microfibril angle (MFA) determination, with the exception of the source theta which was set at 17°. Significant differences from the wild type were determined using a Student t-test.
Cross-sectional staining and microscopy
A 1-cm segment was cut from the base of 18-month-old white spruce stems and submerged in dH2O at room temperature for 1 d. Samples were then radially cut into 20-μm cross-sections using a Leica SM2000r hand sliding microtome (Leica Microsystems, Wetzlar, Germany) and again stored in dH2O until needed. Sections were treated either with 0.01% calcofluor white in 1 × PBS for 3 min, and washed three times for 5 min each in 1 × PBS to remove excess stain (Falconer & Seagull, 1985), or with 10% phloroglucinol with the addition of concentrated HCL.
Four-wk-old A. thaliana stems were also cut 1 cm above the base and further hand-sectioned into c. 500-nm-thick cross-sections. Sections were then stained in 0.05% toluidine blue stain for 5 min. Excess stain was washed away with dH2O. All sections were mounted onto glass slides and examined with a Leica DRM microscope (Leica Microsystems) fitted with epifluorescence optics. Photographs were taken with a QICAM camera (QImaging, Surrey, Canada) and OpenLab software (PerkinElmer Inc., Waltham, MA, USA).
Sequence analyses of PgKOR
PgKOR encodes a putative 65-kD protein with 617 amino acids (GenBank JF343550). A comparison of the AtKOR, PgKOR and PtrKOR genomic sequence structures shows that each gene consists of the same number of exons and introns, but the sizes of the introns vary between species (Fig. 1a). A comparison of the protein sequence with that of seven other putative KOR homologues from other plant species (Fig. 1b) demonstrates the strong overall similarity among these proteins (Libertini et al., 2004). Overall, PgKOR is most similar to the KOR from loblolly pine (PtaKOR1; Nairn et al., 2008), sharing > 94% sequence identity. A comparison of all known KOR genes revealed conserved polarized targeting signals, predicted glycosylation sites, and residues essential for catalytic activity; however, the predicted transmembrane domains vary substantially, with only 40% sequence similarity between PgKOR and AtKOR (Fig. 1b).
The A. thaliana genome contains c. 25 genes that encode endoglucanases that belong to glycosyl hydrolase family 9 (GH9) proteins. These 25 genes separate into at least nine different classes according to Molhoj et al. (2002), who compared the relationships between the 25 endoglucanases in A. thaliana and their indentified homologues in other plant species. Class IX appears to be the most unique class in that members of this class do not contain an endoplasmic reticulum import sequence and are not secreted directly into the apoplast, but rather contain sequences that encode an N-terminal membrane-anchoring domain (Brummell et al., 1997b). In a phylogenetic comparison of PgKOR with all of the known A. thaliana glycosyl transferase family 9 proteins and the class IX proteins from other plant species, PgKOR clusters closest to the class IX proteins that are known to contain sequences that encode an N-terminal membrane-anchoring domain (Fig. 2).
Complementation of the A. thaliana kor1-1 mutant
Having established phylogenetic relatedness of the PgKOR gene and the KOR from A. thaliana, we hypothesized that if KOR was functionally conserved in gymnosperms and angiosperms, then the PgKOR gene would rescue the dwarf phenotype of A. thaliana kor mutants. Arabidopsis thaliana plants homozygous for the T-DNA insertion kor1-1 mutation (Nicol et al., 1998) were transformed with either the PgKOR gene or the endogenous AtKOR gene under the control of the cauliflower mosaic virus (CaMV) 35S promoter. None of the transformants displayed the elongation-deficient phenotype typical of the kor1-1 mutant. The T2 progeny, obtained after selfing, were again able to grow on PESTANAL selection media, and had a wild-type growth phenotype, as shown in Fig. 3(a) for the positive control AtKOR line and two representative PgKOR lines called PgKOR4 and PgKOR5. PCR-amplification of the T-DNA insertion (Fig. 3b) and real-time quantitative PCR analyses for the presence of the endogenous AtKOR transcript (Fig. 4a) confirmed that the representative lines were homozygous for the T-DNA insertion and had significantly lower native gene expression than the wild type. Furthermore, genomic DNA screening for the PgKOR gene construct (Fig. 3c) and real-time quantitative PCR analyses (Fig. 4b) confirmed that the PgKOR transgene is indeed expressed in the mutant homozygous for the T-DNA insertion. Toluidine blue staining of 4-wk-old stem cross-sections revealed that the irregular xylem phenotype common to the kor1-1 mutant was not seen in the transgenic lines and the wild-type xylem morphology phenotype was recovered (Fig. 5). To test if complementation was reflected in the cell wall chemistry, structural cell wall carbohydrates were analysed using anion exchange high-performance liquid chromatography. Structural carbohydrate analyses show that, while the structural glucose levels in the mutant were low when compared with the wild type, they returned to wild-type levels in all complemented lines (Table 1). X-ray diffraction data showed that the degree of cellulose crystallization did not differ between any of the lines, but that the MFA of the kor1-1 mutant was significantly increased when compared with the wild type. Furthermore, the MFAs of the complemented lines did not increase as much as that of the kor1-1 mutant (Table 2), but were still significantly higher than that of the wild type.
Table 1. Structural cell wall carbohydrate analysis of (a) wild-type Arabidopsis thaliana, Korrigan 1-1 (kor1-1) mutants and complemented A. thaliana plants (values represent pools of 20 individual plant stems), and (b) wild-type white spruce (Picea glauca) and PgKOR-RNA interference (RNAi) white spruce lines 2, 7 and 10
WT WS, wild-type A. thaliana (ecotype Wassilewskija); kor1-1, mutant kor1-1 A. thaliana; At, kor1-1 mutants with inserted AtKOR; Pg4 and Pg5, representative lines of kor1-1 mutants with inserted PgKOR. Values in bold have a P-value ≤ 0.05 when compared with wild-type white spruce using a Student t-test; the SE of the mean is shown in parentheses; n =7. na, not applicable.
(a) WT WS
(b) WT spruce
Table 2. Microfibril angle (MFA) and cell wall crystallinity as measured by X-ray diffraction of (a) 8-wk-old Arabidopsis thaliana stems, and (b) 18-month-old wild-type and PgKOR-RNA interference (RNAi) lines 2, 7 and 10 of white spruce (Picea glauca) stems
WT WS, wild-type A. thaliana (ecotype Wassilewskija); kor1-1, A. thaliana mutant kor1-1; At, kor1-1 mutants with inserted AtKOR; Pg4 and Pg5, representative lines of kor1-1 mutants with inserted PgKOR. SE of the mean is shown in parentheses; values in bold have a P-value ≤ 0.05 compared with wild type using a Student t-test. (a) n =10; (b) n =7.
(a) WT WS
(b) WT spruce
Endogenous PgKOR expression
Using 5-yr-old white spruce trees, spatial endogenous PgKOR expression was measured in eight tissue types from four individual trees (Fig. 6). PgKOR expression was highest in the young unexpanded and young expanded needles that correspond to locations of rapid primary growth. Expression from samples taken from the secondary growth of the stem was highest in the developing xylem.
RNAi suppression of endogenous PgKOR expression
Arabidopsis thaliana mutants and transgenic Populus spp. with decreased KOR expression possess irregular xylem phenotypes, in which the vessels of the stem collapse (Szyjanowicz et al., 2004; Maloney & Mansfield, 2010). To explore the PgKOR gene function in the context of gymnosperm wood, we characterized transgenic white spruce trees that had KOR activity suppressed, using a 400-bp hairpin RNAi construct designed from the putative full-length cDNA-encoding PgKOR. Agrobacterium tumefaciens-mediated transformation of white spruce somatic embryo tissue (Klimaszewska et al., 2001) yielded numerous independent embryonic cell lines that were confirmed to be transgenic through PCR amplification of the transgene, from which seven PgKOR-RNAi lines were grown into seedlings and transferred to the glasshouse (Fig. 7). Based on preliminary expression data, three lines that covered the range of suppression were chosen for growth and cell wall analyses. Endogenous PgKOR expression was analysed by real-time PCR, which showed a substantial reduction in transcript abundance in all three PgKOR-RNAi lines when compared with the wild-type trees grown under similar conditions (Fig. 8a). Student t-tests indicated that there was a significant reduction in the plant height and diameter (Fig. 8b) of the three PgKOR-RNAi lines, which can be associated with a reduction in gene expression (Fig. 8b). All trees were harvested after 18 months of growth and were used for all further analyses. Histochemical staining of stem cross-sections with the cellulose stain calcofluor white or the lignin-specific stain phloroglucinol did not reveal any obvious cell wall anomalies or any apparent gross changes in the cell wall chemistry of the PgKOR-RNAi lines (Fig. 9). A detailed structural carbohydrate analyses of the tree cell walls revealed that the RNAi lines do indeed have a significant reduction in glucose levels, indicating a reduction in cellulose compared with the corresponding wild-type trees (Table 1). These findings are consistent with recent findings in poplar where KOR was RNAi-suppressed (Maloney & Mansfield, 2010). Additionally, X-ray diffraction was used to characterize the cellulose produced by the spruce trees; however, only a slight reduction in MFA was evident, while there was no measurable change in cell wall crystallinity (Table 2).
The identification of the white spruce KOR gene has important implications for cell wall evolution and the impact PgKOR RNAi has on cellulose biosynthesis indicates that there are conserved mechanisms of cellulose deposition that require the KOR endoglucanase in tracheophytes. The comparison of genomic and coding sequences of KOR genes from a variety of different species revealed sequence features common to all. These homologues are characterized by a region rich in hydrophobic amino acids located in the N-terminus indicative of a single transmembrane domain as well as the polarized targeting signals, glycosylation sites, and residues essential for catalytic activity predicted by Nicol et al. (1998). These sequence analyses were an indication that an endoglucanase gene isolated from the gymnosperms was a candidate for an angiosperm KOR gene orthologue.
Endogenous PgKOR expression
To test our hypothesis that the white spruce PgKOR gene is an orthologue of the A. thaliana KOR gene, we sought evidence that the two genes have similar expression profiles. Ideally we wanted to show that PgKOR expression is associated with tissues undergoing rapid cell wall development and in fact endogenous PgKOR expression was highest in the young and young expanded needles which are tissues that correspond to locations of increased primary growth. By contrast, the transcript abundance of samples taken from the stem tissue was highest in the developing secondary xylem. These results are consistent with previous findings in loblolly pine where the transcript abundance for the loblolly pine KOR orthologue, PtaKor1, was either similar or up to 4.4-fold higher in xylem than in phloem (Nairn et al., 2008). Expression analyses of a membrane-anchored endo-1,4-β-glucanase cDNA from oilseed rape (Brassica napus; Cel16), which exhibits 94% sequence similarity to the A. thaliana KOR protein, shows that Cel16 is expressed in all tissue types, but it was highest in young roots and flowering stems. Furthermore, a Cel16 promoter-β-glucuronidase (GUS) fusion construct closely paralleled the pattern of abundance of Cel16 mRNA transcript profiling (Molhoj et al., 2001a). Similarly, a recent study by Takahashi et al. (2009) analysed the activity of the A. thaliana KOR promoter, again using a GUS-promoter construct, and found that it showed high activity in young plants, especially in actively expanding cells and vascular tissues, as well as being consistently active in the stems undergoing secondary cell wall thickening. These data provide solid, yet indirect, evidence that KOR functions in similar locations in a variety of species, suggesting that the genes are functionally conserved.
RNAi suppression of endogenous PgKOR expression
In coniferous trees, the water-conducting cells of the xylem are generally tracheids, whereas in the xylem of seed plants, vessel elements are the predominate water-conducting cell type. In the A. thaliana KOR knock-out mutants (Szyjanowicz et al., 2004) as well as KOR-RNAi hybrid poplar trees (Maloney & Mansfield, 2010) the altered phenotype is evident primarily in the vessel elements, manifesting itself in an irregular xylem (irx) phenotype and presumably disrupting water transport and ultimately growth. In the current study evaluating 18-month-old PgKOR-RNAi white spruce, histochemical staining and X-ray diffraction did not reveal any visible cell well anomalies. However, there was a substantial reduction in height and radial growth as well as in structural cell wall glucose content, suggesting a lower cellulose content. The lack of an irregular xylem phenotype in the secondary growth of white spruce may indicate that loss of KOR function produces a secondary cell wall phenotype in a cell-specific manner in vessel elements, but not tracheids or fibres. During development, vessels undergo rapid and extensive radial expansion, compared with tracheids. If KOR works in the expansion phase of xylogenesis during morphogenesis, then vessels could be disproportionately affected compared with fibres or tracheids. Furthermore, complementation of the kor1-1 mutant with AtKOR restored the MFA back to the wild-type level, but the PgKOR-expressing lines exhibited an MFA that was intermediate between those of the wild type and the mutant (Table 2). These observations indicate that, while the PgKOR gene is able to rescue the A. thaliana kor1-1 growth phenotype, gene function may not be completely equivalent. An alternative hypothesis is that the collapse of the xylem in plants displaying an irx phenotype only occurs in cells that are involved in water transport and are subjected to negative hydraulic pressures (Turner & Somerville, 1997). Cells that have higher inherent lignin content, such as tracheids in white spruce, have a greater resistance to negative hydraulic pressure (Boyce et al., 2004; Akiyama et al., 2005), which may help to explain why the tracheids did not collapse in the PgKOR-RNAi white spruce plants, but the plants produced less structural glucose, therefore still leading to reduced growth.
Previous reports on the function of KOR suggest that it is required for development of both the primary and secondary cell walls and affects cellulose biogenesis (reviewed by Molhoj et al., 2002). Additionally, we previously reported that the lack of KOR function in hybrid poplar can manifest in an increase in cell wall xylan content (Maloney & Mansfield, 2010). Xylans are known to be the major cellulose-linking polysaccharide in the secondary cell walls of higher plants. Furthermore, Carafa et al. (2005) demonstrated that xylans are present in the cell walls of all vascular plants tested but not in mosses. The occurrence of xylans in plants with tracheid cell types suggests that the appearance of these polysaccharides has also been an important evolutionary event in the development of vascular and mechanical tissues. KOR therefore may play a role in modulating cellulose properties to optimize interactions with xylans, and this architecture may represent a major evolutionary milestone in the transition of plants from an aquatic to a terrestrial environment.
Significance of the evolutionary conservation of KOR
The evolution of land plants necessitated the need for water-conducting tissues in order to avoid excessive water stress. These water-conducting tissues came in the form of a compression-resistant cell wall skeleton that continues to function following programmed cell death of the protoplast. These evolutionarily important features arose early, with water-conducting cells present in the fossil record 410 Mya (Sperry, 2003). The tracheid cell type, which evolved first, is a narrow single cell with an intact axial cell wall. Vessels, in contrast, are water-conducting tubes with either partially or completely open axial cell walls (Tyree & Zimmermann, 2002). Vessel elements evolved through a series of modifications to the tracheid developmental process and, in general, provide greater xylem hydraulic capacity than tracheids (Tyree & Zimmermann, 2002; Sperry, 2003). The fossil records provide evidence for bryophytes being some of the earliest land plants, and the draft genome sequence of the moss Physcomitrella patens provides extensive information on the bryophyte genome (Rensing et al., 2008). An extensive interrogation of this recently published Physcomitrella patens draft genome sequence with both the AtKOR and the PgKOR gene did not reveal any significant matches. While a previous study on the genes required for cellulose synthesis in Agrobacterium tumefaciens C58 showed that a membrane-anchored endoglucanase (CelC) was involved in the transfer of cellulose oligomers from a lipid carrier to the growing cellulose chain (Matthysse et al., 1995), this gene only shares 11% identity with AtKOR. Furthermore, a screen of the Agrobacterium tumefaciens C58 genome database (Goodner et al., 2001; Wood et al., 2001) as well as the cellulose-producing Oomycetes Phytophthora spp. genome database (Gajendran et al., 2006) did not reveal any significant matches to these membrane-bound endoglucanases. These data and the current evidence for the partial functional conservation of KOR in both tracheid- and vessel-containing plants provide evidence for the evolution of KOR and the secondary cell wall-specific cellulose synthases occurring some time before the evolution of the tracheid cell type, but after the conquest of land by plants.
The authors acknowledge the technical assistance of Yoichiro Watanabe and Hanh Ly. The authors also gratefully acknowledge financial support from the NSERC of Canada Discovery Grants programme for individual grants held by S.D.M. and A.L.S.