Skip to main content
NCBI home page
NIHPA Author Manuscripts logoLink to NIHPA Author Manuscripts
. Author manuscript; available in PMC: 2025 Oct 1.
Published in final edited form as: Plant J. 2024 Aug 15;120(1):234–252. doi: 10.1111/tpj.16983

Identification of glycosyltransferases mediating 2-O-arabinopyranosyl and 2-O-galactosyl substitutions of glucuronosyl side chains of xylan

Ruiqin Zhong 1, Dayong Zhou 2, Dennis R Phillips 3, Earle R Adams 3, Lirong Chen 2, John P Rose 2, Bi-Cheng Wang 2, Zheng-Hua Ye 1,*
PMCID: PMC11424249  NIHMSID: NIHMS2016151  PMID: 39145524

SUMMARY

Xylan is one of the major hemicelluloses in plant cell walls and its xylosyl backbone is often decorated at O-2 with glucuronic acid (GlcA) and/or methylglucuronic acid (MeGlcA) residues. The GlcA/MeGlcA side chains may be further substituted with 2-O-arabinopyranose (Arap) or 2-O-galactopyranose (Gal) residues in some plant species, but the enzymes responsible for these substitutions remain unknown. During our endeavor to investigate the enzymatic activities of Arabidopsis MUR3-clade members of the GT47 glycosyltransferase family, we found that one of them was able to transfer Arap from UDP-Arap onto O-2 of GlcA side chains of xylan, and thus it was named xylan 2-O-arabinopyranosyltransferase 1 (AtXAPT1). The function of AtXAPT1 was verified in planta by its T-DNA knockout mutation showing a loss of the Arap substitution on xylan GlcA side chains. Further biochemical characterization of XAPT close homologs from other plant species demonstrated that while the poplar ones had the same catalytic activity as AtXAPT1, those from Eucalyptus, lemon-scented gum, sea apple, ‘Ohi’a lehua, duckweed and purple yam were capable of catalyzing both 2-O-Arap and 2-O-Gal substitutions of xylan GlcA side chains albeit with differential activities. Sequential reactions with XAPTs and glucuronoxylan methyltransferase 3 (GXM3) showed that XAPTs acted poorly on MeGlcA side chains, whereas GXM3 could efficiently methylate arabinosylated or galactosylated GlcA side chains of xylan. Furthermore, molecular docking and site-directed mutagenesis analyses of Eucalyptus XAPT1 revealed critical roles of several amino acid residues at the putative active site in its activity. Together, these findings establish that XAPTs residing in the MUR3 clade of family GT47 are responsible for 2-O-arabinopyranosylation and 2-O-galactosylation of GlcA side chains of xylan.

Keywords: Arabinopyranosyltransferase, cell wall, galactosyltransferase, glucuronic acid, GT47, xylan

INTRODUCTION

Plant cell walls provide a structural framework to support individual cells and collectively the whole plant body as well as act as the first line of physical barriers against biotic and abiotic stresses. They are mainly made of polysaccharide polymers, including cellulose, hemicelluloses (xylan, mannan, xyloglucan and mixed-linkage β-1,3/1,4-glucan) and pectins, and in some specialized cells, such as xylem and fibers, the cell walls are also impregnated with lignin, a complex polyphenolic polymer (McFarlane, 2023). Cell walls constitute the bulk of plant biomass and secondary walls in the form of wood and fibers are widely used in our daily life. Recently, plant cell walls, coined as lignocellulosic biomass, have also been exploited as a renewable source for biofuel production (Marriot et al., 2016). Therefore, study of how plant cell walls are synthesized and constructed will not only enrich our fundamental knowledge of plant biology but also provide molecular tools for genetic modifications of plant biomass composition suited for our diverse end uses.

The biosynthesis of plant cell wall polysaccharide polymers is carried out by glycosyltransferases (GTs) that catalyze the transfer of a monosaccharide from a nucleotide sugar donor to a moiety on an acceptor substrate forming a specific glycosidic linkage. GTs are classified into different families based on their sequences and structures (Coutinho et al. 2003) and many GT families have been shown to be involved in plant cell wall synthesis (Amo and Mohnen, 2019). Of particular note is family GT47 whose members have been implicated in the synthesis of different cell wall polysaccharides, including xylan, xyloglucan, mannan and pectins. Among the GT47 members, IRX10 and IRX10L/XYS1 are xylan synthases catalyzing xylan backbone elongation (Brown et al., 2009; Wu et al., 2009; Jensen et al. 2014; Urbanowicz et al. 2014; Zeng et al. 2016), FRA8 and F8H are xylan reducing end sequence synthesis-related GTs (Zhong et al. 2005; Lee et al. 2009), MUR3 and XLT2 are xyloglucan position-specific β-galactosyltransferases (Madson et al. 2003; Jensen et al. 2012), XUT1, XST1, XDT and XBT1 are xyloglucan β-galacturonosyltransferase, α-arabinofuranosyl transferase, α-arabinopyranosyl transferase and β-xylosyltransferase, respectively (Pena et al., 2012; Schultink et al., 2013; Zhu et al., 2018; Immelmann et al., 2023; Wilson et al., 2023), MBGT1 is a mannan β-galactosyltransferase (Yu et al., 2022), XGD1 is a pectic xylogalacturonan β-xylosyltransferase (Jensen et al. 2008), and ARAD1 and ARAD2 are putative α-arabinosyltransferases involved in the synthesis of arabinan side chains of the pectic rhamnogalacturonan-I (Harholt et al., 2012). It is intriguing that although MUR3, XLT2, XUT1, XST1, XDT, XBT1 and MBGT1 all belong to the MUR3 clade of family GT47, they exhibit diverse activities acting on various nucleotide sugar donors, including UDP-α-D-galactose (UDP-α-D-Gal), UDP-α-D-galacturonic acid (UDP-α-D-GalA), UDP-β-L-arabinofuranose (UDP-β-L-Araf), UDP-β-L-arabinopyranose (UDP-β-L-Arap) and UDP-α-D-xylose (UDP-α-D-Xyl), and two different hemicellulose acceptors, i.e., xyloglucan and mannan. In Arabidopsis, there are 11 members in the MUR3 clade and five of them, MUR3, XLT2, XUT1, MBGT1 and AtGT11, have been genetically or biochemically studied (Madson et al. 2003; Jensen et al. 2012; Pena et al., 2012; Wei et al., 2021; Yu et al., 2022), but the remaining six members await to be functionally characterized.

Xylan is a hemicellulose consisting of β-1,4-linked Xyl residues that are often substituted with 2-O-α-glucuronic acid (GlcA) and/or methylglucuronic acid (MeGlcA) residues (Timell 1967; Avic 2022). In gymnosperms and grass species, xylan is also substituted with 2-O-/3-O-α-Araf and 2-O-Xyl, and the 3-O-α-Araf residues in grass xylan can be further substituted with 2-O-β-Xyl or 2-O-α-Araf forming Xyl-Araf or Araf-Araf disaccharide side chains (Wende and Fry, 1997; Jacob et al., 2001; Hoije et al., 2006; Bowman et al., 2014; Busse-Wicher et al., 2016). In non-commelinid monocots, such as Spirodela polyrhiza (giant duckweed), Dioscorea alata (purple yam), Crinum americanum (Southern swamp lily), Agapanthus africanus (African lily), Allium cepa (onion), Asparagus officinalis (asparagus) and Orontium aquaticum (Golden club), the GlcA/MeGlcA side chains of xylan were found to be further substituted with 2-O-α-Arap forming Arap-GlcA/MeGlcA side chains (Pena et al., 2016). Xylan from the dicot Eucalyptus grandis was shown to contain both 2-O-β-Arap- and 2-O-β-Gal-substituted MeGlcA side chains (Togashi et al., 2009; Pena et al., 2016). In Arabidopsis, the GlcA side chains of xylan were also found to be decorated with a pentose residue but its exact identity has not been determined (Chong et al., 2015; Mortimer et al., 2015). Genetic and biochemical studies have uncovered a number of GTs involved in addition of side chains onto xylan, including xylan glucuronosyltransferases, arabinosyltransferases and xylosyltransferases (Mortimer et al., 2010; Anders et al., 2012; Chiniquy et al., 2012; Lee et al., 2012a; Zhong et al., 2018, 2021, 2022). However, glycosyltransferases responsible for the addition of 2-O-α-Arap or 2-O-β-Gal onto GlcA side chains of xylan remain elusive.

In this study, we focused on biochemical characterization of an Arabidopsis MUR3-clade GT47 member, At1g68470. Activity assays of its recombinant protein produced in human embryonic kidney (HEK) 293 cells demonstrated that it was a xylan 2-O-arabinopyranosyltransferase (AtXAPT1) transferring Arap residues onto O-2 of GlcA side chains of xylan. We determined the presence of α-Arap-(1,2)-α-GlcA/MeGlcA side chains in Arabidopsis xylan and confirmed the function of AtXAPT1 in planta by its T-DNA knockout mutation. XAPT1 close homologs from a number of other plant species were shown to mediate 2-O-Arap and/or 2-O-Gal transfer onto GlcA side chains of xylan. Further molecular docking and site-directed mutagenesis revealed several functionally critical amino acids at the putative active site of Eucalyptus XAPT1. Our findings have uncovered the functions of a subgroup of MUR3-clade GT47 members in catalyzing the transfer of 2-O-Arap and/or 2-O-Gal onto GlcA side chains of xylan, which expands the scope of the diverse catalytic activities of MUR3-clade members in substituting the side chains of not only xyloglucan and mannan but also xylan.

RESULTS

An Arabidopsis MUR3-clade GT47 member catalyzes 2-O-Arap transfer onto GlcA side chains of xylan

Members of the MUR3-clade of family GT47, including MUR3, XLT2, XUT1, XST1, XDT, XBT1 and MBGT1 from several plant species, have been shown to mediate various 2-O-glycosyl substitutions of Xyl side chains of xyloglucan or 2-O-galactosyl substitutions of Gal side chains of mannan (Madson et al. 2003; Jensen et al. 2012; Pena et al., 2012; Schultink et al., 2013; Zhu et al., 2018; Yu et al., 2022; Immelmann et al., 2023; Wilson et al., 2023). There exist 11 members in the Arabidopsis MUR3 clade (Figure 1a) and only five of them have been previously studied (Madson et al. 2003; Jensen et al. 2012; Pena et al., 2012; Wei et al., 2021; Yu et al., 2022). To explore the functions of the remaining Arabidopsis MUR3-clade members, we set out to express their recombinant proteins in HEK293 cells and investigate their enzymatic activities. Here, we present the functional characterization of one of them, AT1G68470. The recombinant AT1G68470 protein (Figure 1b) was first tested for its ability to transfer Gal, GalA, Xyl, Arap or Araf onto xyloglucan because most of the known MUR3-clade members were shown to mediate the transfer of one of these glycosyl residues onto xyloglucan. To do so, we incubated the recombinant AT1G68470 protein with the xyloglucan oligomer XXXG (X denotes a Xyl-substituted glucose and G denotes an unsubstituted glucose; Fry et al., 1993) and various nucleotide sugar donors, including UDP-Gal, UDP-GalA, UDP-Xyl, UDP-Arap and UDP-Araf, and analyzed the reaction products with matrix-assisted laser desorption ionization-time-of-flight (MALDI-TOF) mass spectrometry. Examination of the MALDI-TOF mass spectra showed no additional ion species besides the one corresponding to the XXXG acceptor at m/z 1085 in all of the reaction products (Figure 1d), indicating that the AT1G68470 protein is incapable of transferring any of these glycosyl residues onto xyloglucan.

Figure 1.

Figure 1.

Open in a new tab

Production and activity assays of recombinant AtXAPT1 protein.

(a) Phylogenetic relationships of the 11 Arabidopsis MUR3-clade GT47 members together with their close homologs from other plant species that have been functionally characterized. AtXAPT1/AT1G68470 studied in this report is highlighted in yellow. The phylogenetic tree was constructed using the maximum likelihood method and the numbers at the nodes are percentages of bootstrap values from 1,000 replicates. The 0.1 scale denotes 10% change. At, Arabidopsis; Cc, Coffea canephora; Eg, Eucalyptus grandis; Os, Oryza sativa; Pp, Physcomitrium patens; Sl, Solanum lycopersicum; Vc, Vaccinium corymbosum.

(b) SDS-PAGE and Coomassie Blue staining of recombinant AtXAPT1 protein expressed in HEK293 cells. Shown on the left are the molecular masses (kDa) of protein markers.

(c) Diagram showing the addition of an Arap residue onto the GlcA side chain of xylan catalyzed by AtXAPT1.

(d,e,f) MALDI-TOF mass spectra of the reaction products of AtXAPT1 incubated with XXXG (d), (GlcA)Xyl6-AA (e) or Xyl6-AA (f) and various nucleotide sugar donors as indicated. The control was AtXAPT1 incubated with the corresponding acceptor without any nucleotide sugar donors. Each ion species is denoted with its oligomer composition and mass ([M+Na]+). Note the appearance of a new ion species at m/z 1262 corresponding to (GlcA)Xyl6-AA substituted with an Arap residue in the reaction products of AtXAPT1 incubated with (GlcA)Xyl6-AA and UDP-Arap.

Considering that the known MUR3-clade members mediate 2-O-glycosyl substitutions of side chains of two different cell wall polysaccharides, xyloglucan and mannan, we hypothesized that AT1G68470 might catalyze 2-O-glycosyl substitutions of side chains of another cell wall polysaccharide. A survey of cell wall polysaccharide structures indicated that like Xyl side chains of xyloglucan and Gal side chains of mannan, GlcA/MeGlcA side chains of xylan could also be substituted at O-2 with a glycosyl residue, Arap or Gal, as shown in several monocots and Eucalyptus (Togashi et al., 2009; Pena et al., 2016). Therefore, we proposed that AT1G68470 might be involved in 2-O-glycosyl substitutions of GlcA/MeGlcA side chains of xylan. To test this proposition, the AT1G68470 protein was incubated with anthranilic acid (AA)-labeled GlcA-substituted xylohexaose [(GlcA)Xyl6-AA] and UDP-Gal, UDP-Arap or UDP-Araf. Examination of the MALDI-TOF mass spectra of the reaction products revealed that when the protein was incubated with (GlcA)Xyl6-AA and UDP-Arap, there appeared a new ion species at m/z 1262 with a mass increase of 132 Da (corresponding to one Arap residue minus water) over the mass of the (GlcA)Xyl6-AA acceptor (m/z 1130) (Figures 1e and S1a), demonstrating that AT1G68470 is capable of transferring Arap onto (GlcA)Xyl6-AA. No new ion species was observed when the protein was incubated with (GlcA)Xyl6-AA and UDP-Gal or UDP-Araf (Figure 1e). In addition, incubation of the protein with Xyl6-AA and UDP-Arap did not produce any new ion species (Figure 1f), indicating that AT1G68470 does not transfer Arap onto backbone Xyl residues. These results suggest that AT1G68470 mediates the transfer of Arap onto GlcA side chains of xylan (Figure 1c), and thus it was named xylan 2-O-arabinopyronosyltransferase 1 (AtXAPT1; see additional proof below).

To substantiate the finding that AtXAPT1 catalyzes the transfer of Arap onto GlcA side chains of xylan, we next applied 1H NMR spectroscopy to analyze the reaction products of AtXAPT1 incubated with (GlcA)Xyl6 and UDP-Arap. Inspection of the 1H NMR spectra showed that compared with the control, the AtXAPT1-catalyzed reaction products had a new resonance peak at 5.41 ppm (Figures 2a and S1b), which is diagnostic of H1 of α-GlcA substituted at O-2 with α-Arap (Pena et al., 2016). These data provide compelling evidence that AtXAPT1 is an arabinosyltransferase catalyzing 2-O-Arap transfer onto GlcA side chains of xylan (Fig. 2b).

Figure 2.

Figure 2.

Open in a new tab

Examination of the AtXAPT1-catalyzed reaction products by 1H-NMR spectroscopy.

(a) 1H-NMR spectra of the reaction products of AtXAPT1 incubated with (GlcA)Xyl6 and UDP-Arap. The resonances are denoted with the proton positions and the corresponding sugar residue identities. The control was heat-denatured AtXAPT1 incubated with (GlcA)Xyl6 and UDP-Arap. Note that the reaction products of AtXAPT1 exhibited a new resonance peak at 5.41 ppm corresponding to GlcA substituted at O-2 with Arap.

(b) Diagram of the structural units of the acceptor and the reaction product of AtXAPT1. Note that AtXAPT1 catalyzes the addition of Arap onto O-2 of GlcA side chains of xylan.

The GlcA side chains of Arabidopsis xylan are substituted at O-2 with Arap

Arabidopsis xylan is decorated with GlcA/MeGlcA residues, but it is currently unknown whether the GlcA/MeGlcA side chains are further substituted at O-2 with Arap. The finding that AtXAPT1 is an arabinosyltransferase transferring 2-O-Arap onto GlcA side chains of xylan prompted us to examine the structure of Arabidopsis xylan side chains. 1H NMR spectroscopy of Arabidopsis xylan revealed the presence of resonances at 5.41 ppm in addition to the resonances corresponding to backbone Xyl residues, GlcA/MeGlcA side chains and the unique reducing end tetrasaccharide sequence (α-GalA and α-Rhamose) (Figures 3a and S2). Because the resonances for 2-O-Arap-substituted GlcA/MeGlcA side chains of xylan isolated from several monocot species were shown to occur at 5.41 ppm (Pena et al., 2016), we reasoned that the observed resonances at 5.41 ppm in Arabidopsis xylan were also attributed to GlcA/MeGlcA substituted at O-2 with Arap. To test this hypothesis, we carried out 2-dimentional (2D) NMR correlation spectroscopy (COSY) to determine the identity of the observed resonances at 5.41 ppm in Arabidopsis xylan. 2D COSY, which provides correlations between neighboring protons within the analyte molecule, allows the detection of 2-O-Arap-substituted α-GlcA side chains of xylan at the diagnostic cross peak between H1 at 5.41 ppm and H2 at 3.76 ppm (Pena et al., 2016). Examination of the COSY spectrum of Arabidopsis xylan showed the presence of the cross peak between H1 at 5.41 ppm and H2 at 3.76 ppm, a coupling pattern characteristic of 2-O-Arap-substituted α-GlcA (Figure 3b). These results indicate that the observed resonances at 5.41 ppm in Arabidopsis xylan correspond to α-GlcA substituted with 2-O-α-Arap (Figure 3c).

Figure 3.

Figure 3.

Open in a new tab

Effect of T-DNA knockout mutation of the AtXAPT1 gene on xylan substitutions.

(a) 1H-NMR spectra of xylooligomers isolated from wild-type Arabidopsis, the xapt1 mutant and xapt1 complemented with the wild-type AtXAPT1 gene. The resonances are labeled with the proton positions and the corresponding sugar residue identities. Note the presence of resonances at 5.41 ppm corresponding to 2-O-Arap-substituted GlcA/MeGlcA in the wild type, their loss in the xapt1 mutant, and their restoration in xapt1 complemented with the wild-type AtXAPT1 gene. Inset is a diagram of the AtXAPT1 gene showing the position of T-DNA insertion in the xapt1 mutant. HDO, hydrogen deuterium.

(b) 2D COSY NMR spectrum of xylooligomers from wild-type Arabidopsis showing the cross peak (H1 at 5.41 ppm and H2 at 3.76 ppm) corresponding to the correlation between H1 and H2 of GlcA/MeGlcA substituted at O-2 with Arap. Cross peaks corresponding to the correlation between H1 and H2 of un-substituted GlcA/MeGlcA (H1 at 5.29 ppm and H2 at 3.54 ppm), reducing α-Xyl (H1 at 5.18 ppm and H2 at 3.51 ppm), reducing β-Xyl (H1 at 4.63 ppm and H2 at 3.42 ppm) and branched Xyl (H1 at 4.70 ppm and H2 at 3.54 ppm) were also marked.

(c) Diagram illustrating the further substitution of Arabidopsis xylan 2-O-GlcA/MeGlcA side chain with 2-O-Arap.

(d) Expression of the AtXAPT1 gene in different organs of Arabidopsis plants. The expression data of AtXAPT1 were retrieved from Arabidopsis eFP Browser (http://bar.utoronto.ca/efp//cgi-bin/efpWeb.cgi). The error bars denote the SD of three replicates.

Mutation of AtXAPT1 results in a loss of Arap substitutions of GlcA side chains of xylan

The AtXAPT1 gene was shown to be expressed ubiquitously in different Arabidopsis organs (Figure 3d), suggesting that AtXAPT1 may be responsible for the observed 2-O-Arap substitutions of xylan GlcA side chains in Arabidopsis. To investigate whether this is the case, we obtained two mutant lines with T-DNA insertions in the AtXAPT1 gene and applied 1H NMR spectroscopy to examine xylan isolated from these mutants. The xapt1 mutant had a T-DNA insertion in the first exon (Figure 3a) and the xapt1–2 mutant had a T-DNA insertion at 38 bp upstream of the start codon (Figure S3a). RT-PCR analysis showed that the xapt1 mutant lacked AtXAPT1 transcripts, whereas the xapt1–2 mutant still had AtXAPT1 transcripts (Figure S3c). Examination of the 1H NMR spectra revealed that the xapt1 mutant xylan had a complete loss of resonances at 5.41 ppm corresponding to 2-O-Arap-substituted GlcA/MeGlcA compared with the wild-type xylan (Figures 3a and S2). Furthermore, these resonances were restored by complementation of the xapt1 mutant with the wild-type AtXAPT1 gene (Figure 3a and S2), indicating that the AtXAPT1 gene is responsible for the 2-O-Arap substitutions of xylan GlcA/MeGlcA side chains. These observations provide in planta evidence for the function of AtXAPT1 in catalyzing 2-O-Arap substitutions of xylan GlcA side chains in Arabidopsis. Consistent with the presence of AtXAPT1 transcripts in the xapt1–2 mutant (Figure S3c), the xapt1–2 mutant xylan still had resonances at 5.41 ppm corresponding to 2-O-Arap-substituted GlcA/MeGlcA (Figure S2). Mutation of the AtXAPT1 gene did not cause any apparent alterations in plant growth and secondary wall thickening in xylem vessels and interfascicular fibers compared with the wild type (Figure S3e–h).

XAPT close homologs from a number of plant species exhibit activities catalyzing the transfer of 2-O-Arap and/or 2-O-Gal onto GlcA side chains of xylan

Previous studies have shown that the GlcA/MeGlcA side chains of xylan from Eucalyptus and several monocots, such as duckweed (Spirodela polyrhiza) and purple yam (Dioscorea alata), are further substituted at O-2 with Arap and/or Gal (Togashi et al., 2009; Pena et al., 2016). Therefore, we next aimed to find out whether close homologs of AtXAPT1 from these and several other plant species could mediate 2-O-Arap or 2-O-Gal substitutions of xylan GlcA side chains. The amino acid sequence of AtXAPT1 was used to BLAST-search for its close homologs from poplar (Populus trichocarpa), Eucalyptus (Eucalyptus grandis), duckweed, purple yam, and three other species related to Eucalyptus in the myrtle family [lemon-scented gum (Corymbia citriodora), sea apple (Syzygium grande) and ‘Ohi’a lehua (Metrosideros polymorpha)], leading to identification of one or two XAPT close homologs from each of these species (Figure 4a). These XAPT close homologs were expressed in HEK293 cells for generation of recombinant proteins and most of them were successfully produced (Figure 4b). These recombinant proteins were incubated with (GlcA)Xyl6-AA and UDP-Arap or UDP-Gal and the reaction products were analyzed by MALDI-TOF mass spectrometry. It was found that like AtXAPT1, all of these XAPT close homologs except EgXAPT2 and MpoXAPT2 efficiently transferred Arap from UDP-Arap onto (GlcA)Xyl6-AA (Figures 4c,e and S4), and hence they were named XAPTs. Interestingly, all of these XAPTs except PtrXAPT1/2 were also able to transfer Gal from UDP-Gal onto (GlcA)Xyl6-AA (Figures 4d,f and S4). These enzymatic activity analyses demonstrated that like AtXAPT1, PtrXAPT1/2 only arabinosylated the (GlcA)Xyl6-AA acceptor, EgXAPT2 and MpoXAPT2 preferentially catalyzed galactosylation of the acceptor, whereas EgXAPT1, SgXAPT1, CCiXAPT2, SpXAPT1 and DaXAPT1 exhibited dual activities catalyzing both arabinosylation and galactosylation of the acceptor (Figures 4c–f and S4).

Figure 4.

Figure 4.

Open in a new tab

Production and activity assays of recombinant XAPT homologs from several other plant species.

(a) Phylogenetic relationships of AtXAPT1 and its close homologs identified from seven other plant species. Highlighted in yellow are the XAPT homologs whose recombinant proteins were successfully produced and studied in this report. The phylogenetic tree was constructed using the maximum likelihood method and the numbers at the nodes are percentages of bootstrap values from 1,000 replicates. The 0.1 scale denotes 10% change. AtMUR3 and AtXLT2 were used as an outgroup. Cci, Corymbia citriodora; Da, Dioscorea alata; Eg, Eucalyptus grandis; Mpo, Metrosideros polymorpha; Ptr, Populus trichocarpa; Sg, Syzygium grande; Sp, Spirodela polyrhiza.

(b) SDS-PAGE and Coomassie Blue staining of recombinant XAPT proteins expressed in HEK293 cells.

(c,d.e.f) MALDI-TOF mass spectra of the reaction products of XAPTs incubated with the (GlcA)Xyl6-AA acceptor and the nucleotide sugar donor UDP-Arap (c,e) or UDP-Gal (d,f). The control reactions were heat-denatured PtrXAPT1 incubated with (GlcA)Xyl6-AA and the corresponding nucleotide sugar donor. Each ion species is denoted with its oligomer composition and mass ([M+Na]+). Note the new ion species at m/z 1262 corresponding to (Arap-GlcA)Xyl6-AA in the reaction products of XAPTs incubated with (GlcA)Xyl6 and UDP-Arap and the new ion species at m/z 1292 corresponding to (Gal-GlcA)Xyl6-AA in those incubated with (GlcA)Xyl6 and UDP-Gal.

To further verify the arabinosylation and galactosylation of GlcA side chains by these XAPTs, we used 1H NMR spectroscopy to examine the reaction products of several representative XAPTs. It was shown that the reaction products of EgXAPT1 and SpXAPT1 incubated with (GlcA)Xyl6 and UDP-Arap displayed a resonance peak at 5.41 ppm (Figures 5a and S5a), which is diagnostic of α-GlcA substituted with 2-O-α-Arap (Pena et al., 2016). The reaction products of CciXAPT2 incubated with (GlcA)Xyl6 and UDP-Gal exhibited a resonance peak at 5.43 ppm (Figures 5a and S5a), which is characteristic of α-GlcA substituted with 2-O-β-Gal (Pena et al., 2016). Taken together, these results confirmed that these XAPTs mediate 2-O-α-Arap and 2-O-β-Gal substitutions of xylan GlcA side chains.

Figure 5.

Figure 5.

Open in a new tab

NMR spectroscopy of XAPT-catalyzed reaction products and investigation of the sequential methylation and glycosyl substitutions of (GlcA)Xyl6 by AtGXM3 and XAPTs.

(a) 1H-NMR spectra of the reaction products of XAPTs incubated with (GlcA)Xyl6 and UDP-Arap or UDP-Gal. The control was heat-denatured EgXAPT1 incubated with (GlcA)Xyl6 and UDP-Arap. The resonances are labeled with the proton positions and the corresponding sugar residue identities. Note that the reactions products of EgXAPT1 and SpXAPT1 incubated with (GlcA)Xyl6 and UDP-Arap displayed a new resonance peak at 5.41 ppm corresponding to 2-O-Arap-substituted GlcA and those of CciXAPT2 incubated with (GlcA)Xyl6 and UDP-Gal exhibited a new resonance peak at 5.43 ppm corresponding to 2-O-Gal-substituted GlcA. HDO, hydrogen deuterium oxide.

(b) MALDI-TOF mass spectra of the reaction products catalyzed sequentially by AtGXM3 followed by XAPTs. Inset is a diagram of the sequential reactions depicting methylation of (GlcA)Xyl6-AA first by AtGXM3 and then its arabinosylation or galactosylation by XAPTs. Each ion species is denoted with its mass ([M+Na]+) and oligomer composition. Note that the methylated acceptor (MeGlcA)Xyl6-AA was barely further arabinosylated or galactosylated by XAPTs.

(c) MALDI-TOF mass spectra of the reaction products catalyzed sequentially by AtXAPT1 followed by AtGXM3. Inset is a diagram of the sequential reactions depicting arabinosylation of (GlcA)Xyl6-AA first by AtXAPT1 and then its methylation by AtGXM3. Note the complete methylation of (Arap-GlcA)Xyl6-AA converting it into (Arap-MeGlcA)Xyl6-AA by AtXAPT1.

(d) MALDI-TOF mass spectra of the reaction products catalyzed sequentially by SpXAPT1 followed by AtGXM3. Inset is a diagram of the sequential reactions depicting galactosylation of (GlcA)Xyl6-AA first by SpXAPT1 and then its methylation by AtGXM3. Note the complete methylation of (Gal-GlcA)Xyl6-AA converting it into (Gal-MeGlcA)Xyl6-AA by SpXAPT1.

XAPTs act poorly on methylated GlcA side chains of xylan but arabinosylated or galactosylated GlcA side chains of xylan can be efficiently methylated

The GlcA side chains of xylan were shown to be partially methylated in some species such as Arabidopsis, duckweed and purple yam, or completely methylated in other species such as poplar and Eucalyptus (Teleman et al., 2000; Zhong et al., 2005; Togashi et al., 2009; Pena et al., 2016). The 4-O-methylation of xylan GlcA side chains is catalyzed by glucuronoxylan 4-O-methyltransferases (GXMs) belonging to the DUF579 family (Lee et al., 2012b; Urbanowicz et al., 2012). To find out whether XAPTs could arabinosylate and/or galactosylate methylated GlcA side chains of xylan, we carried out sequential methylation and arabinosylation/galactosylation of the (GlcA)Xyl6-AA acceptor by AtGXM3 and several representative XAPTs. MALDI-TOF mass spectrometry of the reaction products revealed that when the GlcA residues of (GlcA)Xyl6-AA were first methylated by AtGXM3, EgXAPT1 and SpXAPT1 could barely arabinosylate the methylated GlcA side chains (Figures 5b and S5b), which is in sharp contrast to the efficient arabinosylation of the nonmethylated GlcA side chains by these XAPTs (Figure 4e). Similarly, EgXAPT2 and SpXAPT1 were unable to galactosylate the methylated GlcA side chains (Figures 5b and S5b) although they effectively galactosylated nonmethylated GlcA side chains (Figure 4f). However, when the GlcA residues of (GlcA)Xyl6-AA were first arabinosylated or galactosylated by AtXAPT1 or SpXAPT1, respectively, AtGXM3 could completely methylate the arabinosylated or galactosylated GlcA side chains, resulting in formation of (Arap-MeGlcA)-Xyl6-AA and (Gal-MeGlcA)-Xyl6-AA (Figures 5c,d and S5c). It should be noted that the AtXAPT1- and SpXAPT1-calayzed reaction products used as substrates for AtGXM3 still contained (GlcA)Xyl6-AA in addition to the arabinosylated or galactosylated (GlcA)Xyl6-AA, and as expected, (GlcA)Xyl6-AA was also completely methylated by AtGXM3 (Figures 5c,d and S5c). These findings imply that the GlcA side chains of xylan are first arabinosylated or galactosylated by XAPTs and then methylated by GXMs in planta.

Structural modeling and molecular docking of XAPTs predict candidate amino acid residues involved in substrate binding at the putative active site

To decipher the structure-function relationship of XAPTs, we generated structural models of the catalytic domains of AtXAPT1 and EgXAPT1 using AlphaFold2, a computational method capable of predicting protein structures with atomic accuracy (Jumper et al., 2021). The predicted structural models of AtXAPT1 and EgXAPT1 displayed two distinct N-terminal and C-terminal Rossmann-like domains and an interdomain putative active site cleft (Figure 6a,b), which are structural features characteristic of the GT-B fold (Lairson et al., 2008; Chang et al., 2011). Superimposition of the AtXAPT1 and EgXAPT1 structural models showed similar global folds and structural similarity with a root-mean-square deviation (RMSD) of 0.604 Å between 360 out of 385 amino acid residues (Figure 6c). Protein structure comparison using DALI Server (Holm, 2020) revealed that AtXAPT1 and EgXAPT1 shared the highest structural similarities with human exostosin 1 and 2 (PDB: 7SCJ; Li et al., 2023) and human exostosin-like 3 (PDB: 8OG1; Sammon et al., 2023), all of which contain an N-terminal GT47 domain (Figure S6). Specifically, 106 amino acid residues (out of 350) of AtXAPT1 structure aligned well with exostosin-1 N-terminal GT47 domain structure with an RMSD of 1.19 Å and a Z-score of 22.0. These results indicate that AtXAPT1 and EgXAPT1 exhibit structural features typical of the GT-B fold.

Figure 6.

Figure 6.

Open in a new tab

Structural models of the catalytic domains of AtXAPT1 and EgXAPT1 predicted by AlphaFold2. β-sheets and α-helices are colored in cyan and yellow, respectively, in (a) and (b).

(a,b) Predicted structural models of the catalytic domains of AtXAPT1 (a) and EgXAPT1 (b) showing two separate N- and C-terminal Rossmann-like domains and a cleft, where the predicted catalytic site resides, between them. (c) Superimposed structures of the catalytic domains of AtXAPT1 (green) and EgXAPT1 (blue) showing similar global folds and structural similarity.

To pinpoint candidate amino acid residues at the putative active site of XAPTs involved in substrate binding, we next performed molecular docking of the predicted structural models of AtXAPT1 and EgXAPT1 with the (GlcA)Xyl4 acceptor and the UDP-Arap or UDP-Gal donor. The docked structures showed that both donor and acceptor substrates were positioned in the interdomain cleft of AtXAPT1 and EgXAPT1 (Figure 7a,c,e), where the putative active site is located for GT-B fold proteins (Lairson et al., 2008; Chang et al., 2011). Multiple amino acid residues at the putative active sites of AtXAPT1 and EgXAPT1 were predicted to be involved in interacting with the nucleotide sugar donors. Specifically, the docked AtXAPT1 structure with UDP-Arap showed hydrogen (H)-bonding of Thr205, Tyr251 and Thr345 with the uridine base, Arg286, Arg295, Asp342 and Arg346 with the diphosphate group, and Ser348 with the Arap moiety (Figure 7b). The docked EgXAPT1 structure with UDP-Arap showed H-bonding of Thr234, Tyr282, Thr377 with the uridine base, Arg326 and Arg378 with the diphosphate group, and Cys352 and Arg378 with the Arap moiety of UDP-Arap (Figure 7d), and the docked EgXAPT1 structure with UDP-Gal indicated H-bonding of Thr377 and Arg326 with the uridine base, Gln153 and Arg378 with the diphosphate group, and Tyr285, Ile313, Gly314 and Cys344 with the Gal moiety (Figure 7f). It is of note that several Arg residues in AtXAPT1 and EgXAPT1 are among the predicted amino acid residues forming H-bonds with the diphosphate group of UDP-sugar donors, which is congruent with GT-B fold proteins having positively charged residues binding to the diphosphate group of UDP-sugars (Lairson et al., 2008; Chang et al., 2011). Furthermore, an Asp residue in several inverting GT-B fold proteins was suggested to function as the catalytic base (Lairson et al., 2008). In AtXAPT1, an Asp residue (Asp342) was also predicted to form H-bond with UDP-Arap, and the corresponding Asp374 residue in EgXAPT1 is within the possible interaction distance (4.36 Å from UDP-Arap and 4.21 Å from UDP-Gal) although it was not predicted to have H-bond with UDP-Arap/UDP-Gal.

Figure 7.

Figure 7.

Open in a new tab

Molecular docking of AtXAPT1 and EgXAPT1 with donor and acceptor substrates. Nucleotide sugars, (GlcA)Xyl4 and the side chains of amino acid residues are represented with sticks. Hydrogen bonds are denoted with dotted lines.

(a) The cleft region of docked AtXAPT1 showing the positioning of (GlcA)Xyl4 and UDP-Arap.

(b) The putative active site of docked AtXAPT1 showing hydrogen-bonding between amino acid residues and UDP-Arap.

(c) The cleft region of docked EgXAPT1 showing the positioning of (GlcA)Xyl4 and UDP-Arap.

(d) The putative active site of docked EgXAPT1 showing hydrogen-bonding between amino acid residues and UDP-Arap.

(e) The cleft region of docked EgXAPT1 showing the positioning of (GlcA)Xyl4 and UDP-Gal.

(f) The putative active site of docked EgXAPT1 showing hydrogen-bonding between amino acid residues and UDP-Gal.

Mutagenesis analysis demonstrates the essential roles of several amino acid residues at the putative active site of EgXAPT1 in its activity

To determine the significance of the predicted amino acid residues involved in H-bonding with the nucleotide sugar donor at the putative active site, we chose EgXAPT1 as a representative for site-directed mutagenesis. Each of the amino acid residues predicted to be involved in H-binding with UDP-Arap or UDP-Gal, including Gln153, Thr234, Tyr282, Tyr285, Ile313, Gly314, Arg326, Cys344, Cys352, Thr377 and Arg378, was substituted with alanine (Figure 8a) and the resulting 11 mutants of EgXAPT1 were expressed in HEK293 cells for production of recombinant proteins. We successfully generated eight mutant proteins of EgXAPT1 (Figure 8b), but the other three mutants, including I313A, C344A and C352A, failed to be expressed probably due to their improper folding. Activity assays revealed that when incubated with (GlcA)Xyl6 and UDP-Arap, these mutant proteins had a more than 90% reduction in their xylan 2-O-arabinopyranosyltransferase activity compared with the wild-type EgXAPT1 protein, except that the Y285A mutant had a 50% reduction (Figure 8c). Likewise, a more than 80% reduction in their xylan 2-O-galactosyltransferase activity was observed when these mutant proteins were incubated with (GlcA)Xyl6 and UDP-Gal (Figure 8d). These results indicate that these amino acid residues located at the putative active site play important roles in the enzymatic activity of EgXAPT1.

Figure 8.

Figure 8.

Open in a new tab

Effect of mutations of amino acid residues at the predicted active site of EgXAPT1 on its activity. (a) Diagram of the EgXAPT1 protein showing the positions of the transmembrane helix (TM) and the catalytic domain. Marked below are the mutated amino acid residues located at the predicted active site and all of them are substituted with alanine. (b) SDS-PAGE and Coomassie Blue staining of recombinant EgXAPT1 mutant proteins expressed in HEK293 cells. (c,d) Enzymatic activities of EgXAPT1 mutant proteins. The EgXAPT1 wild-type and mutant proteins were incubated with the acceptor (GlcA)Xyl6 and the donor UDP-Arap (c) or UDP-Gal (d) and their activities were examined using the UDP-Glo glycosyltransferase assay. The activity of wild-type EgXAPT1 was taken as 100% and that of the mutants was shown as percentage of the wild-type. Error bars denote the SD of three independent assays.

Based on the amino acid sequence alignment of XAPTs (Figure S7), it was evident that the amino acid residues predicted to be involved in H-binding with UDP-Arap/UDP-Gal were conserved among all of the biochemically characterized XAPTs (Figure 9a), suggesting that these residues may not be directly responsible for the differential preference for UDP-Arap or UDP-Gal exhibited by different XAPTs. We reasoned that the positioning of some of the amino acids involved in H-bonding with UDP-sugar donors in the substrate-binding pockets of XAPTs might be different among different XAPTs, thereby resulting in their differential binding of UDP-Arap and UDP-Gal. Examination of the structural alignment of amino acid residues H-bonding with UDP-Arap in AtXAPT1 and EgXAPT1 showed that their side chains were closely matched in positioning (Figure 9b), consistent with the fact that both of them use UDP-Arap as a sugar donor. Since EgXAPT1 but not AtXAPT1 could also use UDP-Gal as a sugar donor, it is possible that the positioning of amino acid residues H-bonding with UDP-Gal in EgXAPT1 might exhibit some differences from that of the corresponding residues in AtXAPT1. Unexpectedly, structural alignment of these residues also did not reveal any apparent differences in the positioning of their side chains between EgXAPT1 and AtXAPT1 (Figure 9c), indicating that the positioning of these residues in the substrate-binding pocket is unlikely accountable for the divergence of EgXAPT1 and AtXAPT1 in using UDP-Gal as a sugar donor.

Figure 9.

Figure 9.

Open in a new tab

Structural alignment of the amino acid residues of AtXAPT1 and EgXAPT1 predicted to be involved in H-bonding with UDP-Arap/UDP-Gal.

(a) Sequence alignment of motifs containing amino acid residues predicted to be involved in H-bonding with UDP-Arap/UDP-Gal among the biochemically characterized XAPTs. The top panel shows the alignment of seven motifs, each of which contains one or more amino acid residues (red font) H-bonding with UDP-Arap/UDP-Gal in AtXAPT1 and EgXAPT1. The bottom panel depicts the positions of these residues in their respective proteins. The amino acid residues in the bottom panel are aligned with their corresponding ones (red font) in the top panel. Note the conservation of these residues among all of the biochemically characterized XAPTs.

(b) Structural alignment of amino acid residues H-bonding with UDP-Arap in the putative substrate-binding pockets of AtXAPT1 and EgXAPT1. UDP-Arap (red for AtXAPT1 and black for EgXAPT1) and the side chains of amino acid residues are represented with sticks. The structure of AtXAPT1 is colored in cyan and that of EgXAPT1 in tan. Note the closely matched positioning of the side chains of these amino acid residues between AtXAPT1 and EgXAPT1.

(c) Structural alignment of amino acid residues H-bonding with UDP-Gal in EgXAPT1 with the corresponding residues in AtXAPT1. The structure of AtXAPT1 is colored in cyan and that of EgXAPT1 in tan. Note the closely matched positioning of the side chains of these amino acid residues.

(d) Superimposition of the surface structural model of the substrate-binding pocket of EgXAPT1 with that of AtXAPT1. The amino acid residues of EgXAPT1 (tan) within 8 Å from UDP-Gal were superimposed with the corresponding residues of AtXAPT1 (cyan). Note the closer distance of EgXAPT1 His353 side chain (tan) to the Gal moiety of UDP-Gal than that of the corresponding AtXAPT1 His321 (cyan).

To further investigate whether there exist any subtle differences in the substrate-binding pockets of EgXAPT1 and AtXAPT1, we analyzed the positioning of amino acid residues located within 8 Å from UDP-Gal in the substrate-binding pocket of EgXAPT1 together with the corresponding residues of AtXAPT1. It was found that they were all well aligned structurally except the side chain of His353 in EgXAPT1 and that of the corresponding His321 in AtXAPT1 (Figure 9d). The side chain of His353 in EgXAPT1 showed a distance of 3.60 Å from the Gal moiety of UDP-Gal, which is well within the possible interaction distance (< 5 Å), whereas the corresponding His321 side chain in AtXAPT1 was farther away (5.70 Å). Although the significance of His353 of EgXAPT1 in UDP-Gal binding needs to be experimentally determined, it is tempting to propose that EgXAPT1 but not AtXAPT1 might have evolved to adopt such a subtly altered structure in its substrate-binding pocket that could accommodate UDP-Gal in addition to UDP-Arap as a sugar donor. Sequence alignment of previously characterized Arabidopsis MUR3-clade members, including AtMUR3, AtXLT2, AtXUT1 and AtMBGT1, with XAPTs showed that some amino acid residues predicted be involved in H-binding with UDP-sugars in AtXAPT1 and EgXAPT1 were also conserved in these proteins (Figure S8). The involvement of these amino acid residues in binding of UDP-sugar donors in AtMUR3, AtXLT2, AtXUT1 and AtMBGT1 awaits to be studied.

Occurrence of XAPT close homologs in angiosperms but not in bryophytes, seedless vascular plants and gymnosperms

GlcA/MeGlcA side chains are present in xylan from all lineages of land plants, including bryophytes, seedless vascular plants, gymnosperms and angiosperms. To ascertain in which lineages of land plants XAPT close homologs first appeared, we used the amino acid sequence of AtXAPT1 to BLAST-search the genome sequence databases of various lineages of plant species for XAPT close homologs and verified their relationships with AtXAPT1 by phylogenetic analysis. It was found that none of the MUR3-clade GT47 members from the representative plants of bryophytes (Marchantia polymorpha and Physcomitrium patens), seedless vascular plants (Selaginella moellendorffii) and gymnosperms (pine and spruce) were closely grouped together with XAPTs (Figure 10), indicating that these lineages of plants do not have XAPT close homologs. Several pine and spruce MUR3-clade members appeared to be rooted together with XAPTs, but they are only distantly related to the XAPT subclade and thus they are unlikely XAPT close homologs (Figure 10). Besides the dicot and monocot XAPTs studied in this report, many other dicot and monocot species have XAPT close homologs (Table S1; Figure S9). However, XAPT close homologs do not exist in all dicot and monocot species. Of particular note is that no XAPT close homologs were found in grass species (Figure 10; Table S1), which is consistent with the report that no 2-O-Arap-substituted GlcA/MeGlcA side chains were detected in grass xylan (Pena et al., 2016). In addition, many other plant species, such as coffee (Coffea arabica), sunflower (Helianthus annuus), oakleaf hydrangea (Hydrangea quercifolia), white lupin (Lupinus albus), castor bean (Ricinus communis) and papaya (Carica papaya), do not harbor XAPT close homologs (Table S1 Q12SXW z), indicating that like grass xylan, xylan from these species may not have 2-O-Arap- or 2-O-Gal-substituted GlcA/MeGlcA side chains. The presence of XAPT close homologs in some but not all of dicot and monocot species also implies that XAPTs may have arisen independently multiple times or descended from a common ancestral gene but were lost in some plant groups during angiosperm evolution.

Figure 10.

Figure 10.

Open in a new tab

Phylogenetic relationships of the MUR3-clade GT47 members from representatives of various lineages of land plants together with the XAPTs studied in this report (indicated in red font). The phylogenetic tree was constructed using the maximum likelihood method and the numbers at the nodes are percentages of bootstrap values from 1,000 replicates. The 0.1 scale denotes 10% change. AT, Arabidopsis thaliana; MA, Picea abies; Mapoly, Marchantia polymorpha; Os, Oryza sativa; PITA, Pinus taeda; Pp, Physcomitrium patens, Sm, Selaginella moellendorffii.

DISCUSSION

Previously characterized GT47 members of the MUR3 clade were shown to display diverse enzymatic activities with different donor and acceptor specificities. For example, MUR3, XLT2, XUT1, XST1, XDT and XBT1 catalyze the transfer of various glycosyl residues onto Xyl side chain of xyloglucan (Madson et al. 2003; Jensen et al. 2012; Pena et al., 2012; Schultink et al., 2013; Zhu et al., 2018; Immelmann et al., 2023; Wilson et al., 2023), whereas MBGT1 mediates the addition of a Gal residue onto Gal side chains of mannan (Yu et al., 2022). In this report, we have demonstrated that XAPTs in the MUR3 clade possess a function different from the known members of the MUR3 clade, i.e., they are xylan 2-O-arabinopyranosyltransferases/galactosyltransferases transferring 2-O-Arap and/or 2-O-Gal onto GlcA side chains of xylan. Our findings provide an additional example to the growing list of the remarkably diverse functions of the MUR3-clade GT47 members.

It is intriguing that while XAPTs from Arabidopsis and poplar are only able to transfer Arap onto GlcA side chains of xylan, those from Eucalyptus, lemon-scented gum, sea apple, ‘Ohi’a lehua, duckweed and purple yam are capable of transferring both Arap and Gal onto xylan GlcA side chains albeit with differential preferences. This finding indicates that XAPTs from different plant species evolved to utilize one or both of UDP-Arap and UDP-Gal as the nucleotide sugar donors. XAPTs are the first plant GT47 members demonstrated to exhibit dual activities utilizing two different nucleotide sugar donors. Several other plant glycosyltransferases belonging to different GT families have previously been shown to possess dual activities. For example, the Arabidopsis galactan synthase AtGALS1 in family GT92 catalyzes the transfer of both Gal and Arap to galactan chains (Laursen et al., 2018). Another example is the rice xylan xylosyl/arabinosyl transferase OsXXAT1 in family GT61, which mediates both 2-O-Xyl and 2-O-Araf substitutions of xylan (Zhong et al., 2024b).

We have revealed that 2-O-Arap-substituted GlcA side chains are present in Arabidopsis xylan and mutation of AtXAPT1 results in a complete loss of the 2-O-Arap-substituted GlcA side chains, demonstrating that AtXAPT1 is responsible for the 2-O-Arap substitutions of xylan GlcA side chains in Arabidopsis. 2-O-Arap-substituted GlcA side chains have previously been detected in xylan from Eucalyptus and several non-commelinid monocot species (Togashi et al., 2009; Pena et al., 2016). Considering that XAPTs from Eucalyptus, duckweed and purple yam are able to arabinosylate xylan GlcA side chains, it is reasonable to propose that the 2-O-Arap substitutions of xylan GlcA side chains in these species are also mediated by XAPTs. It is interesting to note that two XAPTs exist in each of the four examined species of the myrtle family and they are phylogenetically grouped into two subgroups; one subgroup (EgXAPT1 and SgXAPT1) exhibited a higher activity using UDP-Arap and the other (EgXAPT2, CciXAPT2 and MpoXAPT2) a higher activity using UDP-Gal (Figure 4). Since both 2-O-Arap- and 2-O-Gal-substituents are present on xylan GlcA side chains in Eucalyptus (Pena et al., 2016), it is conceivable that the concerted actions of EgXAPT1 and EgXAPT2 are responsible for these substitutions.

It was surprising to find that while the Arap- or Gal-substituted GlcA side chains of xylan were efficiently methylated by AtGXM3, the methylated GlcA residues could barely be substituted with Arap or Gal by XAPTs. These findings indicate that Arap- or Gal-substitutions of GlcA side chains do not affect their methylation, but 4-O-methylation of the GlcA side chains prevents their further glycosyl substitutions despite the much smaller size of the methyl group than the Arap and Gal residues. The methyl group may affect the binding of the acceptor to the active site of XAPTs or methylation at the O-4 position of GlcA may sterically hinder its further substitution at the O-2 position. On the contrary, glycosyl substitution at the O-2 position of GlcA does not sterically obstruct its further methylation at the O-4 position and the active site of GXMs must be able to accommodate the larger size of the glycosyl-substituted GlcA. In Eucalyptus, the GlcA side chains of xylan are completely methylated and they also have 2-O-Arap or 2-O-Gal substituents attached to them (Togashi et al., 2009; Pena et al., 2016). Since both EgXAPT1 and EgXAPT2 act poorly on methylated GlcA residues, it is probable that the GlcA side chains of xylan are first substituted with 2-O-Arap and 2-O-Gal by XAPTs and then methylated by GXMs.

Our molecular docking and site-directed mutagenesis study of EgXAPT1 identified several amino acid residues at the putative active site involved in H-bonding with UDP-Arap and UDP-Gal and their essential roles in EgXAPT1 activity. Some of these residues, including Gln153, Gly314, Arg326, Cys344, Thr377 and Arg378, are completely conserved not only in XAPTs but also in four previously characterized Arabidopsis MUR3-clade members, AtXUT1, AtXLT2, AtMUR3 and AtMBGT1 (Figures S7 and S8), indicating that they likely play important roles in the catalytic activities of MUR3-clade members in general. It is interesting to note that although EgXAPT1 but not AtXAPT1 evolved to use UDP-Gal in addition to UDP-Arap as a sugar donor, the amino acid residues involved in H-bonding with UDP-Arap/UDP-Gal and their positioning in the substrate-binding pockets were well conserved between EgXAPT1 and AtXAPT1. Detailed examination of the structural alignment of the substrate-binding pockets of EgXAPT1 and AtXAPT1 revealed only a subtle difference; the His353 side chain in EgXAPT1 is positioned within 3.60 Å to the Gal moiety of UDP-Gal, indicating a possible interaction between them, but the corresponding His321 side chain in AtXAPT1 is 5.70 Å away, which is beyond the distance for a possible interaction. We propose that a subtle alteration in the structure of the substrate-binding pocket may alter the differential preference for UDP-Arap/UDP-Gal exhibited by different XAPTs. Future in-depth structural studies of XAPTs together with the xyloglucan and mannan synthesis-related MUR3-clade members are needed to decipher how the regiospecificity for various donors and acceptors is determined in the functionally diverged MUR3 clade of the GT47 family.

EXPERIMENTAL PROCEDURES

Phylogenetic analysis

The amino acid sequences of Arabidopsis MUR3-clade members and the previously characterized MUR3-clade members from other plant species were retrieved from the genome sequence database at Phytozome v13 (https://phytozome-next.jgi.doe.gov) and the GenBank database. MUR3-clade members from different lineages of land plants were identified by using the amino acid sequences of all Arabidopsis MUR3-clade members as queries to BLAST-search the genomes of Marchantia polymorpha, Physcomitrium patens, Selaginella moellendorffii and Oryza sativa at Phytozome v13, and Picea abies and Pinus taeda at PlantGenIE (https://plantgenie.org/). XAPT close homologs from various plant species were identified by using the amino acid sequence of AtXAPT1 as a query to BLAST-search the GenBank database and the genome sequence databases at Phytozome v13 and PlantGenIE. XAPT homologs from ‘Ohi’a lehua were identified from its genome sequence (Izuno et al., 2019). The phylogenetic relationships of the MUR-clade members and XAPTs was evaluated using the MEGA11 software with the maximum likelihood method.

Production of recombinant XAPT proteins

The mammalian HEK293 cells were used for heterologous expression of recombinant XAPT proteins. The cDNA sequences of the putative catalytic domains of XAPTs were cloned in frame between the murine Igκ chain leader sequence (for protein secretion) and the c-myc epitope and six tandem histidine tag in the pSecTag2 mammalian expression vector (Invitrogen) to generate XAPT expression constructs. The amino acids used for recombinant protein production were amino acids 56 to 455 for AtXAPT1, 58 to 487 for PtrXAPT1, 58 to 487 for PtrXAPT2, 64 to 489 for EgXAPT1, 43 to 464 for EgXAPT2, 60 to 487 for SgXAPT1, 59 to 478 for CciXAPT2, 60 to 485 for MpoXAPT2, 43 to 472 for SpXAPT1 and 44 to 441 for DaXAPT1. The expression constructs were transfected into HEK293 cells using the Invitrogen FreeStyle 293 Expression System according to the manufacturer’s protocol. After 5 days of culture, the secreted recombinant XAPT proteins in the culture media were purified by passing through a nickel resin column. The purified proteins were examined by sodium dodecyl sulfate-polyacrylamide gel electrophoresis followed by Coomassie Blue staining.

Activity assays of recombinant XAPT proteins

AtXAPT1 was tested for its possible activity toward xyloglucan by incubation of 20 μg recombinant AtXAPT1 protein with 0.3 mM XXXG oligomers (Megazyme), 1 mM nucleotide sugar donor [UDP-Gal (Sigma), UDP-GalA (CarboSource), UDP-Xyl (CarboSource), UDP-Arap (CarboSource) or UDP-Araf (Peptide Institute)], 1 mM MgCl2 and 50 mM HEPES buffer (pH 7.0) in a 40-μl reaction mixture. The xylan arabinopyranosyltransferase and galactosyltransferase activities of XAPTs were assayed in a reaction mixture (40 μl) containing 20 μg recombinant protein, 1 mM nucleotide sugar donor (UDP-Arap or UDP-Gal), 0.3 mM xylooligomer acceptors [Xyl6-AA, (GlcA)Xyl6-AA or (MeGlcA)Xyl6-AA], 1 mM MgCl2 and 50 mM HEPES buffer (pH 7.0). Glucuronoxylan methyltransferase assays were conducted in a reaction mixture (40 μl) containing 20 μg recombinant AtGXM3 protein (Lee et al., 2012), 0.3 mM xylooligomer acceptors [(GlcA)Xyl6-AA, (Arap-GlcA)Xyl6-AA or (Gal-GlcA)Xyl6-AA)], 1mM S-adenosylmethionine (SAM), 1 mM CoCl2 and 50 mM HEPES buffer (pH 7.0). To carry out the sequential reactions of methylation and arabinosylation/galactosylation of the acceptor, (GlcA)Xyl6-AA was first methylated by AtGXM3 and then used as an acceptor for arabinosylation/galactosylation by XAPTs, or it was first arabinosylated/galactosylated by XAPTs and then used as an acceptor for methylation by AtGXM3. After incubation at 37 °C for 16 hr, the reaction products were analyzed with MALDI-TOF mass spectrometry. For NMR spectroscopic analysis of the reaction products, the reaction mixtures were scaled up to 300 μl with (GlcA)Xyl6 as the acceptor. The XAPT-catalyzed reaction products from three replicates were used for subsequent analyses.

For quantitative analysis of the activities of EgXAPT1 and its mutant forms, we measured the amount of UDP released from UDP-sugars as a by-product of the EgXAPT1-catalyzed reactions. The reactions (20 μl) containing 10 μg recombinant protein, 0.3 mM nucleotide sugar donor (UDP-Arap or UDP-Gal), 0.3 mM (GlcA)Xyl6, 1 mM MgCl2 and 50 mM HEPES buffer (pH 7.0) were incubated for 1 hr at 37 °C and then heat inactivated. The UDP released in the reaction products was quantitated using the Promega UDP-Glo™ Glycosyltransferase Assay according to the manufacturer’s protocol. Briefly, 10 μl reaction products were mixed with 10 μl of UDP Detection Reagent and after incubation for 1 hr at room temperature, the mixture was measured for luminescence with a luminometer. The activity of wild-type EgXAPT1 remained linear during the 1-hr incubation period (Figure S10).

Xylohexaose (Neogen) was labeled at its reducing end with anthranilic acid (AA) to produce Xyl6-AA according to Ishii et al. (2002). (GlcA)Xyl6-AA oligomers were generated by incubating Xyl6-AA with a recombinant rice xylan glucuronyltransferase (OsGUX1; Zhong et al., 2024a) in a reaction mixture containing 1 mM UDP-GlcA (Sigma), 1 mM MgCl2 and 50 mM HEPES (pH 7.0).

Genotyping and complementation analyses of the xapt1 mutants

Arabidopsis T-DNA insertion mutant seeds of the AtXAPT1 gene [CS855841 (xapt1), CS876068 (xapt1–2), CS855819 and CS855786] were obtained from the Arabidopsis Biological Resource Center (ABRC). Genotyping of these lines showed that CS855841 (xapt1) and CS876068 (xapt1–2) had T-DNA insertions in the AtXAPT1 gene, whereas CS855819 and CS855786 did not have any T-DNA insertions in the AtXAPT1 gene and hence they were not further analyzed. The homozygous xapt1 mutant plants were genotyped using AtXAPT1 gene-specific primers (AtXAPT1-FP1: 5’-actttggagctaacatgctcatgc-3’ and AtXAPT1-RP1: 5’-acatcgcatcgaatgttgatctcc-3’) spanning the T-DNA insertion site and a T-DNA primer (T1: 5’-aacgtccgcaatgtgttattaagttgtc-3’) (Figure S3b). The homozygous xapt1–2 mutant plants were genotyped using AtXAPT1 gene-specific primers (AtXAPT1-FP2: 5’-tcactaagacaagttgttaccgac-3’ and AtXAPT1-RP2: 5’-tgaactggtgcgtagagaaccacg-3’) spanning the T-DNA insertion site and a T-DNA primer (T2: 5’-tagcatctgaatttcataaccaatctcgatacac-3’) (Figure S3b). To examine the level of AtXAPT1 transcripts in these mutants, total RNA from 2-week-old seedlings was isolated using a Qiagen RNA isolation kit and then subjected to RT-PCR analysis of AtXAPT1 transcripts using the AtXAPT1-FP1 and -RP1 primers. The expression of the EF1α gene was used as a reference among different samples. For histological analysis, the bottom parts of stems of 7-week-old plants were fixed in 2% formaldehyde and embedded in LR White resin. One-μm-thick sections of embedded stems were cut with a microtome and stained with toluidine blue for anatomy. For complementation analysis, the entire AtXAPT1 gene encompassing the 3-kb 5’ upstream sequence, the exon and intron region and the 2-kb 3’ downstream sequence was cloned into a modified pGPTV binary vector, which was then introduced into the homozygous xapt1 mutant plants by Agrobacterium-mediated transformation. Transgenic plants were selected by their resistance to hygromycin and more than 60 independent transgenic plants were selected for further analysis. RT-PCR analysis showed that the expression of AtXAPT1 was restored in the xapt1 mutant complemented with the wild-type AtXAPT1 gene (Figure S3d).

Cell wall extraction and xylanase digestion of extracted xylan

Alcohol-insoluble cell wall residues were extracted from the inflorescence stems of 6-week-old wild-type Arabidopsis plants, the xapt1 mutant and xapt1 complemented with the wild-type AtXAPT1 gene according to Zhong et al. (2005). The alcohol-insoluble cell wall residues were extracted for xylan by treatments first with ammonium oxalate to remove pectins and then with 4 N KOH in the presence of sodium borohydride. Xylooligomers were isolated from the KOH-extracted xylan by digestion with endo-1,4-β-xylanase M6 from rumen microorganism (Megazyme) and subsequent separation through a Sephadex G25 column (100 × 1 cm) (Zhong et al. 2005). Xylooligomers isolated from three separate pools of plant samples were used for subsequent analysis.

MALDI-TOF mass spectrometry

The reaction products were analyzed by MALDI-TOF mass spectrometry using 12T SolariX Fourier-transform ion cyclotron resonance mass spectrometer (Bruker) and AutoFlex Max MALDI-TOF mass spectrometer (Bruker). The samples were mixed (1:1, v/v) with the MALDI matrix (0.1 M 2,5-dihydroxybenzoic acid in 50% methanol). The spectra of each sample were acquired with the averages of at least 500 laser shots. Samples from three biological replicates were analyzed and representative spectra were shown.

NMR spectroscopy

The XAPT-catalyzed reaction products and xylooligomers isolated from the wild-type Arabidopsis and the xapt1 mutant were analyzed with Varian Inova 400 MHz and Agilent DD2 600 MHz NMR spectrometers. 1D and 2D (COSY) NMR spectra were recorded using standard Varian pulse sequences. The proton positions and residue identities in the NMR spectra were assigned based on our 1D and 2D NMR spectral data and the published NMR spectral data for xylan (Zhong et al., 2005; Pena et al., 2016).

AlphaFold2 structure prediction and molecular docking

The structures of the catalytic domains of AtXAPT1 (amino acids 67–455) and EgXAPT1 (amino acids 91–489) were predicted using a locally installed AlphaFold2 (v2.3.1) (Jumper et al. 2021) and the generated relaxed models were further employed in the molecular docking studies. The 3D structure of the acceptor substrate (GlcA)Xyl4 was modified from a (MeGlcA)Xyl4 structure (PubChem CID: 156618894), and those of the donor substrates UDP-β-L-Arap and UDP-α-D-Gal were obtained from PubChem database (Kim et al. 2022). Molecular docking of the acceptor and donor substrates onto the AlphaFold2 structures of AtXAPT1 and EgXAPT1 catalytic domains was performed using AutoDock Vina (Eberhardt et al. 2021) in Chimera (Pettersen et al. 2004) with the grid box set to cover the whole macromolecules. After docking each acceptor and donor substrate, the top scoring conformations were analyzed together.

Supplementary Material

Supinfo

Figure S1. Biological replicates of the reaction products of AtXAPT1.

Figure S2. 1H-NMR spectra of biological replicates of xylooligomers isolated from wild-type Arabidopsis, the xapt1 mutant and xapt1 complemented with the wild-type AtXAPT1 gene as shown in Figure 3a.

Figure S3. Characterization of Arabidopsis xapt1 mutants.

Figure S4. MALDI-TOF mass spectra of biological replicates of the reaction products of XAPTs incubated with the (GlcA)Xyl6-AA acceptor and UDP-Arap or UDP-Gal as shown in Figure 4c–f.

Figure S5. Biological replicates of reaction products catalyzed by XAPTs or sequentially by XAPTs and AtGXM3.

Figure S6. Structural alignment of the predicted model of AtXAPT1 with the structures of GT47 domains of three characterized human exostosins showing similar global folds and structural similarity.

Figure S7. Amino acid sequence alignment of XAPTs studied in this report with the four previously functionally characterized Arabidopsis MUR3-clade members, AtXUT1, AtXLT2, AtMUR3 and AtMBGT1.

Figure S8. Sequence alignment of motifs containing amino acid residues (red font) predicted to H-bond with UDP-Arap/UDP-Gal in EgXAPT1 and AtXAPT1 among the biochemically characterized XAPTs together with four other Arabidopsis MUR3-clade members.

Figure S9. Phylogenetic relationships of XAPT close homologs identified from various lineages of plant species together with the XAPTs studied in this report.

Figure S10. Time course of UDP release by EgXAPT1 incubated with UDP-Arap (a) or UDP-Gal (b) and (GlcA)Xyl6.

Table S1 Survey of various lineages of plant species for the presence of XAPT close homologs

SIGNIFICANCE STATEMENT.

Xylan is one of the major hemicelluloses in plant cell walls and its GlcA side chains may be further substituted with 2-O-Arap or 2-O-Gal residues. We have uncovered the functions of a subgroup of MUR3-clade GT47 members in catalyzing the transfer of 2-O-Arap and/or 2-O-Gal onto GlcA side chains of xylan, which expands our understanding of glycosyltransferases involved in xylan substitutions.

ACKNOWLEDGMENTS

This work was funded by the U.S. Department of Energy, Office of Science, Office of Basic Energy Sciences [grant No. DE-FG02-03ER15415]. The authors are grateful for the support of the National Institutes of Health [grant No. S10 OD025118] for funding the acquisition of the Solarix XR mass spectrometer, the Arabidopsis Biological Resource Center for providing the Arabidopsis xapt1 mutant seeds (CS855841), and the UGA Proteomics and Mass Spectrometry Core Facility and the UGA Chemistry NMR Facility for instrumentation.

Footnotes

CONFLICT OF INTEREST

The authors declare no conflict of interest.

ACCESSION NUMBERS

The gene locus identifiers or GenBank accession numbers for the genes studied in this work are AT1G68470 for AtXAPT1, Potri.001G381900 for PtrXAPT1, Potri.001G382200 for PtrXAPT2, Eucgr.D00738 for EgXAPT1, Eucgr.H00343 for EgXAPT2, KAI6699165 for SgXAPT1, Cocit.H0064 for CciXAPT2, BCNH02000008.1 for MpoXAPT2, Spipo26G0017500 for SpXAPT1 and Dioal.04G064800 for DaXAPT1.

SUPPORTING INFORMATION

Supplementary Figures S1–S10 and Table S1 are available in the supplementary data files.

DATA AVAILABILITY STATEMENT

All data generated during this study are included in this article and its Supplementary files.

REFERENCES

  1. Amos RA & Mohnen D (2019) Critical review of plant cell wall matrix polysaccharide glycosyltransferase activities verified by heterologous protein expression. Front. Plant Sci. 10, 915. [DOI] [PMC free article] [PubMed] [Google Scholar]
  2. Anders N, Wilkinson MD., Lovegrove A., Freeman J., Tryfona T., Pellny TK., et al. (2012) Glycosyl transferases in family 61 mediate arabinofuranosyl transfer onto xylan in grasses. Proc. Natl. Acad. Sci. USA 109, 989–993. [DOI] [PMC free article] [PubMed] [Google Scholar]
  3. Avci U (2022) Trafficking of xylan to plant cell walls. Biomass 2, 188–194. [Google Scholar]
  4. Bowman MJ, Dien BS, Vermillion KE & Mertens JA (2014) Structural characterization of (1→2)-β-xylose-(1→3)-α-arabinose-containing oligosaccharide products of extracted switchgrass (Panicum virgatum, L.) xylan after exhaustive enzymatic treatment with α-arabinofuranosidase and β-endo-xylanase. Carbohydr. Res. 398, 63–71. [DOI] [PubMed] [Google Scholar]
  5. Brown DM, Zhang Z, Stephens E, Dupree P and Turner SR (2009) Characterization of IRX10 and IRX10-like reveals an essential role in glucuronoxylan biosynthesis in Arabidopsis. Plant J. 57, 732–46. [DOI] [PubMed] [Google Scholar]
  6. Busse-Wicher M, Li A, Silveira RL, Pereira CS, Tryfona T, Gomes, et al. (2016) Evolution of xylan substitution patterns in gymnosperms and angiosperms: implications for xylan interaction with cellulose. Plant Physiol. 171, 2418–3241. [DOI] [PMC free article] [PubMed] [Google Scholar]
  7. Chang A, Singh S, Phillips GN Jr., Thorson JS (2011) Glycosyltransferase structural biology and its role in the design of catalysts for glycosylation. Curr. Opin. Biotechnol. 22, 800–808. [DOI] [PMC free article] [PubMed] [Google Scholar]
  8. Chiniquy D, Sharma V, Schultink A, Baidoo EE, Rautengarten C, Cheng K, et al. (2012) XAX1 from glycosyltransferase family 61 mediates xylosyltransfer to rice xylan. Proc Natl Acad Sci USA 109, 17117–17122. [DOI] [PMC free article] [PubMed] [Google Scholar]
  9. Chong S-L, Koutaniemia S, Juvonena M, Derba-Maceluchb M,Mellerowiczb EJ & Tenkanena M (2015) Glucuronic acid in Arabidopsis thaliana xylans carries a novel pentose substituent. Intl. J. Biol. Macromol. 79, 807–812. [DOI] [PubMed] [Google Scholar]
  10. Coutinho PM, Stam M, Blanc E & Henrissat B (2003) Why are there so many carbohydrate-active enzyme-related genes in plants? Trends Plant Sci. 8, 563–565. [DOI] [PubMed] [Google Scholar]
  11. Eberhardt J, Santos-Martins D, Tillack AF & Forli S (2021) AutoDock Vina 1.2.0: new docking methods, expanded force field, and python bindings. J. Chem. Inf. Model 61, 3891–3898. [DOI] [PMC free article] [PubMed] [Google Scholar]
  12. Fry SC, York WS, Albersheim P, Darvill A, Hayashi T, Joseleau J-P, et al. (1993) An unambiguous nomenclature for xyloglucan-derived oligosaccharides. Physiol. Planta. 89, 1–3. [Google Scholar]
  13. Harholt J, Jensen JK, Verhertbruggen Y, Søgaard C, Bernard S, Nafisi M, et al. (2012) ARAD proteins associated with pectic Arabinan biosynthesis form complexes when transiently overexpressed in planta. Planta 236, 115–128. [DOI] [PubMed] [Google Scholar]
  14. Höije A, Sandstrom C, Roubroeks JP, Andersson R, Gohil S & Gatenholm P (2006) Evidence of the presence of 2-O-β-D-xylopyranosyl-α-L-arabinofuranose side chains in barley husk arabinoxylan. Carbohydr. Res. 341, 2959–2966. [DOI] [PubMed] [Google Scholar]
  15. Holm L (2020) DALI and the persistence of protein shape. Protein Sci. 29, 128–140. [DOI] [PMC free article] [PubMed] [Google Scholar]
  16. Immelmann R, Gawenda N, Ramírez V & Pauly M (2023) Identification of a xyloglucan beta-xylopyranosyltransferase from Vaccinium Corymbosum. Plant Direct 7, e514.10.1002. [DOI] [PMC free article] [PubMed] [Google Scholar]
  17. Ishii T, Ichita J, Matsue H, Ono H and Maeda I (2002) Fluorescent labeling of pectic oligosaccharides with 2-aminobenzamide and enzyme assay for pectin. Carbohydr. Res.337, 1023–1032. [DOI] [PubMed] [Google Scholar]
  18. Izuno A, Wicker T, Hatakeyama M, Copetti D & Shimizu KK (2019) Updated genome assembly and annotation for Metrosideros polymorpha, an emerging model tree species of ecological divergence. G3 (Bethesda) 9, 3513–3520. [DOI] [PMC free article] [PubMed] [Google Scholar]
  19. Jacobs A, Larsson PT & Dahlman O (2001) Distribution of uronic acids in xylans from various species of soft- and hardwood as determined by MALDI mass spectrometry. Biomacromolecules 2, 979–990. [DOI] [PubMed] [Google Scholar]
  20. Jensen JK, Johnson NR & Wilkerson CG (2014) Arabidopsis thaliana IRX10 and two related proteins from psyllium and Physcomitrella patens are xylan xylosyltransferases. Plant J. 80, 207–215. [DOI] [PubMed] [Google Scholar]
  21. Jensen JK, Schultink A, Keegstra K, Wilkerson CG & Pauly M (2012) RNA-Seq analysis of developing nasturtium seeds (Tropaeolum majus): identification and characterization of an additional galactosyltransferase involved in xyloglucan biosynthesis. Mol. Plant 5, 984–992 [DOI] [PMC free article] [PubMed] [Google Scholar]
  22. Jensen JK, Sørensen SO, Harholt J, Geshi N, Sakuragi Y, Møller I, et al. (2008) Identification of a xylogalacturonan xylosyltransferase involved in pectin biosynthesis in Arabidopsis. Plant Cell 20, 1289–1302. [DOI] [PMC free article] [PubMed] [Google Scholar]
  23. Jumper J, Evans R, Pritzel A, Green T, Figurnov M, Ronneberger O et al. (2021) Highly accurate protein structure prediction with AlphaFold. Nature 596, 583–589. [DOI] [PMC free article] [PubMed] [Google Scholar]
  24. Kim S, Chen J, Cheng T, Gindulyte A, He J, He S et al. (2022) PubChem 2023 update. Nucleic Acids Res. 51, D1373–D1380. [DOI] [PMC free article] [PubMed] [Google Scholar]
  25. Lairson LL, Henrissat B, Davies GJ & Withers SG (2008) Glycosyltransferases: structures, functions, and mechanisms. Annu. Rev. Biochem. 77, 521–55. [DOI] [PubMed] [Google Scholar]
  26. Laursen T, Stonebloom SH., Pidatala VR., Birdseye DS., Clausen MH., Mortimer JC. et al. (2018) Bifunctional glycosyltransferases catalyze both extension and termination of pectic galactan oligosaccharides. Plant J. 94, 340–351. [DOI] [PubMed] [Google Scholar]
  27. Lee C, Teng Q, Huang W, Zhong R and Ye Z-H (2009) The F8H glycosyltransferase is a functional paralog of FRA8 involved in glucuronoxylan biosynthesis in Arabidopsis. Plant Cell Physiol. 50, 812–827. [DOI] [PubMed] [Google Scholar]
  28. Lee C, Teng Q, Zhong R & Ye Z-H (2012a) Arabidopsis GUX proteins are glucuronyltransferases responsible for the addition of glucuronic acid side chains onto xylan. Plant Cell Physiol. 53, 1204–1216. [DOI] [PubMed] [Google Scholar]
  29. Lee C, Teng Q, Zhong R, Yuan Y, Haghighat M & Ye Z-H (2012b) Three Arabidopsis DUF579 domain-containing GXM proteins are methyltransferases catalyzing 4-O-methylation of glucuronic acid on xylan. Plant Cell Physiol. 53: 1934–1949. [DOI] [PubMed] [Google Scholar]
  30. Li H, Chapla D, Amos RA, Ramiah A, Moremen KW & Li H (2023) Structural basis for heparan sulfate co-polymerase action by the EXT1–2 complex. Nature Chem. Biol. 19, 565–574. [DOI] [PMC free article] [PubMed] [Google Scholar]
  31. Madson M, Dunand C, Li X, Verma R, Vanzin GF, Caplan J, et al. (2003) The MUR3 gene of Arabidopsis encodes a xyloglucan galactosyltransferase that is evolutionarily related to animal exostosins. Plant Cell 15, 1662–1670 [DOI] [PMC free article] [PubMed] [Google Scholar]
  32. Marriott PE, Gómez LD and McQueen-Mason SJ (2016) Unlocking the potential of lignocellulosic biomass through plant science. New Phytol. 209, 1366–1381. [DOI] [PubMed] [Google Scholar]
  33. McFarlane HE (2023) Open questions in plant cell wall synthesis. J. Exp. Bot. 74, 3425–3448. [DOI] [PubMed] [Google Scholar]
  34. Mortimer JC, Faria-Blanc N, Yu X, Tryfona T, Sorieul M, Yao Z. Ng YZ., et al. (2015) An unusual xylan in Arabidopsis primary cell walls is synthesised by GUX3, IRX9L, IRX10L and IRX14. Plant J. 83, 413–426. [DOI] [PMC free article] [PubMed] [Google Scholar]
  35. Mortimer JC, Miles GP, Brown DM, Zhang Z, Segura MP, Weimar T, et al. (2010) Absence of branches from xylan in Arabidopsis gux mutants reveals potential for simplification of lignocellulosic biomass. Proc. Natl. Acad. Sci. USA 107, 17409–17414. [DOI] [PMC free article] [PubMed] [Google Scholar]
  36. Pena MJ, Kong Y, York WS & O’Neill MA (2012) A galacturonic acid-containing xyloglucan is involved in Arabidopsis root hair tip growth. Plant Cell 24, 4511–4524 [DOI] [PMC free article] [PubMed] [Google Scholar]
  37. Pena MJ, Kulkarni AR, Backe J, Boyd M, O’Neill MA & York WS (2016) Structural diversity of xylans in the cell walls of monocots. Planta 244, 589–606. [DOI] [PubMed] [Google Scholar]
  38. Pettersen EF, Goddard TD, Huang CC, Couch GS, Greenblatt DM, Meng EC et al. (2004) UCSF Chimera -- A visualization system for exploratory research and analysis. J. Comput. Chem. 25, 1605–1612. [DOI] [PubMed] [Google Scholar]
  39. Sammon D, Kruger A, Busse-Wicher M, Morgan RM, Haslam SM, Schumann B, et al. , (2023) Molecular mechanism of decision-making in glycosaminoglycan biosynthesis. Nature Commun. 14, 6425. [DOI] [PMC free article] [PubMed] [Google Scholar]
  40. Schultink A, Cheng K, Park YB, Cosgrove DJ, & Pauly M (2013) The identification of two arabinosyltransferases from tomato reveals functional equivalency of xyloglucan side chain substituents. Plant Physiol. 163, 86–94. [DOI] [PMC free article] [PubMed] [Google Scholar]
  41. Teleman A, Lundqvist J, Tjerneld F, Stålbrand H & Dahlman O (2000) Characterization of acetylated 4-O-methylglucuronoxylan isolated from aspen employing 1H and 13C NMR spectroscopy. Carbohydr. Re. 329, 807–815. [DOI] [PubMed] [Google Scholar]
  42. Timell TE (1967) Recent progress in the chemistry of wood hemicelluloses. Wood Sci. Technol. 1, 45–70. [Google Scholar]
  43. Togashi H, Kato A & Shimizu K (2009) Enzymatically derived aldouronic acids from Eucalyptus globulus glucuronoxylan. Carbohydr. Polym. 78, 247–252. [Google Scholar]
  44. Urbanowicz BR, Peña MJ, Moniz HA, Moremen KW and York WS (2014) Two Arabidopsis proteins synthesize acetylated xylan in vitro. Plant J. 80, 197–206. [DOI] [PMC free article] [PubMed] [Google Scholar]
  45. Urbanowicz BR, Peña MJ, Ratnaparkhe S, Avci U, Backe J, Steet HF, et al. (2012) 4-O-methylation of glucuronic acid in Arabidopsis glucuronoxylan is catalyzed by a domain of unknown function family 579 protein. Proc. Natl. Acad. Sci. USA 109, 14253–14258. [DOI] [PMC free article] [PubMed] [Google Scholar]
  46. Wende G & Fry SC (1997) 2-O-β-D-Xylopyranosyl-(5-O-feruloyl)-L-arabinose, a widespread component of grass cell walls. Phytochemistry 44, 1019–1030. [Google Scholar]
  47. Wei QQ, Yang Y, Li H, Liu ZW, Fu R, Feng HQ. & Li C. (2021) The xyloglucan galactosylation modulates the cell wall stability of pollen tube. Planta 254, 133. [DOI] [PubMed] [Google Scholar]
  48. Wilson LFL, Neun S, Yu L, Tryfona T, Stott K, Hollfelder F, et al. (2023) The biosynthesis, degradation, and function of cell wall β-xylosylated xyloglucan mirrors that of arabinoxyloglucan. New Phytol. 240, 2353–2371. [DOI] [PMC free article] [PubMed] [Google Scholar]
  49. Wu AM, Rihouey C, Seveno M, Hörnblad E, Singh SK, Matsunaga T, et al. (2009) The Arabidopsis IRX10 and IRX10-LIKE glycosyltransferases are critical for glucuronoxylan biosynthesis during secondary cell wall formation. Plant J. 57, 718–731. [DOI] [PubMed] [Google Scholar]
  50. Yu L, Yoshimi Y, Cresswell R, Wightman R, Lyczakowski JJ, Wilson LFL, et al. (2022) Eudicot primary cell wall glucomannan is related in synthesis, structure, and function to xyloglucan. Plant Cell 34, 4600–4622. [DOI] [PMC free article] [PubMed] [Google Scholar]
  51. Zeng W, Lampugnani ER, Picard KL, Song L, Wu AM, Farion IM, et al. (2016) Asparagus IRX9, IRX10, and IRX14A are components of an active xylan backbone synthase complex that forms in the Golgi apparatus. Plant Physiol. 171, 93–109. [DOI] [PMC free article] [PubMed] [Google Scholar]
  52. Zhong R, Cui D, Phillips DR & Ye Z-H (2018) A novel rice xylosyltransferase catalyzes the addition of 2-O-xylosyl side chains onto the xylan backbone. Plant Cell Physiol. 59, 554–565. [DOI] [PubMed] [Google Scholar]
  53. Zhong R, Cui D, Phillips DR, Sims NT & Ye Z-H (2021) Functional analysis of GT61 glycosyltransferases from grass species in xylan substitutions. Planta 254, 131. [DOI] [PubMed] [Google Scholar]
  54. Zhong R, Lee C, Cui D, Phillips DR, Adams ER, Jeong HY, et al. (2022) Identification of xylan arabinosyl 2-O-xylosyltransferases catalyzing the addition of 2-O-xylosyl residue onto arabinosyl side chains of xylan in grass species. Plant J. 112, 193–206. [DOI] [PubMed] [Google Scholar]
  55. Zhong R, Pena MJ, Zhou G-K, Nairn CJ, Wood-Jones A, Richardson EA, et al. (2005) Arabidopsis Fragile Fiber8, which encodes a putative glucuronyltransferase, is essential for normal secondary wall synthesis. Plant Cell 17, 3390–3408. [DOI] [PMC free article] [PubMed] [Google Scholar]
  56. Zhong R, Phillips DR, Clark KD, Adams ER, Lee C & Ye Z-H (2024a) Biochemical characterization of rice xylan biosynthetic enzymes in determining xylan chain elongation and substitutions. Plant Cell Physiol. doi: 10.1093/pcp/pcae028. [DOI] [PubMed] [Google Scholar]
  57. Zhong R, Zhou D, Phillips DR, Adams ER, Chen L, Rose JP, et al. (2024b) A rice GT61 glycosyltransferase possesses dual activities mediating 2-O-xylosyl and 2-O-arabinosyl substitutions of xylan. Planta 259, 115. [DOI] [PubMed] [Google Scholar]
  58. Zhu L, Dama M, & Pauly M (2018) Identification of an arabinopyranosyltransferase from Physcomitrella patens involved in the synthesis of the hemicellulose xyloglucan. Plant Direct 2, e00046. [DOI] [PMC free article] [PubMed] [Google Scholar]

Associated Data

This section collects any data citations, data availability statements, or supplementary materials included in this article.

Supplementary Materials

Supinfo

Figure S1. Biological replicates of the reaction products of AtXAPT1.

Figure S2. 1H-NMR spectra of biological replicates of xylooligomers isolated from wild-type Arabidopsis, the xapt1 mutant and xapt1 complemented with the wild-type AtXAPT1 gene as shown in Figure 3a.

Figure S3. Characterization of Arabidopsis xapt1 mutants.

Figure S4. MALDI-TOF mass spectra of biological replicates of the reaction products of XAPTs incubated with the (GlcA)Xyl6-AA acceptor and UDP-Arap or UDP-Gal as shown in Figure 4c–f.

Figure S5. Biological replicates of reaction products catalyzed by XAPTs or sequentially by XAPTs and AtGXM3.

Figure S6. Structural alignment of the predicted model of AtXAPT1 with the structures of GT47 domains of three characterized human exostosins showing similar global folds and structural similarity.

Figure S7. Amino acid sequence alignment of XAPTs studied in this report with the four previously functionally characterized Arabidopsis MUR3-clade members, AtXUT1, AtXLT2, AtMUR3 and AtMBGT1.

Figure S8. Sequence alignment of motifs containing amino acid residues (red font) predicted to H-bond with UDP-Arap/UDP-Gal in EgXAPT1 and AtXAPT1 among the biochemically characterized XAPTs together with four other Arabidopsis MUR3-clade members.

Figure S9. Phylogenetic relationships of XAPT close homologs identified from various lineages of plant species together with the XAPTs studied in this report.

Figure S10. Time course of UDP release by EgXAPT1 incubated with UDP-Arap (a) or UDP-Gal (b) and (GlcA)Xyl6.

Table S1 Survey of various lineages of plant species for the presence of XAPT close homologs

NIHMS2016151-supplement-Supinfo.pdf (7.5MB, pdf)

Data Availability Statement

All data generated during this study are included in this article and its Supplementary files.

RESOURCES