Considerations in the search for mixed-linkage (1→3),(1→4)-β-d-glucan-active endotransglycosylases

Abstract

Xyloglucan endotransglucosylase, catalyzed by XTH subfamily members, is thought to play crucial roles in plant cell wall physiology. Recent discovery of endotransglycosylases active on other hemicelluloses extend our understanding of the physiological scope of endotransglycosylation in general. Discovery in Poaceaen XTHs of endotransglycosylases which act on Poaceaen-prevalent hemicelluloses, such as MLG, could reconcile the apparent incongruence between the large size of Poaceaen putative XTH families and the low xyloglucan content of their cell walls. Here, I speculate on hypothetical MLG-active endotransglycosylases and highlight potential hindrances to their discovery. It is suggested that because the location of β-(1→3) bonds within MLG oligosaccharides (MLGOs) could define their ability to act as endotranglycosylase acceptor substrates: a) thorough probing of substrate specificities necessitates the use of MLGOs created using different endo-glycanases; and b) endogenous plant exo-glycosidases, which can hinder endotranglycosylase assays by degrading acceptor substrates, might prove particularly troublesome where MLGOs are concerned.

Keywords: :

• mixed-linkage glucan
• endotransglycosylase
• XTH
• XET
• Poaceae
• cell wall

The plant cell wall is a hydrated extraprotoplasmic complex which plays crucial roles in plant life.¹ It is composed largely of polysaccharides: cellulose, hemicellulose and pectin. The presence and fine structure of hemicelluloses vary in a phylogenetically-dependent manner.^2-5 For example, within land plants, xyloglucan appears ubiquitous^4,5 while mixed-linkage (1→3),(1→4)-β-d-glucan (MLG) is known only in the Poales^5,6 and Equisetum.^7,8 Xyloglucan is a β-(1→4)-d-glucan backbone with additional glycosyl substitutions.^3,9 MLG is an unbranched d-glucosyl homopolymer in which straight β-(1→4)-linked regions, typically 3–4 residues, are connected by single β-(1→3) bonds which cause kinks in the structure¹⁰ (Fig. 1A).

Figure 1. Endotransglycosylase reactions and the use of MLGOs as acceptor substrates. (A) MLG polymer structure showing kinks caused by β-(1→3) bonds as well as distinct sites of enzymic attack: lichenase cleaves β-(1→4) bonds after β-(1→3) bonds²⁹ (gray downward triangles), cellulase cleaves β-(1→4) bonds before β-(1→3) bonds³⁰ (black upward triangles); (B) General endotransglycosylase reaction. During assays the acceptor substrate is typically a labeled (e.g., radioactive) oligosaccharide. Labeled polymeric material is detected after the reaction; Major products of lichenase (C) and cellulase (D) digestion of MLG as each might fit into XTH positive (acceptor substrate-binding) subsites; (E) 5 successive β-D-glucosidase digestion reactions (each releasing Glc from the non-reducing terminus) on a MLG-hexasaccharide (produce by incomplete lichenase digestion) demonstrating the alternation of position of the β-(1→3) kink relative to XTH positive subsites. Reducing termini on right; blue squares, D-Glc residues; β3, β-(1→3)-linkage (highlighted in orange); β4, β-(1→4)-linkage; cello-like β-(1→4)-linked regions underlined.

Endotransglycosylases catalyze the endo-cleavage of a glycan (the donor substrate) and the creation of a glycosidic link between the newly formed potentially reducing terminus and the non-reducing terminus of another glycan (the acceptor substrate; Fig. 1B). Xyloglucan endotransglucosylase (XET) activity^11,12—which uses xyloglucan both as donor and acceptor—is catalyzed by xyloglucan endotransglucosylase/hydrolase (XTH) subfamily members, appears ubiquitous in the land plants and is believed to play crucial roles in cell wall physiology.¹³ XTH active sites contain seven subsites (−4, −3, −2, −1, +1, +2, +3), each binding a xyloglucan backbone residue and/or its substitutions; donor cleavage occurs between subsites −1 and +1, before the acceptor substrate occupies the positive subsites.¹⁴

The relative recentness of the discovery of other endotransglycosylase activities^15-18 is perhaps surprising given the crucial role thought to be played by XET and the wide diversity of hemicelluloses within the plant kingdom. The apparent incongruity between the low xyloglucan content of Poaceaen cell walls^3,19 and the large number of putative XTH genes they maintain²⁰ (e.g., 30 and 32 in rice and maize, respectively)¹⁴ could be reconciled if some were shown to catalyze significant levels of non-XET endotransglycosylase reaction/s.²¹ As MLG is one of the most prevalent hemicelluloses, as well as a distinctive feature, of the “Type II” cell walls possessed by the Poaceae,^2,3,5,19 an MLG-active endotransglycosylase—such as mixed-linkage (1→3),(1→4)-β-d-glucan: xyloglucan endotransglucosylase (MXE), which uses MLG as a donor and xyloglucan as an acceptor¹⁶—would be a reasonable focus of attention. However, the only reports of MXE and MXE-like activities outside Equisetum were “side reactions” catalyzed by predominantly XET-active enzymes.^22,23 Recently a novel method to assay endotransglycosylase “action” in vivo (compare “activity” in vitro) was devised,²⁴ in which [³H]xyloglucan oligosaccharitols ([³H]XGO-ols) permeated the cell walls of intact living plant tissues, allowing them to be acted on in situ by endogenous enzymes with endogenous donor substrates. Subsequent enzymic digestion of any resulting [³H]polymers identified the donor substrate (e.g., digestion with lichenase—“MLG-endoglucanase”—indicated MXE). This method showed speculation that Poacean cell walls might contain inextractable MXE^16,22 to be untrue.

Despite these evidences against MXE in the Poaceae, there still remains huge scope for other potential MLG-active endotransglycosylases, which could theoretically use any donor: acceptor combination conceivable. The novel in vivo method described above can stand alongside previously devised high-throughput screens,^25,26 all of which can facilitate the testing of this plethora of combinations. However, digestion of candidate acceptor substrates by endogenous plant exo-glycosidases is a constant threat to such assays, all of which use un-purified enzymes. [³H]XGO-ol resistance to β-D-glucosidase makes them particularly amenable to these methods, but candidate acceptors with free non-reducing terminal β-d-glucoses, such as MLG oligosaccharides (MLGOs), are highly prone to it,¹⁷ potentially rendering them too short to participate in endotransglycosylation. Accommodation or tackling of this might prove vital for identification of predominantly MLG-active endotransglycosylases: because all sequenced endotransglycosylases to date^18,27,28 are homo-endotransglycosylases (use same glycan type as donor and acceptor), a hypothetical endotransglycosylase using MLG both as donor and acceptor is perhaps a more likely candidate for an MLG-active endotransglycosylase than MXE. Some other endotransglycosylase activities using MLGOs as acceptors have been observed, though for now only as side-reactions of predominantly XET-active enzymes²³ or minor activities in crude preparations.²⁶

A vital consideration for the design of endotransglycosylase assays using MLGOs as acceptors is that, just as MLG-hydrolysing enzymes typically cleave at specific linkages relative to β-(1→3) bonds (Fig. 1A, lichenase,²⁹ cellulase³⁰), it is highly likely that an endotransglycosylase capable of utilizing MLGOs as acceptors will exhibit a similar preference for the position of the β-(1→3) bond relative to their non-reducing termini (and thus the enzyme’s catalytic residues). Given that the identity of such a preference is currently unknown, use of MLGOs created by digestion with different enzymes (Fig. 1C and D) would be crucial to enable the full probing of potential MLGO-utilizing endotransglycosylases. If the active site of a hypothetical MLGO-utilizing XTH-homolog is shaped to suit a β-(1→3) kink between subsites +1 and +2, it will favor cellulase-produced MLGOs. Probing substrate specificities with lichenase-produced MLGOs alone^23,26 might overlook some MLG-active endotransglycosylases. Furthermore, because of this, endogenous β-D-glucosidases would confound the use of MLGOs 2-fold: as well as shortening them, the resultant constant interchange of β-(1→3) bond position relative to the non-reducing terminus means that, even when still of sufficient length, MLGOs might rarely be suitable substrates (Fig. 1E). The scope for future work is both encouraging and perhaps daunting. But the consideration of these issues will be necessary before the presence and prevalence of speculated MLG-active endotransglycosylases can be fully assessed.

Abbreviations:
MLG	=	mixed-linkage (1→3),(1→4)-β-d-glucan
XET	=	xyloglucan endotransglucosylase
XTH	=	xyloglucan endotransglucosylase/hydrolase
MXE, mixed-linkage (1→3),(1→4)-β-d-glucan	=	xyloglucan endotransglucosylase
XGO-ol	=	xyloglucan oligosaccharitol

Acknowledgments

I thank the BBSRC for a PhD studentship.

Disclosure of Potential Conflicts of Interest

No potential conflicts of interest were disclosed.