Structural Organization and a Standardized Nomenclature for Plant Endo-1,4-β-Glucanases (Cellulases) of Glycosyl Hydrolase Family 9

Glycosyl, or glycoside, hydrolases (GHs) comprise a structurally diverse group of enzymes that hydrolyze glycosidic bonds between carbohydrates, or between carbohydrates and other noncarbohydrate moieties, and that collectively exhibit a wide range of substrate specificities. GH enzymes from across the taxonomic spectrum were originally named based on substrate specificity, the corresponding International Union of Biochemistry and Molecular Biology (IUBMB) nomenclature system (EC 3.2.1.-), and the chronological order in which they were reported. However, as growing numbers of GH proteins, and later genes, were characterized, this strategy proved to be increasingly unsatisfactory and a complementary nomenclature was developed based on predicted protein sequence (Henrissat et al., 1998). This approach provides important insights into protein structure, evolutionary relationships, and an opportunity to infer mechanistic relationships. A regularly updated database, Carbohydrate-Active Enzymes (CAZY; www.cazy.org), currently lists 108 distinct GH families, a subset of which are further affiliated with 14 clans based on the presence of defined protein folds and conserved catalytic machineries.

This nomenclature initiative was developed largely in response to the rapidly growing numbers of reported microbial GHs, reflecting their numerous important industrial uses. Notable examples are endo-β-1,4-glucanases, or cellulases, which hydrolyze the β-1,4-glucosyl linkages of cellulose and several other plant cell wall polysaccharides, and are used in the generation of sugars from lignocellulosic biomass for ethanol production. Plant biologists are currently facing a similar nomenclature dilemma, since the sequencing of whole plant genomes has revealed many large GH families (Henrissat et al., 2001): a genome-scale assessment of Arabidopsis (Arabidopsis thaliana) in CAZY identified 393 GHs from 34 families (http://www.cazy.org/geno/3702.html), and equivalent families are emerging in many species as larger EST collections develop. GH activities from plants have long been studied in association with various aspects of growth, development, and cell wall metabolism, but plant GH enzymes have often been named with little consideration of their substrate specificity, molecular structure, or enzymatic reaction mechanism.

The current explosion of interest and new research opportunities in biofuels and bioenergy crops will inevitably result in renewed interest in identifying and annotating plant GHs, through their association with lignocellulosic biomass. Thus, this is an opportune time to adopt a new, rational, well-defined nomenclature. Importantly, it is apparent that many of these plant enzymes share similarities to previously classified families of microbial GHs, which is not clear from their original designations. While the IUBMB-recommended nomenclature (www.chem.qmul.ac.uk/iubmb), based on the substrates used and the reaction catalyzed, continues to provide critical information regarding enzyme function, we propose a complementary, standardized nomenclature for plant GHs, based on the rigorous scheme for naming their microbial counterparts, which is categorized by the catalytic domain of the enzyme (Henrissat et al., 1998).

We have selected plant endo-β-1,4-glucanases as a case study, since this family illustrates the problems with imprecise GH nomenclatures that have evolved over decades of research. In addition, this family of plant enzymes has a divergent subfamily structure that can usefully be incorporated into the development of a naming scheme. Early reports described the existence of plant cellulases (e.g. Hall, 1963) and cellulolytic activities have long been associated with both cell wall construction during cell expansion and the wall disassembly that accompanies processes such as fruit ripening and abscission (for review, see del Campillo, 1999; Rose and Bennett, 1999; Mølhøj et al., 2002) and cellulose biosynthesis (Nicol et al., 1998; Lane et al., 2001; Sato et al., 2001). Most of the plant “cellulases” studied to date are typical endoglucanases (EC 3.2.1.4) with low or no activity on crystalline cellulose, but clearly measurable activity on soluble cellulose derivatives, such as carboxymethyl cellulose, noncrystalline phosphoric acid swollen cellulose, and/or a variety of plant polysaccharide substrates, including xylans, 1,3-1,4-β-glucans, and glucomannans (Master et al., 2004; Yoshida and Komae, 2006; Urbanowicz et al., 2007). This is quite distinct from the case of microbial cellulases, whose modular structure and synergistic actions with a number of different enzymes allow effective degradation of crystalline cellulose. The sequencing of the first plant “cellulases”/endo-β-1,4-glucanases revealed that they belong to the GH9 family (Henrissat, 1991). This is one of the larger plant GH families (Henrissat et al., 2001), and studies of microbial GH9 proteins, including both endoglucanases (EC 3.2.1.4) and cellobiohydrolases (EC 3.2.1.91), have shown that they operate via an inverting mechanism to cleave the 1,4-β-glucosidic bond between two unsubstituted Glc units (Gebler, 1992). Analyses of complete plant genome sequences have now shown that this large multigene family can be subdivided into three distinct structural subclasses (Fig. 1), comprising members that are membrane anchored (with or without a cytosolic domain), secreted, or those that are secreted and have a carbohydrate binding module, CBM49 (Mølhøj et al., 2002; Libertini et al., 2004; Urbanowicz et al., 2007).

Figure 1.

Schematic representation of the modular plant GH9 family structure. Specific domains comprise the cytosolic domain (white), transmembrane domain (wavy lines), signal sequence (dark grey), GH9 catalytic domain (light grey), linker region (thick black line), and carbohydrate binding module (dots). Structural subclasses are represented by SlGHl9A1/TomCel3 (class A, U78526), SlGH9B1/TomCel1 (class B, U13054), and SlGH9C1/TomCel8 (class C, AF098292).

To standardize the nomenclature for these gene families, we suggest the following, using genes encoding GH9 enzymes as an example: an indication of the genus and species, followed by the designated the GH family (GH9). The letters A to C adjacent to the family number correspond to the domain structure, or subclass, of the corresponding protein (Fig. 1), which will provide additional information about potential function. We note that this nomenclature helps determine the structural subclass with which a particular GH9 gene is associated, rather than specifically suggesting an orthologous sequence. As an example of the naming scheme, we have applied the guidelines to rename the members of the GH9 family from tomato (Solanum lycopersicum), the plant species from which the greatest number of family members has been studied in detail. Historically, members of the tomato GH9 family have been referred to as TomCel1-8 and their new designations are shown in Table I. This nomenclature provides important information since, using SlGH9C1 gene as an example, the name indicates that this encodes the first described tomato (Sl, for S. lycopersicum) member of GH family 9 (GH9) with a class C domain structure, indicating that the protein contains a CBM49. This naming scheme has also been applied to the 25 member GH9 family in Arabidopsis (Table II), where a sequenced genome and ease of genetic manipulation is yielding information about the biological roles of many members of this family. Once the biochemical activity and substrate specificity of a particular protein are determined, the enzyme (as opposed to gene) can be given an appropriate EC number and a trivial name, followed by the GH family number, such as “Cel9” for a 1,4-β-endo-glucanase (EC 3.2.1.4) or “Xyn9” for an endo-1,4-β-xylanase (EC 3.2.1.8), based on the dominating activity of that protein. This approach has the advantage of allowing enzymologists to provide accurate and useful information when describing a protein that has been characterized in detail, but still enables facile gene annotation by nonexperts.

Table I.

GH9 proteins of tomato

Current Designation	GenBank ID	New Designation	Featuresa
TomCel1	U13054	SlGH9B1	SP, GH9
TomCel2	U13055	SlGH9B2	SP, GH9
TomCel3	U78526	SlGH9A1	CT, TM, GH9
TomCel4	U20590	SlGH9B3	SP, GH9
TomCel5	AF077339	SlGH9B4	SP, GH9
TomCel6	AAB46829b	SlGH9B5	SP, GH9
TomCel7	Y11268	SlGH9B6	SP, GH9
TomCel8	AF098292	SlGH9C1	SP, GH9, CBM49

^aFeatures noted are cytosolic domain (CT), transmembrane domain (TM), signal peptide (SP), glycosyl hydrolase family 9 catalytic domain (GH9), and a family 49 carbohydrate binding module (CBM49).

^bPeptide fragment.

Table II.

GH9 proteins of Arabidopsis

Chromosome Locus	GenBank ID	New Designation	Featuresa	Synonyms
At5g49720	U37702	AtGH9A1	CT, TM, GH9	Endo-1,4-β-glucanase 25, KORRIGANb, KOR1b
At1g65610	AC001229	AtGH9A2	CT, TM, GH9	Endo-1,4-β-glucanase 7, KOR2c
At4g24260	AL078637	AtGH9A3	CT, TM, GH9	Endo-1,4-β-glucanase 21, KOR3c
At3g43860	AY072099	AtGH9A4	TM, GH9	Endo-1,4-β-glucanase 16
At1g70710	AY048283	AtGH9B1	SP, GH9	Endo-1,4-β-glucanase 8, cellulase 1d, AtCEL1d
At1g02800	AF034573	AtGH9B2	SP, GH9	Endo-1,4-β-glucanase 1, cellulase 2e, AtCEL2e
At1g71380	U17888.1	AtGH9B3	SP, GH9	Endo-1,4-β-glucanase 9, cellulase 3f, AtCEL3f
At1g22880	AF000657	AtGH9B4	SP, GH9	Endo-1,4-β-glucanase 3, cellulase 5f, AtCEL5f
At1g19940	AC007797	AtGH9B5	SP, GH9	Endo-1,4-β-glucanase 2
At1g23210	AC002311	AtGH9B6	SP, GH9	Endo-1,4-β-glucanase 4
At1g75680	AC006434	AtGH9B7	SP, GH9	Endo-1,4-β-glucanase 10
At2g32990	AC003033	AtGH9B8	SP, GH9	Endo-1,4-β-glucanase 11
At2g44540	AC004521	AtGH9B9	SP, GH9	Endo-1,4-β-glucanase 12
At2g44550	AC004521	AtGH9B10	SP, GH9	Endo-1,4-β-glucanase 13
At2g44560	AC003672	AtGH9B11	SP, GH9	Endo-1,4-β-glucanase 14
At2g44570	AC003672	AtGH9B12	SP, GH9	Endo-1,4-β-glucanase 15
AT4g02290	AY079162	AtGH9B13	SP, GH9	Endo-1,4-β-glucanase 17
At4g09740	AL161515	AtGH9B14	SP, GH9	Endo-1,4-β-glucanase 18
At4g23560	AL035394	AtGH9B15	SP, GH9	Endo-1,4-β-glucanase 20
At4g38990	AL035679	AtGH9B16	SP, GH9	Endo-1,4-β-glucanase 22
At4g39000	AL035679, AK117850	AtGH9B17	SP, GH9	Endo-1,4-β-glucanase 23
At4g39010	AY059825	AtGH9B18	SP, GH9	Endo-1,4-β-glucanase 24
At1g48930	AC016041	AtGH9C1	SP, GH9, CBM49	Endo-1,4-β-glucanase 5
At1g64390	AC066689	AtGH9C2	SP, GH9, CBM49	Endo-1,4-β-glucanase 6
At4g11050	AF080120	AtGH9C3	SP, GH9, CBM49	Endo-1,4-β-glucanase 19

^aFeatures noted are cytosolic domain (CT), transmembrane domain (TM), signal peptide (SP), glycosyl hydrolase family 9 catalytic domain (GH9), and a family 49 carbohydrate binding module (CBM49).

^bNicol et al. (1998).

^cMølhøj et al. (2001).

^dShani et al. (1997).

^eYung et al. (1999).

^fdel Campillo et al. (2004).

This nomenclature conforms to that used for bacterial endoglucanases with some slight modifications, since many of the families of plant GHs are much larger than those of bacteria (CAZY) and so the information designating the order in which they are reported is more easily presented by a numerical rather than an alphabetical system. For example, in Arabidopsis there are 49 members of family GH1, 33 of GH16, and 69 of GH28, all of which contain more members than there are letters in the alphabet (http://www.cazy.org/geno/3702.html).

The new nomenclature has been presented in the context of the plant GH9 family, but we further suggest that similar guidelines be adopted for naming members of other GH families and that future efforts to standardize nomenclature be coordinated in consultation with CAZY. Similarly, the use of the established naming schemes for carbohydrate binding modules (www.cazy.org; Boraston et al., 2004) will further help advance the study of analogous hydrolytic enzymes and other associated functional domains from different organisms, both within and between taxa.