Carbohydrate-binding modules

This page is currently under construction. This means that the Responsible Curator has deemed that the page's content is not quite up to CAZypedia's standards for full public consumption. All information should be considered to be under revision and may be subject to major changes.

• Author: ^^^Alicia Lammerts van Bueren^^^ and ^^^Elizabeth Ficko-Blean^^^
• Responsible Curator: ^^^Al Boraston^^^ and ^^^Spencer Williams^^^

---

Overview

Figure 1. An example of modularity in a CBM-containing glycoside hydrolase. Sialidase from Micromonospora viridifaciens contains an N-terminal CBM32 (red) X20 linker (yellow) and a C-terminal catalytic GH33 module (green) [1]. Graphical representation of modularity in amino acid sequence (top) and 3D crystal structure (bottom) PDB ID 1eut.

Carbohydrate-binding modules (CBMs) [2, 3, 4, 5, 6, 7] are generally defined as an amino acid sequence within a larger encoded protein sequence that fold into a structurally discreet module, forming part of a larger multi-modular enzyme (Figure 1). The role of a CBM is to bind to carbohydrate ligand and direct the catalytic machinery onto its substrate, thus enhancing the catalytic efficiency of the multimodular carbohydrate-active enzyme; however, there are several key exceptions of divergent evolution in the functions of CBMs [8] which are discussed below. The individual CBMs are themselves devoid of any catalytic activity and are most commonly associated with Glycoside Hydrolases but have also been identified in Polysaccharide Lyases, polysaccharide oxidases, Glycosyltransferases and plant cell wall-binding expansins [9].

CBMs themselves do not generally undergo any conformational changes when binding ligand. Rather, the topography of the carbohydrate-binding site is preformed to be complementary to the shape of the target ligand (see Types). This is achieved by the presence of amino acid side chains and loops within the CBM binding pocket or cleft. However, multimodular enzymes as a whole may be quite flexible and undergo significant conformational changes when binding substrate. Flexible Ser-Thr-Pro sequences, which are often O-glycosylated, link adjacent modules and can allow for shifts in the orientation and direction of the catalytic module with respect to the CBM on the target substrate [10].

History of CBMs

CBMs were initially characterized as cellulose binding domains (CBDs) in cellobiohydrolases CBHI and CBHII from Trichoderma reesei [11, 12] and cellulases CenA and CexA from Cellulomonas fimi [13]. Limited proteolysis experiments on these enzymes yielded truncated enzyme products that showed a reduced or complete loss in their ability to hydrolyze cellulose substrates. The reduction in enzymatic activity was attributed to the loss of ~100 amino acid C-terminal domains which prevented the adsorbption of the enzymes onto cellulose substrate. Thus it was proposed that these independent "domains" are critical for targeting the enzymes onto its substrate and enhancing their hydrolytic activity. It rapidly became evident that CBDs were not only appended to cellulases but were also found in a range of other plant cell wall degrading enzymes [14, 15, 16].

CBDs were previously categorized into 13 Types based on amino acid sequence similarities [17]. This classification system became complicated when similar functional domains from non-cellulolytic carbohydrate-active enzymes were discovered that did not bind cellulose but met all of the criteria of a CBD (for example see [18]). The term carbohydrate-binding module was proposed to solve this problem to be inclusive of all ancillary modules with non-catalytic carbohydrate-binding function (for a review see [2]). Since this time, CBMs have been found appended to enzymes that interact with almost all characterized carbohydrate sources found on Earth (Table 1).

Table 1: List of Carbohydrates and Interacting CBM Familiesa
Cellulose	CBM1, CBM2, CBM3, CBM4, CBM6, CBM8, CBM9, CBM10, CBM16, CBM17, CBM28, CBM30, CBM37, CBM44, CBM46, CBM49, CBM59, CBM63, CBM64
Xylan	CBM2, CBM4, CBM6, CBM9, CBM13, CBM15, CBM22, CBM31, CBM35, CBM36, CBM37, CBM44, CBM54, CBM59, CBM60
Plant Cell Wall - Other (eg: beta-glucans, porphyrans, pectins, mannans, gluco- and galacturonans)	CBM4, CBM6, CBM11, CBM13, CBM16, CBM22, CBM23, CBM27, CBM28, CBM29, CBM32, CBM35, CBM39, CBM42, CBM43, CBM52, CBM56, CBM59, CBM61, CBM62, CBM65, CBM67
Chitin	CBM1, CBM2, CBM3, CBM5, CBM12, CBM14, CBM18, CBM19, CBM37, CBM50, CBM54, CBM55
Alpha-glucans (starch/glycogen, mutan)	CBM20, CBM21, CBM24, CBM25, CBM26, CBM34, CBM41, CBM45, CBM48, CBM53, CBM58
Mammalian Glycans	CBM32, CBM40, CBM47, CBM51, CBM57b
Other	Bacterial cell wall sugars: CBM35, CBM39, CBM50 Fructans: CBM38, CBM66 Yeast cell wall glucans: CBM54
aBased on the Carbohydrate Active enZyme database. CBM7 is a deleted entry and CBM33 is now reclassified as Auxiliary Activities family AA10. bonly human lectin malectin has been characterized, however a search based on amino acid sequence similarities found that similar modules are appended to many uncharacterized glycoside hydrolases [19].

Classification

Sequence Based Classification

Carbohydrate-binding modules are classified into many tens of families based on amino acid sequence similarities (a continually updated list is available in the Carbohydrate Active enZyme database). These families often cluster modules with similar structural folds and carbohydrate-binding function. However, there are several families that exhibit diversity in the carbohydrate ligands they target (Table 1).

Fold

Figure 2. Classical CBM beta-sandwich fold. C-terminal family CBM27 from Thermotoga maritimamannanase, a Type B CBM (A)(side and front view, PDB ID 1OF4) [20] and C-terminal family CBM6 fromClostridium stercorarium xylanase (B) (PDB ID 1NAE) [21] showing binding sites on the face (A) and edge (B) of the beta sandwich fold respectively.

CBMs fall into one of 7 fold families [2]. The most common fold exhibited by CBMs is the beta-sandwich fold which is comprised of two overlapping beta-sheets consisting of three to six antiparallel beta strands (Figure 2). The ligand binding site is located primarily on the same face of a beta-sheet (Figure 2A), but may also be positioned on the edge of the beta-sheet within the joining loop region (Figure 2B). There are examples of CBMs in the beta-sandwich fold family exhibiting dual binding sites such as CBM6 [22] and dual starch-binding sites in CBM20 [23]. Other fold families include the beta-trefoil fold, cysteine knot, OB fold, the hevein and hevein-like and unique folds [2]. CBMs of the beta-trefoil fold family (CBM13, CBM42) present multivalent sugar-binding sites, as demonstrated for their interaction with xylan and arabinoxylan respectively [24].

Types

Figure 3. CBM Types. (A) Schematic of different CBM Types binding with different regions of a polysaccharide substrate. (B) Type A CBM2b from Pyrococcus furiosis GH18 chitinase(PDB ID 2CRW) [25]. Aromatic side chains of Type A CBMs form the planar binding surface.

CBMs are classified into three main Types defined by the shape and degree of polymerization of their target ligand (Figure 3A). The architecture of the binding site determines what region within a polysaccharide the enzyme will target. A review on CBM plant cell wall recognition [6] has modified the classification of CBM Types to be as follows:

• Type A: bind to crystalline surfaces of cellulose and chitin (example families CBM1, CBM2, CBM3, CBM5, CBM10). Their binding sites are planar and rich in aromatic amino acid residues creating a flat platform to bind to the planar polycrystalline chitin/cellulose surface (Figure 3B). Type A CBMs are unique and differ significantly from Type B or C.
• Type B: bind internal glycan chains (endo-type). Type B are the most abundant form of CBMs reported to date. Type B binding sites appear as extended grooves or clefts comprised of binding subsites to generally accommodate longer sugar chains with four or more monosaccharide units (see Figure 2A for an example). There are some examples in CBM6, CBM36 and CBM60 that contain only two subsites.
• Type C: bind termini of glycans (reducing/non-reducing ends, exo-type). Type C binding sites are short pockets for recognizing short sugar ligands containing one to three monosaccharide units (example families CBM9, CBM13, CBM32, CBM47, CBM66, CBM67). Families containing Type C CBMs are considered 'lectin-like' and may include lectins and CBMs with no appended catalytic modules as members.

A table relating CBM type to CBM family is available in a 2017 review [7]; see also the classic CBM review by Boraston et al. [2].

Properties of CBM Carbohydrate Binding Interactions

Functional Roles of CBMs

CBMs carry out four main functional roles:

• Targeting Effect: CBMs target the enzyme to distinct regions within a larger macromolecular polysaccharide substrate (reducing end, non-reducing end, internal polysaccharide chains), depending on the architecture of its binding site (see Types).

• Proximity Effect: CBMs increase the concentration of enzyme in close proximity to its polysaccharide substrate. This leads to more rapid and efficient substrate degradation.

An excellent example demonstrating targeting and proximity effects of plant cell wall specific CBMs is available [26].

• Disruptive Effect: Some CBMs have been shown to disrupt the surface of tightly packed polysaccharides, such as cellulose fibres and starch granules, causing the substrate to loosen and become more exposed to the catalytic module for more efficient degradation. Disruptive roles have been described for cellulose binding CBM2a [27] and CBM44 [28]. Dual starch-binding domains of family CBM20 from Aspergillus niger glucoamylase have been shown to disrupt the surface of starch [29] while dual-associated CBM41 modules may have a disruptive role in degrading glycogen granules [30]. CBM33 was thought to have a disruptive effect on chitin, however these have now been reclassified as copper-dependent lytic polysaccharide monooxygenases [31] and are found in CAZy Auxiliary Activity Family 10.

• Adhesion: CBMs have been shown to adhere enzymes onto the surface of bacterial cell wall components while exhibiting catalytic activity on an external neighboring carbohydrate substrate. For example, CBM35 modules have been shown to interact with the surface glucuronic acid containing sugars in the cell wall of Amycolatopsis orientalis while the catalytic module is active on external chitosan likely originating from the cell wall of competing soil fungal species [32].

Driving Forces of CBM-Carbohydrate Interactions

There are two key features that drive CBM/carbohydrate interactions. Extensive hydrogen bonding occurs between the hydroxyl groups of carbohydrate ligands and polar amino acid residues within the binding site. Additional water-mediated hydrogen bonding networks between these groups can also be found in the binding site. By far the most important characteristic driving force mediating protein-carbohydrate interactions is the position and orientation of aromatic amino acid residues (Try, Tyr and sometimes Phe) within the binding site. These essential planar residues provide a hydrophobic platform for the planar face of sugar rings, an interaction resembling hydrophobic stacking interactions. Weak intermolecular electrostatic interactions occur between C-H and pi electrons in the planar ring systems and contribute 1.5 - 2.5 kcal/mol energy to the binding reaction [33].

CBMs may also use coordinated metal ions within the binding site to directly interact with their target ligand. For example, families CBM36 [34] and CBM60 [35] exhibit calcium-dependent binding to xylooligosaccharides.

CBM-carbohydrate interactions in general are quite weak (K_a affinities in mM^-1 to uM^-1 range) making the interaction easily reversible. This feature allows for "recycling" of the appended enzyme to bind to a new region on the substrate once catalysis has been completed at a given site. Multivalent effects (more than one saccharide-binding site or multiple CBMs within the polypeptide) may act to increase the overall affinity relative to a single binding site interaction.

CBM Promiscuity

Because of the diversity of carbohydrate structures and motifs found in plant and mammalian glycans, some CBMs have become adapted to recognize more than one type of monosaccharide or glycosidic bond linkage within the binding pocket, a feature called CBM promiscuity. For example a family CBM32 from Clostridium perfringens NagH binds N-acetyl-glucosamine in the primary subsite but can accommodate N-acetyl-galactosamine or mannose in the secondary site [36]. There are several examples of ligand promiscuity within family CBM32. In plant cell wall recognizing CBMs, they are often able to accommodate both cellulose and hemicelluloses. For example, several family CBM6 members interact with cellulose, xylose or laminarin [21, 37]. Family CBM37 exhibit broad binding specificity for xylan, chitin and cellulose (ref). Family CBM41 appended to a GH13 pullulanase can accommodate both alpha-1,4- and alpha-1,6-linked glucose found in amylopectin (from starch/glycogen) [38]. The flexibility in carbohydrate recognition by CBMs contributes to the targeting efficiency of carbohydrate-active enzymes in environments where there is diverse range of polysaccharides present (such as the plant cell wall or mammalian tissues).

CBMs and Multivalency

Multivalency is the collective strength of several interactions with a given ligand. Because CBM-carbohydrate interactions are relatively weak, some carbohydrate-active enzymes, mainly glycoside hydrolases, have developed ways to increase their interaction with substrate via a multivalent effect. Individually, some CBMs may contain multiple binding sites to form a multivalent interaction with their target ligand, although this form of multivalency is quite rare with only a few examples (CBM6, CBM13, CBM20). More commonly, glycoside hydrolases may contain more than one CBM within its modular architecture, either arranged in tandem or at opposing N and C terminal ends of the protein sequence, or both. These CBMs may target the same carbohydrate ligand, different regions within the same ligand, or different ligands within a larger polysaccharide amalgam. A multivalent interaction enhances the overall affinity of an enzyme for its substrate but more importantly, tandem CBMs will cooperatively target the enzyme towards specific regions within a larger polysaccharide substrate based on the orientation and position of binding sites with respect to one another. (Insert some examples here).

Blurred Lines: CBMs, Lectins and Outliers

Several lectins [39, 40] are classified as CBMs in the Carbohydrate Active enZyme database as they share amino acid sequence similarity, exhibit similar folds and display similar carbohydrate binding properties. For example, ricin toxin B chain from Ricinus communis resides in family CBM13, while wheat germ agglutinin (WGA) can be found in family CBM18. The human lectin malectin is classified as family CBM57 and plays a role in N-linked glycan processing of polypeptides in the endoplasmic reticulum [19, 41]. CBMs also share properties with lectins that are not (yet) incorporated in the Carbohydrate Active enZyme database. For example the fucose-specific Anquila anguila lectin AAA is similar to Type C CBMs found in family CBM6 and CBM32 [21]. Lectins which are classified as CBMs are incorporated into a family because they were found to share amino acid sequence identity with a known CBM appended to a carbohydrate-active enzyme.

The biological reaction of agglutination, according to the Merriam-Webster dictionary, is when particles suspended in a liquid collect into clumps, such as that occuring as a serologic response to a specific antibody. The most prominent feature that is genarally considered to separate CBMs from lectins is the involvement of lectins in agglutination of sugar-containing molecules or glycoconjugates. Lectins exploit multivalency, often forming quaternary structures as homodimers, trimers or tetramers with several binding sites which then agglutinate the target glycocongugate [39, 40]. Few studies have been done on the agglutinating effects of CBMs or CBM tandems; however, a CBM26/CBM25 pair from Bacillus halodurans is described as strongly agglutinating on soluble amylopectin (and pullulan), suggesting binding of the individual CBM modules to sites on separate glucan chains [42]. CBMs individually are not known to be directly involved in the formation of quaternary structures and are not known to have agglutinating properties - in common with sugar-recognition modules of all glycan-binding proteins including lectins [8]. Other examples of CBMs participating but not directly implicated in quaternary structure formation are found in cellulosome complexes [43, 44, 45] and in some secreted pathogenic bacterial enzymes complexes [10, 46] where complex formation is mediated through specific cohesin-dockerin module interactions.

Amino acid sequence-based classification of a CBM family may lead to the incorporation of other non-catalytic-associated CBMs within a given family. Some examples of families containing CBMs without appended catalytic modules include those with lectins (such as tachycitin (CBM14), wheat germ agglutinin (CBM18), fucolectin (CBM47), and malectin (CBM57)), and those with periplasmic solute binding proteins (such as within CBM32). Interestingly, the ricin B chain (CBM13), while not on the same polypeptide chain, is covalently linked through a disulfide bond, to the ricin A chain with its N-glycosidase activity. The ricin A chain N-glycosidase cleaves a specific adenine from the pentose ribose in ribosomal RNA. Finally, CBM29 is a family, with only two members, that has no appended catalytic modules; however, the function of these CBMs is to target the catalytic cellulosome machinery to substrate [43].

A brief historical overview of the discovery and characterization of lectins is available [39] as is a review describing the convergent and divergent mechanisms of sugar recognition across the kingdoms of life [8].

Studying CBM-ligand Interactions

A review on laboratory approaches to studying the binding function of carbohydrate-binding modules is available [47]. Typically, molecular biology techniques are used to overproduce a CBM protein in a host strain such as Escherichia coli which is then isolated and purified. Initial screening for carbohydrate binding interactions can be performed using screening techniques such as microarrays [30] or fluorescence microscopy techniques [26, 30, 48]. Several approaches can be taken to verify and quantify CBM-polysaccharide interaction, including affinity gel electrophoresis, UV difference and fluorescence spectroscopy, solid state depletion assay and isothermal titration calorimetry [49].

Overall demonstration of carbohydrate binding function by CBMs is essential to understanding the biological role of these non-catalytic modules.

Biotechnological applications of CBMs

CBMs and their carbohydrate-binding properties are used for many different biological applications. Below is a list of several examples.

• Features of CBMs are currently being exploited to create designer CAZymes with enhanced or modified carbohydrate recognition functions [50, 51, 52, 53].

• Family CBM9 can be used as an affinity tag to purify tagged proteins on a cellulose-based affinity column [54].

• CBMs are used as molecular probes to detect presence of specific carbohydrate motifs in plant [26, 48] and mammalian tissues [38, 42].

• CBMs are used in fibre modification. Engineered CBMs have been shown to increase the strength of cellulose pulp in paper-making processes [55, 56], in crosslinking polysaccharide fibres for biomaterials [57] and cotton fibre modification [58].

• There are several examples of CBMs being used to immobilize whole cells onto carbohydrate surfaces [59, 60, 61].

• CBMs are used to enhance bioprocessing enzymes for industrial uses in pulp processing and biofuel production [28, 62, 63].

• Starch binding CBMs added onto transglucosylating enzyme CGTase from GH13 created a fusion enzyme with more efficient transglucosylating activity with soluble starch, important for industrial biotransformation processes [64].

References

Error fetching PMID 3338453:
Error fetching PMID 3134347:
Error fetching PMID 15214846:
Error fetching PMID 17131061:
Error fetching PMID 16760304:
Error fetching PMID 19908036:
Error fetching PMID 12645054:
Error fetching PMID 22608728:
Error fetching PMID 17187076:
Error fetching PMID 15223327:
Error fetching PMID 16537424:
Error fetching PMID 19193644:
Error fetching PMID 23769966:
Error fetching PMID 20696902:
Error fetching PMID 23213210:
Error fetching PMID 8591030:
Error fetching PMID 21454649:
Error fetching PMID 15010454:
Error fetching PMID 12791255:
Error fetching PMID 12634060:
Error fetching PMID 19218457:
Error fetching PMID 22828270:
Error fetching PMID 18582475:
Error fetching PMID 21298103:
Error fetching PMID 15229195:
Error fetching PMID 23832347:
Error fetching PMID 15242594:
Error fetching PMID 20659893:
Error fetching PMID 10218582:
Error fetching PMID 22492980:
Error fetching PMID 15177165:
Error fetching PMID 16987809:
Error fetching PMID 14738848:
Error fetching PMID 7763519:
Error fetching PMID 23354445:
Error fetching PMID 16391137:
Error fetching PMID 23819686:
Error fetching PMID 23741390:
Error fetching PMID 19414054:
Error fetching PMID 23503312:
Error fetching PMID 8107143:
Error fetching PMID 19422833:
Error fetching PMID 17095014:
Error fetching PMID 15501830:
Error fetching PMID 2125205:
Error fetching PMID 2115772:
Error fetching PMID 8373350:
Error fetching PMID 2481445:
Error fetching PMID 18524852:
Error fetching PMID 23769966:
Error fetching PMID 28547780:
Error fetching PMID 18716000:
Error fetching PMID 11560933:
Error fetching PMID 1490597:
Error fetching PMID 7646033:
Error fetching PMID 25102772:
Error fetching PMID 16230347:

11. Van Tilbeurgh, H., Tomme P., Claeyssens M., Bhikhabhai R., Pettersson G.(1986) Limited proteolysis of the cellobiohydrolase I from Trichoderma reesei. FEBS Lett. 204,223–227. DOI:10.1016/0014-5793(86)80816-X

17. Tomme, P., Warren, R.A., Miller, R.C., Jr., Kilburn, D.G. & Gilkes, N.R. (1995) in Enzymatic Degradation of Insoluble Polysaccharides (Saddler, J.N. & Penner, M., eds.), Cellulose-binding domains: classification and properties. pp. 142-163, American Chemical Society, Washington.

27. Din, N., Gilkes, N.R., Tekant, B., Miller, R.C., Jr., Warren, R.A., and Kilburn, D.G. (1991) Non-Hydrolytic Disruption of Cellulose Fibres by the Binding Domain of a Bacterial Cellulase. Nat. Biotech. 9, 1096 - 1099. DOI:10.1038/nbt1191-1096

31. Vaaje-Kolstad G, Westereng B, Horn SJ, Liu Z, Zhai H, Sørlie M, and Eijsink VG. (2010). An oxidative enzyme boosting the enzymatic conversion of recalcitrant polysaccharides. Science. 2010;330(6001):219-22. DOI:10.1126/science.1192231 |
40. Nathan Sharon and H. Lis. (2007-10) Lectins. Springer Science & Business Media.

55. Levy, I., Paldi, T., Siegel, D., and Shoseyov, O. (2003) Cellulose binding domain from Clostridium cellulovorans as a paper modification reagent. Nordic Pulp Paper Res. J. 18:421-428.

56. Yokota, S., Matuso, K., Kitaoka, T., and Wariishi, H. (2009) Retention and paper strength characteristics of anionic polyacrylamides conjugated with carbohydrate-binding modules. "Carbohydrate-binding anionic PAM". BioResources 4(1):234-244 Article.

58. Zhang, Y., Chen, S., He, M., Wu, J., Chen, J., and Wang, Q. (2011) Effects of Thermobifida fusca Cutinase-carbohydrate-binding Module Fusion Proteins on Cotton Bioscouring. Biotechnology and Bioprocess Engineering. 16,645-653 DOI:10.1007/s12257-011-0036-4