Hydrolytic scission of the β(1→4) glucosidic bond in cellulose, leading to the formation of glucose (Glc) and short cellodextrins, is carried out mainly by cellulases, a group of enzymes comprising cellobiohydrolase (CBH), endo-1,4-β-D-glucanase (EG), and β-glucosidase (BG). Although cellulose is relatively simple in terms of composition (anhydro-Glc units only) and morphology (mainly amorphous and monoclinic I_β or triclinic I_α crystalline), there is a vast natural diversity of cellulases with catalytic modules belonging to ~14 GH families to accommodate four major reactions modes and different synergisms (for recently studied examples, see [33,34,35]).

In plant cell walls, cellulose is entangled with and shielded by hemicellulose, a group of complex polysaccharides made by different glyco-units and glycosidic bonds. Degradation of hemicellulose, which not only “liberates” cellulose for cellulases but also converts hemicellulose into valuable saccharides, is carried out mainly by an array of interdependent and synergistic hemicellulases.

Common hemicelluloses include β-glucan, xylan, xyloglucan, arabinoxylan, mannan, galactomannan, arabinan, galactan, polygalacturonan, etc., which are targets of β-glucanase, xylanase, xyloglucanase, mannanase, arabinase, galactanase, polygalacturonase, glucuronidase, acetyl xylan esterase, and other enzymes [22,52]. Among hemicellulases, glycoside hydrolases (belonging to about 29 GH families) hydrolyze glycosidic bonds, carbohydrate esterases (belonging to about 9 CE families) hydrolyze ester bonds, polysaccharide lyases (belonging to about 5 PL families) cleave glycosidic bonds. endo-Hemicellulases cleave internal/backbone glycosidic bonds, whereas other glycosidases remove mainly the chain’s substituents or side chains. Cellulolytic microbes produce many hemicellulases along with cellulases for effective lignocellulose degradation (for recently studied examples see [36,37,38]).

Different plants have different hemicelluloses: acetylated (galacto)glucomannan (as well as arabinoglucuronoxylan), glucuronoxylan, and arabinoxylan are major hemicellulose in softwood, hardwood, and grass, respectively [52]. Hence different hemicellulase combinations are needed for different biomass feedstocks in industrial biomass conversion. Synergism of hemicellulases is found both amongst hemicellulases themselves and between hemicellulases and cellulases [2,35,53,54,55,56].

The secretomes of most cellulolytic microbes (particularly white and brown rots) contain oxidoreductases, in some cases at quite significant levels, whose co-presence with hydrolytic or lytic enzymes indicates the importance of having an oxidoreductive system as part of effective biological lignocellulose degradation [5,32,37,76,77]. The main task of these oxidoreductases is likely aimed at degradation of lignin, a highly heterogeneous and recalcitrant aromatic polymer (consisted of various syringyl, guaiacyl, or other hydroxyphenyl units) entangled with hemicellulose or cellulose and inactive/inhibitory to (hemi)cellulases. Lignin degradation is imperative for industrial enzymatic biomass-conversion, because it not only increases (hemi)cellulose accessibility for (hemi)cellulase but also diminishes (hemi)cellulase inactivation caused by lignin adsorption.

Lignin peroxidase (EC 1.11.1.14), Mn peroxidase (EC 1.11.1.13), and versatile peroxidase (EC 1.11.1.16) are extracellular fungal heme peroxidases (belonging to LO2 family) with high potency to oxidatively degrade lignin. Upon interaction with H₂O₂, these enzymes form highly reactive Fe(V) or Fe(IV)-oxo species, which abstract electrons from lignin (to cause oxidation or radicalization) either directly or via Mn(III) species. Laccase (EC 1.10.3.2) is a multi-copper oxidase (belonging to LO1 family) secreted by numerous lignocellulolytic fungi. This enzyme can directly oxidize phenolic parts of lignin, or indirectly oxidize non-phenolic lignin parts with the aid of suitable redox-active mediator.

Aryl-alcohol oxidase (EC 1.1.3.7), glyoxal oxidase (EC 1.1.3.-), and various carbohydrate oxidases (EC 1.1.3.4, 9, 10) are also involved in natural lignocellulose degradation. These enzymes, belonging to LDA1–6 families, can generate H₂O₂ from O₂, with concomitant oxidation of aromatic alcohol, glyoxal, or reducing carbohydrates. The generated H₂O₂ may support lignin-degrading peroxidases or power Fenten-type chemistry that degrades lignin non-enzymatically.

Cellobiose dehydrogenase (EC 1.1.99.18) is produced by many lignocellulotic fungi. The flavoheme enzyme (belonging to LO3 family) may dehydrogenate or oxidize cellobiose or other cellodextrins to corresponding aldonolactones, with concomitant quinone reduction to phenol or O₂ reduction to H₂O₂. Until recently, the role of CDH in enzymatic degradation of lignocellulose was largely considered only from a Fenton chemistry perspective [78] or mitigation of cellulases’ product inhibition by cellobiose [79]; however, the recent discovery of the CDH stimulation of GH61 enzyme activity may shed new light upon its function [80,81,82] (see Section 3.1).

The lignin-degrading oxidative species generated by these oxidoreductases may attack inhibitors of industrial sugar-to-fuel/chemical microbial conversion, making the process more effective [83]. However, they may also attack (hemi)cellulose or even (hemi)cellulases, potentially hampering industrial enzymatic (hemi)cellulose conversion [84,85]. For the peroxidases, autooxidation or inactivation may severely limit their performance. Applying such enzymes to industrial enzymatic biomass conversion needs further research for overall benefit.

Many cellulolytic fungi encode and express proteins which are classified as members of the Glycoside Hydrolase 61 (GH61) family [36,37,38], but due to the unusual characteristics of the GH61s, it took more than a decade after the initial discovery of this group of enzymes to begin to understand their function. The first of these proteins was found during a cDNA screen for the Trichoderma reesei complement to the Saccharomyces cerevisiae gene SEC1, which is involved in secretion. Unexpectedly, several of the isolated T. reesei cDNA clones, which were shown to suppress the temperature sensitive sec1-1 mutant of S. cerevisiae, contained a consensus sequence for a CBM. Based on this CBM homology, the protein encoded by this cDNA was expressed and tested for cellulase activity. The prepared protein sample was observed to have weak activity in the breakdown of β-glucan, carboxymethyl cellulose (CMC), and phosphoric acid-swollen cellulose (PASC), and was classified as T. reesei EG-IV [86]. A further characterization of this protein reported that this enzyme could cause the release of small amounts of soluble cellooligosaccharides from cellulosic substrates [87]. Similar results, showing weak if any EG activity, were reported for a homologous GH61 encoded by Aspergillus kawachii [88].

Suspicion that GH61 was not an authentic EG got strong support from GH61 crystal structures. EGs usually have a cleft in which the cellulose strand binds to the enzyme, typically through interactions with aromatic residues, and a hydrolase active site near this cleft containing a catalytic glutamic or aspartic acid. The structure of T. reesei GH61B [89] contains neither a cleft nor an obvious hydrolase active site, but rather displays a surface-exposed divalent metal binding site surrounded by hydrophobic residues, suggesting a possible carbohydrate binding surface and active site of unknown function. Structurally, T. reesei GH61B is homologous to chitin binding protein 21 (CBP21, belonging to CBM33 family) from the bacterium Serratia marcescens, a molecule which at the time was thought to be a non-catalytic protein which enhanced chitin breakdown by chitin hydrolases [90]. Another publication reported the structure of Thielevia terrestris GH61E, which also includes a surface exposed metal binding site; this report also included significant biochemical analysis of T. terrestris GH61E, T. terrestris GH61B, and Thermoascus aurantiocus GH61A activity upon both model and industrial cellulosic substrates [91]. In this analysis, the GH61 molecules tested displayed only negligible ability to cleave any of the cellulosic or hemicellulosic substrates tested. However, addition of GH61s to Trichoderma cellulase cocktails was shown to greatly enhance the activity of these cellulases in lignocellulose degradation, lowering the required enzyme concentration for substrate breakdown by a factor of two. Confirming the role of metal in GH61 activity, the GH61 cellulase-boosting effect was inhibited by addition of EDTA, and metal binding site mutants of GH61 displayed greatly reduced or absent cellulase boosting activity. Surprisingly, the addition of GH61 has no stimulatory effect on Trichoderma cellulases’ breakdown of “pure” cellulose substrates (e.g., microcrystalline cellulose) suggesting that a non-cellulose component might be a requirement for or the target of GH61 activity [91]. While the industrial significance of GH61 as a critical component in cellulase mixtures was established, the biochemical activity of these molecules remained mysterious.

Recently, the functions of members of the fungal GH61 family, and members of the related bacterial CBM33 family, were revealed. Several of these enzymes have been shown to act by cleaving cellulose [80,81,82,92,93,94,95,82,92] or the structurally related polysaccacharide chitin [96]; and while some differences in data and analysis exist amongst these publications it is clear that GH61 and CBM33 enzymes do not function as typical glycoside hydrolases.

Unlike canonical glycoside hydrolases, GH61 and CBM33 enzymes require a redox-active factor for activity. This cofactor requirement explains the formerly puzzling results which show little or no GH61 activity upon model cellulose substrates, but a substantial increase in complex biomass degradation by Trichoderma cellulases or mixtures upon GH61 supplementation [91]; when the soluble phase of dilute acid pretreated biomass was added to model cellulose substrates, GH61 was shown to act upon those substrates, indicating that requisite GH61 cofactor was present in soluble (and perhaps also in insoluble) fractions of pretreated biomass [94]. GH61 and CBM33 polysaccharide cleaving-activity has been demonstrated upon addition of a small molecule reducing-agent such as ascorbate, glutathione, or gallate [92,93,94,95,96] or the fungal enzyme cellobiose dehydrogenase [80,81,82]. Analysis of the polysaccharide cleavage products formed by GH61 and CBM33 molecules reveals that these enzymes release oxidized cellooligosaccharides, though significant quantities of non-oxidized oligosaccharides are also detected in some studies [93,94,95]. Unlike other glycoside hydrolases, which are more active on cellooligosaccharides than crystalline cellulose, GH61 and CBM33 enzymes appear inactive upon cellooligosaccahrides [81,93,94,95,96]. The position of the oxidation on the oligosaccharide products has been reported on the reducing end [81,82,93,95], non-reducing end [81,82], or both [94], which could suggest differences amongst these enzymes. GH61 proteins also require a metal for this cellulose cleaving activity, specifically copper which binds tightly to the protein in a type-2 copper site geometry [82,94,96]. Interestingly, despite the large potential of this enzyme class to promote lignocellulose breakdown, the precise cleavage mechanism of GH61 and CBM33 enzymes is unclear, and is the subject of ongoing study by multiple groups.

Expansins are a class of plant proteins which interact with and modify cell walls and/or cell wall components by an unknown activity, thought to result in expansion, slippage, or lengthening of cell wall structures. These two-domain proteins consist of a domain homologous to the GH45 EG catalytic core and a second domain homologous to Group II grass pollen allergens; both domains have no known catalytic function and display no detectable hydrolytic activity on lignocellulosic or model substrates (expansins are thoroughly reviewed by [97]). T. reesei has been shown to express a protein, named swollenin, which has sequence homology to plant expansins and displays a similar mysterious disruptive effect on cellulosic substrates. Isolated T. reesei swollenin has been shown, without cause to formation of detectable reducing ends, to weaken filter paper and to affect superstructural changes in cotton fibrils by light and atomic force microscopy [98].

Despite the apparent lack of direct lytic activity on cellulose, the addition of T. reesei swollenin significantly increases the breakdown of filter paper by cellulases [99,100]. Swollenin from Aspergillus fumigatus is reported to have weak lytic activity on CMC but no apparent hydrolytic activity on microcrystalline cellulose, though treatment of microcrystalline cellulose with A. fumigatus swollenin reduced apparent microcrystalline cellulose particle size and potentiated breakdown of the cellulose by hydrolytic cellulases [101]. The basidiomycete Bjerkandera adusta produces a similar protein, loosinin, which increases cellulase activity on cotton and agave bagasse [102]. Bacterial species, including Bacillus subtilis [103] and Hahella chejuensis [104], also produce expansin-like molecules. Like fungal swollenins, bacterial expansins alter cellulose fiber structure and promote the breakdown of cellulose by hydrolases without showing detectable direct hydrolase activity [104,105]. Recent structural and mutational studies of expansin EXLX1 from B. substilis have shed some light on these proteins, demonstrating that several clustered residues on GH45-like domain are required for EXLX1’s cell wall modifying activity, and that the second domain is likely a new type of CBM [103,106]. Further study will be required in order to determine the mechanism by which these molecules increase lignocellulose conversion and allow industrial exploitation of this class of protein.

CIP1 and CIP2 (cellulose induced protein-1 and -2, respectively) were first found in a transcriptional analysis of T. reesei. Both contain a CBM and are co-regulated with known cellulases [107]. The function of CIP1 is unknown, though it is claimed that CIP1 from T. reesei has weak activity on p-nitrophenyl β-D-cellobioside [108] and some synergistic activity with both GH61 and swollenin [109]. CIP2, found in both T. reesei and Schizophyllum commune, has recently been shown to be an esterase that cleaves the methyl ester of 4-O-methyl-D-glucuronic acid [110]. This enzyme, now classified as the first member of CE15 family (EC 3.1.1.-), likely acts in the cleavage of hemicellulose-lignin crosslinks. Further investigation of both the functions and the potential of these enzymes in industrial applications are needed.

While many microorganisms secrete biomass degrading enzymes into their environment, other microbes, particularly anaerobic biomass degrading microbes, use cell-surface linked enzymes to break down lignocellulosic materials. Cellulosomes are arrays of multiple cellulase and hemicellulase proteins, assembled by specific interactions between dockerin domains on the enzyme and cohesins bound to structural scaffoldins on the microbial surface (reviewed extensively in [28,29,30,111]). This spatial clustering of multiple lignocellulose degrading enzymes results in an increased synergy between lytic activities [28,30,112]. It has been shown that recombinant cellulosomes can be transplanted to other industrially useful organisms, such as S. cerevisiae [113,114] and B. subtilis [115]. The ability of cellulosomes to cluster activities may present unique capabilities, both in synergistic breakdown of a substrate and in targeted degradation of specific biomass components.