Machine learning reveals sequence-function relationships in family 7 glycoside hydrolases.

Family 7 glycoside hydrolases (GH7) are among the principal enzymes for cellulose degradation in nature and industrially. These enzymes are often bimodular, including a catalytic domain and carbohydrate-binding module (CBM) attached via a flexible linker, and exhibit an active site that binds cello-oligomers of up to ten glucosyl moieties. GH7 cellulases consist of two major subtypes: cellobiohydrolases (CBH) and endoglucanases (EG). Despite the critical importance of GH7 enzymes, there remain gaps in our understanding of how GH7 sequence and structure relate to function. Here, we employed machine learning to gain data-driven insights into relationships between sequence, structure, and function across the GH7 family. Machine-learning models, trained only on the number of residues in the active-site loops as features, were able to discriminate GH7 CBHs and EGs with up to 99% accuracy, demonstrating that the lengths of loops A4, B2, B3, and B4 strongly correlate with functional subtype across the GH7 family. Classification rules were derived such that specific residues at 42 different sequence positions each predicted the functional subtype with accuracies surpassing 87%. A random forest model trained on residues at 19 positions in the catalytic domain predicted the presence of a CBM with 89.5% accuracy. Our machine learning results recapitulate, as top-performing features, a substantial number of the sequence positions determined by previous experimental studies to play vital roles in GH7 activity. We surmise that the yet-to-be-explored sequence positions among the top-performing features also contribute to GH7 functional variation and may be exploited to understand and manipulate function.

The dissociation mechanism of processive cellulases.

Cellulase enzymes deconstruct recalcitrant cellulose into soluble sugars, making them a biocatalyst of biotechnological interest for use in the nascent lignocellulosic bioeconomy. Cellobiohydrolases (CBHs) are cellulases capable of liberating many sugar molecules in a processive manner without dissociating from the substrate. Within the complete processive cycle of CBHs, dissociation from the cellulose substrate is rate limiting, but the molecular mechanism of this step is unknown. Here, we present a direct comparison of potential molecular mechanisms for dissociation via Hamiltonian replica exchange molecular dynamics of the model fungal CBH, Trichoderma reesei Cel7A. Computational rate estimates indicate that stepwise cellulose dethreading from the binding tunnel is 4 orders of magnitude faster than a clamshell mechanism, in which the substrate-enclosing loops open and release the substrate without reversing. We also present the crystal structure of a disulfide variant that covalently links substrate-enclosing loops on either side of the substrate-binding tunnel, which constitutes a CBH that can only dissociate via stepwise dethreading. Biochemical measurements indicate that this variant has a dissociation rate constant essentially equivalent to the wild type, implying that dethreading is likely the predominant mechanism for dissociation.

Biochemical and Structural Characterizations of Two Dictyostelium Cellobiohydrolases from the Amoebozoa Kingdom Reveal a High Level of Conservation between Distant Phylogenetic Trees of Life.

UNLABELLED: Glycoside hydrolase family 7 (GH7) cellobiohydrolases (CBHs) are enzymes commonly employed in plant cell wall degradation across eukaryotic kingdoms of life, as they provide significant hydrolytic potential in cellulose turnover. To date, many fungal GH7 CBHs have been examined, yet many questions regarding structure-activity relationships in these important natural and commercial enzymes remain. Here, we present the crystal structures and a biochemical analysis of two GH7 CBHs from social amoeba: Dictyostelium discoideum Cel7A (DdiCel7A) and Dictyostelium purpureum Cel7A (DpuCel7A). DdiCel7A and DpuCel7A natively consist of a catalytic domain and do not exhibit a carbohydrate-binding module (CBM). The structures of DdiCel7A and DpuCel7A, resolved to 2.1 Å and 2.7 Å, respectively, are homologous to those of other GH7 CBHs with an enclosed active-site tunnel. Two primary differences between the Dictyostelium CBHs and the archetypal model GH7 CBH, Trichoderma reesei Cel7A (TreCel7A), occur near the hydrolytic active site and the product-binding sites. To compare the activities of these enzymes with the activity of TreCel7A, the family 1 TreCel7A CBM and linker were added to the C terminus of each of the Dictyostelium enzymes, creating DdiCel7ACBM and DpuCel7ACBM, which were recombinantly expressed in T. reesei DdiCel7ACBM and DpuCel7ACBM hydrolyzed Avicel, pretreated corn stover, and phosphoric acid-swollen cellulose as efficiently as TreCel7A when hydrolysis was compared at their temperature optima. The Ki of cellobiose was significantly higher for DdiCel7ACBM and DpuCel7ACBM than for TreCel7A: 205, 130, and 29 µM, respectively. Taken together, the present study highlights the remarkable degree of conservation of the activity of these key natural and industrial enzymes across quite distant phylogenetic trees of life. IMPORTANCE: GH7 CBHs are among the most important cellulolytic enzymes both in nature and for emerging industrial applications for cellulose breakdown. Understanding the diversity of these key industrial enzymes is critical to engineering them for higher levels of activity and greater stability. The present work demonstrates that two GH7 CBHs from social amoeba are surprisingly quite similar in structure and activity to the canonical GH7 CBH from the model biomass-degrading fungus T. reesei when tested under equivalent conditions (with added CBM-linker domains) on an industrially relevant substrate.

Glycosylated linkers in multimodular lignocellulose-degrading enzymes dynamically bind to cellulose.

Plant cell-wall polysaccharides represent a vast source of food in nature. To depolymerize polysaccharides to soluble sugars, many organisms use multifunctional enzyme mixtures consisting of glycoside hydrolases, lytic polysaccharide mono-oxygenases, polysaccharide lyases, and carbohydrate esterases, as well as accessory, redox-active enzymes for lignin depolymerization. Many of these enzymes that degrade lignocellulose are multimodular with carbohydrate-binding modules (CBMs) and catalytic domains connected by flexible, glycosylated linkers. These linkers have long been thought to simply serve as a tether between structured domains or to act in an inchworm-like fashion during catalytic action. To examine linker function, we performed molecular dynamics (MD) simulations of the Trichoderma reesei Family 6 and Family 7 cellobiohydrolases (TrCel6A and TrCel7A, respectively) bound to cellulose. During these simulations, the glycosylated linkers bind directly to cellulose, suggesting a previously unknown role in enzyme action. The prediction from the MD simulations was examined experimentally by measuring the binding affinity of the Cel7A CBM and the natively glycosylated Cel7A CBM-linker. On crystalline cellulose, the glycosylated linker enhances the binding affinity over the CBM alone by an order of magnitude. The MD simulations before and after binding of the linker also suggest that the bound linker may affect enzyme action due to significant damping in the enzyme fluctuations. Together, these results suggest that glycosylated linkers in carbohydrate-active enzymes, which are intrinsically disordered proteins in solution, aid in dynamic binding during the enzymatic deconstruction of plant cell walls.

Structural, biochemical, and computational characterization of the glycoside hydrolase family 7 cellobiohydrolase of the tree-killing fungus Heterobasidion irregulare.

Root rot fungi of the Heterobasidion annosum complex are the most damaging pathogens in temperate forests, and the recently sequenced Heterobasidion irregulare genome revealed over 280 carbohydrate-active enzymes. Here, H. irregulare was grown on biomass, and the most abundant protein in the culture filtrate was identified as the only family 7 glycoside hydrolase in the genome, which consists of a single catalytic domain, lacking a linker and carbohydrate-binding module. The enzyme, HirCel7A, was characterized biochemically to determine the optimal conditions for activity. HirCel7A was crystallized and the structure, refined at 1.7 Å resolution, confirms that HirCel7A is a cellobiohydrolase rather than an endoglucanase, with a cellulose-binding tunnel that is more closed than Phanerochaete chrysosporium Cel7D and more open than Hypocrea jecorina Cel7A, suggesting intermediate enzyme properties. Molecular simulations were conducted to ascertain differences in enzyme-ligand interactions, ligand solvation, and loop flexibility between the family 7 glycoside hydrolase cellobiohydrolases from H. irregulare, H. jecorina, and P. chrysosporium. The structural comparisons and simulations suggest significant differences in enzyme-ligand interactions at the tunnel entrance in the -7 to -4 binding sites and suggest that a tyrosine residue at the tunnel entrance of HirCel7A may serve as an additional ligand-binding site. Additionally, the loops over the active site in H. jecorina Cel7A are more closed than loops in the other two enzymes, which has implications for the degree of processivity, endo-initiation, and substrate dissociation. Overall, this study highlights molecular level features important to understanding this biologically and industrially important family of glycoside hydrolases.

The mechanism of cellulose hydrolysis by a two-step, retaining cellobiohydrolase elucidated by structural and transition path sampling studies.

Glycoside hydrolases (GHs) cleave glycosidic linkages in carbohydrates, typically via inverting or retaining mechanisms, the latter of which proceeds via a two-step mechanism that includes formation of a glycosyl-enzyme intermediate. We present two new structures of the catalytic domain of Hypocrea jecorina GH Family 7 cellobiohydrolase Cel7A, namely a Michaelis complex with a full cellononaose ligand and a glycosyl-enzyme intermediate, that reveal details of the 'static' reaction coordinate. We also employ transition path sampling to determine the 'dynamic' reaction coordinate for the catalytic cycle. The glycosylation reaction coordinate contains components of forming and breaking bonds and a conformational change in the nucleophile. Deglycosylation proceeds via a product-assisted mechanism wherein the glycosylation product, cellobiose, positions a water molecule for nucleophilic attack on the anomeric carbon of the glycosyl-enzyme intermediate. In concert with previous structures, the present results reveal the complete hydrolytic reaction coordinate for this naturally and industrially important enzyme family.

Adaptation of Dekkera bruxellensis to lignocellulose-based substrate.

Adaptation of Dekkera bruxellensis to lignocellulose hydrolysate was investigated. Cells of D. bruxellensis were grown for 72 and 192 H in batch and continuous culture, respectively (adapted cells). Cultivations in semisynthetic medium were run as controls (nonadapted cells). To test the adaptation, cells from these cultures were reinoculated in the lignocellulose medium, and growth and ethanol production characteristics were monitored. Cells adapted to lignocellulose hydrolysate had a shorter lag phase, grew faster, and produced a higher ethanol concentration as compared with nonadapted cells. A stability test showed that after cultivation in rich medium, cells partially lost the adapted phenotype but still showed faster growth and higher ethanol production as compared with nonadapted cells. Because alcohol dehydrogenase genes have been described to be involved in the adaptation to furfural in Saccharomyces cerevisiae, an analogous mechanism of adaptation to lignocelluloses hydrolysate of D. bruxellensis was hypothesized. However, gene expression analysis showed that genes homologous to S. cerevisiae ADH1 were not involved in the adaptation to lignocelluloses hydrolysate in D. bruxellensis.

The crystal structure of RsSymEG1 reveals a unique form of smaller GH7 endoglucanases alongside GH7 cellobiohydrolases in protist symbionts of termites.

Glycoside hydrolase family 7 (GH7) cellulases are key enzymes responsible for carbon cycling on earth through their role in cellulose degradation and constitute highly important industrial enzymes as well. Although these enzymes are found in a wide variety of evolutionarily distant organisms across eukaryotes, they exhibit remarkably conserved features within two groups: exo-acting cellobiohydrolases and endoglucanases. However, recently reports have emerged of a separate clade of GH7 endoglucanases from protist symbionts of termites that are 60-80 amino acids shorter. In this work, we describe the first crystal structure of a short GH7 endoglucanase, RsSymEG1, from a symbiont of the lower termite Reticulitermes speratus. A more open flat surface and shorter loops around the non-reducing end of the cellulose-binding cleft indicate enhanced access to cellulose chains on the surface of cellulose microfibrils. Additionally, when comparing activities on polysaccharides to a typical fungal GH7 endoglucanase (Trichoderma longibrachiatum Cel7B), RsSymEG1 showed significantly faster initial hydrolytic activity. We also examine the prevalence and diversity of GH7 enzymes that the symbionts provide to the termite host, compare overall structures and substrate binding between cellobiohydrolase and long and short endoglucanase, and highlight the presence of similar short GH7s in other organisms.

Product binding varies dramatically between processive and nonprocessive cellulase enzymes.

Cellulases hydrolyze ß-1,4 glycosidic linkages in cellulose, which are among the most prevalent and stable bonds in Nature. Cellulases comprise many glycoside hydrolase families and exist as processive or nonprocessive enzymes. Product inhibition negatively impacts cellulase action, but experimental measurements of product-binding constants vary significantly, and there is little consensus on the importance of this phenomenon. To provide molecular level insights into cellulase product inhibition, we examine the impact of product binding on processive and nonprocessive cellulases by calculating the binding free energy of cellobiose to the product sites of catalytic domains of processive and nonprocessive enzymes from glycoside hydrolase families 6 and 7. The results suggest that cellobiose binds to processive cellulases much more strongly than nonprocessive cellulases. We also predict that the presence of a cellodextrin bound in the reactant site of the catalytic domain, which is present during enzymatic catalysis, has no effect on product binding in nonprocessive cellulases, whereas it significantly increases product binding to processive cellulases. This difference in product binding correlates with hydrogen bonding between the substrate-side ligand and the cellobiose product in processive cellulase tunnels and the additional stabilization from the longer tunnel-forming loops. The hydrogen bonds between the substrate- and product-side ligands are disrupted by water in nonprocessive cellulase clefts, and the lack of long tunnel-forming loops results in lower affinity of the product ligand. These findings provide new insights into the large discrepancies reported for binding constants for cellulases and suggest that product inhibition will vary significantly based on the amount of productive binding for processive cellulases on cellulose.

Enzyme kinetics by GH7 cellobiohydrolases on chromogenic substrates is dictated by non-productive binding: insights from crystal structures and MD simulation.

Cellobiohydrolases (CBHs) in the glycoside hydrolase family 7 (GH7) (EC3.2.1.176) are the major cellulose degrading enzymes both in industrial settings and in the context of carbon cycling in nature. Small carbohydrate conjugates such as p-nitrophenyl-ß-d-cellobioside (pNPC), p-nitrophenyl-ß-d-lactoside (pNPL) and methylumbelliferyl-ß-d-cellobioside have commonly been used in colorimetric and fluorometric assays for analysing activity of these enzymes. Despite the similar nature of these compounds the kinetics of their enzymatic hydrolysis vary greatly between the different compounds as well as among different enzymes within the GH7 family. Through enzyme kinetics, crystallographic structure determination, molecular dynamics simulations, and fluorometric binding studies using the closely related compound o-nitrophenyl-ß-d-cellobioside (oNPC), in this work we examine the different hydrolysis characteristics of these compounds on two model enzymes of this class, TrCel7A from Trichoderma reesei and PcCel7D from Phanerochaete chrysosporium. Protein crystal structures of the E212Q mutant of TrCel7A with pNPC and pNPL, and the wildtype TrCel7A with oNPC, reveal that non-productive binding at the product site is the dominating binding mode for these compounds. Enzyme kinetics results suggest the strength of non-productive binding is a key determinant for the activity characteristics on these substrates, with PcCel7D consistently showing higher turnover rates (kcat ) than TrCel7A, but higher Michaelis-Menten (KM ) constants as well. Furthermore, oNPC turned out to be useful as an active-site probe for fluorometric determination of the dissociation constant for cellobiose on TrCel7A but could not be utilized for the same purpose on PcCel7D, likely due to strong binding to an unknown site outside the active site.

Side-by-side biochemical comparison of two lytic polysaccharide monooxygenases from the white-rot fungus Heterobasidion irregulare on their activity against crystalline cellulose and glucomannan.

Our comparative studies reveal that the two lytic polysaccharide monooxygenases HiLPMO9B and HiLPMO9I from the white-rot conifer pathogen Heterobasidion irregulare display clear difference with respect to their activity against crystalline cellulose and glucomannan. HiLPMO9I produced very little soluble sugar on bacterial microcrystalline cellulose (BMCC). In contrast, HiLPMO9B was much more active against BMCC and even released more soluble sugar than the H. irregulare cellobiohydrolase I, HiCel7A. Furthermore, HiLPMO9B was shown to cooperate with and stimulate the activity of HiCel7A, both when the BMCC was first pretreated with HiLPMO9B, as well as when HiLPMO9B and HiCel7A were added together. No such stimulation was shown by HiLPMO9I. On the other hand, HiLPMO9I was shown to degrade glucomannan, using a C4-oxidizing mechanism, whereas no oxidative cleavage activity of glucomannan was detected for HiLPMO9B. Structural modeling and comparison with other glucomannan-active LPMOs suggest that conserved sugar-interacting residues on the L2, L3 and LC loops may be essential for glucomannan binding, where 4 out of 7 residues are shared by HiLPMO9I, but only one is found in HiLPMO9B. The difference shown between these two H. irregulare LPMOs may reflect distinct biological roles of these enzymes within deconstruction of different plant cell wall polysaccharides during fungal colonization of softwood.

Development of minimal enzyme cocktails for hydrolysis of sulfite-pulped lignocellulosic biomass.

Despite recent progress, saccharification of lignocellulosic biomass is still a major cost driver in biorefining. In this study, we present the development of minimal enzyme cocktails for hydrolysis of Norway spruce and sugarcane bagasse, which were pretreated using the so-called BALI™ process, which is based on sulfite pulping technology. Minimal enzyme cocktails were composed using several glycoside hydrolases purified from the industrially relevant filamentous fungus Trichoderma reesei and a purified commercial ß-glucosidase from Aspergillus niger. The contribution of in-house expressed lytic polysaccharide monooxygenases (LPMOs) was also tested, since oxidative cleavage of cellulose by such LPMOs is known to be beneficial for conversion efficiency. We show that the optimized cocktails permit efficient saccharification at reasonable enzyme loadings and that the effect of the LPMOs is substrate-dependent. Using a cocktail comprising only four enzymes, glucan conversion for Norway spruce reached >80% at enzyme loadings of 8mg/g glucan, whereas almost 100% conversion was achieved at 16mg/g.

Structural and biochemical studies of GH family 12 cellulases: improved thermal stability, and ligand complexes.

In this review we will describe how we have gathered structural and biochemical information from several homologous cellulases from one class of glycoside hydrolases (GH family 12), and used this information within the framework of a protein-engineering program for the design of new variants of these enzymes. These variants have been characterized to identify some of the positions and the types of mutations in the enzymes that are responsible for some of the biochemical differences in thermal stability and activity between the homologous enzymes. In this process we have solved the three-dimensional structure of four of these homologous GH 12 cellulases: Three fungal enzymes, Humicola grisea Cel12A, Hypocrea jecorina Cel12A and Hypocrea schweinitzii Cel12A, and one bacterial, Streptomyces sp. 11AG8 Cel12A. We have also determined the three-dimensional structures of the two most stable H. jecorina Cel12A variants. In addition, four ligand-complex structures of the wild-type H. grisea Cel12A enzyme have been solved and have made it possible to characterize some of the interactions between substrate and enzyme. The structural and biochemical studies of these related GH 12 enzymes, and their variants, have provided insight on how specific residues contribute to protein thermal stability and enzyme activity. This knowledge can serve as a structural toolbox for the design of Cel12A enzymes with specific properties and features suited to existing or new applications.

Crystal complex structures reveal how substrate is bound in the -4 to the +2 binding sites of Humicola grisea Cel12A.

As part of an ongoing enzyme discovery program to investigate the properties and catalytic mechanism of glycoside hydrolase family 12 (GH 12) endoglucanases, a GH family that contains several cellulases that are of interest in industrial applications, we have solved four new crystal structures of wild-type Humicola grisea Cel12A in complexes formed by soaking with cellobiose, cellotetraose, cellopentaose, and a thio-linked cellotetraose derivative (G2SG2). These complex structures allow mapping of the non-covalent interactions between the enzyme and the glucosyl chain bound in subsites -4 to +2 of the enzyme, and shed light on the mechanism and function of GH 12 cellulases. The unhydrolysed cellopentaose and the G2SG2 cello-oligomers span the active site of the catalytically active H.grisea Cel12A enzyme, with the pyranoside bound in subsite -1 displaying a S31 skew boat conformation. After soaking in cellotetraose, the cello-oligomer that is found bound in site -4 to -1 contains a beta-1,3-linkage between the two cellobiose units in the oligomer, which is believed to have been formed by a transglycosylation reaction that has occurred during the ligand soak of the protein crystals. The close fit of this ligand and the binding sites occupied suggest a novel mixed beta-glucanase activity for this enzyme.

Towards a molecular-level theory of carbohydrate processivity in glycoside hydrolases.

Polysaccharide depolymerization in nature is primarily accomplished by processive glycoside hydrolases (GHs), which abstract single carbohydrate chains from polymer crystals and cleave glycosidic linkages without dissociating after each catalytic event. Understanding the molecular-level features and structural aspects of processivity is of importance due to the prevalence of processive GHs in biomass-degrading enzyme cocktails. Here, we describe recent advances towards the development of a molecular-level theory of processivity for cellulolytic and chitinolytic enzymes, including the development of novel methods for measuring rates of key steps in processive action and insights gained from structural and computational studies. Overall, we present a framework for developing structure-function relationships in processive GHs and outline additional progress towards developing a fundamental understanding of these industrially important enzymes.

Enhanced ethanol production from wheat straw by integrated storage and pre-treatment (ISP).

Integrated storage and pre-treatment (ISP) combines biopreservation of moist material under airtight conditions and pre-treatment. Moist wheat straw was inoculated with the biocontrol yeast Wickerhamomyces anomalus, the xylan degrading yeast Scheffersomyces stipitis or a co-culture of both. The samples and non-inoculated controls were stored at 4 or 15 °C. The non-inoculated controls were heavily contaminated with moulds, in contrast to the samples inoculated with W. anomalus or S. stipitis. These two yeasts were able to grow on wheat straw as sole source of nutrients. When ethanol was produced from moist wheat straw stored for four weeks at 4 °C with S. stipitis, an up to 40% enhanced yield (final yield 0.15 g ethanol per g straw dry weight) was obtained compared to a dry sample (0.107 g/g). In all other moist samples, stored for four weeks at 4 °C or 15 °C, 6-35% higher yields were obtained. Thus, energy efficient bio-preservation can improve the pre-treatment efficiency for lignocellulose biomass, which is a critical bottleneck in its conversion to biofuels.

The putative endoglucanase PcGH61D from Phanerochaete chrysosporium is a metal-dependent oxidative enzyme that cleaves cellulose.

Many fungi growing on plant biomass produce proteins currently classified as glycoside hydrolase family 61 (GH61), some of which are known to act synergistically with cellulases. In this study we show that PcGH61D, the gene product of an open reading frame in the genome of Phanerochaete chrysosporium, is an enzyme that cleaves cellulose using a metal-dependent oxidative mechanism that leads to generation of aldonic acids. The activity of this enzyme and its beneficial effect on the efficiency of classical cellulases are stimulated by the presence of electron donors. Experiments with reduced cellulose confirmed the oxidative nature of the reaction catalyzed by PcGH61D and indicated that the enzyme may be capable of penetrating into the substrate. Considering the abundance of GH61-encoding genes in fungi and genes encoding their functional bacterial homologues currently classified as carbohydrate binding modules family 33 (CBM33), this enzyme activity is likely to turn out as a major determinant of microbial biomass-degrading efficiency.