Dataset Information

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

ABSTRACT: 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.

SUBMITTER: Gado JE

PROVIDER: S-EPMC8329511 | biostudies-literature |

REPOSITORIES: biostudies-literature

ACCESS DATA

Similar Datasets

Project description:As part of a program to discover improved glycoside hydrolase family 12 (GH 12) endoglucanases, we have studied the biochemical diversity of several GH 12 homologs. The H. schweinitzii Cel12A enzyme differs from the T. reesei Cel12A enzyme by only 14 amino acids (93% sequence identity), but is much less thermally stable. The bacterial Cel12A enzyme from S. sp. 11AG8 shares only 28% sequence identity to the T. reesei enzyme, and is much more thermally stable. Each of the 14 sequence differences from H. schweinitzii Cel12A were introduced in T. reesei Cel12A to determine the effect of these amino acid substitutions on enzyme stability. Several of the T. reesei Cel12A variants were found to have increased stability, and the differences in apparent midpoint of thermal denaturation (T(m)) ranged from a 2.5 degrees C increase to a 4.0 degrees C decrease. The least stable recruitment from H. schweinitzii Cel12A was A35S. Consequently, the A35V substitution was recruited from the more stable S. sp. 11AG8 Cel12A and this T. reesei Cel12A variant was found to have a T(m) 7.7 degrees C higher than wild type. Thus, the buried residue at position 35 was shown to be of critical importance for thermal stability in this structural family. There was a ninefold range in the specific activities of the Cel12 homologs on o-NPC. The most and least stable T. reesei Cel12A variants, A35V and A35S, respectively, were fully active. Because of their thermal tolerance, S. sp. 11AG8 Cel12A and T. reesei Cel12A variant A35V showed a continual increase in activity over the temperature range of 25 degrees C to 60 degrees C, whereas the less stable enzymes T. reesei Cel12A wild type and the destabilized A35S variant, and H. schweinitzii Cel12A showed a decrease in activity at the highest temperatures. The crystal structures of the H. schweinitzii, S. sp. 11AG8, and T. reesei A35V Cel12A enzymes have been determined and compared with the wild-type T. reesei Cel12A enzyme. All of the structures have similar Calpha traces, but provide detailed insight into the nature of the stability differences. These results are an example of the power of homolog recruitment as a method for identifying residues important for stability.

Project description:Family GH16 glycoside hydrolases can be assigned to five subgroups according to their substrate specificities, including xyloglucan transglucosylases/hydrolases (XTHs), (1,3)-beta-galactanases, (1,4)-beta-galactanases/kappa-carrageenases, "nonspecific" (1,3/1,3;1,4)-beta-D-glucan endohydrolases, and (1,3;1,4)-beta-D-glucan endohydrolases. A structured family GH16 glycoside hydrolase database has been constructed (http://www.ghdb.uni-stuttgart.de) and provides multiple sequence alignments with functionally annotated amino acid residues and phylogenetic trees. The database has been used for homology modeling of seven glycoside hydrolases from the GH16 family with various substrate specificities, based on structural coordinates for (1,3;1,4)-beta-D-glucan endohydrolases and a kappa-carrageenase. In combination with multiple sequence alignments, the models predict the three-dimensional (3D) dispositions of amino acid residues in the substrate-binding and catalytic sites of XTHs and (1,3/1,3;1,4)-beta-d-glucan endohydrolases; there is no structural information available in the databases for the latter group of enzymes. Models of the XTHs, compared with the recently determined structure of a Populus tremulos x tremuloides XTH, reveal similarities with the active sites of family GH11 (1,4)-beta-D-xylan endohydrolases. From a biological viewpoint, the classification, molecular modeling and a new 3D structure of the P. tremulos x tremuloides XTH establish structural and evolutionary connections between XTHs, (1,3;1,4)-beta-D-glucan endohydrolases and xylan endohydrolases. These findings raise the possibility that XTHs from higher plants could be active not only on cell wall xyloglucans, but also on (1,3;1,4)-beta-D-glucans and arabinoxylans, which are major components of walls in grasses. A role for XTHs in (1,3;1,4)-beta-D-glucan and arabinoxylan modification would be consistent with the apparent overrepresentation of XTH sequences in cereal expressed sequence tags databases.

Dataset Information

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

Similar Datasets

OmicsDI is part of the ELIXIR infrastructure

Tweets