Protein representations: Encoding biological information for machine learning in biocatalysis.

Enzymes offer a more environmentally friendly and low-impact solution to conventional chemistry, but they often require additional engineering for their application in industrial settings, an endeavour that is challenging and laborious. To address this issue, the power of machine learning can be harnessed to produce predictive models that enable the in silico study and engineering of improved enzymatic properties. Such machine learning models, however, require the conversion of the complex biological information to a numerical input, also called protein representations. These inputs demand special attention to ensure the training of accurate and precise models, and, in this review, we therefore examine the critical step of encoding protein information to numeric representations for use in machine learning. We selected the most important approaches for encoding the three distinct biological protein representations - primary sequence, 3D structure, and dynamics - to explore their requirements for employment and inductive biases. Combined representations of proteins and substrates are also introduced as emergent tools in biocatalysis. We propose the division of fixed representations, a collection of rule-based encoding strategies, and learned representations extracted from the latent spaces of large neural networks. To select the most suitable protein representation, we propose two main factors to consider. The first one is the model setup, which is influenced by the size of the training dataset and the choice of architecture. The second factor is the model objectives such as consideration about the assayed property, the difference between wild-type models and mutant predictors, and requirements for explainability. This review is aimed at serving as a source of information and guidance for properly representing enzymes in future machine learning models for biocatalysis.

Action and cooperation in alginate degradation by three enzymes from the human gut bacterium Bacteroides eggerthii DSM 20697.

Alginate is a polysaccharide consumed by humans in edible seaweed and different foods where it is applied as a texturizing hydrocolloid or in encapsulations of drugs and probiotics. While gut bacteria are found to utilize and ferment alginate to health-beneficial short-chain fatty acids, knowledge on the details of the molecular reactions is sparse. Alginates are composed of mannuronic acid (M) and its C-5 epimer guluronic acid (G). An alginate-related polysaccharide utilization locus (PUL) has been identified in the gut bacterium Bacteroides eggerthii DSM 20697. The PUL encodes two polysaccharide lyases (PLs) from the PL6 (BePL6) and PL17 (BePL17) families as well as a KdgF-like metalloprotein (BeKdgF) known to catalyze ring-opening of 4,5-unsaturated monouronates yielding 4-deoxy-l-erythro-5-hexoseulose uronate (DEH). B. eggerthii DSM 20697 does not grow on alginate, but readily proliferates with a lag phase of a few hours in the presence of an endo-acting alginate lyase A1-I from the marine bacterium Sphingomonas sp. A1. The B. eggerthii lyases are both exo-acting and while BePL6 is strictly G-block specific, BePL17 prefers M-blocks. BeKdgF retained 10-27% activity in the presence of 0.1-1 mM EDTA. X-ray crystallography was used to investigate the three-dimensional structure of BeKdgF, based on which a catalytic mechanism was proposed to involve Asp102, acting as acid/base having pKa of 5.9 as determined by NMR pH titration. BePL6 and BePL17 cooperate in alginate degradation with BeKdgF linearizing producing 4,5-unsaturated monouronates. Their efficiency of alginate degradation was much enhanced by the addition of the A1-I alginate lyase.

Enzymatic glycosylation of aloesone performed by plant UDP-dependent glycosyltransferases.

Aloesone is a bioactive natural product and biosynthetic precursor of rare glucosides found in rhubarb and some aloe plants including Aloe vera. This study aimed to investigate biocatalytic aloesone glycosylation and more than 400 uridine diphosphate-dependent glycosyltransferase (UGT) candidates, including multifunctional and promiscuous enzymes from a variety of plant species were assayed. As a result, 137 selective aloesone UGTs were discovered, including four from the natural producer rhubarb. Rhubarb UGT72B49 was further studied and its catalytic constants (kcat = 0.00092 ± 0.00003 s-1, KM = 30 ± 2.5 µM) as well as temperature and pH optima (50 °C and pH 7, respectively) were determined. We further aimed to find an efficient aloesone glycosylating enzyme with potential application for biocatalytic production of the glucoside. We discovered UGT71C1 from Arabidopsis thaliana as an efficient aloesone UGT showing a 167-fold higher catalytic efficiency compared to that of UGT72B49. Interestingly, sequence analysis of all the 137 newly identified aloesone UGTs showed that they belong to different phylogenetic groups, with the highest representation in groups B, D, E, F and L. Finally, our study indicates that aloesone C-glycosylation is highly specific and rare, since it was not possible to achieve in an efficient manner with any of the 422 UGTs assayed, including multifunctional GTs and 28 known C-UGTs.

GASP: A Pan-Specific Predictor of Family 1 Glycosyltransferase Acceptor Specificity Enabled by a Pipeline for Substrate Feature Generation and Large-Scale Experimental Screening.

Glycosylation represents a major chemical challenge; while it is one of the most common reactions in Nature, conventional chemistry struggles with stereochemistry, regioselectivity, and solubility issues. In contrast, family 1 glycosyltransferase (GT1) enzymes can glycosylate virtually any given nucleophilic group with perfect control over stereochemistry and regioselectivity. However, the appropriate catalyst for a given reaction needs to be identified among the tens of thousands of available sequences. Here, we present the glycosyltransferase acceptor specificity predictor (GASP) model, a data-driven approach to the identification of reactive GT1:acceptor pairs. We trained a random forest-based acceptor predictor on literature data and validated it on independent in-house generated data on 1001 GT1:acceptor pairs, obtaining an AUROC of 0.79 and a balanced accuracy of 72%. The performance was stable even in the case of completely new GT1s and acceptors not present in the training data set, highlighting the pan-specificity of GASP. Moreover, the model is capable of parsing all known GT1 sequences, as well as all chemicals, the latter through a pipeline for the generation of 153 chemical features for a given molecule taking the CID or SMILES as input (freely available at https://github.com/degnbol/GASP). To investigate the power of GASP, the model prediction probability scores were compared to GT1 substrate conversion yields from a newly published data set, with the top 50% of GASP predictions corresponding to reactions with >50% synthetic yields. The model was also tested in two comparative case studies: glycosylation of the antihelminth drug niclosamide and the plant defensive compound DIBOA. In the first study, the model achieved an 83% hit rate, outperforming a hit rate of 53% from a random selection assay. In the second case study, the hit rate of GASP was 50%, and while being lower than the hit rate of 83% using expert-selected enzymes, it provides a reasonable performance for the cases when an expert opinion is unavailable. The hierarchal importance of the generated chemical features was investigated by negative feature selection, revealing properties related to cyclization and atom hybridization status to be the most important characteristics for accurate prediction. Our study provides a GT1:acceptor predictor which can be trained on other data sets enabled by the automated feature generation pipelines. We also release the new in-house generated data set used for testing of GASP to facilitate the future development of GT1 activity predictors and their robust benchmarking.

A growth selection system for sucrose synthases (SuSy): design and test.

High throughput screening (HTS) methods of enzyme variants are essential for the development of robust biocatalysts suited for low impact, industrial scale, biobased synthesis of a myriad of compounds. However, for the majority of enzyme classes, current screening methods have limited throughput, or need expensive substrates in combination with sophisticated setups. Here, we present a straightforward, high throughput selection system that couples sucrose synthase activity to growth. Enabling high throughput screening of this enzyme class holds the potential to facilitate the creation of robust variants, which in turn can significantly impact the future of cost effective industrial glycosylation.

The sugar donor specificity of plant family 1 glycosyltransferases.

Plant family 1 glycosyltransferases (UGTs) represent a formidable tool to produce valuable natural and novel glycosides. Their regio- and stereo-specific one-step glycosylation mechanism along with their inherent wide acceptor scope are desirable traits in biotechnology. However, their donor scope and specificity are not well understood. Since different sugars have different properties in vivo and in vitro, the ability to easily glycodiversify target acceptors is desired, and this depends on our improved understanding of the donor binding site. In the aim to unlock the full potential of UGTs, studies have attempted to elucidate the structure-function relationship governing their donor specificity. These efforts have revealed a complex phenomenon, and general principles valid for multiple enzymes are elusive. Here, we review the studies of UGT donor specificity, and attempt to group the information into key concepts which can help shape future research. We zoom in on the family-defining PSPG motif, on two loop residues reported to interact with the C6 position of the sugar, and on the role of active site arginines in donor specificity. We continue to discuss attempts to alter and expand the donor specificity by enzyme engineering, and finally discuss future research directions.

A life cycle assessment of early-stage enzyme manufacturing simulations from sustainable feedstocks.

Enzyme-catalyzed reactions have relatively small environmental footprints. However, enzyme manufacturing significantly impacts the environment through dependence on traditional feedstocks. With the objective of determining the environmental impacts of enzyme production, the sustainability potential of six cradle-to-gate enzyme manufacturing systems focusing on glucose, sea lettuce, acetate, straw, and phototrophic growth, was thoroughly evaluated. Human and ecosystem toxicity categories dominated the overall impacts. Sea lettuce, straw, or phototrophic growth reduces fermentation-based emissions by 51.0, 63.7, and 79.7%, respectively. Substituting glucose-rich media demonstrated great potential to reduce marine eutrophication, land use, and ozone depletion. Replacing organic nitrogen sources with inorganic ones could further lower these impacts. Location-specific differences in electricity result in a 14% and a 27% reduction in the carbon footprint for operation in Denmark compared to the US and China. Low-impact feedstocks can be competitive if they manage to achieve substrate utilization rates and productivity levels of conventional enzyme production processes.

Chemoenzymatic indican for light-driven denim dyeing.

Blue denim, a billion-dollar industry, is currently dyed with indigo in an unsustainable process requiring harsh reducing and alkaline chemicals. Forming indigo directly in the yarn through indican (indoxyl-ß-glucoside) is a promising alternative route with mild conditions. Indican eliminates the requirement for reducing agent while still ending as indigo, the only known molecule yielding the unique hue of blue denim. However, a bulk source of indican is missing. Here, we employ enzyme and process engineering guided by techno-economic analyses to develop an economically viable drop-in indican synthesis technology. Rational engineering of PtUGT1, a glycosyltransferase from the indigo plant, alleviated the severe substrate inactivation observed with the wildtype enzyme at the titers needed for bulk production. We further describe a mild, light-driven dyeing process. Finally, we conduct techno-economic, social sustainability, and comparative life-cycle assessments. These indicate that the presented technologies have the potential to significantly reduce environmental impacts from blue denim dyeing with only a modest cost increase.

Glucuronan lyases from family PL7 use a Tyr/Tyr syn ß-elimination catalytic mechanism for glucuronan breakdown.

TpPL7A and TpPL7B, members of CAZy family PL7, act as ß-glucuronan lyases. TpPL7A diverges by lacking the catalytic histidine, identified as the Brønsted base in PL7 alginate lyases. Our research, including TpPL7A's crystal structure, and mutagenesis studies, reveals a shared syn-ß-elimination mechanism with a single tyrosine serving as both base and acid catalyst. This mechanism may extend to subfamily PL7_4 glucuronan lyases.

Regioselective Glycosylation of Polyphenols by Family 1 Glycosyltransferases: Experiments and Simulations.

Family 1 glycosyltransferases (GT1s, UGTs) form natural product glycosides with exquisite control over regio- and stereoselectivity, representing attractive biotechnological targets. However, regioselectivity cannot be predicted and large-scale activity assessment efforts of UGTs are commonly performed via mass spectrometry or indirect assays that are blind to regioselectivity. Here, we present a large high performance liquid chromatography screening discriminating between regioisomeric products of 40 diverse UGTs (28.6% average pairwise sequence identity) against 32 polyphenols, identifying enzymes able to reach high glycosylation yields (≥90% in 24 h) in 26/32 cases. In reactions with >50% yield, we observed perfect regioselectivity for 47% (75/158) on polyphenols presenting two hydroxyl groups and for 30% (43/143) on polyphenols presenting ≥3 hydroxyl groups. Moreover, we developed a nuclear magnetic resonance-based procedure to identify the site of glycosylation directly on enzymatic mixtures. We further selected seven regiospecific reactions catalyzed by four enzymes on five dihydroxycoumarins. We characterized the four enzymes, showing that temperature optima are functions of the acceptor substrate, varying by up to 20 °C for the same enzyme. Furthermore, we performed short molecular dynamics simulations of 311 ternary complexes (UGT, UDP-Glc, and glycosyl acceptor) to investigate the molecular basis for regioselectivity. Interestingly, it appeared that most UGTs can accommodate acceptors in configurations favorable to the glycosylation of either hydroxyl. In contrast, evaluation of hydroxyl nucleophilicity appeared to be a strong predictor of the hydroxyl predominantly glycosylated by most enzymes.

Promiscuous Yet Specific: A Methionine-Aromatic Interaction Drives the Reaction Scope of the Family 1 Glycosyltransferase GmUGT88E3 from Soybean.

Family 1 glycosyltransferases (GT1s, UGTs) catalyze the regioselective glycosylation of natural products in a single step. We identified GmUGT88E3 as a particularly promising biocatalyst able to produce a variety of pure, single glycosidic products from polyphenols with high chemical yields. We investigated this particularly desirable duality toward specificity, i.e., promiscuous toward acceptors while regiospecific. Using high-field NMR, kinetic characterization, molecular dynamics simulations, and mutagenesis studies, we uncovered that the main molecular determinant of GmUGT88E3 specificity is a methionine-aromatic bridge, an interaction often present in protein structures but never reported for enzyme-substrate interactions. Here, mutating Met127 led to inactive proteins or 100-fold reduced activity.

The function of UDP-glycosyltransferases in plants and their possible use in crop protection.

Glycosyltransferases catalyse the transfer of a glycosyl moiety from a donor to an acceptor. Members of this enzyme class are ubiquitous throughout all kingdoms of life and are involved in the biosynthesis of countless types of glycosides. Family 1 glycosyltransferases, also referred to as uridine diphosphate-dependent glycosyltransferases (UGTs), glycosylate small molecules such as secondary metabolites and xenobiotics. In plants, UGTs are recognised for their multiple functionalities ranging from roles in growth regulation and development, in protection against pathogens and abiotic stresses and in adaptation to changing environments. In this study, we review UGT-mediated glycosylation of phytohormones, endogenous secondary metabolites, and xenobiotics and contextualise the role this chemical modification plays in the response to biotic and abiotic stresses and plant fitness. Here, the potential advantages and drawbacks of altering the expression patterns of specific UGTs along with the heterologous expression of UGTs across plant species to improve stress tolerance in plants are discussed. We conclude that UGT-based genetic modification of plants could potentially enhance agricultural efficiency and take part in controlling the biological activity of xenobiotics in bioremediation strategies. However, more knowledge of the intricate interplay between UGTs in plants is needed to unlock the full potential of UGTs in crop resistance.

Functional characterization of the phosphotransferase system in Parageobacillus thermoglucosidasius.

Parageobacillus thermoglucosidasius is a thermophilic bacterium characterized by rapid growth, low nutrient requirements, and amenability to genetic manipulation. These characteristics along with its ability to ferment a broad range of carbohydrates make P. thermoglucosidasius a potential workhorse in whole-cell biocatalysis. The phosphoenolpyruvate:carbohydrate phosphotransferase system (PTS) catalyzes the transport and phosphorylation of carbohydrates and sugar derivatives in bacteria, making it important for their physiological characterization. In this study, the role of PTS elements on the catabolism of PTS and non-PTS substrates was investigated for P. thermoglucosidasius DSM 2542. Knockout of the common enzyme I, part of all PTSs, showed that arbutin, cellobiose, fructose, glucose, glycerol, mannitol, mannose, N-acetylglucosamine, N-acetylmuramic acid, sorbitol, salicin, sucrose, and trehalose were PTS-dependent on translocation and coupled to phosphorylation. The role of each putative PTS was investigated and six PTS-deletion variants could not grow on arbutin, mannitol, N-acetylglucosamine, sorbitol, and trehalose as the main carbon source, or showed diminished growth on N-acetylmuramic acid. We concluded that PTS is a pivotal factor in the sugar metabolism of P. thermoglucosidasius and established six PTS variants important for the translocation of specific carbohydrates. This study lays the groundwork for engineering efforts with P. thermoglucosidasius towards efficient utilization of diverse carbon substrates for whole-cell biocatalysis.

An improved integrative GFP-based vector for genetic engineering of Parageobacillus thermoglucosidasius facilitates the identification of a key sporulation regulator.

Parageobacillus thermoglucosidasius is a thermophilic Gram-positive bacterium, which is a promising host organism for sustainable bio-based production processes. However, to take full advantage of the potential of P. thermoglucosidasius, more efficient tools for genetic engineering are required. The present study describes an improved shuttle vector, which speeds up recombination-based genomic modification by incorporating a thermostable sfGFP variant into the vector backbone. This additional selection marker allows for easier identification of recombinants, thereby removing the need for several culturing steps. The novel GFP-based shuttle is therefore capable of facilitating faster metabolic engineering of P. thermoglucosidasius through genomic deletion, integration, or exchange. To demonstrate the efficiency of the new system, the GFP-based vector was utilised for deletion of the spo0A gene in P. thermoglucosidasius DSM2542. This gene is known to be a key regulator of sporulation in Bacillus subtilis, and it was therefore hypothesised that the deletion of spo0A in P. thermoglucosiadius would produce an analogous sporulation-inhibited phenotype. Subsequent analyses of cell morphology and culture heat resistance suggests that the P. thermoglucosidasius ∆spo0A strain is sporulation-deficient. This strain may be an excellent starting point for future cell factory engineering of P. thermoglucosidasius, as the formation of endospores is normally not a desired trait in large-scale production.

Sequence mining yields 18 phloretin C-glycosyltransferases from plants for the efficient biocatalytic synthesis of nothofagin and phloretin-di-C-glycoside.

C-glycosyltransferases (C-GTs) offer selective and efficient synthesis of natural product C-glycosides under mild reaction conditions. In contrast, the chemical synthesis of these C-glycosides is challenging and environmentally harmful. The rare occurrence of C-glycosylated compounds in Nature, despite their stability, suggests that their biosynthetic enzymes, C-GTs, might be scarce. Indeed, the number of characterized C-GTs is remarkably lower than O-GTs. Therefore, discovery efforts are crucial for expanding our knowledge of these enzymes and their efficient application in biocatalytic processes. This study aimed to identify new C-GTs based on their primary sequence. 18 new C-GTs were discovered, 10 of which yielded full conversion of phloretin to its glucosides. Phloretin is a dihydrochalcone natural product, with its mono-C-glucoside, nothofagin, having various health-promoting effects. Several of these enzymes enabled highly selective production of either nothofagin (UGT708A60 and UGT708F2) or phloretin-di-C-glycoside (UGT708D9 and UGT708B8). Molecular docking simulations, based on structural models of selected enzymes, showed productive binding modes for the best phloretin C-GTs, UGT708F2 and UGT708A60. Moreover, we characterized UGT708A60 as a highly efficient phloretin mono-C glycosyltransferase (kcat  = 2.97 s-1 , KM  = 0.1 µM) active in non-buffered, dilute sodium hydroxide (0.1-1 mM). We further investigated UGT708A60 as an efficient biocatalyst for the bioproduction of nothofagin.

Family 1 glycosyltransferases (GT1, UGTs) are subject to dilution-induced inactivation and low chemo stability toward their own acceptor substrates.

Glycosylation reactions are essential but challenging from a conventional chemistry standpoint. Conversely, they are biotechnologically feasible as glycosyltransferases can transfer sugar to an acceptor with perfect regio- and stereo-selectivity, quantitative yields, in a single reaction and under mild conditions. Low stability is often alleged to be a limitation to the biotechnological application of glycosyltransferases. Here we show that these enzymes are not necessarily intrinsically unstable, but that they present both dilution-induced inactivation and low chemostability towards their own acceptor substrates, and that these two phenomena are synergistic. We assessed 18 distinct GT1 enzymes against three unrelated acceptors (apigenin, resveratrol, and scopoletin-respectively a flavone, a stilbene, and a coumarin), resulting in a total of 54 enzymes: substrate pairs. For each pair, we varied catalyst and acceptor concentrations to obtain 16 different reaction conditions. Fifteen of the assayed enzymes (83%) displayed both low chemostability against at least one of the assayed acceptors at submillimolar concentrations, and dilution-induced inactivation. Furthermore, sensitivity to reaction conditions seems to be related to the thermal stability of the enzymes, the three unaffected enzymes having melting temperatures above 55°C, whereas the full enzyme panel ranged from 37.4 to 61.7°C. These results are important for GT1 understanding and engineering, as well as for discovery efforts and biotechnological use.

Structure-guided engineering of key amino acids in UGT85B1 controlling substrate and stereo-specificity in aromatic cyanogenic glucoside biosynthesis.

Cyanogenic glucosides are important defense molecules in plants with useful biological activities in animals. Their last biosynthetic step consists of a glycosylation reaction that confers stability and increases structural diversity and is catalyzed by the UDP-dependent glycosyltransferases (UGTs) of glycosyltransferase family 1. These versatile enzymes have large and varied substrate scopes, and the structure-function relationships controlling scope and specificity remain poorly understood. Here, we report substrate-bound crystal structures and rational engineering of substrate and stereo-specificities of UGT85B1 from Sorghum bicolor involved in biosynthesis of the cyanogenic glucoside dhurrin. Substrate specificity was shifted from the natural substrate (S)-p-hydroxymandelonitrile to (S)-mandelonitrile by combining a mutation to abolish hydrogen bonding to the p-hydroxyl group with a mutation to provide steric hindrance at the p-hydroxyl group binding site (V132A/Q225W). Further, stereo-specificity was shifted from (S) to (R) by substituting four rationally chosen residues within 6 Å of the nitrile group (M312T/A313T/H408F/G409A). These activities were compared to two other UGTs involved in the biosynthesis of aromatic cyanogenic glucosides in Prunus dulcis (almond) and Eucalyptus cladocalyx. Together, these studies enabled us to pinpoint factors that drive substrate and stereo-specificities in the cyanogenic glucoside biosynthetic UGTs. The structure-guided engineering of the functional properties of UGT85B1 enhances our understanding of the evolution of UGTs involved in the biosynthesis of cyanogenic glucosides and will enable future engineering efforts towards new biotechnological applications.

Family 1 Glycosyltransferase UGT706F8 from Zea mays Selectively Catalyzes the Synthesis of Silibinin 7-O-ß-d-Glucoside.

Regioselective glycosylation is a chemical challenge, leading to multistep syntheses with protecting group manipulations, ultimately resulting in poor atom economy and compromised sustainability. Enzymes allow eco-friendly and regioselective bond formation with fully deprotected substrates in a single reaction. For the selective glucosylation of silibinin, a pharmaceutical challenged with low solubility, enzyme engineering has previously been employed, but the resulting yields and k cat were limited, prohibiting the application of the engineered catalyst. Here, we identified a naturally regioselective silibinin glucosyltransferase, UGT706F8, a family 1 glycosyltransferase from Zea mays. It selectively and efficiently (k cat = 2.1 ± 0.1 s-1; K M = 36.9 ± 5.2 µM; TTN = 768 ± 22) catalyzes the quantitative synthesis of silibinin 7-O-ß-d-glucoside. We solved the crystal structure of UGT706F8 and investigated the molecular determinants of regioselective silibinin glucosylation. UGT706F8 was the only regioselective enzyme among 18 glycosyltransferases found to be active on silibinin. We found the temperature optimum of UGT706F8 to be 34 °C and the pH optimum to be 7-8. Our results indicate that UGT706F8 is an efficient silibinin glycosyltransferase that enables biocatalytic production of silbinin 7-O-ß-d-glucoside.

Oligomerization engineering of the fluorinase enzyme leads to an active trimer that supports synthesis of fluorometabolites in vitro.

The fluorinase enzyme represents the only biological mechanism capable of forming stable C-F bonds characterized in nature thus far, offering a biotechnological route to the biosynthesis of value-added organofluorines. The fluorinase is known to operate in a hexameric form, but the consequence(s) of the oligomerization status on the enzyme activity and its catalytic properties remain largely unknown. In this work, this aspect was explored by rationally engineering trimeric fluorinase variants that retained the same catalytic rate as the wild-type enzyme. These results ruled out hexamerization as a requisite for the fluorination activity. The Michaelis constant (KM ) for S-adenosyl-l-methionine, one of the substrates of the fluorinase, increased by two orders of magnitude upon hexamer disruption. Such a shift in S-adenosyl-l-methionine affinity points to a long-range effect of hexamerization on substrate binding - likely decreasing substrate dissociation and release from the active site. A practical application of trimeric fluorinase is illustrated by establishing in vitro fluorometabolite synthesis in a bacterial cell-free system.

Discovery of fungal oligosaccharide-oxidising flavo-enzymes with previously unknown substrates, redox-activity profiles and interplay with LPMOs.

Oxidative plant cell-wall processing enzymes are of great importance in biology and biotechnology. Yet, our insight into the functional interplay amongst such oxidative enzymes remains limited. Here, a phylogenetic analysis of the auxiliary activity 7 family (AA7), currently harbouring oligosaccharide flavo-oxidases, reveals a striking abundance of AA7-genes in phytopathogenic fungi and Oomycetes. Expression of five fungal enzymes, including three from unexplored clades, expands the AA7-substrate range and unveils a cellooligosaccharide dehydrogenase activity, previously unknown within AA7. Sequence and structural analyses identify unique signatures distinguishing the strict dehydrogenase clade from canonical AA7 oxidases. The discovered dehydrogenase directly is able to transfer electrons to an AA9 lytic polysaccharide monooxygenase (LPMO) and fuel cellulose degradation by LPMOs without exogenous reductants. The expansion of redox-profiles and substrate range highlights the functional diversity within AA7 and sets the stage for harnessing AA7 dehydrogenases to fine-tune LPMO activity in biotechnological conversion of plant feedstocks.