Genome mining of amylases and amylase inhibitors from Streptomyces.

Genome mining of amylases and amylase inhibitors from Streptomyces - PMC Skip to main content An official website of the United States government Here's how you know Here's how you know Search Log in Primary site navigation Search Log in Search PMC Full-Text Archive Search in PMC Journal List User Guide As a library, NLM provides access to scientific literature. Inclusion in an NLM database does not imply endorsement of, or agreement with, the contents by NLM or the National Institutes of Health. Learn more: PMC Disclaimer | PMC Copyright Notice Microb Genom . 2026 Jun 5;12(6):001747. doi: 10.1099/mgen.0.001747 Genome mining of amylases and amylase inhibitors from Streptomyces Andreas C Lawaetz 1, * , Alexa Gannon 1 , Paul A Hoskisson 1, * Author information Article notes Copyright and License information 1 Strathclyde Institute of Pharmacy and Biomedical Sciences, University of Strathclyde, 161 Cathedral Street, Glasgow, G4 0RE, UK * Andreas C. Lawaetz, andreas.lawaetz@strath.ac.uk * Paul A. Hoskisson, paul.hoskisson@strath.ac.uk The authors declare that there are no conflicts of interest. Supplementary material: All supporting data, code and protocols have been provided within the article or through supplementary data files. Five supplementary figures are available with the online version of this article. Received 2026 Jan 14; Accepted 2026 May 17; Collection date 2026. Copyright © 2026 The Authors This is an open-access article distributed under the terms of the Creative Commons Attribution License. This article was made open access via a Publish and Read agreement between the Microbiology Society and the corresponding author’s institution. PMC Copyright notice PMCID: PMC13240717  PMID: 42247255 Abstract Most antibiotics are derived from Streptomyces bacteria and produced industrially via fermentations relying on food-grade feedstocks, which have a substantial environmental impact. Efficient utilization of starch-rich organic wastes, such as bread and potato waste, as alternative feedstocks, remains limited by incomplete knowledge of starch metabolism in Streptomyces . Here, a genus-wide analysis of 295 Streptomyces strains was performed, identifying 645 amylases grouped into pullulanases, α -1,4-amylases and cyclomaltodextrin-like amylases, alongside biosynthetic gene clusters encoding amylase inhibitors such as acarbose and tendamistat. Structural and sequence analyses revealed that inhibitor resistance may arise from subtle modifications in the α -amylase catalytic pocket. This comprehensive analysis of amylases and inhibitors provides a foundation for engineering inhibitor-resistant enzymes tailored to diverse starch substrates, facilitating the development of sustainable, starch-based Streptomyces fermentations. Keywords: acarbose, amylase, pullulanase, starch, Streptomyces , tendamistat Impact Statement. Pharmaceutical manufacturing contributes substantially to global carbon emissions, with feedstocks used in natural product fermentations being a major factor. Replacing conventional feedstocks with organic waste streams, such as starch-rich bread or potato waste, could reduce the environmental impact, but two key challenges remain: production strains may lack the enzymes required to efficiently degrade alternative substrates, and nutrient changes may trigger carbon catabolite repression, reducing product yields. While previous studies have linked carbohydrate-active enzymes with biosynthetic gene clusters, no comprehensive analysis has focused specifically on starch degradation in Streptomyces . To fill this gap, a genus-wide analysis of amylases and amylase inhibitors across 295 Streptomyces genomes was carried out. This created an atlas of 645 amylases, which may serve as the foundation for the targeted selection of enzymes optimized for catabolizing organic waste from various sources, thereby improving the sustainability of industrial natural product production, such as antibiotics. Data Summary Scripts used along with supplementary files and figures can be accessed through GitHub at https://github.com/ALawaetz/Amylases_and_amylase_inhibitors_in_Streptomycetes . Introduction The majority of antibiotics and other bioactive natural products are generated by complex biosynthetic pathways found naturally in Streptomyces bacteria and are produced industrially in large-scale fermentations [ 1 ]. Pharmaceutical manufacturing has a substantial impact on the environment and accounts for 5% of the UK’s carbon footprint with a large contribution coming from the carbon-intensive processed sugars and lipids currently used as feedstocks in antibiotic production systems [ 2 ]. One strategy to enhance the sustainability of pharmaceutical manufacturing is to replace current feedstocks with organic waste products. Globally, ~1.3 billion tons of food waste is produced each year, representing a cheap and sustainable energy source that could become part of a circular bioeconomy instead of ending up in landfills [ 3 , 4 ]. Bread waste, for example, is comprised of 50–70% starch, which could potentially be fed into antibiotic fermentations [ 4 , 5 ]. Changing feedstocks for bioactive natural products, such as antibiotics in production strains, however, is not straightforward, as primary metabolism is intimately linked to biosynthesis, which is in turn subject to carbon catabolite repression [ 6 ]. To rationally engineer cells to metabolize desired carbon sources while maintaining high antibiotic titres, a thorough understanding of metabolic networks is required. Hence, substantial efforts have gone into correlating the repertoire of biosynthetic gene clusters (BGCs) with that of carbohydrate-active enzymes (CAZymes) in many Streptomyces genomes [ 7 , 9 ]. While CAZyme analyses use resources such as the Carbohydrate-Active Enzyme database [ 10 ] and annotation tools like dbCAN [ 11 ] and provide valuable insights into the metabolic breadth of organisms, they are limited by the level of functional detail they can offer. Many CAZyme categories are broad and encompass proteins with diverse enzymatic activities. For example, the glycoside hydrolase category 13 (GH13) family contains enzymes with activities as diverse as amylases, sucrases and trehalose synthases. Whilst efforts are underway to further subdivide CAZyme families into more homogeneous, functionally coherent groups [ 10 ], these refinements remain incomplete. As a result, previous efforts of genome mining constricted to CAZyme-based approaches do not resolve specific carbohydrate utilization pathways. Therefore, to bioinformatically assess the capacity of streptomycetes to metabolize starch specifically, further characterization beyond standard CAZyme annotation is required. Starch is a naturally occurring polymer and many bacteria, including species of Streptomyces , encode and secrete amylases, which enable them to break down and catabolize starch [ 5 ]. Starch consists of long chains of glucose residues covalently linked through α -1,4-glycosidic bonds and may further be branched through α -1,6 linkages (amylose and amylopectin, respectively) [ 5 ]. In the natural microenvironments of streptomycetes, other starch-like polysaccharides may be encountered, including pullulan (composed of maltotriose units linked by α -1,6 bonds) and cyclomaltodextrins (cyclic oligomers of α -1,4-linked glucose residues). Concordantly, Streptomyces genomes have been annotated as encoding amylases, pullulanases and cyclomaltodextrinases [ 5 ]. Starch is among the most abundant and energy-dense polymers in the environment [ 5 ], and fierce competition between microbes to access this resource is likely. Unsurprisingly, species of Streptomyces synthesize amylase inhibitors, including carbohydrate-based compounds like acarbose, acarviostatin and trestatin and small peptide inhibitors like tendamistat, Gaim and Scaim [ 12 , 16 ]. While amylase inhibitors and inhibitor-resistant amylases have been characterized in some species, their prevalence and correlation with the repertoire of amylases across the Streptomyces genus have not been systematically assessed. It remains unclear whether inhibitor production and inhibitor-resistant amylases represent isolated traits or whether they form part of a broader ecological strategy linked to starch utilization. Addressing this gap is important for the development of sustainable fermentation processes based on starch-rich waste streams, where efficient substrate utilization is critical. The ability of Streptomyces strains to metabolize such feedstocks optimally may depend not only on the presence of amylases but on their exact substrate specificities and susceptibility to inhibition. Resolving these factors is, therefore, essential for the rational selection or engineering of strains optimized for the sustainable manufacturing of specialized metabolites. The saprophytic nature and abundance of starch in the natural environment of Streptomyces [ 5 ], coupled with the ability to produce numerous specialized metabolites, led to the hypothesis that starch utilization is a common trait in streptomycetes, which is intricately linked to the production of amylase inhibitors and the evolution of inhibitor-resistant amylases. To test this, bioinformatic resources, including dbCAN [ 17 ], SignalP [ 18 ], hmmer [ 19 ] and GATOR [ 20 ], were combined to assess the distribution of amylases and amylase inhibitors across 295 Streptomyces genomes [ 21 ], providing a broad representation of the starch catabolic diversity of the genus. It was found that pullulanases and α -1,4-acting amylases are almost ubiquitous, whereas only a subset of strains encodes a third, cyclomaltodextrin-like amylase. The total number of amylases per genome ranged from zero to six, and there was a positive correlation between strains encoding higher numbers of amylases and those encoding amylase inhibitors. BGCs encoding amylase inhibitors were typically found alongside multiple amylases, including variants with amino acid motifs known to provide resistance to inhibition [ 16 ] as well as putative resistance variants. Methods Phylogenetic analysis of amylases in Streptomyces A database of 295 Streptomyces genomes [ 21 ] (accession numbers found in File S6) was used for the analysis and the protein fasta files for each genome were downloaded with the National Center for Biotechnology Information (NCBI) datasets command-line tools (script: File S10, available in the online Supplementary Material). In brief, the Streptomyces genomes dataset [ 21 ] was selected as follows: 873 publicly available Streptomyces genomes containing single-copy variants of all 6 MLST markers used in the pubMLST Streptomyces scheme [ 22 ] were analysed using a minimum spanning tree approach [ 21 ]. This analysis resolved the genomes into 295 sequence type (ST) groups. A single representative genome from each ST group was then selected for downstream phylogenomic analysis, which was conducted using a concatenated alignment of 137 single-copy orthologue genes [ 21 ]. Amylases were found by using dbCAN (version: 5.1.2) [ 11 ] with the option CAZyme_annotation (script: File S7). Amylases belong to the CAZyme family GH13, so results were subsequently filtered for GH13 proteins where this was the recommended result (i.e. agreement between HMMER and DIAMOND). Proteins whose recommended result included multiple CAZyme groups, e.g. GH13 and/or CBM20, were also included in the analysis. Results that included subgroups, e.g. GH13_32, were included in the analysis simply as GH13 (script: File S8). SignalP (version: 6.0) [ 18 ] was then used to select only proteins with secretion signals (script: File S9), yielding a total of 645 amylases. A multiple sequence alignment of all amylases was carried out using Mafft (version: 7.525) [ 23 ] (script: File S11), and their phylogenetic relationship was determined using RAxML-NG (version: 1.2.2) [ 24 ] with 100 bootstrap replicates (script: File S12). A circular tree rooted at the midpoint was drawn from the resulting output using R (script: File S13). Domain analysis was carried out using NLM’s conserved domain database [ 25 ] using representative proteins from each clade. Representative proteins were defined as the proteins most similar to the consensus sequence in each clade (script: File S22). These were WP_062143504.1 (Clade 1), WP_218063397.1 (Clade 2) and WP_150491328.1 (Clade 3). Amylase inhibitors Previously characterized BGCs encoding amylase inhibitors acarbose [ 16 ], acarviostatin [ 13 ] and trestatin [ 16 ] were used to search Streptomyces genomes using GATOR-GC (version: 0.9.0) [ 20 ] (script: File S14). GATOR-GC takes two lists of genes as input: a list of required genes and a list of optional genes. Required genes to search for acarbose BGC were the core genes required for aminocyclitol synthesis: gac (A, V, W, X, Y, U, S, R, K, I, Q, C, J, M, O). The homologous genes in Streptomyces coelicoflavus ZG0656 and Streptomyces dimorphogenes ATCC 31484 were used to search for acarviostatin and trestatin BGCs, respectively. A combined list of optional genes was used to search all clusters. This included all amylases and proteinaceous amylase inhibitors from the characterized BGCs in Streptomyces glaucescens [ 15 , 16 ], S. coelicoflavus ZG0656 [ 13 ] and S. dimorphogenes ATCC 31484 [ 16 ], all 645 amylases identified in this study and tendamistat from Streptomyces tendae ( M28478.1 ). A genus-wide search for tendamistat-like peptides was carried out by building a HMMER profile (HMMER version: 3.4) from a tendamistat seed alignment (smart domain smart00783 from the NLM’s conserved domain database) and then using hmmsearch (script: Files S15) with an E -value cutoff of 10 −10 to find occurrences across the Streptomyces genus. Occurrences of amylases and amylase inhibitors were plotted using R around a circular tree rooted at the midpoint, displaying the evolutionary relationship between 295 Streptomyces genomes based on whole-genome phylogeny [ 21 ] (script: File S16). Multiple sequence alignment Multiple sequence alignments of each identified clade of amylases were carried out with Mafft (version 7.525) [ 23 ] using the seed option (seed alignment from conserved α-amylase domain cd00551) (File S11). Aligned sequences were visualized with Jalview (version: 2.11.5.1) [ 26 ]. Statistical analysis The distribution of the number of amylases per genome was tested using scipy.stats shapiro package and found not to be normally distributed. A Mann–Whitney U test was hence applied to test whether the distributions of the number of amylases in genomes with and without acarbose BGC were significantly different from each other. All genomes carried either one or zero cyclomaltodextrin-like amylases and one or zero tendamistat genes, and with two binary variables, a chi-square test was used to test the correlation between the two. To test for enrichment of aligned amino acids in amylases found in acarbose BGCs, a Fisher’s exact test was applied and corrected for multiple testing using FDR (Benjamini–Hochberg) and Bonferroni. A P -value of 0.05 was used as a cutoff for significance. All scripts used for statistical analysis can be found in Files S17 and S18. Protein structures and ligand binding modelling AlphaFold 3 [ 27 ] was used to model the 3-dimensional structure of proteins. CB-DOCK2 [ 28 ] and ClusPro 2.0 [ 29 ] were used to predict the binding of acarbose (PubChem CID 9811704) and Gaim to amylases, respectively. Default parameters were used in all cases. For simplicity, the modelling of binding interactions between Gaim and pullulanases was done using the α -amylase domains only. As such, it cannot be excluded that amylase inhibitor resistance may be achieved through interdomain interactions obstructing Gaim binding. Open source PyMOL was installed using conda and used to visualize and annotate proteins and protein–ligand interactions (script: Files S19–S21). Results To assess the prevalence of amylases in Streptomyces species, dbCAN [ 11 ] was used to search 295 genomes, broadly encompassing the genetic diversity of the genus [ 21 ], for proteins belonging to the CAZyme family GH13 [ 10 ]. Amylases are secreted enzymes that hydrolyse starch in the extracellular environment (Janec̆ek, 1997). The presence of a signal peptide, therefore, allows them to be distinguished from other non-secreted GH13 family members, such as enzymes involved in trehalose and glycogen metabolism [ 30 , 32 ]. A total of 4,173 GH13 proteins were analysed with SignalP [ 18 ], which identified 645 enzymes with secretion signal peptides. Conserved domain analysis [ 25 ] showed that all 645 contained either an amylase domain, a pullulanase domain, or both (File S23). An additional 24 GH13 proteins contained amylase domains but lacked identifiable signal peptides and were hence omitted from further analysis. Phylogenetic analysis carried out on the 645 identified amylases revealed three discrete clades ( Fig. 1a ). Analysis of conserved domains found that the three clades constituted pullulanases (clade 1), α -1,4-amylases (clade 2) and cyclomaltodextrin-like amylases (clade 3, Fig. 1b ). To assess the distribution of each type of amylase across the Streptomyces genus, the presence of each was mapped to each genome using a phylogenetic tree displaying the evolutionary relationship between all 295 genomes ( Fig. 2 ). It was found that most strains encoded pullulanases and α -1,4-amylases (87% and 89%, respectively), whereas only 22% of strains had a cyclomaltodextrin-like amylase in their genome. The total number of amylases per genome varied between zero and six. Fig. 1. Streptomycetes possess three types of amylases. A total of 645 amylases were found across 295 Streptomyces genomes. Phylogenetic analysis showed that amylases in streptomycetes cluster into three major clades. Branch support values (bootstrap) are shown as Felsenstein’s bootstrap proportions scores, with coloured circles indicating values ≥70% at internal nodes (a). Conserved domain analysis of representative proteins from each clade found that clade 1 consisted of pullulanases, clade 2 of α -1,4-acting amylases and clade 3 of cyclomaltodextrin (CMD)-like amylases (b). AmyAc_bac_CMD_like_2, α -amylase catalytic domain found in bacterial cyclomaltodextrinases; Aamy_C, maltogenic C-terminal amylase domain; PUD, pullulanase-associated domain; CBM, carbohydrate-binding module; pullul_strch, pullulanase type α -1,6-glycosidase domain; AmyAc_bac_euk_AmyA, α -1,4 amylase catalytic domain found in bacterial and eukaryotic α -amylases; FN3, fibronectin type 3 domain. Domains not drawn to scale. Open in a new tab Fig. 2. Amylases frequently co-occur with amylase inhibitors in Streptomyces genomes. The evolutionary relationship between 295 Streptomyces species was inferred from whole-genome phylogeny, and the prevalence of amylases and amylase inhibitors is visualized using a colour scale (white indicates absence; darker colours indicate higher prevalence). Annotations shown in the legend (right panel), listed from top to bottom, correspond to the concentric circles around the phylogenetic tree from the innermost to the outermost. Pullulanases (innermost, red) and α -1,4-amylases (green) are found almost ubiquitously throughout the Streptomyces genus, whereas only a subset of species encodes cyclomaltodextrin-like amylases (blue). Genomes with higher numbers of amylases frequently coincide with those encoding amylase inhibitors, including acarbose (black), acarviostatin (dark grey), trestatin (grey) and tendamistat (purple). Branch support values (bootstrap), computed as transfer bootstrap expectation scores, are represented by coloured circles at internal nodes. Open in a new tab Previous studies have reported BGCs encoding amylase inhibitory molecules acarbose, acarviostatin and trestatin located adjacent to amylase gene arrays [ 13 , 15 , 16 ]. To test if strains encoding amylase inhibitors generally possess more amylase genes, an hmmsearch [ 19 ] with a tendamistat HMM profile was used to analyse the genomes and found 38 species encoding tendamistat homologs. Using GATOR [ 20 ], 12 species were identified having BGCs encoding acarbose, acarviostatin or trestatin ( Figs2 3 ). These clusters were typically located adjacent to multiple amylase genes (numbers ranging from one to four, Fig. 3 ). Acarbose, acarviostatin and trestatin are all pseudooligosaccharides, but they vary in the number of C 7 N-aminocyclitol units they contain [ 16 ]. It was found that all acarbose-like gene clusters harboured a common set of genes required for C 7 N-aminocyclitol synthesis (see the ‘Methods’ section), but differed in the domain architecture of a pseudoglycosyltransferase responsible for the non-glycosidic C-N bonding in Fig. 3. Acarbose-like BGCs possess between one and four amylases in their gene neighbourhoods. GATOR identified 15 genomes (12 from the representative list of 295 genomes plus the acarviostatin, acarbose and trestatin BGCs already described in literature). Core genes required for the synthesis of aminocyclitol are shown in blue, amylases are shown in orange and other genes in green. Tendamistat-like peptide inhibitors are shown in orange and marked with a red star. Gene-level conservation between clusters is shown by colour intensity (more intense colour for a higher level of conservation). Genes with red edges indicate contig edges. CMD, cyclomaltodextrinase. Open in a new tab pseudooligosaccharides [ 33 ]. Further variation was seen in their neighbouring genes including homologs of the acarbose importer AcbFGH [ 34 , 35 ], which was absent in some clusters ( Fig. 3 ). A Mann–Whitney U test revealed that genomes encoding acarbose BGCs had significantly higher numbers of amylases and a large effect size compared to those without such clusters ( r =0.78; P =4.95×10 −7 ) ( Fig. 4a ). A similar analysis found a significant correlation between higher number of amylases and genomes encoding tendamistat but with a smaller effect size ( r =0.27; P =2.93×10 −3 ) ( Fig. 4b ). A chi-square test further revealed a positive correlation between cyclomaltodextrin-like amylases and tendamistat with a small effect size (phi=0.16; P =0.009). No significant difference in genome size was found between strains encoding acarbose ( P =0.40) or tendamistat ( P =0.29) and those that did not. Fig. 4. Streptomyces species encoding amylase inhibitors possess more amylase genes. Violin-boxplots showing significant correlations between higher numbers of amylase per genome and genomes encoding acarbose BGCs (a) ( r =0.78; P =4.95×10 −7 ) or tendamistat (b) ( r =0.27; P =2.93×10 −3 ). Open in a new tab Previous studies in S. glaucescens and S. dimorphogenes pairing in silico analysis with mutational screening identified a catalytic pocket in pullulanase domain containing α -amylases where substitution of a histidine with alanine or asparagine was sufficient to confer resistance to acarbose [ 16 ]. To evaluate whether this might be a general feature of amylases associated with acarbose-like BGCs, a multiple sequence alignment (MSA) was constructed (File S5) and found seven amylases across six genomes with acarbose-like BGCs having a histidine to asparagine substitution at this catalytic site. Although 11 genomes lacking such BGCs also encoded a pullulanase with the histidine to asparagine substitution (Fig. S5), a two-sided Fisher’s exact test found it 30-fold enriched in species harbouring acarbose-like clusters ( P =5.56×10 −6 ). The substitution was only found in pullulanase domain-containing amylases. Interestingly, not all species encoding acarbose had an amylase with the described histidine to asparagine substitution (Fig. S1), suggesting that acarbose resistance may be achieved through other mechanisms. To find possible acarbose-resistant amylases, an MSA was constructed for each clade of amylases (pullulanases, α -1,4-amylases and cyclomaltodextrin-like amylases), identifying numerous amino acid motifs significantly enriched in species harbouring acarbose-like BGCs (Files S2–S4). In acarbose BGC-associated α -1,4-amylases, it was found that they possess several amino acid changes including a H/Y to G and a D to T substitution ( Fig. 5a ). Using AlphaFold3 [ 27 ] to model the structures of an acarbose-associated α -1,4-amylase ( WP_077796428.1 ) and a non-acarbose-associated amylase from the same clade ( WP_103530170.1 ), a noticeable difference in the architecture of the catalytic pocket [ 16,36 , 38 ] was found between the two groups ( Fig. 5b ). Importantly, using CB-DOCK2 [ 28 ] to model the binding of acarbose, it was found that the inhibitor is predicted to bind both amylases but in different orientations ( Fig. 5b ). To assess, in silico , the effect of the H235G and D293T substitutions in the acarbose-associated amylase, the amino acids were reverted to H and D, respectively, and it was found that this alone was enough to widen the catalytic pocket and change the binding configuration of acarbose similar to that predicted for the non-acarbose-associated amylase ( Fig. 5b ). A similar conversion was observed by making the reciprocal substitution in the non-acarbose-associated amylase (Fig. S2). Importantly, modelling of an α -1,4-amylase ( WP_150242730 ) from an acarbose BGC without a pullulanase having the H to N acarbose-resistance-substitution [ 16 ] also displayed an inhibitor-resistant acarbose binding phenotype that could be reversed through a single G238H substitution (Fig. S5). It was hypothesized that reorienting the binding of acarbose within the catalytic pocket might be a general way of acquiring acarbose resistance, and, therefore, the binding of acarbose to an acarbose-associated pullulanase ( WP_064455463 ) was modelled. It was found that acarbose was predicted to bind similarly (Fig. S3) to that determined for the putatively acarbose-resistant α -1,4-amylase, with the imino moiety of acarbose oriented away from the region of the binding pocket known to be instrumental in establishing acarbose resistance [ 16 ]. Notably, these predictions are consistent with previously reported structural modelling and docking analyses of acarbose and pullulanases [ 16 ]. The asparagine to histidine conversion appears to result in the reversion of this group of pullulanases from an inhibitor-resistant to an inhibitor-sensitive conformation [ 16 ], and the modelling predicts a reorientation of acarbose binding (Fig. S3). A lysine residue directly adjacent to this site is conserved in both pullulanases and α -1,4- amylases (File S5), and as described previously, the orientation of this residue’s side chain varies considerably between resistant and sensitive amylases [ 16 ]. Similar observations were found in α -1,4-amylases (Fig. S4), likely attributable to altered hydrogen bonding networks in the catalytic pocket ( Fig. 5 ). Fig. 5. Predicted binding sites of acarbose differ between amylases that are located in acarbose BGCs and amylases that are not. MSA shows an enrichment of GLY235 and THR293 (+) in alpha-1,4-amylases located in acarbose BGCs. Both residues are found within the conserved catalytic pocket close to the catalytic residue glutamate (*) (a). The catalytic pockets (residues highlighted in green) are shown for a putatively acarbose-sensitive amylase [Amy (S) – accession: WP_103530170 ] found in a genome without acarbose and a putatively resistant amylase [Amy (R) – accession: WP_077796428 ] found within an acarbose BGC [proteins marked by black squares in (a)]. Amy (R)* shows the structure of Amy (R), where GLY235 and THR293 have been converted to HIS and ASP, identical to the residues found at these positions in Amy (S). Hydrogen bonds between residues within the catalytic pocket are shown with dashed cyan lines, and hydrogen bonds between acarbose and the proteins are shown with dashed yellow lines (bottom panel). LYS213/208, previously described as involved in inferring acarbose resistance, is highlighted (b). Open in a new tab The acarbose and acarviostatin BGCs of S. glaucescens and S. coelicoflavus encode in addition to pseudooligosaccharides, also small proteinaceous amylase inhibitors, Gaim and Scaim [ 13 , 16 ]. These peptides contain the conserved amylase inhibitor domain (smart00783) also found in tendamistat. Tendamistat is known to inhibit eukaryotic amylases [ 39 , 40 ]. Little is known, however, about its activity against prokaryotic amylases, and to the authors’ knowledge, no studies have investigated potential resistance mechanisms to tendamistat or tendamistat-like peptides. GATOR identified three additional Streptomyces species that harbour acarbose BGCs encoding small peptides with amylase inhibitor domains (Fig. 3). To assess potential inhibition of Streptomyces amylases, interactions between Gaim and three pullulanases from Streptomyces sp . JHA26 (GCF_002005565.1) , two encoded within the acarbose BGC and one outside the cluster were modelled using ClusPro [ 29 ]. Both BGC-associated pullulanases contain the histidine to asparagine substitution known to confer resistance to acarbose [ 16 ], while the non-BGC-associated pullulanase does not. Gaim was predicted to bind in the α-amylase catalytic pocket of the non-BGC-associated pullulanase as well as one of the two BGC-associated pullulanases, resembling the resolved interaction between tendamistat and porcine pancreatic amylase ( Fig. 6 [ 40 ]). In contrast, no comparable binding was predicted for the second BGC-associated pullulanase. This appears to be due to an obstructed binding pocket formed by protruding side chains of MET78, GLY80 and LYS217, which sterically hinders Gaim binding (distance between protruding residues and Gaim<1 Å). Furthermore, a weak electrostatic interaction between LYS217 and ASP411 is predicted to promote a closed pocket formation that further prevents Gaim binding ( Fig. 6 ). Beyond these three pullulanases, Streptomyces sp. JHA26 encodes two alpha-1,4-amylases and one cyclomaltodextrin-like amylase. Gaim was predicted to bind all three of these enzymes (data not shown), suggesting that the Gaim occlusion phenotype predicted for WP_077796429.1 ( Fig. 6 ) is unique to this protein within Streptomyces sp . JHA26 . Fig. 6. Structure predictions and protein-protein docking analyses suggest obstructed binding of proteinaceous amylase inhibitor Gaim to an acarbose-associated pullulanase from Streptomyces sp. JHA26. Gaim is predicted to bind two pullulanases from Streptomyces sp. JHA26: one encoded within the acarbose BGC [Aca-1 (magenta); WP_077796430.1 ] and one located outside the BGC [non-Aca (cyan); WP_077796080.1 ]. Gaim (green) encoded within the acarbose BGC is predicted to bind in the starch binding pocket containing the catalytic aspartate and glutamate residues (blue), like the binding of tendamistat (green) to porcine pancreatic amylase (orange, Pig). Alignment of the second acarbose BGC-encoded pullulanase [Aca-2 (transparent yellow); WP_077796429.1 ] with the Gaim-bound non-acarbose-associated pullulanase reveals three protruding amino acid residues (red) that are predicted to obstruct Gaim binding. An overlay (Overlay) of the three Streptomyces amylases highlights these residues (MET78, GLY80 and LYS217), including weak electrostatic interactions between ASP411 and LYS217, which creates a closed pocket confirmation that sterically clashes with Gaim binding (Over+Gaim). Open in a new tab Discussion Streptomycetes are known to inhabit a wide range of ecological niches and exhibit extensive genome plasticity, particularly in their repertoires of BGCs and CAZymes [ 41 ]. In this study, it was found that Streptomyces species frequently harbour BGCs encoding amylase inhibitors, such as acarbose and tendamistat, alongside inhibitor-resistant amylases. It is known that acarbose not only inhibits amylases but may also function as a ‘carbophor’ for the producing organism [ 35 , 42 ]. In this role, acarbose is covalently linked to glucose or dextrins through the acarviosyltransferase AcbD and then transported into the cytoplasm through the ABC transporter AcbFGH-MsiK [ 34 , 35 ]. Notably, this work indicates that acarbose resistance is more likely achieved through reconfiguration of acarbose binding within the alpha-amylase catalytic pocket rather than exclusion of the inhibitor. This suggests that acarbose-resistant amylases may participate in the carbophor cycle, enabling acarbose-producing organisms to sequester oligosaccharides, which might otherwise be freely accessible to competing organisms, by covalently linking them to acarbose, thereby restricting uptake to organisms possessing the dedicated importer. This work further reveals substantial diversity in both the number and types of amylases found across the Streptomyces genus. It was found that amylases broadly cluster into three major clades that are homologous in their α -amylase domains but differ in their adjacent carbohydrate-binding domains. This likely confers substrate specificity towards starches of varying composition, such as those derived from different plant species or fungal sources. Importantly, these conclusions are based on bioinformatic predictions, whereby enzymatic function, substrate specificity and inhibitor interactions were inferred from sequence homology, domain architecture and previously characterized motifs, and remain to be experimentally validated. Despite these limitations, the atlas of amylases presented in this work may provide a useful resource for the targeted selection of enzymes for heterologous overexpression in industrial antibiotic production strains. As such, it may facilitate the optimization of sustainable antibiotic fermentations by deploying amylases with substrate specificities tailored to organic waste streams of different origins. Moreover, this approach may provide a useful source of bacterial strains that produce novel alpha-glucosidase inhibitors for therapeutic use in diabetes mellitus [ 43 ]. In the context of engineering starch-fuelled industrial cell factories, it is important to consider the potential repertoire of endogenous amylase inhibitors encoded by a given production strain. Amylase inhibitors are frequently found in Streptomyces bacteria, and care must be taken to ensure that any host strain does not produce inhibitors that act against a heterologously expressed amylase, as this could compromise manufacturing efficiency. Streptomyces hygroscopicus , for example, is used industrially for the large-scale production of a range of bioactive compounds, including rapamycin [ 44 ]. Improving the sustainability of fermentations involving this organism using organic waste substrates, such as bread or potato waste, would be of considerable environmental value [ 4 , 45 ]. However, as this analysis shows, S. hygroscopicus (GCF_001652895.1) encodes an acarbose BGC, and this may potentially interfere with amylase activity during fermentation. Strategies to overcome this challenge could be to either delete this cluster or, alternatively, overexpress an acarbose-resistant amylase. The latter approach may be more broadly applicable, as overexpressing genes often is easier than deleting them. Moreover, considering the ongoing antibiotic resistance crisis and the need to discover and industrially deploy new antibiotics, production pipelines should be readily adaptable to newly identified producer strains. Rather than repeatedly facing the challenging task of deleting gene clusters encoding amylase inhibitors in newly characterized strains, it may be more efficient to develop a small set of broadly applicable overexpression vectors. Furthermore, as this and previous work suggest, creating inhibitor-resistant amylases may be achieved through minimal sequence modifications, including single amino acid substitutions. It may hence be possible to develop a toolbox of inhibitor-resistant amylases with optimized substrate specificities towards starches from diverse sources, which could be readily integrated into a wide range of natural product-producing Streptomyces strains. In summary, this work highlights the coordinated evolution of amylases and amylase inhibitors in Streptomyces , revealing how starch utilization and inhibitor resistance are tightly integrated. These findings provide insight into how producing organisms balance nutrient acquisition with competitive exclusion and offer a framework for engineering inhibitor-resistant, starch-utilizing strains for sustainable antibiotic fermentations. Supplementary material Supplementary Material 1. mgen-12-01747-s001.pdf (1.6MB, pdf) DOI: 10.1099/mgen.0.001747 Acknowledgements We would like to thank Professor Iain S. Hunter, Dr John T. Munnoch and Dr Steve Kendrew (GSK) for helpful discussions. Abbreviations BGCs biosynthetic gene clusters CAZymes carbohydrate-active enzymes GH13 glycoside hydrolase category 13 MSA multiple sequence alignment ST sequence type Footnotes Funding: P.A.H. would also like to acknowledge funding from BBSRC (BB/Y007611/1; BB/T001038/1) and the Royal Academy of Engineering Research Chair Scheme for long-term personal research support (RCSRF2021\11\15). Author contributions: Conceptualization: A.C.L. and P.A.H. Data curation: A.C.L. Formal analysis: A.C.L. Funding acquisition: P.A.H. Investigation: A.C.L. Methodology: A.C.L. and P.A.H. Project administration: A.C.L. and P.A.H. Supervision: P.A.H. Validation: A.C.L. Visualization: A.C.L. Writing – original draft: A.C.L. Writing – review and editing: A.C.L., A.G. and P.A.H. All authors gave final approval for publication and agreed to be held accountable for the work performed therein. Ethical statement: This work did not require ethical approval from a human subject or animal welfare committee. Contributor Information Andreas C. Lawaetz, Email: andreas.lawaetz@strath.ac.uk. Alexa Gannon, Email: alexa.gannon@strath.ac.uk. Paul A. Hoskisson, Email: paul.hoskisson@strath.ac.uk. References 1. Hoskisson PA, Seipke RF. Cryptic or Silent? The known unknowns, unknown knowns, and unknown unknowns of secondary metabolism. mBio. 2020;11:e02642-20. doi: 10.1128/mBio.02642-20. [ DOI ] [ PMC free article ] [ PubMed ] [ Google Scholar ] 2. Lenzen M, Malik A, Li M, Fry J, Weisz H, et al. The environmental footprint of health care: a global assessment. Lancet Planet Health. 2020;4:e271–e279. doi: 10.1016/S2542-5196(20)30121-2. [ DOI ] [ PubMed ] [ Google Scholar ] 3. Fritsch C, Staebler A, Happel A, Cubero Márquez M, Aguiló-Aguayo I, et al. Processing, valorization and application of bio-waste derived compounds from potato, tomato, olive and cereals: a review. Sustainability . 2017;9:1492. doi: 10.3390/su9081492. [ DOI ] [ Google Scholar ] 4. Kumar V, Brancoli P, Narisetty V, Wallace S, Charalampopoulos D, et al. Bread waste - A potential feedstock for sustainable circular biorefineries. Bioresour Technol. 2023;369:128449. doi: 10.1016/j.biortech.2022.128449. [ DOI ] [ PubMed ] [ Google Scholar ] 5. Tomasik P, Horton D. Enzymatic conversions of starch. Adv Carbohydr Chem Biochem. 2012;68:59–436. doi: 10.1016/B978-0-12-396523-3.00001-4. [ DOI ] [ PubMed ] [ Google Scholar ] 6. Fernández-Martínez LT, Hoskisson PA. Expanding, integrating, sensing and responding: the role of primary metabolism in specialised metabolite production. Curr Opin Microbiol . 2019;51:16–21. doi: 10.1016/j.mib.2019.03.006. [ DOI ] [ PubMed ] [ Google Scholar ] 7. Otani H, Udwary DW, Mouncey NJ. Comparative and pangenomic analysis of the genus Streptomyces. Sci Rep . 2022;12:18909. doi: 10.1038/s41598-022-21731-1. [ DOI ] [ PMC free article ] [ PubMed ] [ Google Scholar ] 8. Ríos-Fernández P, Caicedo-Montoya C, Ríos-Estepa R. Genomic Diversity of Streptomyces clavuligerus: implications for clavulanic acid biosynthesis and industrial hyperproduction. Int J Mol Sci . 2024;25:10992. doi: 10.3390/ijms252010992. [ DOI ] [ PMC free article ] [ PubMed ] [ Google Scholar ] 9. Vela Gurovic MS, Díaz ML, Gallo CA, Dietrich J. Phylogenomics, CAZyome and core secondary metabolome of Streptomyces albus species. Mol Genet Genomics . 2021;296:1299–1311. doi: 10.1007/s00438-021-01823-9. [ DOI ] [ PubMed ] [ Google Scholar ] 10. Drula E, Garron M-L, Dogan S, Lombard V, Henrissat B, et al. The carbohydrate-active enzyme database: functions and literature. Nucleic Acids Res . 2022;50:D571–D577. doi: 10.1093/nar/gkab1045. [ DOI ] [ PMC free article ] [ PubMed ] [ Google Scholar ] 11. Zheng J, Ge Q, Yan Y, Zhang X, Huang L, et al. dbCAN3: automated carbohydrate-active enzyme and substrate annotation. Nucleic Acids Res. 2023;51:W115–W121. doi: 10.1093/nar/gkad328. [ DOI ] [ PMC free article ] [ PubMed ] [ Google Scholar ] 12. Aschauer H, Vértesy L, Nesemann G, Braunitzer G. The primary structure of the alpha-amylase inhibitor Hoe 467A from Streptomyces tendae 4158. A new class of inhibitors. Hoppe Seylers Z Physiol Chem . 1983;364:1347–1356. [ PubMed ] [ Google Scholar ] 13. Guo X, Geng P, Bai F, Bai G, Sun T, et al. Draft genome sequence of Streptomyces coelicoflavus ZG0656 reveals the putative biosynthetic gene cluster of acarviostatin family α-amylase inhibitors. Lett Appl Microbiol. 2012;55:162–169. doi: 10.1111/j.1472-765X.2012.03274.x. [ DOI ] [ PubMed ] [ Google Scholar ] 14. Rehm S, Han S, Hassani I, Sokocevic A, Jonker HRA, et al. The high resolution NMR structure of parvulustat (Z-2685) from Streptomyces parvulus FH-1641: comparison with tendamistat from Streptomyces tendae 4158. Chembiochem . 2009;10:119–127. doi: 10.1002/cbic.200800547. [ DOI ] [ PubMed ] [ Google Scholar ] 15. Rockser Y, Wehmeier UF. The gac-gene cluster for the production of acarbose from Streptomyces glaucescens GLA.O: identification, isolation and characterization. J Biotechnol . 2009;140:114–123. doi: 10.1016/j.jbiotec.2008.10.016. [ DOI ] [ PubMed ] [ Google Scholar ] 16. Tanoeyadi S, Tsunoda T, Ito T, Philmus B, Mahmud T. Acarbose May function as a competitive exclusion agent for the producing bacteria. ACS Chem Biol . 2023;18:367–376. doi: 10.1021/acschembio.2c00795. [ DOI ] [ PMC free article ] [ PubMed ] [ Google Scholar ] 17. Zheng J, Ge Q, Yan Y, Zhang X, Huang L, et al. dbCAN3: automated carbohydrate-active enzyme and substrate annotation. Nucleic Acids Res. 2023;51:W115–W121. doi: 10.1093/nar/gkad328. [ DOI ] [ PMC free article ] [ PubMed ] [ Google Scholar ] 18. Teufel F, Almagro Armenteros JJ, Johansen AR, Gíslason MH, Pihl SI, et al. SignalP 6.0 predicts all five types of signal peptides using protein language models. Nat Biotechnol . 2022;40:1023–1025. doi: 10.1038/s41587-021-01156-3. [ DOI ] [ PMC free article ] [ PubMed ] [ Google Scholar ] 19. Eddy SR. Accelerated profile HMM searches. PLoS Comput Biol . 2011;7:e1002195. doi: 10.1371/journal.pcbi.1002195. [ DOI ] [ PMC free article ] [ PubMed ] [ Google Scholar ] 20. Cediel-Becerra JDD, Cumsille A, Guerra S, Ding Y, de Crécy-Lagard V, et al. Targeted genome mining with GATOR-GC maps the evolutionary landscape of biosynthetic diversity. Nucleic Acids Res. 2025;53:gkaf606. doi: 10.1093/nar/gkaf606. [ DOI ] [ PMC free article ] [ PubMed ] [ Google Scholar ] 21. Kiepas AB. University of Strathclyde; 2025. Resolving Streptomyces taxonomy: conflicts between whole-genome, core gene, MLST and 16S classifications. [ Google Scholar ] 22. Jolley KA, Bray JE, Maiden MCJ. Open-access bacterial population genomics: BIGSdb software, the PubMLST.org website and their applications. Wellcome Open Res. 2018;3:124. doi: 10.12688/wellcomeopenres.14826.1. [ DOI ] [ PMC free article ] [ PubMed ] [ Google Scholar ] 23. Madeira F, Madhusoodanan N, Lee J, Eusebi A, Niewielska A, et al. The EMBL-EBI job dispatcher sequence analysis tools framework in 2024. Nucleic Acids Res. 2024;52:W521–W525. doi: 10.1093/nar/gkae241. [ DOI ] [ PMC free article ] [ PubMed ] [ Google Scholar ] 24. Kozlov AM, Darriba D, Flouri T, Morel B, Stamatakis A. RAxML-NG: a fast, scalable and user-friendly tool for maximum likelihood phylogenetic inference. Bioinformatics. 2019;35:4453–4455. doi: 10.1093/bioinformatics/btz305. [ DOI ] [ PMC free article ] [ PubMed ] [ Google Scholar ] 25. Wang J, Chitsaz F, Derbyshire MK, Gonzales NR, Gwadz M, et al. The conserved domain database in 2023. Nucleic Acids Res. 2023;51:D384–D388. doi: 10.1093/nar/gkac1096. [ DOI ] [ PMC free article ] [ PubMed ] [ Google Scholar ] 26. Waterhouse AM, Procter JB, Martin DMA, Clamp M, Barton GJ. Jalview Version 2--a multiple sequence alignment editor and analysis workbench. Bioinformatics . 2009;25:1189–1191. doi: 10.1093/bioinformatics/btp033. [ DOI ] [ PMC free article ] [ PubMed ] [ Google Scholar ] 27. Jumper J, Evans R, Pritzel A, Green T, Figurnov M, et al. Highly accurate protein structure prediction with AlphaFold. Nature . 2021;596:583–589. doi: 10.1038/s41586-021-03819-2. [ DOI ] [ PMC free article ] [ PubMed ] [ Google Scholar ] 28. Liu Y, Yang X, Gan J, Chen S, Xiao Z-X, et al. CB-Dock2: improved protein–ligand blind docking by integrating cavity detection, docking and homologous template fitting. Nucleic Acids Res. 2022;50:W159–W164. doi: 10.1093/nar/gkac394. [ DOI ] [ PMC free article ] [ PubMed ] [ Google Scholar ] 29. Kozakov D, Hall DR, Xia B, Porter KA, Padhorny D, et al. The ClusPro web server for protein-protein docking. Nat Protoc . 2017;12:255–278. doi: 10.1038/nprot.2016.169. [ DOI ] [ PMC free article ] [ PubMed ] [ Google Scholar ] 30. Preiss J. Glycogen: biosynthesis and regulation. EcoSal Plus. 2014;6 doi: 10.1128/ecosalplus.esp-0015-2014. [ DOI ] [ PubMed ] [ Google Scholar ] 31. Bobek J, Šmídová K, Čihák M. A waking review: old and novel insights into the spore germination in Streptomyces. Front Microbiol. 2017;8:2205. doi: 10.3389/fmicb.2017.02205. [ DOI ] [ PMC free article ] [ PubMed ] [ Google Scholar ] 32. Schumacher MA, Wörmann ME, Henderson M, Salinas R, Latoscha A, et al. Allosteric regulation of glycogen breakdown by the second messenger cyclic di-GMP. Nat Commun. 2022;13:5834. doi: 10.1038/s41467-022-33537-w. [ DOI ] [ PMC free article ] [ PubMed ] [ Google Scholar ] 33. Tsunoda T, Samadi A, Burade S, Mahmud T. Complete biosynthetic pathway to the antidiabetic drug acarbose. Nat Commun . 2022;13:3455. doi: 10.1038/s41467-022-31232-4. [ DOI ] [ PMC free article ] [ PubMed ] [ Google Scholar ] 34. Brunkhorst C, Wehmeier UF, Piepersberg W, Schneider E. The acbH gene of Actinoplanes sp. encodes a solute receptor with binding activities for acarbose and longer homologs. Res Microbiol. 2005;156:322–327. doi: 10.1016/j.resmic.2004.10.016. [ DOI ] [ PubMed ] [ Google Scholar ] 35. Wehmeier UF, Piepersberg W. Biotechnology and molecular biology of the alpha-glucosidase inhibitor acarbose. Appl Microbiol Biotechnol. 2004;63:613–625. doi: 10.1007/s00253-003-1477-2. [ DOI ] [ PubMed ] [ Google Scholar ] 36. Buisson G, Duée E, Haser R, Payan F. Three dimensional structure of porcine pancreatic alpha-amylase at 2.9 A resolution. Role of calcium in structure and activity. EMBO J . 1987;6:3909–3916. doi: 10.1002/j.1460-2075.1987.tb02731.x. [ DOI ] [ PMC free article ] [ PubMed ] [ Google Scholar ] 37. Matsuura Y, Kusunoki M, Harada W, Kakudo M. Structure and possible catalytic residues of Taka-amylase A. J Biochem . 1984;95:697–702. doi: 10.1093/oxfordjournals.jbchem.a134659. [ DOI ] [ PubMed ] [ Google Scholar ] 38. Uitdehaag JC, Mosi R, Kalk KH, van der Veen BA, Dijkhuizen L, et al. X-ray structures along the reaction pathway of cyclodextrin glycosyltransferase elucidate catalysis in the alpha-amylase family. Nat Struct Biol . 1999;6:432–436. doi: 10.1038/8235. [ DOI ] [ PubMed ] [ Google Scholar ] 39. Vertesy L, Oeding V, Bender R, Zepf K, Nesemann G. Tendamistat (HOE 467), a tight-binding alpha-amylase inhibitor from Streptomyces tendae 4158. isolation, biochemical properties. Eur J Biochem . 1984;141:505–512. doi: 10.1111/j.1432-1033.1984.tb08221.x. [ DOI ] [ PubMed ] [ Google Scholar ] 40. Wiegand G, Epp O, Huber R. The crystal structure of porcine pancreatic alpha-amylase in complex with the microbial inhibitor Tendamistat. J Mol Biol. 1995;247:99–110. doi: 10.1006/jmbi.1994.0125. [ DOI ] [ PubMed ] [ Google Scholar ] 41. Norte DM, Avitia-Dominguez LA, Rozen DE. Evolution and ecology of Streptomyces. Annu Rev Microbiol. 2025;79:383–403. doi: 10.1146/annurev-micro-051524-032254. [ DOI ] [ PubMed ] [ Google Scholar ] 42. Hemker M, Stratmann A, Goeke K, Schröder W, Lenz J, et al. Identification, cloning, expression, and characterization of the extracellular acarbose-modifying glycosyltransferase, AcbD, from Actinoplanes sp. Strain SE50. J Bacteriol . 2001;183:4484–4492. doi: 10.1128/JB.183.15.4484-4492.2001. [ DOI ] [ PMC free article ] [ PubMed ] [ Google Scholar ] 43. Moelands SV, Lucassen PL, Akkermans RP, De Grauw WJ, Van de Laar FA. Alpha-glucosidase inhibitors for prevention or delay of type 2 diabetes mellitus and its associated complications in people at increased risk of developing type 2 diabetes mellitus. Cochrane Database Syst Rev . 2018;12:CD005061. doi: 10.1002/14651858.CD005061.pub3. [ DOI ] [ PMC free article ] [ PubMed ] [ Google Scholar ] 44. Park SR, Yoo YJ, Ban Y-H, Yoon YJ. Biosynthesis of rapamycin and its regulation: past achievements and recent progress. J Antibiot . 2010;63:434–441. doi: 10.1038/ja.2010.71. [ DOI ] [ PubMed ] [ Google Scholar ] 45. Ribeiro AR, Silva EL. Potato waste as feedstock to produce biohydrogen and organic acids: a comparison of acid and alkaline pretreatments using response surface methodology. J Environ Manage. 2022;323:116308. doi: 10.1016/j.jenvman.2022.116308. [ DOI ] [ PubMed ] [ Google Scholar ] Associated Data This section collects any data citations, data availability statements, or supplementary materials included in this article. Supplementary Materials Supplementary Material 1. mgen-12-01747-s001.pdf (1.6MB, pdf) DOI: 10.1099/mgen.0.001747 Articles from Microbial Genomics are provided here courtesy of Microbiology Society ACTIONS View on publisher site PDF (7.3 MB) Cite Collections Permalink RESOURCES Similar articles Cited by other articles Links to NCBI Databases Follow NCBI NCBI on X (formerly known as Twitter) NCBI on Facebook NCBI on LinkedIn NCBI on GitHub NCBI RSS feed Connect with NLM NLM on X (formerly known as Twitter) NLM on Facebook NLM on YouTube National Library of Medicine 8600 Rockville Pike Bethesda, MD 20894 Web Policies FOIA HHS Vulnerability Disclosure Help Accessibility Careers NLM NIH HHS USA.gov Back to Top

Related documents

• Gapped BLAST and PSI-BLAST: a new generation of protein database search programs#6260
• Draft genome sequence of blaIMP-1-carrying Phytobacter ursingii OUH-01, isolated from the feces of an acute lymphoblastic leukemia patient#348397
• Sparsomycin inhibits translation through a conserved mechanism across all domains of life#922972
• Searching the Druggable Genome using Large Language Models#484907
• Jalview Version 2—a multiple sequence alignment editor and analysis workbench#6502