Machine Learning Guided Rational Design of a Non-Heme Iron-Based Lysine Dioxygenase Improves its Total Turnover Number

Highly selective C-H functionalization remains an ongoing challenge in organic synthetic methodologies. Biocatalysts are robust tools for achieving these difficult chemical transformations. Biocatalyst engineering has often required directed evolution or structure-based rational design campaigns to improve their activities. In recent years, machine learning has been integrated into these workflows to improve the discovery of beneficial enzyme variants. In this work, we combine a structure-based self-supervised machine learning framework, MutComputeX, with classical molecular dynamics simulations to down select mutations for rational design of a non-heme iron-dependent lysine dioxygenase, LDO. This approach consistently resulted in functional LDO mutants and circumvents the need for extensive study of mutational activity before-hand. Our rationally designed single mutants purified with up to 2-fold higher expression yields than WT and displayed higher total turnover numbers (TTN). Combining five such single mutations into a pentamutant variant, LPNYI LDO, leads to a 40 % improvement in the TTN (218±3) as compared to WT LDO (TTN=160±2). Overall, this work offers a low-barrier approach for those seeking to synergize machine learning algorithms with pre-existing protein engineering strategies.

Machine learning guided rational design of a non-heme iron-based lysine dioxygenase improves its total turnover number

Highly selective C-H functionalization remains an ongoing challenge in organic synthetic methodologies. Biocatalysts are robust tools for achieving these difficult chemical transformations. Biocatalyst engineering has often required directed evolution or structure-based rational design campaigns to improve their activities. In recent years, machine learning has been integrated into these workflows to improve the discovery of beneficial enzyme variants. In this work, we combine a structure-based machine-learning algorithm with classical molecular dynamics simulations to down select mutations for rational design of a non-heme iron-dependent lysine dioxygenase, LDO. This approach consistently resulted in functional LDO mutants and circumvents the need for extensive study of mutational activity before-hand. Our rationally designed single mutants purified with up to 2-fold higher yields than WT and displayed higher total turnover numbers (TTN). Combining five such single mutations into a pentamutant variant, LPNYI LDO, leads to a 40% improvement in the TTN (218±3) as compared to WT LDO (TTN = 160±2). Overall, this work offers a low-barrier approach for those seeking to synergize machine learning algorithms with pre-existing protein engineering strategies.

Metabolic and proteomic signatures of type 2 diabetes subtypes in an Arab population

Type 2 diabetes (T2D) has a heterogeneous etiology influencing its progression, treatment, and complications. A data driven cluster analysis in European individuals with T2D previously identified four subtypes: severe insulin deficient (SIDD), severe insulin resistant (SIRD), mild obesity-related (MOD), and mild age-related (MARD) diabetes. Here, the clustering approach was applied to individuals with T2D from the Qatar Biobank and validated in an independent set. Cluster-specific signatures of circulating metabolites and proteins were established, revealing subtype-specific molecular mechanisms, including activation of the complement system with features of autoimmune diabetes and reduced 1,5-anhydroglucitol in SIDD, impaired insulin signaling in SIRD, and elevated leptin and fatty acid binding protein levels in MOD. The MARD cluster was the healthiest with metabolomic and proteomic profiles most similar to the controls. We have translated the T2D subtypes to an Arab population and identified distinct molecular signatures to further our understanding of the etiology of these subtypes.

Metabolomic Profiling Reveals That 5-Hydroxylysine and 1-Methylnicotinamide Are Metabolic Indicators of Keloid Severity

Background: Keloid is a skin fibroproliferative disease with unknown pathogenesis. Metabolomics provides a new perspective for revealing biomarkers related to metabolites and their metabolic mechanisms. Method: Metabolomics and transcriptomics were used for data analysis. Quality control of the data was performed to standardize the data. Principal component analysis (PCA), PLS-DA, OPLS-DA, univariate analysis, CIBERSORT, neural network model, and machine learning correlation analysis were used to calculate differential metabolites. The molecular mechanisms of characteristic metabolites and differentially expressed genes were identified through enrichment analysis and topological analysis. Result: Compared with normal tissue, lipids have a tendency to decrease in keloids, while peptides have a tendency to increase in keloids. Significantly different metabolites between the two groups were identified by random forest analysis, including 1-methylnicotinamide, 4-hydroxyproline, 5-hydroxylysine, and l-prolinamide. The metabolic pathways which play important roles in the pathogenesis of keloids included arachidonic acid metabolism and d-arginine and d-ornithine metabolism. Metabolomic profiling reveals that 5-hydroxylysine and 1-methylnicotinamide are metabolic indicators of keloid severity. The high-risk early warning index for 5-hydroxylysine is 4 × 10⁸-6.3×10⁸ (p = 0.0008), and the high-risk predictive index for 1-methylnicotinamide is 0.95 × 10⁷-1.6×10⁷ (p = 0.0022). Conclusion: This study was the first to reveal the metabolome profile and transcriptome of keloids. Differential metabolites and metabolic pathways were calculated by machine learning. Metabolomic profiling reveals that 5-hydroxylysine and 1-methylnicotinamide may be metabolic indicators of keloid severity.

Transcriptomic Analysis Reveals a Sex-Dimorphic Influence of GAT-2 on Murine Liver Function

Accumulating evidence shows that the γ-amino butyric acid (GABA)ergic system affects the functions of different organs, and liver is one of the most sex-dimorphic organs in animals. However, whether and how the GABAergic system influences liver function in a sex-specific manner at the intrinsic molecular level remains elusive. In this study, firstly, we find that the levels of GABA are significantly increased in the livers of female mice with GABA transporter (GAT)-2 deficiency (KO) whereas it only slightly increased in male GAT-2 KO mice. Apart from the amino acid profiles, the expressions of toll-like receptors (TLRs) also differ in the livers of female and male KO mice. Moreover, RNA-seq results show 2,227 differentially expressed genes (DEGs) in which 1,030 are upregulated whereas 1,197 that are downregulated in the livers of female KO mice. Notably, oxidative phosphorylation, non-alcoholic fatty liver disease, Huntington's disease, and peroxisome proliferator-activated receptor (PPAR) signaling pathways are highly enriched by GAT-2 deficiency, indicating that these pathways probably meditate the effects of GAT-2 on female liver functions, on the other hand, only 1,233 DEGs, including 474 are upregulated and 759 are downregulated in the livers of male KO mice. Interestingly, retinol metabolism, PPAR signaling pathway, and tuberculosis pathways are substantially enriched by GAT-2 deficiency, suggesting that these pathways may be responsible for the effects of GAT-2 on male liver functions. Collectively, our results reveal the sex-dimorphic effects of GAT-2 in guiding liver functions, and we propose that targeting the GABAergic system (e.g., GATs) in a sex-specific manner could provide previously unidentified therapeutic opportunities for liver diseases.

Chemical Synthesis and Biological Evaluations of Adiponectin Collagenous Domain Glycoforms

The homogeneously glycosylated 76-amino acid adiponectin collagenous domains (ACDs) with all of the possible 15 glycoforms have been chemically and individually synthesized using stereoselective glycan synthesis and chemical peptide ligation. The following biological and pharmacological studies enabled correlating glycan pattern to function in the inhibition of cancer cell growth as well as the regulation of systemic energy metabolism. In particular, hAdn-WM6877 was tested in detail with different mouse models and it exhibited promising in vivo antitumor, insulin sensitizing, and hepatoprotective activities. Our studies demonstrated the possibility of using synthetic glycopeptides as the adiponectin downsized mimetic for the development of novel therapeutics to treat diseases associated with deficient adiponectin.

Precursor-Guided Mining of Marine Sponge Metabolomes Lends Insight into Biosynthesis of Pyrrole-Imidazole Alkaloids

Pyrrole-imidazole alkaloids are natural products isolated from marine sponges, holobiont metazoans that are associated with symbiotic microbiomes. Pyrrole-imidazole alkaloids have attracted attention due to their chemical complexity and their favorable pharmacological properties. However, insights into how these molecules are biosynthesized within the sponge holobionts are scarce. Here, we provide a multiomic profiling of the microbiome and metabolomic architectures of three sponge genera that are prolific producers of pyrrole-imidazole alkaloids. Using a retrobiosynthetic scheme as a guide, we mine the metabolomes of these sponges to detect intermediates in pyrrole-imidazole alkaloid biosynthesis. Our findings reveal that the nonproteinogenic amino acid homoarginine is a critical branch point that connects primary metabolite lysine to the production of pyrrole-imidazole alkaloids. These insights are derived from the polar metabolomes of these sponges which additionally reveal the presence of zwitterionic betaines that may serve important ecological roles in marine habitats. We also establish that metabolomic richness does not correlate with microbial diversity of the sponge holobiont for neither the polar nor the nonpolar metabolomes. Our findings now provide the biochemical foundation for genomic interrogation of the sponge holobiont to establish biogenetic routes for pyrrole-imidazole alkaloid production.

Comparison of Targeted (HPLC) and Nontargeted (GC-MS and NMR) Approaches for the Detection of Undeclared Addition of Protein Hydrolysates in Turkey Breast Muscle

The adulteration of fresh turkey meat by the undeclared addition of protein hydrolysates is of interest for fraudsters due to the increase of the economic gain by substituting meat with low cost ingredients. The aim of this study was to compare the suitability of three different analytical techniques such as GC-MS and ¹H-NMR with HPLC-UV/VIS as a targeted method, for the detection of with protein hydrolysates adulterated turkey meat. For this, turkey breast muscles were treated with different plant- (e.g., wheat) and animal-based (e.g., gelatin, casein) protein hydrolysates with different hydrolyzation degrees (15–53%: partial; 100%: total), which were produced by enzymatic and acidic hydrolysis. A water- and a nontreated sample (REF) served as controls. The data analyses revealed that the hydrolysate-treated samples had significantly higher levels of amino acids (e.g., leucine, phenylalanine, lysine) compared with REF observed with all three techniques concordantly. Furthermore, the nontargeted metabolic profiling (GC-MS and NMR) showed that sugars (glucose, maltose) and/or by-products (build and released during acidic hydrolyses, e.g., levulinic acid) could be used for the differentiation between control and hydrolysates (type, degrees). The combination of amino acid profiling and additional compounds gives stronger evidence for the detection and classification of adulteration in turkey breast meat.

Modulation of receptor binding to collagen by glycosylated 5-hydroxylysine: Chemical biology approaches made feasible by Carpino's Fmoc group

The creation of the 9-fluorenylmethoxycarbonyl (Fmoc) group by the Carpino laboratory facilitated the synthesis of peptides containing acid-sensitive groups, such as O-linked glycosides. To fully investigative collagen biochemistry, one needs to assemble peptides that possess glycosylated 5-hydroxylysine (Hyl). A convenient method for the synthesis of Fmoc-Hyl(ε-tert-butyloxycarbonyl (Boc),O-tert-butyldimethylsilyl (TBDMS)) and efficient methods for the synthesis of Fmoc-Hyl[ε-Boc,O-(2,3,4,6-tetra-O-acetyl-β-D-galactopyranosyl)] have been developed. Glycosylated Fmoc-Hyl derivatives were used to construct a series of types I-IV collagen-model triple-helical peptides (THPs) that incorporated known or proposed receptor binding sites. Glycosylation of Hyl was found to strongly down-regulate the binding of CD44 and the α3β1 integrin to collagen, while the impact on α2β1 integrin binding was more modest. Molecular modeling of integrin binding indicated that Hyl glycosylation directly impacted the association between the α3β1 integrin metal ion-dependent adhesion site (MIDAS) and the receptor binding site within type IV collagen. The Fmoc solid-phase strategy ultimately allowed for chemical biology approaches to be utilized to study tumor cell interactions with glycosylated collagen sequences and document the modulation of receptor interactions by Hyl posttranslational modification.

In silico prediction reveals the existence of potential bioactive neuropeptides produced by the human gut microbiota

This work reports on a large-scale potential neuropeptide activity screening in human gut microbiomes deposited in public databases. In our experimental approach, the sequences of the bioactive peptides collected in the MAHMI database, mainly predicted as immunomodulatory or antitumoral, were crossed with those of the neuroactive/digestive peptides. From 91,325,790 potential bioactive peptides, only 581 returned a match when crossed against the 5949 neuroactive peptides from the NeuroPep database and the 15 digestive hormones. Relevant bacterial taxa, such as Ruminococcus sp., Clostridium sp. were found among the main producers of the matching sequences, and many of the matches corresponded to adiponectin and the hormone produced by adipocites, which is involved in glucose homeostasis. These results show, for the first time, the presence of potentially bioactive peptides produced by gut microbiota members over the nervous cells, most notably, peptides with already predicted immunomodulatory or anti-inflammatory activity. Classical (Lactobacillus sp.) and next-generation (Faecalibacterium sp.) probiotics are shown to produce these peptides, which are proposed as a potential mechanism of action of psychobiotics. Our previous experimental results showed that many of these peptides were active when incubated with immune cells, such as dendritic cells, so their effect over the nervous system innervating the gut mucosa holds significant potential and should be explored.

Discovery of Lysine Hydroxylases in the Clavaminic Acid Synthase-Like Superfamily for Efficient Hydroxylysine Bioproduction

Hydroxylation via C-H bond activation in the absence of any harmful oxidizing reagents is technically difficult in modern chemistry. In this work, we attempted to generate pharmaceutically important hydroxylysine from readily available l-lysine with l-lysine hydroxylases from diverse microorganisms. Clavaminic acid synthase-like superfamily gene mining and phylogenetic analysis led to the discovery of six biocatalysts, namely two l-lysine 3S-hydroxylases and four l-lysine 4R-hydroxylases, the latter of which partially matched known hydroxylases. Subsequent characterization of these hydroxylases revealed their capacity for regio- and stereoselective hydroxylation into either C-3 or C-4 positions of l-lysine, yielding (2S,3S)-3-hydroxylysine and (2S,4R)-4-hydroxylysine, respectively. To determine if these factors had industrial application, we performed a preparative production of both hydroxylysines under optimized conditions. For this, recombinant l-lysine hydroxylase-expressing Escherichia coli cells were used as a biocatalyst for l-lysine bioconversion. In batch-scale reactions, 531 mM (86.1 g/liter) (2S,3S)-3-hydroxylysine was produced from 600 mM l-lysine with an 89% molar conversion after a 52-h reaction, and 265 mM (43.0 g/liter) (2S,4R)-4-hydroxylysine was produced from 300 mM l-lysine with a molar conversion of 88% after 24 h. This report demonstrates the highly efficient production of hydroxylysines using lysine hydroxylases, which may contribute to future industrial bioprocess technologies.IMPORTANCE The present study identified six l-lysine hydroxylases belonging to the 2-oxoglutarate-dependent dioxygenase superfamily, although some of them overlapped with known hydroxylases. While the substrate specificity of l-lysine hydroxylases was relatively narrow, we found that (2S,3S)-3-hydroxylysine was hydroxylated by 4R-hydroxylase and (2S,5R)-5-hydroxylysine was hydroxylated by both 3S- and 4R-hydroxylases. Moreover, the l-arginine hydroxylase VioC also hydroxylated l-lysine, albeit to a lesser extent. Further, we also demonstrated the bioconversion of l-lysine into (2S,3S)-3-hydroxylysine and (2S,4R)-4-hydroxylysine on a gram scale under optimized conditions. These findings provide new insights into biocatalytic l-lysine hydroxylation and thus have a great potential for use in manufacturing bioprocesses.

d-Glucose based syntheses of β-hydroxy derivatives of l-glutamic acid, l-glutamine, l-proline and a dihydroxy pyrrolidine alkaloid

The β-hydroxy derivatives of l-glutamic acid, l-glutamine and l-proline, useful for peptide/protein studies, were synthesized starting from d-glucose.

Facile synthesis of 5-hydroxy-L-lysine from D-galactose as a chiral-precursor

A concise synthesis of (2S,5R) and (2S,5S)-5-hydroxy-lysine was achieved by utilizing D-galactose as a chiral-precursor with stereo retention. This synthetic strategy showcased the potential of utilizing carbohydrates as starting materials to prepare amino acids. Using the diazido intermediate, the derived β-D-galactopyranosyl and α-D-glucopyranosyl-(1→2)-β-D-galactosyl moieties were synthesized.

Rare and unusual glycosylation of peptides and proteins

Glycosylation represents the most complex co- and post-translational modification of proteins. In addition to N- and O-glycans, almost all combinations, including the nature of the carbohydrate moiety and the amino-acid involved, but also the type of the chemical linkage, can be isolated from natural glycoconjugates. This diversity correlates with the importance and the variety of the biological processes (and consequently the diseases) glycosides are involved in. This review focuses on rare and unusual glycosylation of peptides and proteins.