Systematic Mining And Engineering Of Epothilone Glycosyltransferases And Their Applications In Glycosylation Modification

Macrolide compounds are key components of active natural products.In clinical trials,many macrolide drug candidates face pharmaceutical problems such as high cytotoxicity,poor water solubility,and low bioavailability.Epothilone is an example.Epothilones are a kind of 16-membered polyketides produced by the myxobacterium Sorangium cellulosum,with great antitumor application prospects.Epothilone A and B are the two major products,compared to epothilone A.epothilone B contains an additional methyl group at C-12,and shows stronger potency in antitumor activity.Ixabepilone,a semi-synthetic analogue of epothilone B has been approved by the Food and Drug Administration(FDA)for the treatment of cancer patients in 2008.and Utidelone,a Chinese independent research and development epothilone product,has come into the market recently in 2021,epothilone always attracts significant attentions in the treatment of tumors.Similar antitumor mechanism to that of Paclitaxel,epothilone is able to bind to the β-tubulin subunit of microtubule to inhibit microtubule depolymerization,thus preventing mitosis and resulting in apoptosis of cells.Epothilone is considered to be an updated alternative of paclitaxel.However,its clinical usage was limited by their significant neuro-and haematological toxicities due to compatibility of normal proliferating cells and cancer cells in the division process.Glycosylation improves the water solubility,stability and bioactivity of compounds,and plays important roles in pharmacological and pharmacokinetic properties.There are many precedents of optimized glycosylated macrolides,and the undesired toxicity of epothilone might be decreased by glycosylation.providing a broader prospect for the practical application.Epothilone glycosylation is mediated by glycosyltransferases(GTs),especially the promiscuous GTs for the glycosylation of bioactive molecules in secondary metabolism.In this thesis.we systematically mined promiscuous GTs and performed epothilone glycosylation,a star drug molecule with high application values but restricted clinical application,which are significant for enriching chemical diversity and optimizing candidate drugs of epothilone.GTs,as natural biocatalysts,can catalyze the transfer of the glycosyl moieties from activated sugar donors to proteins,nucleic acids,and secondary metabolites.GTs constitute a superfamily,and are classified into 114 families(up to May 20,2021),whereas the number of GTs with resolved structures are still few.Based on the current structure types,GT-A and GT-B folds are the two mainly predominate topologies.The GT-A fold possesses a central β-sheet surrounded by α-helices and two tightly abutting β/α/β domains,and a highly conserved DXD motif(ASP-x-ASP)coordinates with a divalent cation.The GT-B fold consists of two separate and flexibly linked Rossmann-like domains,and the catalytic cleft is located between these two domains,GT-B fold contains a relatively conserved donor binding motif at the C-terminal domain,whereas the N-terminal domain is highly divergent,exhibiting the greater plasticity of topology.Thus,the GT-B structure shows flexible substrates binding pattern compared to the GT-A fold.Many GT families adopt GT-B fold,such as GT1,GT9,GT10 and GT28.Remarkably,the GTs in GT1 family generally exhibited excellent glycosylation capacities toward numerous valued small bioactive molecules,meanwhile,the number of GTs with characterized and structure-resolved in GT1 family is largest among all GT families,which makes GT1 family become one of the most compelling families.However,compared to the large numbers of sequenced GTs,the number of characterized and structure-resolved GTs in GT1 family is still few.According to Carbohydrate-Active enZYmes(CAZy)database,GT1 family contains more than 20,000 members,distributed in nearly 8,000 different life domains.Although the GT1 family members display an uneven distribution,they cover almost all major life taxonomies,whereas only 1.6%of GTs were characterized,which implies that GTs in GT1 family have great potential for mining.In this study,all the GT1 family members deposited in the CAZy database were phylogenetically analyzed,and we found that divergent evolution of GTs showed typical species-dependent manner,and revealed many unexplored branches.Further,the substrates coverage of the characterized GTs from different taxonomic life domains were summarized,and bacteria-derived GTs were found to accept broader chemical skeletons as substrates spectrum compared to the other species(such as plants and animals).In addition,the analysis of the sequence length and 28 complete crystal structures of GTs in GT1 family revealed two important regions(Region A and Region B)located at the N-terminal domain,the two regions exhibited extreme variation in the sequences of different species,and played an important role in substrate binding or/and catalysis.The systematic analysis of substrates spectrum,sequences length and key motifs of GTs in GT1 family would contribute to the mining of unknown GTs,the rational engineering of known enzymes and high-efficiency enzymatic synthesis of important drug glycosides.Through systematic discussion of GTs in GT1 family,we elaborated bacteria-derived GTs exhibit a higher substrates promiscuity and wider substrates spectrum,and further applied to epothilone glycosylation.Phylogeny-guided approach was employed to explore efficient promiscuous GTs for epothilone glycosylation.Previously,YjiC was only the GT reported with the glycosylation activity for epothilone.We phylogenetically analyzed 8261 bacterial GT sequences from GT1 family available in the CAZy database,with YjiC as a probe.YjiC was located at a subbranch containing 161 GT members in Macrolide I branch,and the subbranch was defined as YjiC-subbranch.10 GTs were selected from either the YjiC-subbranch or non-YjiC-subbranches in the Macrolide I group to evaluate their glycosylation efficiencies on epothilone A.We found that 5 of the 7 selected YjiC-subbranch GTs were able to efficiently glycosylate epothilone A,and two were in low efficiency,while the 3 selected GTs from non-YjiC-subbranches in the Macrolide I branch had low or almost no epothilone glycosylation activity.Our studies showed that the GTs in YjiC-subbranch usually glycosylated epothilone A with higher or lower catalytic activity.It is worth noting that BsGT-1 exhibited excellent glycosylation activity on epothilone A,with 100%conversion rate after 2 h at 40℃ and displayed wider temperature tolerance,possessing more than 70%relative activity after 2 h within the range of 15 and 50℃.In terms of acid-alkali tolerance,the optimum pH value at 8.0 and having more than 60%relative activity at pH 6.0-9.0.Furthermore,we found that two low-active GTs in the YjiC-subbranch,BssGT and BamGT had high sequence identity with BsGT-1,and the sequence identity between BsGT-1 and BssGT even reached 88%.whereas they exhibited different glycosylation efficiency.Sequence alignment showed that the 66th and 77th amino acids in the docking models were close to the catalytic cavity and conserved in five high-active GTs.but variational in the two low-active GTs.Amino acids swapping between two low-active GTs(BssGT amd BamGT)and BsGT-1 demonstrated that the 66th and 77th residues played an important role in the catalytic efficiency of epothilone A.We phylogenetically analyzed bacteria-derived GTs in GT1 family for epothilone glycosylation,and preliminarily investigated amino acid residues that affected epothilone catalysis through swapping mutation.We mined the highest efficient GT BsGT-1 through phylogenetic analysis and only explored the transfer of glucose motif from UDP-Glc to,epothilone A,however,the catalytic activities of BsGT-1 toward epothilone B,a better antitumor compound and other sugar donors are still unknown.Our experiments further enunciated that the highly promiscuous BsGT-1 exhibited potent catalytic activity for epothilone B,and generated two novel epothilone B di-glycosides,while it only exhibited high conversion rate toward NDP-Glc.It is likely that BsGT-1 has catalytic promiscuity toward acceptors,but possesses stronger selectivity for sugar donors.Whereupon,based on guidance of crystal structure and molecular docking of BsGT-1,we performed glycosylation modification of epothilone B and found "hot spots" that could significantly alter the sugar donor selectivity of BsGT-1.Gratifyingly,we revealed pivotal amino acid residues that obviously affect the interaction between epothilone B and sugar donors through the alanine-scanning and site-directed saturated mutations of "hot spots" regions.In the C-terminal,the mutation of F231 of C2 loop region,M296 and Q318 of PSPG-like motif significantly affected the sugar donor preference of BsGT-1,and improved glycosylation activity toward UDP-GlcNAz,UDP-Gal and UDP-NAc,generating several novel epothilone B glycosides.In N-terminal,to our surprise,the scope of the N3 loop region is basically consistent with Region A that previously identified as an important GT region,and alariine-scanning screening of the N3 loop region showed the mutations in N3 loop region usually result in fluctuation of conversion rates of epothilone B glycosides.It is noteworthy that the I62A mutant in N3 loop region dramatically enhanced the conversion rates of two epothilone B di-glycosides and produced a novel epothilone B tri-glycoside.Sequence alignment and site-directed mutagenesis revealed that "hot spots" characterized in BsGT-1 also existed in the homologous GTs(YjiC,BaGT,BgGT and BpGT),which further confirmed the importance and universality of these"hot spots".In vitro cytotoxicity experiments of 7 epothilone B glycosides showed vast drop in cytotoxicity,and the antitumor activities were also decreased and varied depending on the properties of sugar moieties.In brief,the epothilone B galactose showed the highest antitumor activity(IC50 value,1.09 μM)among the epothilone B glycosides.The guidance of crystal structure and molecular docking for protein engineering of BsGT-1 would contribute to the enzyme tools application for drug derivatives development.The substrates promiscuity of BsGT-1 may apply to the glycosylation for wide range of drug molecules.Epothilone and rapamycin(RAP)are both macrolides,and RAP mainly binds to the FK506-binding protein 12(FKBP12)and the FKBP12-rapamycin binding(FRB)domain of the serine/threonine kinase(The Mammalian Target of RAP,mTOR),which affects the vital life activities of eukaryotic cells,thus exhibiting versatile bioactivity.The clinical usage of RAP is also compromised by its poor water solubility and low bioavailability,and glycosylation is an effective way to optimize the pharmacological properties and biological activities of RAP.We explored glycosylation activities of GTs in YjiC-subbranch toward RAP,BsGT-1 showed the highest catalytic activity with UDP-Glc as sugar donor,generating three RAP glucosides:rapamycin-40-O-β-D-glucoside(RG1),rapamycin-28-O-β-D-glucoside(RG2)and rapamycin-28,40-O-β-D-diglucoside.The glycosylation greatly improved the water solubility,eliminated the cytotoxicity,but attenuated the antifungal,antitumor,and immunodepression activities of RAP.Our studies showed that C-40 glycosylation had less impact on bioactivities compared to C-28 glycosylation.The docking model showed that the glycosylation at C-28 hydroxyl group hampered hydrophobic interactions and hydrogen bonds contacts between RAP glycosides and FK506-binding protein 12(FKBP12)and the FKBP12-rapamycin binding(FRB)domain.Expanding chemical diversities of the RAP by using promiscuous GTs could be beneficial to producing and optimizing novel rapamycin derivatives.Overall,in this thesis,through the view from macroscale to microscale,from the groups to individuals,we phylogenetically analyzed GTs of GT1 family,screened high-efficiency GT BsGT-1 for epothilone glycosylation,expanding and engineering BsGT-1 for sugar donors promiscuity,and explored promiscuous enzyme tools for drug molecule glycosylation.We elaborated the characteristics of the sequences,structures,and "hot spots" regions of GTs,which contributed to rational design of promiscuous GTs in deeper level,and prompted the efficient synthesis of drugs limited in clinical usage,chemicals and biomaterials with industrial valuables.