Enzyme-like C–H Oxidation of Glucosides Promoted by Visible Light

Abstract

Open AccessCCS ChemistryRESEARCH ARTICLE7 Nov 2022Enzyme-like C–H Oxidation of Glucosides Promoted by Visible Light Mingzhong Wu, Qian Jiang, Qiong Tian, Tianyun Guo, Feng Cai, Shouchu Tang, Jian Liu and Xiaolei Wang Mingzhong Wu State Key Laboratory of Applied Organic Chemistry, Department of Chemistry and School of Pharmacy, Lanzhou University, Lanzhou 730000 , Qian Jiang State Key Laboratory of Applied Organic Chemistry, Department of Chemistry and School of Pharmacy, Lanzhou University, Lanzhou 730000 , Qiong Tian State Key Laboratory of Applied Organic Chemistry, Department of Chemistry and School of Pharmacy, Lanzhou University, Lanzhou 730000 State Key Laboratory of Veterinary Etiological Biology, OIE/National Foot and Mouth Disease Reference Laboratory, Lanzhou University, Lanzhou 730000 , Tianyun Guo State Key Laboratory of Applied Organic Chemistry, Department of Chemistry and School of Pharmacy, Lanzhou University, Lanzhou 730000 , Feng Cai The National Glycoengineering Research Center, Shandong Provincial Key Laboratory of Carbohydrate Chemistry and Glycobiology, Shandong University, Qingdao 266237 , Shouchu Tang State Key Laboratory of Applied Organic Chemistry, Department of Chemistry and School of Pharmacy, Lanzhou University, Lanzhou 730000 , Jian Liu *Corresponding authors: E-mail Address: [email protected] E-mail Address: [email protected] State Key Laboratory of Applied Organic Chemistry, Department of Chemistry and School of Pharmacy, Lanzhou University, Lanzhou 730000 and Xiaolei Wang *Corresponding authors: E-mail Address: [email protected] E-mail Address: [email protected] State Key Laboratory of Applied Organic Chemistry, Department of Chemistry and School of Pharmacy, Lanzhou University, Lanzhou 730000 State Key Laboratory of Veterinary Etiological Biology, OIE/National Foot and Mouth Disease Reference Laboratory, Lanzhou University, Lanzhou 730000 https://doi.org/10.31635/ccschem.022.202101621 SectionsSupplemental MaterialAboutAbstractPDF ToolsAdd to favoritesDownload CitationsTrack Citations ShareFacebookTwitterLinked InEmail To match nature's prowess at using enzymes to make desired motifs in a regioselective fashion, we explore the use of visible light for the selective oxidation of the hydroxyl group to afford the corresponding keto-saccharide. This highly admirable approach offers several advantages over the enzymatic approach in terms of yields and the scope of substrates. Herein, we report the development of a simple visible-light-promoted selective oxidation of unprotected glucosides that allows for inexpensive access to valuable keto-saccharide building blocks. The method is employed on a variety of different natural and artificial glucosides, is operationally simple and scalable, and can be used to access keto-saccharides rapidly and inexpensively. Download figure Download PowerPoint Introduction Carbohydrates are the most abundant class of organic compounds found in living organisms. They have a diverse, critical role in life processes, including providing the energy for the most obvious functions of the organisms, acting as a key storage form of energy in the body, and serving as an important motif in cellular networks.1–4 However, functional group transformation (FGT) of carbohydrates mainly depends on the use of protecting groups. In carbohydrate chemistry, even a simple straightforward transformation often requires a considerable number of steps in protecting/deprotecting the functional groups.5,6 To avoid these tedious procedures, and to allow the application of easily-available saccharides, a lot of significant strategies for the selective modification of unprotected carbohydrates have been developed in recent years.7–12 In particular, the Minnaard group13 realized regioselective oxidation of several types of glycosides via a palladium system with benzoquinone as an oxidant without using protecting groups. As the development of photoredox catalysis in recent years,14–16 the Wendlandt group17 recently reported the preparation of rare sugar isomers directly from biomass carbohydrates through site-selective epimerization reactions. The Minnaard and Taylor groups18–20 reported site-selective C–H alkylations of carbohydrates under photoredox catalysis. Related mechanistic studies show that the C–H bond on carbohydrates can be selectively activated through the site-selective hydrogen-atom transfer (HAT) process. Carbohydrates play an essential role in various life processes, and carbohydrate-based drugs potentially have high specificity in novel therapeutic approaches.21,22 So far, more than 170 commercial carbohydrate-based drugs have been successfully approved to treat various diseases (Figure 1).23,24 For example, aminoglycosides such as kanamycin and amikacin, containing a glucoside motif, have been widely used for the treatment of bacterial infections.25 Sodium-glucose cotransporter-2 (SGLT-2) inhibitors, for example, canagliflozin, dapagliflozin, and empagliflozin, have been used extensively clinically to treat patients with type II diabetes mellitus to improve glycemic, cardiovascular, and renal outcomes and reduce blood glucose levels.26 Glycopeptide antibiotics, specifically vancomycin, are extensively used for the treatment of a number of bacterial infections.27,28 For glycoside chemistry, carbonyl groups are basically in the form of hydration and might work in a dual-type saccharides manner.29 For FGT, ketones are among the most versatile building blocks in organic synthesis.30 This variable feature of ketones allows the straightforward preparation of new carbohydrate motifs from unprotected keto-saccharides. Thus, the development of site-selective oxidation of unprotected glycosides to afford keto-saccharides, using abundant air as an oxidant, is still quite demanding and challenging due to the extreme similarity of the hydroxyl groups and radical intermediate tolerance of air and water. To deal with this challenge, one has to face the density of functional groups of glycosides, the judicious selection of appropriate catalyst combinations, the site selectivity of C–H activation, the compatibility of HAT reagents, photocatalysts, in situ-generated hydroperoxide, and the stability of the related keto-saccharides. Herein, we report site-selective oxidation of glucosides using cheap and readily available riboflavin tetraacetate (RFTA) as a photocatalyst source and quinuclidine as a HAT reagent. Under these developed conditions, we also prepared several series of glycolipids, which can be further used to explore a new type of drug delivery system. Figure 1 | Representative of carbohydrate-based drugs. Download figure Download PowerPoint Experimental Methods General procedure for regioselective oxidation Into a well-dried 10-mL sealed tube, α-d-methylglucoside 1a (19.4 mg, 0.1 mmol, 1.0 equiv), RFTA (2.8 mg, 0.005 mmol, 5 mol %), quinuclidine (1.1 mg, 0.01 mmol, 10 mol %), and (Bu)4NH2PO4 (8.6 mg, 0.025 mmol, 25 mol %), MnO2 (4.4 mg, 0.05 mmol, 50 mol %) were added, and a 10 × 3 mm polytetrafluoroethylene magnetic stir bar was used to stir the solution. A mixture solution of acetonitrile (MeCN)/dimethylsulfoxide (DMSO) (1/1 v/v, 0.5 mL) was added, and the vial was tightly capped in the air balloon. The reaction tube was left stirring under blue light-emitting diode (LED) irradiation with a cooling fan. The reaction was monitored by thin-layer chromatography. After about 16 h, the crude reaction mixture was concentrated under a vacuum to furnish a wet residue. The residue was purified by flash column chromatography (CH2Cl2/MeOH) to afford the desired products. Results and Discussion Initially, the minimally protected and commercially available α-d-methylglucoside 1a was used as a model substrate. After an extensive investigation of reaction conditions, α-d-methylglucoside was found to be reactive under photochemical conditions to afford ketone 2a as the major product in 78% yield for 16 h (Table 1, entry 1, see Supporting Information Tables S1–S6 for additional details). The optimal reaction conditions utilize catalytic quantities of RFTA, quinuclidine, and tetrabutylammonium dihydrogen phosphate at room temperature under air and blue light irradiations. No oxidative product was observed in the absence of RFTA, quinuclidine, ammonium salt, or light respectively (entries 2–5), and only a trace amount of product (<5%) was observed in the absence of air (entry 6). No desired or only trace product was observed when using other oxidants, such as 1,4-benzoquinone (entry 7), nor while selecting other HAT regents (see Supporting Information Table S5 for additional details). The reaction yield and efficiency was considerably diminished in the absence of MnO2 (entry 8) or by using other hydroperoxide scavengers (entry 9). 4-CzIPN and Ir[(dF(CF3)ppy)2(dtbpy)]PF6 also promoted this transformation, giving the desired product in 61% and 34% yields respectively (entries 10 and 11). However, due to its efficiency and the cost difference, we exploited the cheaper RFTA as a preferred photocatalyst. Table 1 | Reaction Discovery and Condition Optimizationa Entry Variations from Standard Conditions Yield (%)b 1a None 78 2 No RFTA 0 3 No quinuclidine 0 4 No Bu4NH2PO4 0 5 No blue LED 0 6 Under argon <5 7c 1,4-Benzoquinone as the oxidant 0 8 No MnO2 47 9d [Fe(TPA)(MeCN)2](ClO4)2 44 10 4-CzlPN instead of RFTA 61 11 Ir[(dF(CF3)ppy)2(dtbpy)]PF6 instead of RFTA 34 aStandard reaction conditions: substrates 1a (0.1 mmol), photocatalyst (5 mol %), quinuclidine (10 mol %), (Bu)4NH2PO4 (25 mol %), MnO2 (50 mol %), MeCN/DMSO (1/1, 0.5 mL), air, the 24 W blue LED, 23 °C, 16 h. bIsolated yield. c1,4-Benzoquinone (1.2 equiv). d[Fe(TPA)(MeCN)2](ClO4)2 (50 mol %). With the above established catalytic system in hand, we wanted to further explore the substrate scope for the reaction (Figure 2). First, various glucosides bearing a different functional group at the C6-position were tested, affording the corresponding products 2b– 2d in high to moderate yields with excellent regioselectivity. Other C6-functionalized glucosides, such as a fluoro- and a triazole-motif, also underwent the reaction conversion smoothly to provide desired products 2e and 2f. Further, 6-deoxy-d-glucoside can also undergo regioselective oxidation under standard conditions with a lower yield ( 2g). Benzyl and propargyl glucosides were also tolerant in this reaction system, providing corresponding products 2h and 2i in good yields. Other types of glucosides, especially those containing primary alcohol or heterocyclic rings, also afforded the related products with excellent regioselectivity ( 2j– 2l). 2-Deoxy-glucosides can also perform regioselective oxidation without any problems ( 2m and 2n). β-d-glucosides were also investigated under these reaction conditions to form corresponding β-ketoglucosides ( 2o and 2p), although the reaction efficiency was a little bit lower. Furthermore, to test the lipid tolerance, we tested the glucolipid ( 1q) as the substrate and obtained the desired keto-glucoside 2q in 43% yield. To broaden the applications of this reaction, dapagliflozin, used as an inhibitor of SGLT-2 for the treatment of type II diabetes, also generated the associated keto-Dapaglifozin ( 2r), which might work as a dual glycoside. To further test the influence of the 4-hydroxy group, 1s was prepared. To our delight, 2s and 2t can be also synthesized under the optimized condition. Based on the outstanding regioselectivity of the reaction on monosaccharides, we further examined the scope of this strategy on much more complex and challenging substrates, including natural oligosaccharides d-(+)-sucrose, d-(+)-trehalose, methyl-β-d-maltopyranoside, astragaloside A, and trisaccharides raffinose, or artificial disaccharides, containing other glycoside motifs. The desired products ( 2u– 2aa) were obtained with good yields and regioselectivity. Figure 2 | Site- and stereoselective oxidation of glucosides. Standard reaction conditions: all reactions were run with an initial substrate concentration of 0.2 M, 16 h, isolated yield. Download figure Download PowerPoint These results indicate that glucosides can be selectively oxidized in the presence of other types of glycosides and a highly reactive cyclopropane motif. To further prove the selectivity of this methodology on glucosides, we next examined the reactivity of other types of glycosides. Consistent with the selectivity shown in Figure 2, monosaccharide galactoside ( 1ab), 1,6-anhydro-d-glucose ( 1ac), and d-xylopyranoside ( 1ad) were almost untouched under this reaction system. Only a complex mixture was obtained while using mannoside ( 1ae) and rhamnoside ( 1af) as the reactants. The gram-scale reaction While scaling up the reaction, the conversion percentage decreased significantly due to various reasons such as insufficient light contact area and the decomposition of the photocatalyst. Thus, we assembled a continuous-flow reactor using crocheted fluorinated ethylene propylene tubing placed inside a 24 W blue LED light strip at 40 mL/min (Figure 3). Under this system, 1.587 g of ketone 2c (79% yield) was prepared easily through the flow reactor in one flask (see the Supporting Information for details). Figure 3 | A continuous-flow approach for the gram-scale synthesis of 2c. Download figure Download PowerPoint To demonstrate the benefits of our regioselective oxidation reaction, further functionalization of the unprotected keto-glucoside was performed to access various biologically relevant carbohydrate building blocks (Figure 4).31 However, the stereoselectivity of the functionalization reaction of keto-glucoside is still quite difficult to achieve. To our delight, alloside-lipid 3 was smoothly generated while treated with NaBH4 at 0 °C in 89% yield. Oxime 4 was also synthesized while treated with methoxyamine. Olefin 5 can be obtained under the Nysted reagent after screening for Wittig and Peterson olefination. Glycolipid 6 was successfully achieved under Barbier conditions in 72% yield with high stereoselectivity without influencing the ester group. The epoxide 7 can be readily synthesized under diazomethane reagent, a key synthetic intermediate that can be applied to synthesis other functionalized glycolipids. A rare sugar 8 containing an amine group at the C3 position was also generated in two steps. These chemical modifications further demonstrate the variable feature of these unprotected keto-saccharides. Figure 4 | Chemical transformations of the lipid saccharides 2q. (a) NaBH4 (1.5 equiv), MeOH (0.5 mL), 0 °C to r.t., 20 min. (b) MeONH2·HCl (1.5 equiv), NaHCO3 (1.5 equiv), MeOH (2.5 mL), reflux, 2 h. (c) Nysted reagent (20 wt %, 4.0 equiv), tetrahydrofuran (THF), −78 °C to r.t., 18 h. (d) Indium podwer (1.5 equiv), Allyl bromide (1.5 equiv), H2O/THF (9/1 v/v), r.t., 24 h. (e) CH2N2 (0.5 M in Et2O, 2.5 equiv), MeOH/H2O (1/1, 0.5 mL), −40 °C, 1 h. (f) HONH2·HCl (3.0 equiv), NaOAc (3.0 equiv), EtOH (1.5 mL), reflux, 2 h. (g) H2 (5 bar), PtO2 (10 mol %), AcOH (2 mL), 24 h. Download figure Download PowerPoint Mechanistic investigations We performed a series of mechanistic studies to investigate the mechanism of oxidation reaction (Figure 5, Supporting Information Table S7 and Figures S2–S5 for additional details). When 2,2,6,6-tert-methyl-1-piperidinyloxy and 2,6-di-tert-butyl-4-methylphenol were added to the reaction mixture as a radical trap, only trace amounts of oxidation product were formed (Figure 5a), which showed that a radical pathway can be involved in the transformation. In addition, the electron paramagnetic resonance (EPR) experiments also indicated that the involvement of the singlet oxygen 1O2 was generated by the photocatalyst (Figure 5b). The Stern–Volmer fluorescence-quenching experiments revealed that the excited state of the photocatalyst RFTA was quenched by quinuclidine. Fluorescence intensity was dramatically decreased with the increase of quinuclidine concentration (Figure 5c). According to the crude nuclear magnetic resonance (NMR) analysis of the reaction mixture, we detected the formation of dimethyl sulfone, which might be oxidized from DMSO by in situ-generated reactive oxygen species. Thus, we set up a control experiment as shown in Figure 5d. The result indicated that hydrogen peroxide can oxidize DMSO to dimethyl sulfone. Meanwhile, we can also conclude that hydrogen peroxide cannot oxidize the α-d-methylglucoside 1a to the corresponding keto-saccharide 2b. Figure 5 | Mechanistic experiments. (a) The radical experiments. (b) EPR spectroscopic studies. (c) Stern–Volmer fluorescence-quenching studies. (d) Control reaction: hydrogen peroxide as the oxidant. Download figure Download PowerPoint It is well known that the rate of hydrogen abstraction from a C–H bond depends not only on the C–H bond dissociation enthalpy (BDE), but also on polar effects in the transition state.32 The generality of this concept has subsequently been delineated through the broad application of polarity reversal catalysis, which takes advantage of favorable polar effects to control the regioselectivity of HAT from similar C–H bonds.33 In particular, the interaction of the hydroxyl group of carbohydrates with a hydrogen-bond acceptor catalyst should increase n–σ* delocalization of the oxygen lone pair, thereby rendering the α C–H bond more hydridic (i.e., more polarized).17,34–36 The regioselective HAT strategy Although the basis for site selectivity is not yet fully understood, the C3 selectivity observed here is congruent with the substrate-controlled selectivity previously noted by both Minnaard, Waymouth and with Wendlandt in oxidation reactions.13,17,18,37 NMR titration experiments reveal an equilibrium interaction between 1a and tetrabutylammonium phosphate (TBAP) with alteration of substrate 1J(C–H) coupling constants, implicating the presence of hydrogen-bonding interactions between 1a and the TBAP with the concomitant weakening of α C–H bonds. The proposed mechanism Based on mechanistic experiments (see Supporting Information Figures S2–S6 for additional details) and literature reports,38–43 herein we propose a plausible mechanism for the photochemical redox process in Figure 6. First, irradiation of photocatalyst RFTA with blue LEDs generated the excited RFTA* [E*1/2red = +1.67 V vs saturated calomel electrode (SCE)],44 which is sufficient for the oxidation of quinuclidine (E*1/2ox = +1.1 V vs SCE)45,46 to its radical cation and intermediate reduced riboflavin tetraacetate radical (RFTAH•) after grabbing a proton. Furthermore, this can also be demonstrated by fluorescence quenching experiments. Concomitantly, the quinuclidine radical cation 4 can site-selective undergo HAT process at the C3-position of α-methylglucose, which provides α-hydroxy radical ( I) and quinuclidinium ion 5 (BDEH–N+ = 100 kcal/mol),36 a kinetically favorable process due to the increased hydridic character of the α C–H bonds under the action of TBAP, which is a hydrogen-bonding catalyst.17 Then, RFTA-H2O2 and the desired product 2a were generated in the presence of oxygen or singlet oxygen.47,48 Subsequently, desorption of hydroperoxide from RFTA-H2O2 provided the ground-state catalyst RFTA. In conclusion, the related mechanistic studies presented here support this regioselective oxidation reaction proceeding through two steps: regioselective HAT by quinuclidinium radical cation from glucosides, followed by oxygen quenching to the incipient glucosides radical. However, it is still quite difficult to explain the low reactivity of similar types of glycosides. Computational modeling cannot give any useful information due to the super similarity of the hydroxyl groups of different types of glycosides. Ongoing studies are focusing on the synthesis and biological evaluation of keto-saccharides as to which might function as a dual saccharide towards some disease-related proteins. Figure 6 | Proposed mechanism. (a) H-bond–assisted C–H activation of a saccharide. (b) A proposed mechanistic pathway. Download figure Download PowerPoint Conclusion We realized regioselective C3 oxidation of glucosides, a long-standing challenge in carbohydrate chemistry. The excellent regioselectivity was achieved via a simple visible-light-promoted strategy. The method was employed on a variety of different natural and artificial glucosides, is operationally simple and scalable, and was applied to access keto-saccharides rapidly and inexpensively. Finally, mechanistic interrogation suggests that this unusual regio- and glycoside-type selectivity was achieved via a photoinduced radical-based reaction pathway. Supporting Information Supporting Information is available and includes experimental procedures and compound characterization data. Conflict of Interest There is no conflict of interest to report. Funding Information Financial support for this work was provided by the National Science Foundation of China (grant no. 22071087), the Fundamental Research Funds for the Central Universities (grant no. lzujbky-2021-ct05), and the Open Projects Funds of Shandong Key Laboratory of Carbohydrate Chemistry and Glycobiology, Shandong University (grant no. 2019CCG05). X.W. also thanks the Thousand Young Talents Program and Longyuan Talents Program for financial support. Acknowledgments The authors thank Prof. Qiang Liu, Dr. Xiaofen Chen, and Fengming Qi for their assistance in the mechanism study. The authors also thank Prof. Decai Xiong for directing our attention to the role of substrate scope and Sumit O. Bajaj (Corden Pharma Boulder) for language polishing. References 1. Varki A., Richard R. D., Esko J. D., Stanley P., Hart G. W., Aebi M., Darvill A. G., Kinoshita T., Packer N. H., Prestegard J. H., Schnaar R. L., Seeberger P. H., Eds.; Essentials of Glycobiology, 3rd ed.; Cold Spring Harbor Laboratory Press: Cold Spring Harbor, NY, 2017. Google Google P. R. and the Google Chemical of Google Guo in Carbohydrate on of Google 6. J. G. in Carbohydrate R. Google and Oxidation of Carbohydrates to under Google Minnaard A. of Google N. N. Minnaard A. Carbohydrate A (NMR) on and Google Taylor of in Carbohydrate Google G. Taylor of a Google of by Google J. Minnaard A. Oxidation of Google J. R. of in J. Google in Organic Google J. A. in the of Organic Google Wang Wendlandt A. of through Google Minnaard A. in Google G. Taylor and C–H of Carbohydrates via and Google J. Taylor and C–H of Carbohydrates by and Google R. and on General of Carbohydrates in Discovery and In Carbohydrates in Discovery and Google Cai Liu Liu R. and in Google Experimental for and Other Google P. in Google A. Spring Harbor in Google for with 2 A and Google A of Google A of and Google by Google N. N. Minnaard A. of and Related Google Minnaard A. of Google of Google of and in Organic Google A. P. in in the of and Google N. A. with Google J. J. from α C–H for of Google Waymouth R. Oxidation of Google in Google R. NMR of Control of the Google Liu of the J. Google Google for Oxidation of as an Google Wendlandt A. of Carbohydrates by Google and Oxidation of Google P. Oxidation of of Google of by and Google of and of Google A. to and Google Information Chemical authors thank Prof. Qiang Liu, Dr. Xiaofen Chen, and Fengming Qi for their assistance in the mechanism study. The authors also thank Prof. Decai Xiong for directing our attention to the role of substrate scope and Sumit O. Bajaj (Corden Pharma Boulder) for language polishing.