C-glycoside synthesis through radical cross-coupling of glycohydrazides Authors: Yinliang Guo, Yiheng Li, Benedikt Buchberger, Yixin Liu, Carla Capone, Tapas Adak, Shubham Ojha, Jasper L Tyler, Philipp Neigenfind, Molhm Nassir, Yu Kawamata, Varinder K Aggarwal, Phil S Baran Published: 2026 ABSTRACT AbstractCarbohydrates are among the most abundant and structurally diverse biomolecules in nature, playing central roles in energy storage, molecular recognition and cell signalling. Within this domain, C-glycosides1–3, in which the oxygen atom of the glycosidic bond in O-glycosides is replaced by carbon, have emerged as valuable motifs in medicinal chemistry due to their resistance to enzymatic hydrolysis2,4. Of particular importance are C-aryl glycosides, exemplified by the SGLT2 inhibitors dapagliflozin, canagliflozin and empagliflozin, which are frontline therapies for type 2 diabetes5–7. However, scalable syntheses of C-aryl glycosides have relied traditionally on protected sugar derivatives, lengthy sequences or conventional cross-couplings that often suffer from poor selectivity, limited scope and extensive protecting-group manipulation6. Herein, we report a practical approach to C-aryl glycosides using glycosyl sulfonyl hydrazides as redox-neutral radical precursors for cross-coupling. Prepared directly from unprotected native sugars, these reagents generate glycosyl radicals under mild conditions and enable efficient access to diverse C-aryl glycosides, including all approved SGLT2 inhibitors, natural products such as salmochelins and neopetrosins, and medicinally relevant probes. Beyond anomeric functionalization, this platform enables C–C bond formation at several positions on carbohydrate scaffolds and supports stereoretentive radical coupling that can override inherent stereochemical biases, expanding practical access to carbohydrate-derived therapeutics and chemical tools.Subject terms: Carbohydrate chemistry, Synthetic chemistry methodology Main Carbohydrates are ubiquitous in biology, where they mediate processes ranging from energy storage to molecular recognition and cell signalling8,9. Among carbohydrate-based motifs, C-glycosides, where the anomeric oxygen of an O-glycoside is replaced by carbon, have attracted sustained interest due to enhanced resistance to hydrolytic and enzymatic degradation2,5. This feature has made them particularly important pharmacophores, most notably in sodium-glucose cotransporter 2 (SGLT2) inhibitors such as dapagliflozin (1) and empagliflozin (2), which are used widely for the treatment of type 2 diabetes5,10 (Fig. 1a). Five medicines of this type are currently United States (US) Food and Drug Administration (FDA)-approved, with combined annual sales exceeding US$20 billion. Although the clinical success of SGLT2 inhibitors reflects many pharmacological, pharmacokinetic and clinical factors, these compounds also exemplify how a robust C(sp³)–C(sp²) linkage can contribute to favourable metabolic stability and durable biological activity, underscoring the need for efficient and scalable synthetic strategies to access diverse C-aryl glycoside architectures. General access to C-glycosides. a, C-aryl glycoside drugs: privileged structures for the treatment of type 2 diabetes. b, Cross-coupling approach: previous art for 2e– and 1e– mode. c, Glycohydrazides: opening up new chemical space through redox-neutral RCC. Historically, the synthesis of C-aryl glycosides (5) has relied on methods wherein a glycosyl electrophile or nucleophile equivalent is used (6)3. For example, Friedel–Crafts-type glycosylations with electron-rich arenes or nucleophilic additions of organometallic aryl reagents (for example, aryl lithium or Grignard species) to glycosyl electrophiles. Although these approaches have enabled the construction of key C-glycosidic bonds, they are often plagued by notable limitations including extensive protecting-group manipulations, poor stereoselectivity at the anomeric centre, limited substrate scope and multi-step preparations of activated donors from native sugars. For instance, the general synthesis of 5 relies on polar bond disconnections through a multi-step sequence with only one strategic C–C bond-forming step6,7,11. Recent advances since 2007 in the arena of conventional cross-coupling (cf. 7 and 8), including Pd- and Ni-catalysed glycosyl cross-couplings3,6,7,12–14, have addressed some of these issues but still frequently demand protected sugars, specialized catalysts or organotin intermediates, hindering late-stage diversification and large-scale synthesis (Fig. 1b). In contrast, radical cross-coupling (RCC) can, in principle, offer a more direct and chemoselective approach to such targets by means of a glycosyl radical (4). Yet, practical implementation remains limited, as the preparation of these intermediates still typically relies on lengthy sequences involving protecting-group manipulations, particularly in the cases of 9–12 (refs. 15–19). Against this backdrop, glycosyl donor 14, introduced by Koh, stands out for its direct preparation from an unprotected native sugar20. However, its synthetic utility along with 13 has so far been confined to radical addition reactions (for example, Giese addition), and the application to more versatile RCC has yet to be realized21–23. In this work, a transformative approach to C-aryl glycoside synthesis is presented using easily accessible glycosyl sulfonyl hydrazides as redox-neutral radical precursors for RCC24–28 (Fig. 1c). Derived in a single step from unprotected native sugars without the need for protecting groups, stable crystalline glycohydrazides (16) enable straightforward generation of glycosyl radicals under mild and scalable conditions, facilitating couplings with a variety of aryl partners. This methodology not only streamlines access to bioactive molecules, including all currently approved SGLT2 inhibitor C-aryl glycosides (5), but also unlocks new possibilities for radical-mediated sugar modifications, including cross-couplings at non-anomeric sites (across positions C2–C6, 20–23). Moreover, the unique reactivity of hydrazides can override inherent stereochemical biases that have represented unanswered challenges for radical-based cross-coupling (cf. 18 and 19) by enabling a stereoretentive RCC approach26,28. Collectively, glycohydrazides address longstanding challenges in efficiency, selectivity and scalability in C-glycoside synthesis and provide a practical platform for the rapid exploration of carbohydrate-derived therapeutics, natural products and chemical probes. Unprotected glycohydrazide C-arylation The development of suitable conditions for RCC-arylation of unprotected glycohydrazides required extensive optimization, the highlights of which are summarized graphically in Fig. 2a (for a more complete summary, see Supplementary Information). Glycohydrazides such as 25 can be prepared easily on a decagram scale by simply stirring a 1:1 mixture of sugar (glucose 24 in the case of 25) and sulfonylhydrazine in acetic acid followed by simple crystallization29. Screening commenced with glucose-derived hydrazide 25 and aryl iodide 26 using the conditions reported previously24 for redox-neutral RCC with the identity of the sulfonyl hydrazide, solvent, base, temperature and ligand being evaluated systematically. Sulfonyl hydrazides are uniquely tunable radical precursors; the electronic nature of the sulfonyl group can dictate the rate of radical formation to reach an optimal level of kinetic matching in the coupling process27. Glycohydrazide 25, when subjected to standard conditions at 70 °C, provided only trace quantities of product with a slight increase in temperature (100 °C) leading to 12% yield of coupled product 1 as a 1:1 mixture of diastereoisomers. Although the use of more electron-deficient hydrazides such as 27, 28 and 29 led to a modest increase in yield at 70 °C, glycohydrazide 25 was chosen for further optimization owing to its lower cost, easy accessibility and suitability for large-scale synthesis of gliflozin-based medicines. Next, a solvent screen was performed (0.1 M), resulting in a higher yield using dimethylsulfoxide (DMSO; 27% yield) but unfortunately still as a 1:1 mixture of isomers. The most striking breakthrough came from an exploration of base. A variety of organic and inorganic bases were evaluated, differing in basicity, leading to an increase in yield (up to 45% with K2CO3) but still with a 1:1 mixture of diastereomers. Surprisingly, tetramethylguanidine (TMG) was singularly successful in boosting both the yield (81%) and selectivity (β:α = 11:1) as well as reducing the temperature to 70 °C. Although this outcome is difficult to rationalize given the lack of a clear correlation between base strength and either yield or diastereomeric ratio (see Supplementary Information for exhaustive listing of bases screened), we speculate that hydrogen bonding between TMG and the glucose hydroxyl groups may be important, as β-selectivity drops substantially with the corresponding Bn-protected glucose substrate (44%, β:α = 3.6:1). Consistent with this observation, the Niu group also observed high β-selectivity in RCCs with unprotected glucose in the presence of TMG, whereas the corresponding protected substrate showed much lower β-selectivity17. Finally, a number of ligands (L1–L5) were screened to see whether that parameter could be simplified as L6 (dNH2bpy) is more expensive. L1 was chosen considering the balance of yield (88%) and diastereomeric ratio (β:α > 19:1) observed, but other inexpensive ligands such as bpy (L3) or dtbbpy (L4) can also be used. Ultimately, a simple protocol using these conditions was developed that involves mixing all reaction components and heating to 70 °C for 1 h followed by a standard aqueous workup. Reaction development and generality. a, Reaction development. b, Protecting-group-free glucose C1 coupling: Substrate scope. aReaction condition A: 20 mol% Ni(dme)Cl2, 20 mol% L1, 3.0 eq. TMG, 0.1 M DMSO, 70 °C. bReaction condition B: 20 mol% Ni(dme)Cl2, 20 mol% L6, 3.0 eq. TMG, 0.1 M DMSO, 70 °C. cStep counts were calculated from the sugar source rather than from the coupling partner. Diastereomeric ratio was determined from the crude reaction mixture by 1H NMR or liquid chromatography–mass spectrometry (LCMS). Ar1 = 1-chloro-2-[(4-ethoxyphenyl)methyl]-benzene; ArBr, 4-bromo-1-chloro-2-[(4-ethoxyphenyl)methyl]benzene; DBU, 1,8-diazabicyclo[5.4.0]undec-7-ene; DMF, dimethylformamide; EtOH, ethanol; MeCN, acetonitrile; eq., equivalents; PMG, pentamethylguanidine; PMP, 1,2,2,6,6-pentamethylpiperidine; THF, tetrahydrofuran. As depicted in Fig. 1b, glycohydrazide-based RCC could be used to access nearly all currently known FDA-approved gliflozin-based medicines (1, 2, 30, 32, 33) as well as several analogues that are currently in clinical trials (34, 35, 36). In addition, the known ketone-containing intermediate 37 (ref. 30) was prepared, which previously necessitated protecting groups due to the harsh conditions of Grignard addition. Accessing these structures using the less expensive aryl bromide was also possible (31, 32, 33, 34, 35, 36) using L6. Remarkably, di- and tri-saccharide aryl-linked structures such as 38, 39, 40 and 41 could also be accessed directly from inexpensive lactose, cellobiose, maltose and maltotriose, respectively. The general procedure could be applied to a decagram-scale synthesis of 1. For this scale-up, glycohydrazide 25 was prepared directly from commercially available dextrose powder (purchased at Walmart for US$5 per pound) by treatment with TsNHNH2 in 70% household vinegar and stirring for 6 h, followed by simple ether precipitation to afford the crystalline intermediate. The key cross-coupling took place smoothly to provide 1 in 61% isolated yield with high diastereomeric ratio (β:α > 19:1) after extraction and chromatography. On larger scales, the product should be purified easily through direct recrystallization without the need for column chromatography31. To further investigate the substrate scope, a series of glycohydrazides were prepared from native hexoses and pentoses using a common protocol (TsNHNH2/acetic acid), as shown in Fig. 3. Notably, all of the resulting glycohydrazides adopt the pyranose form, as confirmed by X-ray crystallography and two-dimensional nuclear magnetic resonance (NMR) spectroscopy analyses, despite the conventional assumption that pentoses preferentially exist as furanoses whereas hexoses predominantly adopt pyranose structures. More intriguingly, these glycohydrazides exhibit markedly divergent reactivity under different reaction conditions: hexose-derived hydrazides can furnish either pyranose or furanose products, and a similar variability is observed for pentose-derived substrates, presumably arising from the unique ring–chain tautomerization behaviour of glycohydrazides (cf. 67–69). Following systematic optimization, it was found that both the base and ligand play decisive roles in controlling the yield and selectivity of the main product. For hexose-derived substrates, pyranose products are formed preferentially. For example, mannose-, N-acetylglucosamine-, rhamnose- and 2-deoxyglucose-derived hydrazides deliver products 47, 49, 51 and 53, respectively, with high diastereoselectivity. In contrast, the galactose-derived hydrazide yields predominantly the furanose product 56. Notably, the reaction of fucose-derived hydrazide exhibits pronounced ligand dependence on the reactive tautomer: with L1 (condition A or B), the furanose product 58 is formed predominantly, whereas using L4 (condition C) switches the selectivity to the pyranose isomer 57 as the main product. For pentose-derived substrates, ribose-, arabinose- and xylose-derived hydrazides furnish mainly the corresponding furanose products 60, 62 and 66, whereas the lyxose-derived hydrazide instead affords the pyranose product 63. Overall, this transformation exhibits broad substrate generality and well-defined chemo- and stereoselectivity, enabling efficient access to either furanose or pyranose C1-glycosylation products in a substrate-dependent manner, with ligand- and base-dependent modulation of selectivity demonstrated in representative cases. These features provide a versatile platform for the rapid construction of structurally diverse C1-glycosides, offering potential utility in synthesis, chemical biology and medicinal chemistry. The ligand-controlled divergency in the ring size of glycosyl RCC described above has not been described previously. PG-free sugar C1 coupling: substrate scope. The structures of ribose, xylose and galactose hydrazides were further confirmed by X-ray crystallography. Diastereomeric ratio was determined from the crude reaction mixture by 1H NMR spectroscopy or LCMS. aToluenesulfonyl hydrazide. b3,5-difluorobenzenesulfonyl hydrazide. c3,4,5-trifluorobenzenesulfonyl hydrazide. d2-(trifluoromethyl)benzenesulfonyl hydrazide. eStructures were assigned using two-dimensional NMR spectroscopy (see Supplementary Information for details). ND, not determined. Application to natural product synthesis The strategic application of redox-neutral glycohydrazide-based RCC to a variety of natural products and medicinally relevant structures beyond gliflozins was next explored, as documented in Fig. 4. Salmochelin-SX (72)—a C-aryl glycosidic natural product in the salmochelin family known as siderophores excreted by Salmonella enterica and uropathogenic Escherichia coli strains under low-iron stress32—was procured previously through a nine-step polar approach relying on conventional cross-coupling, pyrophoric reagents and protecting-group interchanges13. In contrast, a four-step approach commencing with crystalline glycohydrazide 25 led to 72 after coupling with aryl iodide 70 followed by a sequence of hydrolysis, amidation and debenzylation. The critical RCC took place in 61% isolated yield, favouring the desired diastereomer (β:α > 10:1). Beyond glucose-derived hydrazide 25, mannose-derived hydrazide 73 likewise enabled efficient access to C-glycosides. Notably, neopetrosin C (75)—a rare indole C-glycoside alkaloid whose congeners have been reported to exhibit moderate hepatoprotective activity33—was prepared directly from 73 in a single step in 52% isolated yield with α:β > 19:1. In this case, the desired cross-coupling product was obtained under modified conditions using a more electron-deficient sulfonyl group (Ar = 2-CF3C6H4) and Et3N (TEA) in place of the more strongly basic TMG. The pronounced α-selectivity probably reflects a matched scenario in which both steric and anomeric effects converge to favour formation of the α-anomer, consistent with previous studies of RCC in related systems17,34. RCC has been applied to this natural product but it requires the use of expensive peracetyl mannosyl bromide (US$218 per 1 g commercial compound that needs to be stabilized by 2% CaCO3) by means of photoinduced electron transfer involving numerous additives and expensive iridium photocatalysts35. A polar-bond-based disconnection strategy using a Larock coupling of alkynyl glycoside traversing through ten steps is also known36. Similarly, the naturally occurring tryptophan-mannose conjugate 77 can be procured from 73 and 76 in 56% isolated yield as a single diastereomer (α-anomer) after alkaline hydrolysis of the Trp-containing TFA motif. Conventional polar strategies required more than 20 steps37 and redox-reliant RCC could achieve the same transformation, albeit requiring a cocktail of additives and an expensive Ir-catalyst38. Finally, the ribose-derived IMPDH inhibitor 80 could be accessed concisely from ribohydrazide 78 and m-iodobenzamide 79 as a single diastereomer (β-anomer) in 65% isolated yield, representing a substantial simplification relative to the previous five-step route involving extensive protecting-group manipulations39. Synthetic applications towards natural products and drug derivatives. Diastereomeric ratio was determined from the crude reaction mixture by 1H NMR spectroscopy or LCMS. cat., catalyst. Pre-radical stereocontrol Up to this stage, stereochemical control in our redox-neutral platform arises exclusively from post-radical control, wherein product configuration is set during bond formation through the interplay of substrate stereoelectronic bias and catalyst steric effects. To our knowledge, all previously reported RCC approaches (Giese, as in refs. 20,40; Minisci, as in ref. 21 and analogous innate radical additions are not included as they are not considered cross-coupling, which involves catalytic organometallic intermediates) to C-glycosides operate by this same principle17. Although high selectivity and, in some cases, stereodivergence can be achieved, these outcomes are typically confined to particular substrates rather than representing a general solution. This limitation is evident across carbohydrate scaffolds: whereas glucose-derived systems have furnished numerous stereoselective examples by leveraging either steric effects (favouring the β-anomer18,19,38,41) or the anomeric effect (favouring the α-anomer15,34,42–47), several other scaffold classes still lack any stereoselective RCC manifold (Fig. 5). Collectively, these observations underscore the inherent substrate dependence of post-radical stereocontrol and the need for a more general, programmable approach. Stereoretentive RCC: programmable C1 stereochemistry overriding substrate biases. Diastereomeric ratio was determined from the crude reaction mixture by 1H NMR spectroscopy or LCMS. aStereoretentive condition: 20 mol% Ni(dme)Cl2, 20 mol% L12, 3.0 eq. TEA, 0.1 M DMSO, 80 °C. dr, diastereomeric ratio. Departing from this paradigm, programmable pre-radical stereocontrol was envisioned building on the recent discovery of stereoretentive RCC26 (Fig. 5). For these studies, Bn-protected sugars were used to simplify the preparation and purification of either anomeric series (α or β). The divergent preparation of either the β- or α-glycohydrazides could be accomplished commencing with 81 using either a Mitsunobu reaction or through Yu-glycosylation48 with benzaldehyde p-toluenesulfonylhydrazone 82, respectively. With configurationally enriched anomers in hand, stereoretentive RCC was investigated using a variety of ligands, bases and solvents (see Supplementary Information for a more complete listing). In line with previous reports on substrate controlled RCC of protected glucose derivatives, the result is a near 1:1 mixture of products (entry 1) in 32% yield (Fig. 2). Reducing the strength of the base to TEA improved the yield albeit with a poor stereoselectivity (entry 2). After an extensive ligand screen (entries 3–8) and in accordance with results from our studies of stereoretentive RCC arylation26, electron-deficient bipyridine ligand L12 enabled product formation in good yield (68%) with synthetically useful levels of stereocontrol (5:1) favouring the β-product. Turning attention to α-83, the same set of conditions could be applied to deliver 70% of adduct 85 in a 3:1 ratio favouring the α-product, unequivocally demonstrating that the stereochemistry was retained through the process. Unsurprisingly, the use of L1 (non-stereoretentive ligand) delivered a 1:1 mixture of isomers once again, highlighting the critical importance of the ligand. Stereoretentive couplings of this type were explored on three additional substrates, such as 2-deoxyglycohydrazides β-87 and α-87, ribohydrazides β-89 and α-89, and 2-deoxyribohydrazides β-91 and α-91. In most cases, a modest to good level of stereoretention could be observed, thereby enabling the practitioner to enrich diastereochemical outcomes independent from inherent substrate biases. The results presented herein represent an important advance in stereoretentive radical C-glycosylation given past precedent. For instance, ribose systems favour β-products exclusively due to steric considerations; in 2-deoxyglucose and 2-deoxyribose, the stereoelectronic effect dominates to give the α-product44. Notably, several anomeric configurations including β-2-deoxyglucose (93), α-ribose (94) and β-2-deoxyribose (95) have not been accessed previously through RCC. Peripheral C–C bond formation The scope of glycohydrazide-based RCC was explored to enable peripheral C–C bond formation at positions beyond the anomeric centre, providing a systematic exploration of RCC across all alcohol-containing positions of riboses and glucoses. Accessing such chemical space could potentially open many new opportunities for diversification of carbohydrate scaffolds. The results of these investigations are depicted in Fig. 6a and commence with the preparation of 2-, 3-, 4- and 6-substituted glucohydrazides 99, 105, 111 and 117 by means of reductive amination from the corresponding carbonyl compounds. For these studies, additional optimization was conducted leading to the use of difluorophenyl sulfonyl hydrazide, 1,10-phenanthroline as ligand and TEA as base (see Supplementary Information for summary). Using these conditions, a library of 20 RCC products (16 of which are depicted herein, see Supplementary Information for full listing) was obtained in synthetically useful yields with post-radical stereocontrol (stereoretentive RCC was not attempted, nor was optimization of any individual reaction). Notably, the first examples of RCC-based sugar olefinations were also achieved in synthetically useful yields, providing adducts 110, 116, 122. Similarly, ribose- and deoxyribose-derived hydrazides 123, 125, 127, 129 and 131 were prepared and subjected to RCC to produce a library of 19 RCC-derived adducts, 5 of which are depicted in Fig. 6a. Of particular significance, glycohydrazide-based RCC proved amenable to unprotected glucoside 134. Methyl 3-keto-α-d-glucopyranoside was accessed through selective oxidation of α-d-glucopyranoside 133 (refs. 49,50), followed by reductive amination to furnish the corresponding α-hydrazide 134. Subsequent coupling delivered adducts 136, 137 and 138 in good yield with high diastereoselectivity (β:α > 19:1; Fig. 6b). These results further indicate that, in addition to protecting-group strategies that facilitates handling and purification, unprotected glycohydrazides can also exhibit suitable reactivity and selectivity in RCC. Peripheral C–C bond formation. a, Coupling at C2–C6 positions. b, Selective oxidation followed by C3 coupling. c, A redox-neutral platform to access nucleoside analogues. aCondition: 20 mol% Ni(dme)Cl2, 20 mol% L8, 3.0 eq. TEA, 0.1 M DMSO, 50–70 °C. bCondition: 20 mol% Ni(dme)Cl2, 20 mol% L6, 3.0 eq. PMP, 0.1 M DMF, 75 °C. cCondition: 20 mol% Ni(dme)Cl2, 20 mol% L4, 3.0 eq. TEA, 0.1 M DMF, 100 °C. dStep counts for synthesizing the hydrazides: glucose C2 hydrazide 99 (three steps from commercially available SF5); glucose C3 hydrazide 105 (three steps from SF5); glucose C4 hydrazide 111 (four steps from SF5); glucose C6 hydrazide 117 (four steps from commercially available SF9); ribose C2 hydrazide 123 (five steps from commercially available SF12); ribose C3 hydrazide 125 (five steps from SF12); ribose C5 hydrazide 127 (two steps from commercially available SF15); 2-deoxyribose C3 hydrazide 129 (five steps from commercially available SF16); 2-deoxyribose C5 hydrazide 131 (six steps from SF16). eCondition: 20 mol% Ni(dme)Cl2, 20 mol% L1, 3.0 eq. TEA, 0.1 M DMSO, 75 °C. f[(neocuproine)Pd(μ-OAc)]2[OTf]2. See Supplementary Information for detailed procedures and full listing of 17 examples in c. Diastereomeric ratio was determined from the crude reaction mixture by 1H NMR spectroscopy or LCMS. Finally, given the broad importance of unnatural nucleosides across biology and medicine, we sought to demonstrate concise entry into this chemical space (Fig. 6c). Using conditions related closely to those developed for the non-anomeric couplings in Fig. 6a, C1-arylated ribose derivatives 139–142 were obtained in synthetically useful yields, with 13 additional examples provided in the Supplementary Information. The efficiency of this transformation is comparable to that of the recent photochemical approach reported by Britton and colleagues51, yet the ribohydrazide platform offers a practical advantage by avoiding the engineering constraints associated with photochemical reaction setups. Notably, high β-selectivity is achieved without invoking a stereoretentive manifold, as the intrinsic steric bias of the ribose scaffold is sufficient to enforce conventional post-radical stereocontrol. Conclusion The past decade has witnessed a vibrant and growing effort to enable synthetic elaboration of carbohydrates without the protecting-group sequences17,20,23,52–58 and redox manipulations that long dominated the field. The present work joins and extends this trajectory by addressing a distinct and hitherto unmet challenge: the absence of a practical, versatile radical glycosyl donor suitable for modular cross-coupling. By converting unprotected sugars in a single step into stable sulfonyl hydrazides that serve as such donors, a direct, efficient, highly chemoselective and scalable route to C1-glycosides has been unlocked. This chemistry provides rapid access to current FDA-approved SGLT2 inhibitors, challenging natural products and a host of new glyco-architectures—including modifications at non-anomeric positions and stereoisomers previously considered out of reach. In doing so, it shifts carbohydrates from being among the most cumbersome classes of molecules to synthesize into versatile and readily diversifiable building blocks. We anticipate that this platform, amenable to derivatization at several sites on sugars and to stereoretentive coupling, will catalyse a substantial expansion in the exploration of carbohydrate chemical space and help unleash the full therapeutic potential of glycomimetics in the coming years. Online content Any methods, additional references, Nature Portfolio reporting summaries, source data, extended data, supplementary information, acknowledgements, peer review information; details of author contributions and competing interests; and statements of data and code availability are available at 10.1038/s41586-026-10807-x. Supplementary information Experimental procedures, characterization data and NMR spectra. Crystallographic data for compound Ribose-hydrazide 78. Crystallographic data for compound Galactose-hydrazide β-SH12. Crystallographic data for compound Xylose-hydrazide β-SH21. Crystallographic data for compound Mannose-arylated product α-47. Peer Review File Acknowledgements We are grateful to L. Pasternack and G. Kroon (Scripps Research) for assistance with NMR spectroscopy; to J. Chen, B. Orzolek, Q. Nguyen, J. Lee and J. Smith (Scripps Research ASF) for assistance with UPLC, HRMS, preparative high-performance liquid chromatography and preparative supercritical fluid chromatography analysis and purification. We thank J. Bailey and M. Gembicky (University of California San Diego) for X-ray analysis. Author contributions Conceptualization: Y.G., Y. Li, T.A., Y.K., V.K.A., P.S.B. Investigation: Y.G., Y. Li, B.B., Y. Liu, C.C., T.A., S.O., P.N., M.N. Data analysis: Y.G., Y. Li, B.B., Y. Liu, C.C., T.A., S.O., Y.K., P.S.B. Writing–original draft: Y.G., Y. Li, B.B., J.L.T., S.O., T.A., Y.K., V.K.A., P.S.B. Writing–review & editing: Y.G., Y. Li, B.B., J.L.T., S.O., T.A., Y.K., V.K.A., P.S.B. Funding acquisition: P.S.B., V.K.A. Project administration: Y.K., V.K.A., P.S.B. Peer review Peer review information Nature thanks the anonymous reviewers for their contribution to the peer review of this work. Peer reviewer reports are available. Funding Financial support for this work was provided by the National Institutes of Health (GM-118176, P.S.B.) S.O. thanks the Bristol Chemical Synthesis Centre for Doctoral Training, funded by EPSRC (EP/S024107/1) and AstraZeneca for a PhD studentship. We also thank EPSRC UKRI (EP/Y028015/1) for financial support. Data availability The data that support the findings in this work are available within the paper and Supplementary Information. Competing interests The authors declare no competing interests. Footnotes Publisher’s note Springer Nature remains neutral with regard to jurisdictional claims in published maps and institutional affiliations. These authors contributed equally: Yinliang Guo, Yiheng Li Contributor Information Yu Kawamata, Email: yukawama@scripps.edu. Varinder K. Aggarwal, Email: V.Aggarwal@bristol.ac.uk Phil S. Baran, Email: pbaran@scripps.edu Supplementary information The online version contains supplementary material available at 10.1038/s41586-026-10807-x. References Associated Data Supplementary Materials Experimental procedures, characterization data and NMR spectra. Crystallographic data for compound Ribose-hydrazide 78. Crystallographic data for compound Galactose-hydrazide β-SH12. Crystallographic data for compound Xylose-hydrazide β-SH21. Crystallographic data for compound Mannose-arylated product α-47. Peer Review File Data Availability Statement The data that support the findings in this work are available within the paper and Supplementary Information.