An isofagomine analogue with an amidine group inthe 1,6-position

An isofagomine analogue with an amidine group in the1,6-position I. Caroline Vaaland Holmgard1, Idd Andrea Christensen2, Ljupcho Pejov3,4, Monika Moreń1, Nikolai Westlund1, Tereza Cristina Santos Evangelista1, Magne O. Sydnes1,5, Finn L. Aachmann2, Óscar López6 and Emil Lindbäck1 1Department of Chemistry, Bioscience and Environmental Engineering, Faculty of Science and Technology, University of Stavanger, Stavanger, Norway 2Norwegian Biopolymer Laboratory (NOBIPOL), Department of Biotechnology and Food Science, NTNU Norwegian University of Science and Technology, Trondheim, Norway 3Institute of Chemistry, Faculty of Natural Sciences and Mathematics, SS Cyril and Methodius University, Skopje, North Macedonia 4Environmental and Resource Management Program, The Polytechnic School, Ira A. Fulton Schools of Engineering, Arizona State University, Tempe, AZ, USA 5Department of Chemistry, Faculty of Science and Technology, University of Bergen, Bergen, Norway 6Departamento de Química Orgánica, Facultad de Química, Universidad de Sevilla, Seville, Spain EL,0000-0001-5809-7368 The synthesis of an isofagomine analogue with an amidine group in the 1,6-position is described. Density functional theory calculations showed that this compound has a remarkably different charge distribution compared with isofagomine. This may explain why the amidine is a poor glycosidase inhibitor (IC50 > 50 µM against all tested enzymes) compared with isofagomine. 1. Introduction Glycosidases are a ubiquitous group of enzymes in living organisms, catalysing the hydrolysis of glycosidic bonds during the degradation of polysaccharides and glycoconjugates. Iminosugars represent a group of carbohydrate mimetics that are well known for their frequently high activity as glycosidase inhibitors [1]. This activity has made them attractive as lead compounds for the treatment of a wide spectrum of diseases such as cancer [2], diabetes [3], lysosomal storage disorders [4] and lately COVID-19 [5]. © 2025 The Authors. Published by the Royal Society under the terms of the Creative Commons Attribution License http://creativecommons.org/licenses/by/4.0/, which permits unrestricted use, provided the original author and source are credited. Research Cite this article: Vaaland Holmgard IC, Christensen IA, Pejov L, Moreń M, Westlund N, Evangelista TCS, Sydnes MO, Aachmann FL, López Ó, Lindbäck E. 2025 An isofagomine analogue with an amidine group in the 1,6-position. R. Soc. Open Sci. 12: 241877. https://doi.org/10.1098/rsos.241877 Received: 28 October 2024 Accepted: 6 December 2024 Subject Category: Chemistry Subject Areas: organic chemistry Keywords: isofagomine, analogue, amidines, 1,6-position Author for correspondence: Emil Lindbäck e-mail: [email protected] This article has been edited by the Royal Society of Chemistry, including the commissioning, peer review process and editorial aspects up to the point of acceptance. Electronic supplementary material is available online at https://doi.org/10.6084/ m9.figshare.c.7669967. Downloaded from https://royalsocietypublishing.org/ on 04 June 2025 The high glycosidase inhibition activity of basic iminosugars, such as 1-deoxynojirimycin (DNJ) (1) (figure 1), which is a mimic of glucose, has been attributed to their charge resemblance to the oxocarbenium ion (A)-like transition state in their protonated states. This allows them to interact with the catalytic carboxyl and/or carboxylate groups in the active site of glycosidases through ionic bonds [6]. However, because 1 adopts a chair conformation, it is not a perfect shape mimic of oxocarbenium ion A that adopts a more flattened conformation. Therefore, Ganem and co-workers tailored ᴅ-glucoamidine (2) (figure 1), which is an analogue of glucose, assumed to mimic both the shape and charge distribution of the glycosidase transition state [7]. This compound, which is a roughly 7900-fold stronger base than 1, was found to inhibit a broad spectrum of glycosidases [7,8]. The low configurational selectivity of 2 has been argued to indicate that it is not a perfect mimic of the transition state [6]. Azasugars, exemplified by isofagomine (IFG) (3) (figure 1) [9], are a group of glycosidase inhibitors related to iminosugars that contain a nitrogen atom in the pseudo anomeric position. The conjugate acids of such compounds resemble the resonance form B of the oxocarbenium ion A [10]. X-ray crystallographic studies have demonstrated that the attachment of a β-glucopyranosyl group in the 4-position of isofagomine provides an inhibitor that binds to cellulase Cel5A through ionic interactions with the catalytic carboxyl and carboxylate groups in the active site [11]. Insertion of an oxo group in the 2-position of 3 provided a non-basic lactam analogue that is a much weaker glycosidase inhibitor than 3 [12], indicating that a basic nitrogen atom in the anomeric position has an enhancing effect on the glycosidase inhibitory activity. Following this line, compound 4 (figure 1), which is an analogue of isofagomine (3), containing a basic amidine group at the 1,2-position, was prepared [13]. This compound was found to exhibit a very different inhibition profile compared with 3 since it is a much stronger jack bean α-mannosidase inhibitor (Ki = 0.75 µM for 4 [13] and IC50 = 250 µM for 3 [14]) and a much weaker almond β-glucosidase inhibitor (Ki = 450 µM for 4 [13] and Ki = 0.11 µM for 3 [9]). The significantly different inhibition profiles were proposed to be due to variations in the charge distributions in 3 and 4, in which density functional theory (DFT) calculations showed that 3 has a positive group charge at NH2-1, whereas there is a slightly negative group charge at NH-1 in 4 [13]. Herein, we report the synthesis of the regioisomer of 4, namely amidine 5 and its N-benzyl derivative 6 (figure 1), in addition to their glycosidase inhibition activities on a wide panel of commercially available glycosidase and their distribution of atomic and group charges. The retrosynthetic analysis of amidine 5 is presented in Scheme 1. For the synthesis of amidine 4, the amidine group was installed by a Cu(I)-promoted intramolecular addition of an amino group to a cyano group [13,15]. Thus, we anticipated that the amidine group in 5 would be installed through Cu(I)-promoted cyclization of amino nitrile precursor 7. Compound 7 would be synthesized by reductive amination between pyranose 8 and benzylamine (9). Pyranose 8 would be obtained when compound 10 [16] undergoes de-O-benzylation. Figure 1. Oxocarbenium ion A and its resonance form B, and examples of iminosugars and azasugars. Amidines 5 and 6 are the target compounds in this work. Scheme 1. Retrosynthetic analysis of amidine 5. 2 royalsocietypublishing.org/journal/rsos R. Soc. Open Sci. 12: 241877 Downloaded from https://royalsocietypublishing.org/ on 04 June 2025 The synthesis of compounds 5 and 6 is presented in Scheme 2. The synthesis commenced with the palladium-catalysed de-O-benzylation of benzylarabinopyranoside 10 to generate desired pyranose 8 in 33% yield along with compound 11 in 14% yield and a high diastereomeric ratio (dr) after silica gel flash column chromatography. The configuration of the anomeric carbon of 11 could not be determined because the signal of the anomeric proton overlapped with other signals in the 1H nuclear magnetic resonance (NMR) spectra. The formation of 11 was expected to proceed through a de-O-ben- zylation/nitrile hydrogenation/reductive amination sequence of 10 to form a piperidine (i.e. acetal-pro- tected isofagomine) whose primary hydroxyl reacted with 8. Selective BCl3-mediated de-O-benzylation of 10 was explored, which unfortunately only gave a reaction on the acetal ring leaving the O-benzyl group untouched. Amino nitrile 7 was obtained in 67% yield through reductive amination when 8 was treated with benzylamine in the presence of sodium cyanoborohydride. Copper(I) chloride triggered the cyclization of 7 into N-benzylamidine 12 in 85% yield after removing the Cu-ion residues upon precipitation with H2S in methanol followed by subsequent silica gel flash column chromatography. The acetal group in compound 12 was removed upon treatment with aqueous HCl solution to furnish amidine 6 in 72% yield. 70 mol% of palladium on carbon under a hydrogen atmosphere was employed to remove the benzyl group from 12 to afford compound 13 whose acetal group was removed to afford amidine 5 in 96% yield when it was treated with an aqueous solution of HCl. The pKa values for azasugars have been determined by NMR spectroscopy in other studies [17]. This methodology was also applied in this study to determine the pKa value of 6. Specifically, the chemical shifts of the 2-Ha proton were plotted against the pH* (the pH measured in D2O using a H2O-calibrated pH electrode), which provided an inflection point on the resulting best-fit sigmoidal curve at pH* > 11.0, corresponding to the pKa* value for 6 (figure 2). This pKa* value was corrected to pKa > 10.6 when a correction factor was applied [18]. The sigmoidal curve is missing points at high pH* values (greater than 11) because 6 started degrading at roughly pH* 10.5. Recording of both a 1H Scheme 2. Synthesis of amidines 5 and 6. Figure 2. Titration curve of the chemical shifts of 2-Ha versus pH* for amidine 6. 3 royalsocietypublishing.org/journal/rsos R. Soc. Open Sci. 12: 241877 Downloaded from https://royalsocietypublishing.org/ on 04 June 2025 NMR and 13C HSQC spectra at each titration point allowed us to track the changes in chemical shifts at pH levels higher than where the degradation initially started. The pKa of amidine 5 was determined at even lower accuracy than for 6, because it started degrading at roughly pH* 9.4, where there is no indication of 5 being deprotonated. Presumably, the pKa of 5 is higher than that for 6, because iminosugars armed with aromatic moieties via a CH2 group have been found to have lower pKa values than the corresponding iminosugars without such N-substituents [19]. Additionally, a high pKa value of 5 was expected as the pKa value of structurally related ᴅ-glucoamidine (2) is also high (≥10.5) [7]. The glycosidase inhibitory activities (IC50 (µM)) of amidines 5 and 6 on a panel of six glycosidases are presented in table 1 and compared with structurally related isofagomine (3) and amidine 4. Remarkably, even though the structure of amidine 5 is quite similar to those of 3 and 4, which are powerful inhibitors of β-glucosidase and α-mannosidase, respectively, it only showed a very weak inhibition (IC50 > 50 µM) of both of these enzymes. Indeed, amidine 5 is a very poor inhibitor of all enzymes in the panel. The benzyl group on amidine 6 seems to have no positive impact on inhibition potency, as this compound is also a very weak inhibitor for all enzymes tested. The partial charges of all atoms in 3–6 were calculated by the Mulliken, CHelpG and HLY charge partitioning schemes on the basis of DFT electronic densities (electronic supplementary material, tables S1–S4). Selected atomic charges and group charges calculated by the CHelpG scheme for 3–6 are presented in figure 3. Our calculated charges agree with previously calculated atomic and group charges for 3 and 4 [13]. Amidines 4–6 contain a substantial positive charge at the sp2-carbon (i.e. C-2 for 4 and C-6 for 5 and 6). Interestingly, 4 and 5 have more negative charge at the anomeric nitrogen atom (N-1) than isofagomine (3). Additionally, 4 and 5 have a negative group charge at NH-1, in contrast to the positive group charge of NH2-1 in isofagomine (3). This may explain why 4 and 5 are poor β-glucosidase inhibitors and 3 is a very potent β-glucosidase inhibitor; the negative NH-1 group of 4 and 5 prevents charge–charge interactions with the active site carboxylates, whereas the positive NH2-1 group of 3 allows charge–charge interactions with the same residues. Our hypothesis agrees Figure 3. Selected atomic and group charges for 3–6. aFrom Lindbäck et al. [13]. Table 1. IC50 and Ki values for 3–6. IC50 (µM) and Ki (µM) enzyme 3 4 5 6 α-glucosidasea, pH 6.8 — — IC50 > 50 IC50 > 50 β-glucosidaseb, pH 6.8 Ki = 0.11gKi = 450gIC50 > 50 IC50 > 50 α-galactosidasec, pH 6.8 IC50 > 2000hNIg,i IC50 > 50 IC50 > 50 β-galactosidased, pH 6.8 IC50 = 270hKi = 17 400gIC50 > 50 IC50 > 50 β-galactosidasee, pH 6.8 — — IC50 > 50 IC50 > 50 α-mannosidasef, pH 5.6 IC50 = 250hKi = 0.75gIC50 >50 IC50 > 50 aFrom Saccharomyces cerevisiae. bFrom almonds. cFrom green coffee beans. dFrom Aspergillus oryzae. eFrom Escherichia coli. fFrom jack beans. g[13]. h[14]. iNI = no inhibition. 4 royalsocietypublishing.org/journal/rsos R. Soc. Open Sci. 12: 241877 Downloaded from https://royalsocietypublishing.org/ on 04 June 2025 with an earlier study in which two regioisomeric gluco-imidazoles were compared for the inhibition of retaining β-glucosidases [20]. The study showed that the compound with the most positive charge on the anomeric carbon displayed the highest activity. However, our hypothesis fails to explain why N-benzylamidine 6 is a poor β-glucosidase inhibitor, as it hosts a slight positive charge at N-1. Perhaps the benzyl group connected to N-1 interferes with the charge–charge interactions between N-1 and the active site carboxylates and/or is lacking an interaction site nearby the active site of the enzyme. 2. Conclusions We propose that amidine 5 is a poor β-glucosidase inhibitor because its negatively charged anomeric NH group is unable to establish ionic interactions with the active site carboxylate groups. The fact that 6, hosting a slightly positive charge on the anomeric nitrogen atom, is also a poor β-glucosidase inhibitor may be due to the Bn group interfering with the formation of charge–charge interactions with the active site carboxylate groups. Ethics. This work did not require ethical approval from a human subject or animal welfare committee. Data accessibility. The data supporting this article are included as part of the electronic supplementary material [21]. Declaration of AI use. We have not used AI-assisted technologies in creating this article. Authors’ contributions. I.C.V.H.: investigation, methodology, writing—original draft; I.A.C.: investigation, methodology, writing—review and editing; L.P.: investigation, methodology, writing—review and editing; M.M.: investigation, methodology; N.W.: validation; T.C.S.E.: investigation, methodology, writing—review and editing; M.O.S.: writing—review and editing; F.L.A.: investigation, methodology, resources, supervision, writing—original draft; Ó.L.: investigation, methodology, resources, writing—review and editing; E.L.: conceptualization, funding acquisition, methodology, project administration, resources, supervision, writing—original draft. All authors gave final approval for publication and agreed to be held accountable for the work performed therein. Conflict of interest declaration. We declare we have no competing interests. Funding. Financial support from the University of Stavanger, Plogen, and the Research Council of Norway through project 226244 (NNP-Norwegian NMR Platform) is gratefully acknowledged. We thank Grant PID2020-116460RB I00 funded by MCIN/AEI/10.13039/501100011033, and Junta de Andalucía (FQM134) for financial support. References 1. Stütz AE, Wrodnigg TM. 2011 Imino sugars and glycosyl hydrolases: historical context, current aspects, emerging trends. Adv. Carbohydr. Chem. 66, 187–298. (doi:10.1016/b978-0-12-385518-3.00004-3) 2. Goss PE, Reid CL, Bailey D, Dennis JW. 1997 Phase IB clinical trial of the oligosaccharide processing inhibitor swainsonine in patients with advanced malignancies. Clin. Cancer Res. 3, 1077–1086. 3. Ferhati X, Matassini C, Fabbrini MG, Goti A, Morrone A, Cardona F, Moreno-Vargas AJ, Paoli P. 2019 Dual targeting of PTP1B and glucosidases with new bifunctional iminosugar inhibitors to address type 2 diabetes. Bioorg. Chem. 87, 534–549. (doi:10.1016/j.bioorg.2019.03.053) 4. Sánchez-Fernández EM, Fernández JMG, Mellet CO. 2016 Glycomimetic-based pharmacological chaperones for lysosomal storage disorders: lessons from Gaucher, GM1-gangliosidosis and Fabry diseases. Chem. Commun. 52, 5497–5515. (doi:10.1039/C6CC01564F) 5. Karade SS et al. 2023 Structure-based design of potent iminosugar inhibitors of endoplasmic reticulum α-glucosidase I with anti-SARS-CoV-2 activity. J. Med. Chem. 66, 2744–2760. (doi:10.1021/acs.jmedchem.2c01750) 6. Heightman TD, Vasella AT. 1999 Recent insights into inhibition, structure, and mechanism of configuration-retaining glycosidases. Angew. Chem. Int. Ed. 38, 750–770. (doi:10.1002/(SICI)1521-3773(19990315)38:6<750::AID-ANIE750>3.0.CO;2-6) 7. Tong MK, Papandreou G, Ganem B. 1990 Potent, broad-spectrum inhibition of glycosidases by an amidine derivative of D-glucose. J. Am. Chem. Soc. 112, 6137–6139. (doi:10.1021/ja00172a045) 8. Papandreou G, Tong MK, Ganem B. 1993 Amidine, amidrazone, and amidoxime derivatives of monosaccharide aldonolactams: synthesis and evaluation as glycosidase inhibitors. J. Am. Chem. Soc. 115, 11682–11690. (doi:10.1021/ja00078a004) 9. Jespersen TM, Dong W, Sierks MR, Skrydstrup T, Lundt I, Bols M. 1994 Isofagomine, a potent, new glycosidase inhibitor. Angew. Chem. Int. Ed. Engl. 33, 1778–1779. (doi:10.1002/anie.199417781) 10. Bols M. 1998 1-Aza sugars, apparent transition state analogues of equatorial glycoside formation/cleavage. Acc. Chem. Res. 31, 1–8. (doi:10. 1021/ar970058r) 11. Varrot A, Tarling CA, Macdonald JM, Stick RV, Zechel DL, Withers SG, Davies GJ. 2003 Direct observation of the protonation state of an imino sugar glycosidase inhibitor upon binding. J. Am. Chem. Soc. 125, 7496–7497. (doi:10.1021/ja034917k) 12. Lillelund VH, Liu H, Liang X, Søhoel H, Bols M. 2003 Isofagomine lactams, synthesis and enzyme inhibition. Org. Biomol. Chem. 1, 282–287. (doi: 10.1039/b208784g) 5 royalsocietypublishing.org/journal/rsos R. Soc. Open Sci. 12: 241877 Downloaded from https://royalsocietypublishing.org/ on 04 June 2025 13. Lindbäck E, López O, Fernández-Bolaños JG, Sauer SP, Bols M. 2011 An isofagomine analogue with an amidine at the pseudoanomeric position. Org. Lett. 13, 2908–2911. (doi:10.1021/ol200942g) 14. Ichikawa Y, Igarashi Y, Ichikawa M, Suhara Y. 1998 1-N-Iminosugars: potent and selective inhibitors of β-glycosidases. J. Am. Chem. Soc. 120, 3007–3018. (doi:10.1021/ja973443k) 15. Rousselet G, Capdevielle P, Maumy M. 1993 Copper(I)-induced addition of amines to unactivated nitriles: the first general one-step synthesis of alkyl amidines. Tetrahedron Lett. 34, 6395–6398. (doi:10.1016/0040-4039(93)85054-Z) 16. Doherty GG et al. 2023 Synthesis of uronic acid 1-azasugars as putative inhibitors of α-iduronidase, β-glucuronidase and heparanase. ChemBioChem 24, e202200619. (doi:10.1002/cbic.202200619) 17. van Rijssel ER, Janssen APA, Males A, Davies GJ, van der Marel GA, Overkleeft HS, Codée JDC. 2017 Conformational behaviour of azasugars based on mannuronic acid. ChemBioChem 18, 1297–1304. (doi:10.1002/cbic.201700080) 18. Krȩżel A, Bal W. 2004 A formula for correlating pKa values determined in D₂O and H₂O. J. Inorg. Biochem. 98, 161–166. (doi:10.1016/j.jinorgbio. 2003.10.001) 19. Wang B, Olsen JI, Laursen BW, Poulsen JCN, Bols M. 2017 Determination of protonation states of iminosugar–enzyme complexes using photoinduced electron transfer. Chem. Sci. 8, 7383–7393. (doi:10.1039/C7SC01540B) 20. Schröder SP etal. 2018 Gluco-1H-imidazole: a new class of azole-type β-glucosidase inhibitor. J. Am. Chem. Soc. 140, 5045–5048. (doi:10.1021/ jacs.8b02399) 21. Vaaland Holmgard IC, Christensen IA, Pejov L, Moren M, Westlund N, Evangelista TCS etal. 2025. Supplementary Material from: An Isofagomine Analogue with an Amidine Group in the 1,6-Position. FigShare (doi:10.6084/m9.figshare.c.7669967) 6 royalsocietypublishing.org/journal/rsos R. Soc. Open Sci. 12: 241877 Downloaded from https://royalsocietypublishing.org/ on 04 June 2025