Direct synthesis of anomeric tetrazolyl iminosugars from sugar-derived lactams

Herein we present the direct asymmetric synthesis of tetrazole-functionalized 1 - deoxynojirimycin derivatives from simple sugars via a Schwartz’s reagent -mediated reductive amide functionalization followed by a variant of the Ugi-azide multi-component reaction. The anomeric configurations of two products were unambiguously confirmed by X -ray analysis. This work also describes examples of interesting further transformations of the title products. Finally, some surprising observations regarding the mechanism of their formation were made.

amide functionalization; iminosugars; Schwartz’s reagent; tetrazole

Transformation of an amide into another chemical moiety in a controlled manner is not a trivial task. Although the Vilsmeier -Haack reaction1 or amide reduction with LiAlH4 are textbook examples that easily come to mind, there are not many other methods available. Simple alkyl and aryl amides, unlike other carbonyl compounds, typically do not undergo direct addition by a nucleophile, including active organom etallic 2 compounds. For this rea son, it has been chemists’ long -lasting ambition to develop a reliable, mild, and selective methodology for amide functionalization. 2 Even though a tremendous amount of work has been already done towards this matter, it is still a highly active field of research. Several review articles have been written about this topic, enclosing most of the advances made to date.3–5 A fascinating subset of these transformations encompass the reduction of amides to imines, with direct subsequent functionalization. One of th e methodologies for such a modification was developed by Charette et al . In their procedure the combination of triflic anhy dride and pyridine 6 (or its 2 -fluoro derivative 7) was used as an activating agent to transform amides into reactive iminium complexes. Another stoichiometric approach was presented by Georg et al. by utilization of zirconocene chloride hydride, known as Schwartz’s reagent. 8 This reduces an amide moiety, giving a complex that can be readily transformed into an imine or iminium cation.9 There have also been some catalytic protocols developed for the reduction of amides to imines. The most notable examples incorporate iridium complexes and silanes.10,11 Cheng and Brookhart showed that the chlorobis(cyclooctene)iridium dimer ([Ir(coe)2Cl]2) can act as the catalyst in combination with Et 2SiH2.12 Surprisingly, they were able to obtain imines as well as amines using this methodology. Ba sed on the works of Nagashima, 13 an iridium-based protocol for tertiary amides was introduced by Dixon14–16 and Huang. 17,18 Adolfsson expanded this by use of molybdenum -based catalysts.19 The reductive ap proach allows the issues associated with nucleophilic addition to amide carbonyl groups to be overcome and as such is finding its place in a growing number of synthetic applications.20 3 Employment of these methods for modification of lactams is a challenge in its own right — there are hardly any examples of such transformations available in the literature.21 Our group was the first to surmount this challenge by means of Schwartz’s reagent-mediated reductive functionalization. Since then, we have performed a number of diff erent functionalizations of such cyclic systems with various complexity, and with a particular focus on the modification of sugar -derived lactams. As summarized in Scheme 1, this includes simple nuclephile addition to in situ generated imines, 21 the consecutive one-pot Mannich/Michael sequence lead ing to oligocyclic compounds, 22 and employment in subsequent Joulié-Ugi multi-component reactions.23 Scheme 1: Our previous e fforts in the field of functionalization of sugar -derived lactams. This work is an extension of these e ff orts and seeks to inves tigate the possibility of incorporating the Ugi -azide multi -component reaction in this workflow. A molecule incorporating both an iminosugar an d a tetrazole fragment is of particular interest, due to the interesting properties of both moieties (Figure 1) . It is probably hard to overestimate the importance of sugar sca ff olds in nature, and we believe that it speaks for itself, however a significance of iminosugar derivatives may be less obvious. Several pharmaceuticals are based on this sca ff old including the glucose -derived nojirimycin, an antibiotic and glycosidase inhibitor 24 and 1-deoxygalactonojirimycin, known under a 4 trade name Galafold ®, which is utilized for the treatment of Fabry disease, a rare genetic condition. 25 On the other hand, the tetrazole moiety is known to have a bioisosteric relationship to carboxylic acids 26, which also makes them suitable for usage as bi ologically active compounds. Moreover, there are numerous reports of the organocatalytic activity of chiral aminotriaz oles and aminotetrazoles in num ber of reactions, such as the aldol reaction, 27 Michael addition, 28 Mannich reaction, 29 and hydrogenation.30 Figure 1: Key concepts behind the goal of this work. 31

and Discussion Quite recently Xie and Dixon showed that it is possible to synthesize α-tetrazolo amines from simple and linear tertiary amides using an iridium -based catalytic protocol.16 They have, however, only reported one example of lactam functionalization which only proceeded with moderate e fficiency (1-tert -butyl-azepan-2-one, 41 % yield of des ired product). Unfor tunately, this approach cannot be utilized for the functionalization of secondary amides, like sugar -derived lactams, due to the afore - mentioned method’s limitation to tertiary amides. Our previous work shows that Charette’s methodology is also not applicable in this case, as it does not lead to the formation of an imine. 21 Luckily, we we re able to use a formerly estab lished strategy based on Georg’s procedure with standard Ugi -azide32–36 reaction conditions in a one- 5 pot, tandem process. Subjecting glucose-derived lactam 1 to such a procedure gave the desired product in good yield, but with virtually no diastereoselectivity, as shown in Scheme 2. Scheme 2: Preliminary experiment in search of a procedure for the synthesis of 2 -(1H- tetrazol-5-yl)-iminosugars. Optimization and scope An initial optimization study for the proton donor for TMSN 3 activation (shown in Table 1) using commo nly encountered reagents for such reactions was performed. To our surprise, we observed the formation of the Ugi-azide product even in the absence of a protic additive . Moreover, the aprotic conditions proved to provide the highest yield and diastereoselectivity, thus were chosen as optimal (Table 1, entry 9.). We also tried to isolate the imine after the reduction step and carry out the second step in a solvent commonly used for the Ugi -azide reaction alone. For this, we observed a significant decrease in overall yield and suspect that the low stability of imines of type 2 may be the reason for this behaviour. Table 1: Optimization of 2 -(1H-tetrazol-5-yl)-iminosugar synthesis via Schwartz’s reagent-mediated reduction of amides and Ugi-azide reaction. 6 № Additive Solvent Yield /% d.r. a 1. MeOH b THF 65 43:57 2. CF3CO2H THF 24 43:57 3. AcOH THF 47 80:20 4. Et3N · HCl THF 45 74:26 5. H2O THF 34 >95:5 6. (CF3)2CHOH THF 35 >95:5 7. none MeOH 19 c >95:5 8. none DCM 36 c >95:5 9. none THF 73 >95:5 A: 1.6 equiv. Cp 2Zr(H)Cl in THF under argon atmosphere; B: 1.6 equiv. of additive (if applicable), 1.1 equiv. CyNC, and 1.1 equiv. TMSN 3. a 2-(R) to 2-(S), isolated. b Additive used in excess. c Imine was isolated after reduction. The established optimal conditions were applied for the synthesis of selected examples of various 2 -(1H -tetrazol-5-yl)-iminosugars (Table 2). Attempts at using this methodology to synthesize pentose -derived 2 -(tetrazol-5-yl)-iminosugars, using 2,3,5 - tri-O-benzyl-D-ribo-furanose- and -arabinofuranose-derived lactams as substrates were made. Very unexpectedly, we failed to isolate such products although we did observe their formation via mass spec trometry of the reaction mixtures. Employing alternative procedures did not help, and none of the desired products were observed at all when applying iridium complexes- or triflic anhydride-based methods. Table 2: Synthesis of 2-(1H -tetrazol-5-yl)-iminosugars using optimized conditions. Reaction yield and d.r. are given. 7 № Product -R2 Yield /% d.r.a 1. 3a Cy 73 >95:5 2. 3b CH2CO2Et 49 >95:5 3. 3c Bn 18 >95:5 4. 3d PMP 29 79:21 5. 3e PMB 42 >95:5 6. 3f tert-Bu 40 >95:5 7. 3g tert-Oct 48 >95:5 8. 5a Cy 33 >95:5 9. 5b CH2CO2Et 16 >95:5 a 2-(R) to 2-(S), isolated. The methodology described here provides a pathway to new, interesting compounds, containing both an iminosugar and tetrazole moiety. Such compounds have not been seen to date, and their accessibility creates exciting synthetic opportunities. Here we present two examples of possible further transformations of the products obtained over the course of this research directed towar ds novel, attractive molecules. Compound 3b underwent a cyclization reaction in the presence of benzoic acid at an elevated temperature yielding lactam 6 almost quantitatively. Deoxygenative reduction of this compound turned out to be challenging, as the ty pical procedure using LiAlH4 proved ine ff ective. We were able to obtain 7 using a Schwartz’s reagent - mediated amide activation methodology followed by NaBH 4 reduction. This structure with three condensed rings can be seen as a new class of unnatural, chiral alkaloid scaff old, potentially exhibiting pharmacological activity (Scheme 3).37 8 Scheme 3: Synthesis of a new class of alkaloid sca ffold using the presented methodology. Various unsuccessful attempts were made to deprotect compound 3e. Unexpectedly, however, one of those experiments resulted in rearrangement in the tetrazole ring, as shown in Scheme 4. We were able to obtain the desired aminotetrazole 9 by treating 3g with dry HCl at elevated temperature (Scheme 5). The resulting compound is particularly appealing, as similar sca ffolds are widely used as organocatalysts. Such moieties are employed in a number of important synthetic transformations, including the aldol reaction, 27 Michael addition, 28 Mannich reaction, 29 and hydrogenation.30 We plan to test these possibilities in the near future. Scheme 4: Rearrangement of 3e under acidic conditions. Scheme 5: Synthesis of a new, chiral 2 -(tetrazol-5-yl)-iminosugar based potential organocatalyst. 9 Stereochemistry and configuration of products As presented in Table 2, only one diastereomer of the desir ed iminosugar is obtained in almost all cases. This outstanding selectivity has been observed before and is described in our previous works devoted to the functionalization of sugar -derived lactams.21–23 We explain it in light of Woerpel’s model, which characterizes the direction of nucleophilic addition to oxo carbenium ions. 38–40 According to this concept, the conformational stability of the compound in question is the key property to consider when predicting the reaction’s stereoselectivity. When the oxocarbenium ion is substituted, two diastereomeric half-chair conformers are possible: 3H4 and 4H3 (shown for a 4 -substituted pyranose cation in Scheme 6). Both may undergo attack by a nucleophile in two ways: on the axial trajectory from the top or the bottom face. Such an event would result in the formation of the product as a chair (1C4, 4C1) or a skew-boat (1S3, 3S1) conformer, of which the former is favored, as it proceeds via the lower-energetic chair-like transition state. The favored path of action will result in addition syn or anti to the substituent in position 4, depending on the starting conformer. Therefore, once the ground conformer of the oxocarbenium ion is established, this logic may be used to predict the reaction’s stereochemistry. 10 Scheme 6: Principle behind Woerpel’s model for pred iction of the direction of nucleophile addition to oxocarbenium cations. The same principle may be successfully applied to reactions of iminium cations. We have previously shown that in the case of glucose - and galactose-derived, O -benzyl- protected iminos ugars the addition syn to the substituent in position 3 is favored (Scheme 7). This work proves no di ff erent, as the isolated major products were in such configuration. The experimental determination of this, however, was not straig htforward in all cases. Scheme 7: Diff erence in conformational stability of glucose - and galactose -derived iminium cations and the major product of nucleophile attack according to Woerpel’s model.22 We were able to determine the structure of compounds 3a and 3e unambiguously by means of X -ray analysis, as shown in Figure 2. The configuration of the remaining glucose based products 3 was easily determined by the analysis of 1H–1H coupling constants and NOE e ff ects. Unfortunately, the same approach was not possible in the case of compounds 5, as 1H NMR spectroscopy showed indefinite results. In compound 5a the coupling constant between protons H 2 and H3 has a value of 8.5 Hz. This cannot 11 be associated with a particular relative configuration without comparison with the corresponding coupling constant in 2-epi-5a. But, alas, this value is unknown, due to of overlapping and broadening of the relevant signals in the 1H NMR spectrum of the compound in question. For the same reasons NOE e ffects present in 2-epi-5a cannot be accurately interpreted. However, analysis of NOE e ffects in 5a, particularly a small eff ect between protons H 2 and H 7 suggest that it may be the diastereomer 2 -(R), as shown in Figure 3 . This result would be in accordance with the previously mentioned Woerpel’s model. Figure 2: ORTEP structures of compounds 3a and 3e obtained by X -ray analysis. Hydrogen atoms and benzyl groups are omitted for clarity. Full crystallographic data available in Supplementary Information File 2 and 3, and in Cambridge Crystallographic Database under CCDC-2001373 and CCDC-2001372 numbers respectively. 12 Figure 3: Proposed absolute configuration (2-(R)) of compound 5a with selected carbon atoms numbered. We made an attempt at resolving this problem by means of the electronic circular dichroism (ECD) technique. We recorded an ECD spectrum of both compounds and compared it with simulated spectra, generated for both possible diastereomers (2 -(R) and 2-(S)) using computational chemistry software. Unfortunately, we were not able to fit any of the se simulations to the experimental data with sufficient certainty. For the inquisitive readers, this work is fully described in the supporting information section. Mechanism of reaction As mentioned previously, we observed Ugi-azide products, despite the absence of a proton donor in the reaction mixture. Intriguingly, this behaviour is inconsistent with the generally accepted mechanism of this transformation, which assumes hydrolysis of TMSN3 to HN 3 and activation of the imine species by protonation. Scheme 8 presents our proposal for the possible course of the Ugi -azide reaction variant described in this work. We suppose that after reduction of amide I by Schwartz’s reagent, complex II undergoes a slow, spontaneous decomposition, y ielding imine III. III then reacts with TMSN3, which acts as both, an imine activator and an azide anion source. Complex IV undergoes a subsequent addition of an isocyanide moiety (intermediate V), followed by an azide anion addition. Intermediate VI undergoes a cyclization, producing VII, a silylated derivative of the expected product. The hydrolysis of VII most likely occurs during the reaction’s work-up. 13 Scheme 8: Proposed reaction mechanism for the described Ugi-azide reaction variant. Preliminary DTF calculations were performed on a simplified model to provisionally validate this proposed mechanism. The geometry of the intermediate species were optimized with Gaussian 09 software 41, using the B3LYP/LANL2DZ theory leve l for Zr and B3LYP/6 -31G(d,p) for other atoms, with GD3 empirical dispersion correction. Optimization was followed by a single-point energy calculation using the larger basis set Def2TZVP with a PCM solvatation model for THF, as implemented in the Gaussian software. Energy values reported are a sum of electronic and zero-point energies. Scheme 9 shows possible pathways for the spontaneous decomposition of zirconium com plex INT-1-A to free imine species INT-3. This process is much more likely to occur via the 5-memberd cyclic transition state TS-1-A than the alternative TS- 1-B, as the energy barrier of 60.1 kcal·mol−1 is definitely too high for the reaction to take place, even at an elevated temperature. Path A with a barrier of 22.6 kcal·mol−1 is certainly more feasible. We assume that the Cp 2Zr(OH)Cl species just leaves the initial complex, as this seems to be the simplest possibility in absence of any Lewis acid which could catalyze this decomposition. 14 Scheme 9: Possible pathway for spontaneous imine formation. Values reported are in kcal·mol−1. Scheme 10 shows the energy diff erences in the subsequent steps of the examined reaction. The reported energy barriers are reasonably high for a slow process taking place at room temperature. The overall barrier is not considerably di ff erent to those previously published for typical mechanisms o f tetrazole formation by azide addition to nitriles.42 It is important to note that the computational investigation of this reaction’s mechanism was not a primary goal of this work. That said, we consiser this simple, crude DFT research to support our model of the transformation described herein. 15 Scheme 10: A possible path for tetrazole formation in the described conditions. Values reported are in kcal·mol−1.

During the course of this research we have developed a methodology for the synthesis of sugar -derived α-tetrazolyl amines. Such compounds — incorporating both iminosugar and tetrazole fragments — are particularly interesting, thanks to the well - known biological and catalytic activity of these moieties. This work is the first example of using Schwartz’s reagent -mediated partial reduction of lactams and the Ugi -azide multicomponent reacition in a tandem pro cess. Yields of the described products are moderate to good, a satisfying result for such a multi-step process. We have shown that such a reaction does not necessarily requires protic conditions, in opposition to what is 16 generally agreed upon for these type of reactions. An alternative reaction mechanism is proposed and provision ally confirmed with DFT calcula tions. Moreover, selected α- tetrazolyl iminosugars were subjected to further transformations, yielding new, potentially biologically active and organocatalytic compounds. Experimental Experimental procedures and other data are available in Supporting Information File 1. Supporting Information Supporting Information File 1: ESI; pdf; Experimental data and additional details; experimental procedures, characterisation of compounds, ECD analyses for compounds 5a and 2-epi-5a, calculations of appropriate ECD and UV spectra, crystallographic data for compounds 3a and 3e, atomic coordinates, energies, and number of imaginary frequencies for computed stationary points, and copies of 1H NM R and 13C NMR spectra. Supporting Information File 2: CCDC-2001373; cif; X-ray crystallographic data for compound 3a. Supporting Information File 2: CCDC-2001372; cif; X-ray crystallographic data for compound 3e.

We would like to thank Magdalena Jawiczuk for her invaluable tips and guidance in the field of chemical computations. 17 Founding We would like to thank National Science Centre of Poland for the financial support provided with grant PRELUDIUM № 2017/25/N/ST5/00079 (in the fie lds of synthesis and analysis). This research was supported in part by PLGrid Infrastructure (in the f ield of chemical computations).

(1) Vilsmeier, A.; Haack, A. Berichte der Deutschen Chemischen Gesellschaft 1927, 60B, 119–22. (2) Pace, V.; Holzer, W.; Olof sson, B. Advanced Synthesis and Catalysis 2014, 356, 3697–3736. (3) Chaudhari, M. B.; Gnanaprakasam, B. Chemistry - An Asian Journal 2019, 14, 76– 93. (4) Volkov, A.; Tinnis, F.; Slagbrand, T.; Trillo, P.; Adolfsson, H. Chemical Society Reviews 2016, 45, 6685–6697. (5) Kaiser, D.; Bauer, A.; Lemmerer, M.; Maulide, N. Chemical Society Reviews 2018, 47, 7899–7925. (6) Charette, A. B.; Grenon, M. Canadian Journal of Chemistry 2001, 79, 1694–1703. (7) Pelletier, G.; Bechara, W. S.; Charette, A. B. Journal of the American Chemical Society 2010, 132, 12817– 12819. (8) Leighty, M. W.; Spletstoser, J. T.; Georg, G. I. Mild Conversion of Tertiary Amides to Aldehydes using Cp 2Zr(H)Cl (Schwartz's Reagent). In Organic Syntheses; John Wiley and Sons, Inc., 2003. 18 (9) Spletstoser, J. T.; White, J. M.; Tunoori, A. R.; Georg, G. I. Mild and Selective Hydrozirconation of Amides to Aldehydes Using Cp 2Zr(H)Cl: Scope and Mechanistic Insight. Journal of the American Chemical Society 2007, 129, 3408– 3419. (10) Andersson, P. G.; Editor, Iridium Catalysis. In Top. Organomet. Chem., 2011; 34 ; Springer GmbH, 2011. (11) Matheau-Raven, D.; Gabriel, P.; Leitch, J. A.; Almehmadi, Y. A.; Yamazaki, K.; Dixon, D. J. ACS Catal. 2020, 10, 8880–8897. (12) Cheng, C.; Brookhart, M. Journal of the American Chemical Society 2012, 134, 11304–7. (13) Motoyama, Y.; Aoki, M.; Takaoka, N.; Ao to, R.; Nagashima, H. Chemical Communications 2009, 1574– 1576. (14) Gregory, A. W.; Chambers, A.; Hawkins, A.; Jakubec, P.; Dixon, D. J. Chemistry 2015, 21, 111–4. (15) Xie, L.-G.; Dixon, D. J. Chem. Sci. 2017, 8, 7492–7497. (16) Xie, L.-G.; Dixon, D. J. Nature Communications 2018, 9, 2841. (17) Xiao, K. -J.; Luo, J. -M.; Ye, K. -Y.; Wang, Y.; Huang, P. -Q. Angewandte Chemie International Edition 2010, 49, 3037–3040. (18) Huang, P. Q.; Ou, W.; Han, F. Chemical Communications 2016, 52, 11967–11970. (19) Tinnis, F.; Volkov, A.; Slagbrand, T.; Adolfsson, H. Angewandte Chemie International Edition 2016, 55, 4562–4566. (20) Więcław, M. M.; Stecko, S. European Journal of Organic Chemistry 2018, 2018, 6601–6623. 19 (21) Szcześniak, P.; Stecko, S.; Staszewska -Krajewska, O.; Furman, B. Tetrahedron 2014, 70, 1880– 1888. (22) Szcześniak, P.; Stecko, S.; Maziarz, E.; Staszewsk a-Krajewska, O.; Furman, B. The Journal of Organic Chemistry 2014, 79, 10487–10503. (23) Szcześniak, P.; Maziarz , E.; Stecko, S.; Furm an, B. The Journal of Organic Chemistry 2015, 80, 3621– 3633. (24) Inouye, S.; Tsuruoka, T.; Ito, T.; Niida, T. Tetrahedron 1968, 24, 2125–2144. (25) Sánchez-Fernández, E. M.; García Fernández, J. M.; Mellet, C. O. Chemical Communications 2016, 52, 5497– 5515. (26) Malik, M. A.; Wani, M. Y.; Al -Thabaiti, S. A.; Shiekh, R. A. Journal of Inclusion Phenomena and Macrocyclic Chemistry 2014, 78, 15–37. (27) Hartikka, A.; Arvidsson, P. I. Tetrahedron: Asymmetry 2004, 15, 1831–1834. (28) Chen, H.; Zhang, D.; Xue, F.; Qin, Y. Tetrahedron 2013, 69, 3141–3148. (29) Kumar, I.; Ramaraju, P.; Mir, N. A.; Singh, D.; Gupta, V. K.; Rajnikant, Chemical Communications 2013, 49, 5645–5647. (30) Mirabal-Gallardo, Y.; Piérola, J.; Shankaraiah, N.; Santos, L. S. Tetrahedron Letters 2012, 53, 3672–3675. (31) The color palette used in t his document is optimized for color -blind individuals as proposed by Wong: Wong, B. Nature Methods 2011, 8, 441–441. (32) Ugi, I.; Chemische Berichte 1961, 94, 734–742. (33) Ramezanpour, S.; Balalaie, S.; Rominger, F.; Alavijeh, N. S.; Bijanzadeh, H. R. Tetrahedron 2013, 69, 10718–10723. (34) Shmatova, O. I.; Nenajdenko, V. G. The Journal of Organic Chemistry 2013, 78, 9214–9222. 20 (35) Safa, K. D.; Shokri, T.; Abbasi, H.; Teimuri -Mofrad, R. Journal of Het erocyclic Chemistry 2014, 51, 80–84. (36) Santhosh, L.; Nagamangala, S. R.; Thimmalapura, V. M.; Vommina, S. V. ChemistrySelect 2017, 2, 5497–5500. (37) Finsinger D.; Wucherer-Plietker M.; Blume B.; Tetrahydro-tetrazolo[1,5-a]pyrazines as ROR-gamma inhibitors. 2015; Patent WO2015090507A1, Merck Patent GmbH, Germany. (38) Larsen, C. H.; Ridgway, B. H.; Shaw, J. T.; Woerpel, K. A. A Journal of the American Chemical Society 1999, 121, 12208– 12209. (39) Romero, J. A. C.; Tabacco, S. A.; Woerpel, K. A. Journal of the American Chemical Society 2000, 122, 168–169. (40) Woods, R. J.; Andrews, C. W.; Bowen, J. P. Journal of the American Chemical Society 1992, 114, 859–864. (41) Frisch, M. J. et al. Gaussian 09 Revision E.01. 2016; Gaussian Inc. Wallingford CT. (42) Himo, F.; Demko, Z. P.; Noodleman, L.; Sharpless, K. B. Journal of the American Chemical Society 2002, 124, 12210–12216.