Full text
1 sp2-Iminosugar Glycolipids as Inhibitors of Lipopolysaccharide-Mediated Human Dendritic Cell Activation in Vitro and of Acute Inflammation in Mice in Vivo Evelyne Schaeffer,1 Elena M. Sánchez-Fernández,2 Rita Gonçalves-Pereira,2 Vincent Flacher,1 Delphine Lamon,1 Monique Duval,1 Jean-Daniel Fauny,1 José M. García Fernández,3 Christopher G. Mueller,1* Carmen Ortiz Mellet2* 1 Université de Strasbourg, CNRS, Immunopathology and Therapeutic Chemistry, UPR 3572, 67000 Strasbourg, France 2 Department of Organic Chemistry, Faculty of Chemistry, University of Seville, c/ Profesor García González 1, 41012 Seville, Spain 3 Chemical Research Institute (IIQ), CSIC – University of Seville, Avda. Américo Vespucio 49, E-41092 Seville, Spain *Corresponding authors. C. G. M. Tel: +33 (0)3 8841 7027; fax: +33 (0)3 8861 0680. E-mail: [email protected] C.O.M. Tel.: +34 9 5455 9806; fax: +34 9546 24960. E-mail: [email protected]
2 Abstract: Glycolipid mimetics consisting of a bicyclic polyhydroxypiperidine-cyclic carbamate core and a pseudoanomeric hydrophobic tail, termed sp2-iminosugar glycolipids (sp2-IGLs), target microglia during neuroinflammatory processes. Here we have synthesized and investigated new variants of sp2-IGLs for their ability to suppress the activation of human monocyte-derived dendritic cells (DCs) by lipopolysaccharide (LPS) signaling through Tolllike receptor 4. We report that the best lead was (1R)-1-dodecylsulfonyl-5N,6O- oxomethylidenenojirimycin (DSO2-ONJ), able to inhibit LPS-induced TNFα production and maturation of DCs. Immunovisualization experiments, using a mannoside glycolipid conjugate (MGC) that also suppress LPS-mediated DC activation as control, evidenced a distinct mode of action for the sp2-IGLs: unlike MGCs, DSO2-ONJ did not elicit internalization of the LPS co-receptor CD14 or induce its co-localization with the Toll-like receptor 4. In a mouse model of LPS-induced acute inflammation, DSO2-ONJ demonstrated anti-inflammatory activity by inhibiting the production of the pro-inflammatory interleukin 6. The ensemble of the data highlights sp2-IGLs as a promising new class of molecules against inflammation by interfering in Toll-like receptor intracellular signaling. Highlights · sp2-Iminosugar glycolipids (sp2-IGLs) with S-linked aglycones have been synthesized. · The configuration modulates the antiinflammatory response in LPS-activated human DCs. · sp2-IGLs interfere with TLR-4 intracellular signaling, likely by binding to p38α MAPK. · The best candidate proved efficacy in an in vivo model of LPS-mediated inflammation. Keywords: Iminosugar; glycolipid; inflammation; dendritic cell; sulfoxide; sulfone
3 1. Introduction sp2-Iminosugars are an original family of sugar look-alikes possessing a pseudoamide-type nitrogen atom with a high sp2-hybridation character at the position of the endocyclic oxygen [1-3], with the unique property of mimicking not only the structure and function of the natural counterparts, as classical iminosugars do, but also their chemistry [4]. Most significantly, sp2- iminosugars can engage in glycosylation reactions, affording stable glycoconjugate analogs capable of agonizing or antagonizing the biological processes in which the natural counterparts are involved. Examples that illustrate the opportunities that sp2-iminosugar conjugates offer in biology and medicine are the design of pharmacological chaperones targeting lysosomal glycosidases [5], the synthesis of multivalent constructs to explore multivalent enzyme inhibition [6] or the preparation of Tn antigen glycopeptide mimics [7]. Pursuing our interest in the potential of sp2-iminosugars as monosaccharide surrogates for neoglycoconjugate drug development, we recently turned our attention to glycolipids as immunoactive agents for clinical and mechanistic applications, a very active research area in medicinal chemistry. Studies report that glycolipids exert antiinflammatory and antiproliferative activity, as well as beneficial effects on lipid metabolism [8]. The chemical synthesis of glycolipid analogues is often challenging, however, especially when it comes to producing pure anomeric forms of the sugar head group. Remarkably, sp2-iminosugars benefit from an exacerbated anomeric effect, which warrants total -stereoselectivity during O-, S or N- glycosylation steps and imparts chemical and enzymatic stability to the resulting glycosides. Indeed, some sp2-iminosugar glycolipids (sp2-IGLs) have been already prepared and found to exhibit a range of bioactivities including anti-proliferative [9], anti-metastatic [10,11], antiparasitic [12] and anti-neuroinflammatory activity [13,14]. The sp2-IGL representatives (1R,SR)1-dodecylsulfinyl-5N,6O-oxomethylidenenojirimycin (R-DSO-ONJ) and (1R)-1- dodecylsulfonyl-5N,6O-oxomethylidenenojirimycin (DSO2-ONJ) were particularly efficient at inhibiting the lipopolysaccharide (LPS)-induced inflammatory signaling in mouse Bv.2 microglial cells by decreasing inducible nitric oxide synthase (iNOS) and pro-inflammatory cytokine production and simultaneously upregulating production of the anti-inflammatory cytokine IL-10. The latter compound was able to significantly reduce reactive gliosis in retinal explants of the diabetic db/db mice model that recapitulates the progression of diabetic retinopathy [14,15]. Although initially the protein target was hypothesized to be an endoplasmic
4 reticulum -glucosidase, the sp2-IGLs were found not to impact glycoprotein biosynthesis, which is consistent with their lack of toxicity in healthy cells [16]. Instead, they seemed to interfere with the mitogen activated protein kinase (MAPK) cascade, at the crossroad between immunity and cancer, offering new therapeutic opportunities. Dendritic cells (DCs), present in most tissues and body liquids, are key players in the initiation and polarization of the immune response [17]. The investigation of key indicators of the inflammatory response in DCs upon treatment with sp2-IGLs was therefore mandatory to validate the above notion. These immune sentinels are highly sensitive to innate immune stimuli and mature into professional antigen-presenting cells producing inflammatory cytokines, such as TNFα, and upregulating the T cell co-stimulators CD83 and CD86 at the cell surface. Mature DCs are highly efficient in priming the T cell-mediated adaptive immune response [16]. A particularly efficient DC stimulator is LPS, an important toxin found in the outer membrane of Gram-negative bacteria involved in the development of septic shock, characterized by the hyperproduction of inflammatory cytokines and organ damage [18]. LPS-elicited acute inflammatory responses are mediated by the signaling receptor Toll-like receptor (TLR)-4. The signaling cascade is activated by the successive interaction of LPS with the lipopolysaccharide binding protein, the cluster differentiation antigen CD14, and the myeloid differentiation protein (MD-2) that binds LPS and presents it in a monomeric form to TLR-4 [19]. TLR-4 undergoes dimerization and recruits downstream adaptor proteins, which initiates an intracellular signaling cascade that leads to nuclear translocation of transcription factors and the biosynthesis of pro-inflammatory cytokines and interferons [20,21]. Therefore, therapeutic interventions to reduce TLR-4 signaling in response to LPS are of clinical importance. With the objective to gather information on the mode of action of sp2-IGLs and on the structural features that determine their anti-inflammatory properties, we have now generated a series of derivatives 1-8 (Figure 1) that keep the dodecylsulfinyl or dodecylsulfonyl lipidic aglycone moiety but differ in the hydroxylation pattern (D-gluco, D-galacto or D-manno) of the sp2- iminosugar aglycone, and studied their capacity to inhibit the activation of human DCs by LPS. The sulfone derivative DSO2-ONJ (compound 3), having a hydroxylation profile of structural complementarity to -D-glucopyranosides, emerged as the most active member in reducing TNFα production and downregulating the DC maturation markers CD83 and CD86. Interestingly, co-localization studies support that sp2-IGLs act by inhibition of intracellular
5 signaling rather than by modifying LPS receptor cell-surface localization. This behavior is sharply different from that reported for other neoglycolipids such as the mannoside glycolipid conjugate (MGC) 9 (Figure 1), which suppress TLR-4 signaling in human DCs by targeting early membrane steps of the CD14―TLR-4 signaling membrane complex [22-24]. In a mouse model of endotoxin shock, sp2-IGL 3 reduced the production of the inflammatory interleukin 6 (IL-6), supporting its therapeutic potential. Figure 1. Structures of sp2-IGL derivatives (1-8) and the mannoside glycolipid conjugate (MGC) 9 evaluated in this study. 2. Results 2.1. Design Rational and Synthesis.
6 Compounds 1-8 were chosen for the initial cell toxicity and anti-inflammatory activityscreening, primarily because sp2-IGLs bearing sulfur anomeric functional groups and twelvecarbon aliphatic aglycones were previously identified as privileged structures in anticancer/anti-inflammatory tests in different settings [13,14,16]. The sulfinyl and sulfonyl derivatives, differing in their oxidation state, have been suggested [14] to emulate the phosphate group in anti-inflammatory phosphatidylinositol analogs (PIAs) and alkyl ether lipids (e.g., perifosine) for which a mechanism of action involving p38 MAPK binding has been evidenced [25]. Dodecyl-sulfinyl and -sulfonyl conjugates were therefore considered with sp2-iminosugar moieties that mimic D-gluco (1-3), D-galacto (4-6) and D-manno (7 and 8) glycolipids in order to assess the effect of the glycone configuration in the anti-inflammatory activity. Compounds 1-3 were prepared following previously reported procedures [14,15]. The chemical syntheses of the new derivatives 4-8 follow a general scheme involving (i) steroselective -S- glycosidation of the corresponding peracetylated sp2-iminosugar (10 or 14) with dodecanethiol and (ii) partial or full oxidation of the thioglycoside intermediate (11 or 15) with mchloroperbezoic acid (m-CPBA) to afford the target sulfoxide and sulfone sp2-IGLs, respectively (Schemes 1 and 2). In the case of the galacto-configured sulfoxides, both the (SS) and the (RS) diastereomers 12 and 13 were formed in close to 1:1 ratio and could be separated by column chromatography before deacetylation to give 4 and 5, respectively. Their absolute stereochemistry was assigned based on their 1H NMR spectra. Notably, the H-5 resonance was deshielded by 0.67 ppm in 12 (δ 4.96 ppm) compared with 13 (δ 4.29 ppm). Literature data on alkyl -D-glycosyl sulfoxides support then the (SS) assignment for the first one, where H-5 and the sulfoxide oxygen would be located in close proximity in the more favourable exoanomerictype conformation, that is, with C-2 in the ring and the exocyclic methylene carbon in antidisposition [26]. The lower magnetic nonequivalence (Δδ) of the methylene protons vicinal to the chiral sulfur atom (SOCH2) for 12 (57 Hz) as compared with 5 (81 Hz) is also agreement with the NMR properties reported in the literature for (SS) and the (RS) alkyl -D-glycosyl sulfoxides, respectively [27]. The partial oxidation of the manno-configured thioglycoside 15 afforded a major sulfoxide product (16) that was obtained in pure form after column chromatography and assigned the (SS) configuration based on the Δδ value for the SOCH2 protons (48 Hz) [27]. The (RS) diastereomer was detected in the reaction mixture as a trace compound but could not be isolated in pure form. Finally, deacetylation of 17 by treatment with sodium methylate afforded the fully unprotected target sp2-IGL 7.
7 Scheme 1. Synthesis of the D-galacto-configured sp2-IGLs 4-6. Scheme 2. Synthesis of the D-manno-configured sp2-IGLs 7 and 8. 2.2. Toxicity of sp2-IGLs on DCs. Human monocyte-derived dendritic cells (DCs) were differentiated from monocytes purified from the blood of different donors. The cytotoxicity of the sp2-IGL derivatives was evaluated by incubation of the individual components 1-8 at 25 and 50 µM concentrations for 24 h, followed by exposure to 7-aminoactinomycin D, a DNA-intercalating fluorescent dye that only penetrates the cell upon cell death. As control, DMSO solvent was added. Assays were performed in parallel in the absence and presence of 100 ng mL-1 LPS. In its absence, the mean cell viability did not fall below an 80% threshold at 25 M or a 70% limit at 50 M upon treatment with the sp2-IGLs (Figure 2a, b). However, in the presence of LPS, cell viability was reduced and fall below 70% for compounds 5 and 6 at 50 M (Figure 2a, b). DMSO had no
8 significant impact on cell viability irrespective of LPS. To be in a relatively safe window, we settled the dose limit to 50 M for future experimentation with the sp2-IGLs. Figure 2. Effect of sp2-IGLs on the viability of human DCs. The different sp2-IGL compounds were used at (a) 25 M and (b) 50 M. Solvent controls were performed with DMSO at 0.5 and 1%. Cell viability in the absence or presence of LPS (100 ng·mL-1) was measured after 24 h by 7-aminoactinomycin D (7-AAD) staining, followed by flow cytometry analysis. Each point corresponds to a different monocyte blood donor. 2.3. sp2-IGLs Inhibit LPS Activation of Human DCs. We next examined the capacity of the selected sp2-IGLs representatives to inhibit the activation of DCs by LPS. By stimulating TLR-4-signalling, LPS is a potent activator of DCs leading to the production of TNFα and the upregulation of the T cell co-stimulators CD83 and CD86. A 24 h exposure time to LPS in the absence or presence of 25 µM and 50 µM of sp2-IGLs was established to warrant significant expression of endogenous CD83 and CD86 [28,29]. LPS induced the production of TNFα, as determined by ELISA, but there was a significant reduction of TNFα by compound 3 at 25 µM and by compounds 3-8 at 50 µM (Figure 3a, b). The expression of CD83 and CD86 on the cell surface was subsequently assessed by fluorescenceactivated flow cytometry (FACS). At 25 µM, compounds 1 and 4 significantly reduced CD83 levels (Figure 3c). As for CD86, only compound 3 was active at both 25 µM and 50 µM (Figure 3d). These results showed that most sp2-IGLs inhibited to some extent LPS-mediated DC maturation, compound 3 being the most active. Subsequent analysis of its TNFproduction dose-dependent inhibition on LPS-activated DCs (Figure 4a) allowed us to establish the dose-
9 response curve for compound 3 and its IC50 at 20 10 M (Figure 4b). Its efficient downregulatory effect of CD86 expression was also confirmed (Figure 4c), while there was a significant reduction of CD83 expression only at 50 µM (Figure 4d). Taken together, the sp2- IGL 3 displayed a significant anti-LPS activity using human DCs as test cells. Figure 3. sp2-IGLs are anti-inflammatory for LPS-activated DCs. Cells were stimulated with LPS (100 ng·mL-1) in the absence or presence of sp2-IGLs for 24 h. (a, 25 M; b, 50 M) Effect on the pro-inflammatory cytokine TNF- released in the media. (c, d) Effect on the relative expression of CD83 and CD86 cell surface maturation markers. Non-treated cells (n.t.; control) or treated with the solvent (DMSO) are indicated. Data correspond to at least 3 cell donors / independent assays. Bars represent means SEM.
16 BF3.OEt2 (134 μL, 1.09 mmol) and dodecylthiol (0.61 mmol) were added under Ar atmosphere. The mixture was stirred for 30 min (TLC monitoring), diluted with DCM (50 mL) and washed with water (10 mL), aq NaHCO3 (10 mL) and brine (10 mL), dried (MgSO4) and concentrated under reduced pressure. Purification by column chromatography (1:3 EtOAc-cyclohexane) yielded 11 (118 mg, 79%). Rf 0.68 (1:1 EtOAc-cyclohexane). [α]D +161.7 (c 1.0 in DCM).1H NMR (500 MHz, CDCl3): δ 5.78 (d, 1 H, J1,2 = 5.1 Hz, H-1), 5.43 (t, 1 H, J3,4 = J4,5 = 1.6 Hz, H-4), 5.26-5.16 (m, 2 H, H-2, H-3), 4.43-4.34 (m, 2 H, H-5, H-6a), 4.04 (dd, 1 H, J6a,6b = 13.5 Hz, J5,6b = 10.0 Hz, H-6b), 2.57 (ddd, 1 H, 2JH,H = 12.7 Hz, 3JH,H = 8.3 Hz, 3JH,H = 6.0 Hz, SCH2), 2.43 (ddd, 1 H, SCH2), 2.17-2.00 (3 s, 9 H, MeCO), 1.70-1.20 (m, 20 H, CH2), 0.87 (t, 3 H, 3JH,H = 7.0 Hz, CH3).13C NMR (125.7 MHz, CDCl3): δ 170.4-169.8 (MeCO), 155.7 (CO), 68.9 (C-3), 68.4 (C-4), 67.2 (C-2), 62.8 (C-6), 58.1 (C-1), 50.4 (C-5), 32.0-22.8 (CH2), 30.4 (SCH2), 20.8-20.7 (MeCO), 14.2 (CH3). ESIMS: m/z 538.5 [M + Na]+. HRFABMS Calcd for C25H41NO8SNa [M + Na]+ 538.2451, found 538.2444. (1R,SS)- and (1R,SR)-2,3,4-Tri-O-acetyl-1-dodecylsulfinyl-5N,6O- oxomethylidenegalactonojirimycin (12 and 13). To a solution of 11 (131 mg, 0.23 mmol) in DCM (6 mL), mCPBA (70%, 40 mg, 0.23 mmol) was added at 0 ºC. The reaction mixture was stirred for 10 min, diluted with DCM (50 mL), washed with aqueous NaHCO3 (10 mL), brine (10 mL), dried (MgSO4) and concentrated under reduced pressure. The resulting crude was purified by column chromatography (1:2 EtOAc-cyclohexane) to afford the pure diastereomers 12 and 13. Yield of 12: 44 mg (35%). Rf 0.53 (3:2 EtOAc-cyclohexane). [α]D +68.4 (c 0.8 in DCM). 1H NMR (300 MHz, CDCl3): δ 5.72 (dd, 1 H, J2,3 = 10.7 Hz, J3,4 = 2.6 Hz, H-3), 5.49 (t, 1 H, J4,5 = 2.6 Hz, H-4), 5.45 (dd, 1 H, J1,2 = 7.5 Hz, H-2), 5.05 (d, 1 H, H-1), 4.96 (ddd, 1 H, J5,6a = 9.2 Hz, J5,6b = 3.8 Hz, H-5), 4.50 (t, 1 H, J6a,6b = 9.2 Hz, H-6a), 4.07 (dd, 1 H, H-6b), 2.84 (ddd, 1 H, 2JH,H = 13.0 Hz, 3JH,H = 8.1 Hz, 3JH,H = 6.8 Hz, SOCH2), 2.65 (ddd, 1 H, SOCH2), 2.17-2.01 (2 s, 9 H, MeCO), 1.78-1.20 (m, 20 H, CH2), 0.87 (t, 3 H, 3JH,H = 6.6 Hz, CH3). 13C NMR (75.5 MHz, CDCl3): δ 170.3-169.5 (MeCO), 157.3 (CO), 68.9-68.7 (C-3, C-4), 67.0-66.8 (C-1, C-2), 63.9 (C-6), 54.2 (C-5), 49.3 (SOCH2), 32.0-22.8 (CH2), 20.7 (MeCO), 14.2 (CH3). ESIMS: m/z 554.3 [M + Na]+. Anal. Calcd for C25H41NO9S: C, 56.48; H, 7.77; N, 2.63; S, 6.03. Found: C, 56.63; H, 7.90; N, 2.44; S, 5.79. Yield of 13: 39 mg (32%). Rf 0.47 (3:2 EtOAc-cyclohexane). [α]D +75.4 (c 1.0 in DCM). 1H NMR (300 MHz, CDCl3): δ 5.65 (dd, 1 H, J2,3 = 10.8 Hz, J3,4 = 2.6 Hz, H-3), 5.58-5.48 (m, 2 H, H-2, H-4), 5.23 (d, 1 H, J1,2 = 6.2 Hz, H-1), 4.42 (t, 1 H, J5,6a = J6a,6b = 9.0 Hz, H-6a), 4.29
17 (ddd, 1 H, J5,6b = 4.1 Hz, J4,5 = 1.6 Hz, H-5), 4.07 (dd, 1 H, H-6b), 2.98 (ddd, 1 H, 2JH,H = 13.0 Hz, 3JH,H = 9.3 Hz, 3JH,H = 5.3 Hz, SOCH2), 2.71 (m, 1 H, SOCH2), 2.18-2.02 (3 s, 9 H, MeCO), 1.90-1.18 (m, 20 H, CH2), 0.87 (t, 3 H, 3JH,H = 6.6 Hz, CH3). 13C NMR (75.5 MHz, CDCl3): δ 170.3-169.5 (MeCO), 157.3 (CO), 68.9 (C-3), 68.7 (C-4), 67.0 (C-1), 66.8 (C-2), 63.9 (C-6), 52.2 (C-5), 49.3 (SOCH2), 32.0-22.7 (CH2), 20.7 (MeCO), 14.2 (CH3). ESIMS: m/z 554.3 [M + Na]+. Anal. Calcd for C25H41NO9S: C, 56.48; H, 7.77; N, 2.63; S, 6.03. Found: C, 56.55; H, 7.63; N, 2.31; S, 5.67. (1R,SS)-1-Dodecylsulfinyl-5N,6O-oxomethylidenegalactonojirimycin (4). Compound 4 was obtained from 12 (30 mg, 0.06 mmol) by conventional O-deacetylation (see General Methods) and purification by column chromatography (4:1 EtOAc-cyclohexane → EtOAc). Yield: 24 mg (98%). Rf 0.09 (EtOAc). [α]D +64.1 (c 1.0 in MeOH). 1H NMR (300 MHz, CD3OD): δ 4.51 (t, 1 H, J6a,6b = J5,6a = 8.6 Hz, H-6a), 4.42 (dd, 1 H, J5,6b = 5.4 Hz, H-6b), 4.35 (ddd, 1 H, J4,5 = 2.0 Hz, H-5), 4.29 (dd, 1 H, J2,3 = 9.9 Hz, J1,2 = 6.8 Hz, H-2), 3.99 (dd, 1 H, J3,4 = 2.5 Hz, H-3), 3.88 (t, 1 H, H-4), 3.00-2.82 (m, 2 H, SOCH2), 1.84-1.73 (m, 2 H, SOCH2CH2), 1.58-1.21 (m, 18 H, CH2), 0.90 (t, 3 H, 3JH,H = 6.5 Hz, CH3). 13C NMR (75.5 MHz, CD3OD): δ 160.2 (CO), 72.8-72.4 (C-1, C-3), 70.1 (C-4), 67.7 (C-2), 65.5 (C-6), 57.3 (C-5), 51.9 (SOCH2), 33.0-23.7 (CH2), 14.4 (CH3). ESIMS: m/z 428.2 [M + Na]+. Anal. Calcd for C19H35NO6S: C, 56.27; H, 8.70; N, 3.45; S, 7.91. Found: C, 56.16; H, 8.65; N, 3.13; S, 7.60. (1R,SR)-1-Dodecylsulfinyl-5N,6O-oxomethylidenegalactonojirimycin (5). Compound 5 was obtained from 13 (21 mg, 0.04 mmol) by conventional O-deacetylation (see General Methods) and final purification by column chromatography (4:1 EtOAc-cyclohexane → EtOAc). Yield: 15 mg (96%). Rf 0.09 (EtOAc). [α]D +65.0 (c 1.0 in MeOH). 1H NMR (300 MHz, CD3OD): δ 4.96 (d, 1 H, J1,2 = 6.5 Hz, H-1), 4.50-4.40 (m, 2 H, H-6a, H-6b), 4.37 (dd, 1 H, J2,3 = 10.0 Hz, H-2), 4.16 (ddd, 1 H, J5,6a = 8.7 Hz, J5,6b = 5.2 Hz, J4,5 = 2.0 Hz, H-5), 4.05 (dd, 1 H, J3,4 = 2.5 Hz, H-3), 3.90 (t, 1 H, H-4), 3.12 (ddd, 1 H, 2JH,H = 13.0 Hz, 3JH,H = 8.7 Hz, 3JH,H = 5.8 Hz, SOCH2), 2.98 (ddd, 1H, SOCH2), 1.87-1.70 (m, 2 H, SO2CH2CH2), 1.60-1.21 (m, 18 H, CH2), 0.90 (t, 3 H, 3JH,H = 6.8 Hz, CH3). 13C NMR (75.5 MHz, CD3OD): δ 159.2 (CO), 73.1-72.3 (C-1, C-3), 70.0 (C-4), 69.5 (C-2), 65.3 (C-6), 56.5 (C-5), 50.8 (SOCH2), 33.1- 23.7 (CH2), 14.4 (CH3). ESIMS: m/z 428.2 [M + Na]+. Anal. Calcd for C19H35NO6S: C, 56.27; H, 8.70; N, 3.45; S, 7.91. Found: C, 56.30; H, 8.90; N, 3.17; S, 7.75. (1R)-1-Dodecylsulfonyl-5N,6O-oxomethylidenegalactonojirimycin (6). To a solution of 11 (51 mg, 0.09 mmol) in DCM (3 mL), mCPBA (70%, 31 mg, 0.18 mmol) was added at 0 ºC. The reaction mixture was stirred for 20 min, diluted with DCM (50 mL), washed
18 with aqueous NaHCO3 (10 mL), brine (10 mL), dried (MgSO4), concentrated under reduced pressure. Further purification by column chromatography (4:1 EtOAc-cyclohexane → EtOAc) and conventional O-deacetylation (see General Methods) afforded 6. Yield: 18 mg (47% global yield). Rf 0.20 (EtOAc). [α]D +23.2 (c 1.0 in MeOH). 1H NMR (300 MHz, CD3OD): δ 5.16 (d, 1 H, J1,2 = 6.0 Hz, H-1), 4.50 (t, 1 H, J6a,6b = J5,6a = 8.5 Hz, H-6a), 4.42 (dd, 1 H, J5,6b = 5.2 Hz, H-6b), 4.31 (ddd, 1 H, J4,5 = 1.7 Hz, H-5), 4.28-4.24 (m, 2 H, H-2, H-3), 3.88 (t, 1 H, J3,4 = 1.7 Hz, H-4), 3.39-3.34 (m, 2 H, SO2CH2), 1.84 (m, 2 H, 3JH,H = 7.2 Hz, SO2CH2CH2), 1.48-1.29 (m, 18 H, CH2), 0.90 (t, 3 H, 3JH,H = 6.7 Hz, CH3). 13C NMR (75.5 MHz, CD3OD): δ 159.2 (CO), 72.3 (C-1), 70.7 (C-2), 70.2 (C-4), 67.8 (C-3), 65.4 (C-6), 56.3 (C-5, CH2SO2), 33.0-22.5 (CH2), 14.4 (CH3). ESIMS: m/z 444.2 [M + Na]+. Anal. Calcd for C19H35NO7S: C, 54.14; H, 8.37; N, 3.32; S, 7.61. Found: C, 53.90; H, 8.18; N, 3.11; S, 7.28. (1R)-2,3,4-Tri-O-acetyl-1-dodecylthio-5N,6O-oxomethylidenemannonojirimycin (15). To a stirred solution of (1R)-1,2,3,4-tetra-O-acetyl-5N,6O- oxomethylidenemannojirimycin (14, 0.29 mmol) in anhydrous DCM (4 mL) at 0 ºC, BF3.OEt2 (134 μL, 1.09 mmol) and dodecylthiol (0.61 mmol) were added under Ar atmosphere. The mixture was stirred for 30 min (TLC monitoring), diluted with DCM (50 mL) and washed with water (10 mL), aq NaHCO3 (10 mL) and brine (10 mL), dried (MgSO4) concentrated under reduced pressure and subjected to column chromatography (1:4 → 1:2 EtOAc-cyclohexane) to give 15. Yield: 101 mg (68%). Rf 0.72 (1:1 EtOAc-cyclohexane). [α]D +18.2 (c 1.0 in DCM). 1H NMR (300 MHz, CDCl3): δ 5.33-5.28 (m, 2 H, H-2, H-3), 5.22 (t, 1 H, J3,4 = J4,5 = 9.9 Hz, H-4), 5.15 (d, 1 H, J1,2 = 2.0 Hz, H-1), 4.41 (dd, 1 H, J6a,6b = 9.1 Hz, J5,6a = 8.2 Hz, H-6a), 4.32 (dd, 1 H, J5,6b = 5.1 Hz, H-6b), 4.08 (ddd, 1 H, H-5), 2.73-2.51 (m, 2 H, CH2S), 2.10-1.99 (3 s, 9 H, MeCO), 1.67-1.51 (m, 2 H, CH2CH2S), 1.41-1.17 (m, 18 H, CH2), 0.86 (t, 3 H, 3JH,H = 6.6 Hz, CH3).13C NMR (75.5 MHz, CDCl3): δ 170.3-169.6 (MeCO), 156.3 (CO), 70.8, 69.1 (C-2, C-3), 69.5 (C-4), 66.1 (C-6), 58.3 (C-1), 52.8 (C-5), 31.9-28.7 (CH2), 20.8-20.6 (MeCO), 14.2 (CH3). ESIMS: m/z 538.3 [M + Na]+. Anal. Calcd for C25H41NO8S: C, 58.23; H, 8.01; N, 2.72; S, 6.22. Found: C, 58.32; H, 8.20; N, 2.60; S, 6.04. (1R,SS)-2,3,4-Tri-O-acetyl-1-dodecylsulfinyl-5N,6O- oxomethylidenemannonojirimycin (16). To a solution of 15 (131 mg, 0.23 mmol) in DCM (6 mL), mCPBA (70%, 40 mg, 0.23 mmol) was added at 0 ºC. The reaction mixture was stirred for 10 min, diluted with DCM (50 mL), washed with aqueous NaHCO3 (10 mL), brine (10 mL), dried (MgSO4) and concentrated under reduced pressure. The resulting crude was purified by column chromatography (Et2O) to give 16. Yield: 48 mg (39%). Rf 0.59 (Et2O). [α]D - 4.1 (c
19 1.0 in DCM). 1H NMR (300 MHz, CDCl3): δ 5.73 (t, 1 H, J1,2 = J2,3= 2.5 Hz, H-2), 5.61 (dd, 1 H, J3,4 = 9.0 Hz, H-3), 5.27 (t, 1 H, J4,5 = 9.0 Hz, H-4), 4.62 (d, 1 H, H-1), 4.60-4.50 (m, 1 H, H-5), 4.52 (t, 1 H, J6a,6b = J5,6a = 8.3 Hz, H-6a), 4.38 (dd, 1 H, J5,6b = 3.7 Hz, H-6b), 2.95 (m, 1 H, SOCH2), 2.79 (m, 1 H, SOCH2), 2.16-2.01 (3 s, 9 H, MeCO), 1.79-1.18 (m, 20 H, CH2), 0.87 (t, 3 H, 3JH,H = 6.9 Hz, CH3). 13C NMR (75.5 MHz, CDCl3): δ 170.3-169.4 (MeCO), 157.7 (CO), 69.6 (C-2), 69.0 (C-3), 68.8 (C-4), 68.4 (C-1), 66.9 (C-6), 55.8 (C-5), 50.8 (SOCH2), 31.9-22.4 (CH2), 20.7-20.6 (MeCO), 14.1 (CH3). ESIMS: m/z 554.5 [M + Na]+. Anal. Calcd for C25H41NO9S: C, 56.48; H, 7.77; N, 2.63; S, 6.03. Found: C, 56.13; H, 7.44; N, 2.29; S, 5.76. (1R,SS)-1-Dodecylsulfinyl-5N,6O-oxomethylidenemannonojirimycin (7). Compound 7 was obtained from 16 (20 mg, 0.04 mmol) by conventional O-deacetylation. Column chromatography (5:1 EtOAc-MeOH). Yield: 14 mg (86%). Rf 0.52 (5:1 EtOAc- MeOH). [α]D +23.8 (c 0.6 in MeOH). 1H NMR (300 MHz, CD3OD): δ 4.67 (d, 1 H, J1,2 = 2.0 Hz, H-1), 4.63 (t, 1 H, J6a,6b = J5,6a = 8.7 Hz, H-6a), 4.38 (t, 1 H, J2,3 = 2.0 Hz, H-2), 4.34 (dd, 1 H, J5,6b = 4.9 Hz, H-6b), 4.11 (ddd, 1 H, J4,5 = 9.3 Hz, H-5), 3.90 (dd, 1 H, J2,3 = 2.0 Hz, J3,4 = 9.3 Hz, H-3), 3.75 (t, 1 H, H-4), 2.98-2.81 (m, 2 H, SOCH2), 1.82-1.71 (m, 2 H, SOCH2CH2), 1.52-1.24 (m, 18 H, CH2), 0.90 (t, 3 H, 3JH,H = 6.7 Hz, CH3). 13C NMR (75.5 MHz, CD3OD): δ 160.5 (CO), 73.6-73.0 (C-1, C-3), 71.9- 71.2 (C-2, C-4), 68.7 (C-6), 58.9 (C-5), 51.1 (SOCH2), 33.1-23.6 (CH2), 14.5 (CH3). ESIMS: m/z 428.3 [M + Na]+. HRFABMS Calcd for C19H35NO6SNa [M + Na]+ 428.2077, found 428.2081. (1R)-2,3,4-Tri-O-acetyl-1-dodecylsulfonyl-5N,6O-oxomethylidenemannojirimycin (17). To a solution of 15 (51 mg, 0.09 mmol) in DCM (3 mL), mCPBA (70%, 31 mg, 0.18 mmol) was added at 0 ºC. The reaction mixture was stirred for 20 min, diluted with DCM (50 mL), washed with aqueous NaHCO3 (10 mL), brine (10 mL), dried (MgSO4) and concentrated under reduced pressure. Final purification by column chromatography (1:3 EtOAccyclohexane) afforded 17. Yield: 33 mg (71%). Rf 0.66 (1:1 EtOAc-cyclohexane). [α]D -6.5 (c 1.0 in DCM). 1H NMR (300 MHz, CDCl3): δ 5.93 (t, 1 H, J1,2 = J2,3 = 2.6 Hz, H-2), 5.54 (dd, 1 H, J3,4 = 9.0 Hz, H-3), 5.21 (t, 1 H, J4,5 = 9.0 Hz, H-4), 5.06 (d, 1 H, H-1), 4.56-4.49 (m, 1 H, H-6a), 4.43-4.37 (m, 2 H, H-5, H-6b), 3.05 (m, 2 H, SO2CH2), 2.13-2.02 (3 s, 9 H, MeCO), 1.95-1.78 (m, 2 H, SO2CH2CH2), 1.46-1.19 (m, 18 H, CH2), 0.87 (t, 3 H, 3JH,H = 6.7 Hz, CH3). 13C NMR (75.5 MHz, CDCl3): δ 170.4-169.2 (MeCO), 156.6 (CO), 69.4 (C-1), 69.0 (C-4), 68.9 (C-3), 67.2 (C-6), 65.0 (C-2), 54.1 (C-5), 52.2 (CH2SO2), 32.0-21.7 (CH2), 20.8-20.6 (MeCO), 14.2 (CH3). ESIMS: m/z 570.4 [M + Na]+. Anal. Calcd for C25H41NO10S: C, 54.83; H, 7.55; N, 2.56; S, 5.85. Found: C, 55.01; H, 7.69; N, 2.29; S, 5.62.
20 (1R)-1-Dodecylsulfonyl-5N,6O-oxomethylidenemannonojirimycin (8). Compound 8 was obtained by conventional O-deacetylation (see General Methods) of 17. Column chromatography (4:1 EtOAc-cyclohexane → EtOAc). Yield: 12 mg (64%). Rf 0.20 (EtOAc). [α]D +23.2 (c 1.0 in MeOH). 1H NMR (300 MHz, CD3OD): δ 5.03 (d, 1 H, J1,2 = 1.6 Hz, H-1), 4.65-4.59 (m, 2 H, H-2, H-6a), 4.36 (dd, 1 H, J6a,6b = 9.0 Hz, J5,6b = 5.0 Hz, H-6b), 4.08 (td, 1 H, J4,5 = J5,6a = 9.0 Hz, H-5), 3.86 (dd, 1 H, J3,4 = 9.5 Hz, J2,3 = 3.1 Hz, H-3), 3.73 (t, 1 H, H- 4), 3.25-3.06 (m, 2 H, SO2CH2), 1.89-1.73 (m, 2 H, SO2CH2CH2), 1.47-1.26 (m, 18 H, CH2), 0.89 (t, 3 H, 3JH,H = 6.7 Hz, CH3). 13C NMR (75.5 MHz, CD3OD): δ 159.5 (CO), 74.0 (C-1), 73.0 (C-3), 70.7 (C-4), 68.6 (C-6), 66.9 (C-2), 57.0 (C-5), 52.0 (CH2SO2), 32.0-22.7 (CH2), 14.4 (CH3). ESIMS: m/z 444.3 [M + Na]+. Anal. Calcd for C19H35NO7S: C, 54.14; H, 8.37; N, 3.32; S, 7.61. Found: C, 53.88; H, 8.12; N, 3.04; S, 7.32. 4.3.Cell Culture. Human primary cells were cultured at 37 °C and 5% CO2 in RPMI 1640 supplemented with gentamycin and 10% (v/v) heat-inactivated fetal calf serum (complete medium). Human monocytes were purified from buffy coats by successive Ficoll and Percoll gradients. Monocyte-derived DCs were differentiated from monocytes by 5 days of culture with GM-CSF (50 ng/ml; ImmunoTools) and IL-4 (10 ng/ml; ImmunoTools), as described.19 Proper differentiation was characterized by low CD14 and high CD1a and DC-SIGN expression levels. 4.4. Flow Cytometry. The cells phenotype was analyzed using the following antibodies from BD Bioscience: HLADR-FITC (Tü39 or L243), DC-SIGN/CD209-PerCP-Cy5.5 (DCN46), CD1a-APC (HI149), CD86-FITC (FUN-1), CD83-APC (HB15e), and CD14-PE (MEM-15, ImmunoTools). Data were acquired on a FACSCalibur (Becton-Dickinson) or Gallios (Beckman-Coulter) after exclusion of dead cells by Sytox Red (Molecular Probes, Invitrogen) or 7-AAD (BD- Pharmingen) DNA intercalating dyes. Data were analysed using the Cell Quest Pro software (BD Bioscience) or FlowJo (Treestar). 4.5. Cell Activation. DCs (105 cells/ well) were cultured in complete medium (200 L) and incubated in the presence of Ultra pure E. coli K12 lipopolysaccharide (LPS; 100 ng mL-1). Compounds were usually added at the same time as LPS, or, where indicated, 2, 4 or 6 h after LPS addition. 4.6. Enzyme-Linked ImmunoabSorbent Assay (ELISA).
21 After 24 h of stimulation with LPS, cell supernatants were collected and stored at -20°C. Levels of TNF-α released in the media, or in the case of in vivo assays, levels of IL-6 in the mouse sera, were measured by ELISA with the corresponding BD OptEIA Set (BD-Pharmingen). 4.7. Effect on Cell Surface Proteins. DCs (105 cells) in complete medium (200 L) were plated in a 96-well plate. They were left untreated or treated with sp2-IGLs (25 and 50 M) or conjugate MGC 9 (100 M) for 30 min at 37 °C. They were washed, kept on ice, labeled with anti-CD14-PE, and analyzed by flow cytometry. 4.8. Co-localization Analysis of CD14 and TLR-4 by Confocal Microscopy. Assays were performed according to the previously described procedure [22]. Briefly, DCs (105 cells per chamber) were cultured on poly-L-Lysine-coated, 8-chambers slides (Lab-Tek, Nunc) in complete medium (300 L). Cells were left untreated or treated with sp2-IGLs (50 M) or MGC 9 (100 M) for 30 min. Cells were washed, fixed, permeabilized and labeled with anti- TLR-4 polyclonal goat antibody (R&D Systems AF1478, 5 g mL-1) followed by AF488-donkey anti-goat (Molecular Probes, Invitrogen), and anti-CD14-APC (ImmunoTools). DAPI was used as nuclear counterstaining. Slides were mounted using Fluoromount (Dako). Images were acquired on a Zeiss LSM 780 confocal microscope with GaAsP detector and Zen acquisition software. Images were processed using the ImageJ software. 4.9. In Vivo Assays in C57BL/6J Mice. All animals received humane care in compliance with the guidelines formulated by the French Ministry of Agriculture and of Higher Education and Research, and all procedures were reviewed by the Regional Ethical Committee for Animal Experimentation of the University of Strasbourg. sp2-IGL 3 was prepared in sterile solvent (7.25 mg mL-1) by solubilization in DMSO (10.5 mg in 1 mL) followed by addition of H2O (0.5 mL) and PBS x10 (150 L). LPS (0.5 mg/kg; strain O111.B4; Sigma-Aldrich) was prepared in sterile PBS. To evaluate the in vivo toxicity of the compound, 8-week old male and female mice received intraperitoneal (i.p.) administrations of 200 µl of sp2-IGL 3 (30 mg/kg) or of control DMSO/PBS on day 0, 1, 2, 3. Evaluation of their well-being and weight were performed every day up to day 12. To assess the compounds for their capacity to block LPS-induced IL-6 production, C57BL/6J male 8- week old mice received a single intraperitoneal (i.p.) administration of sp2-IGL 3 (50 mg/kg) 1 h before LPS (0.5 mg/kg) challenge. Control animals received equivalent volumes of solvent
22 (PBS or PBS/DMSO 2:1) (200 L) by the intraperitoneal route. Blood (150 L) was taken 2 h after LPS injection. Serum was prepared and IL-6 measured by ELISA (see above). Statistical Analysis. Statistical analysis was performed using paired Student’s t-test. Data were considered significantly different when p was less than 0.05. Supporting information available Additional in vitro assays on DCs of sp2-IGLs and 1H and 13C NMR spectra of all new compounds. This material is available free of charge via the Internet at http://xxxxxxxx.xxx Notes The authors declare no competing financial interest. Acknowledgements This work was supported by the “Centre National de la Recherche Scientifique” (CNRS) and the Agence Nationale pour la Recherche (Program "Investissements d'Avenir", ANR-10- LABX-0034 MEDALIS; ANR-11-EQPX-022). This work was also supported by grants from the Spanish Ministry of Economy and Competitiveness: SAF2016-76083-R (MINECOFEDER) and CTQ2015-64425-C2-1-R (MINECO-FEDER). We thank the Strasbourg/Esplanade imaging platform. Technical assistance from the research support services of the University of Seville (CITIUS) is also acknowledged. ABBREVIATIONS USED DC, dendritic cell; DMSO, dimethyl sulfoxide; DCM, dichloromethane; mCPBA, mchloroperoxybenzoic acid; sp2-IGL, sp2-iminosugar glycolipid; R-DSO-ONJ, (1R)-1- dodecylsulfinyl-5N,6O-oxomethylidenenojirimycin; DSO2-ONJ, (1R)-1-dodecylsulfonyl- 5N,6O-oxomethylidenenojirimycin; LPS, lipopolysaccharide; MGC, mannoside glycolipid conjugate; IL-10, interleukin-10; IL-6, interleukin-6; TNF-α, Tumor Necrosis Factor-α. MAPK, mitogen activated protein kinase.
23 REFERENCES [1] T. Mena-Barragán, M. I. García-Moreno, A. Sevšek, T. Okazaki, E. Nanba, K. Higaki, N. I. Martin, R. J. Pieters, J. M. García Fernández, C. Ortiz Mellet, Probing the inhibitor versus chaperone properties of sp2-Iminosugars towards human β-glucocerebrosidase: a picomolar chaperone for Gaucher disease, Molecules 23 2018 927. [2] T. Mena-Barragn, M. I. García-Moreno, E. Nanba, K. Higaki, A. L. Concia, P. Claps, J. M. García Fernndez, C. Ortiz Mellet, Inhibitor versus chaperone behaviour of D-fagomine, DAB and LAB sp2-iminosugar conjugates against glycosidases: a structure-activity relationship study in Gaucher fibroblasts, Eur. J. Med. Chem. 121 (2016) 880-891. [3] A. de la Fuente, R. Rísquez-Cuadro, X. Verdaguer, J. M. García Fernández, E. Nanba, K. Higaki, C. Ortiz Mellet, A. Riera, A, Efficient stereoselective synthesis of 2-acetamido-1,2- dideoxyallonojirimycin (DAJNAc) and sp2-iminosugar conjugates: novel hexosaminidase inhibitors with discrimination capabilities between the mature and precursor forms of the enzyme, Eur. J. Med. Chem. 121 (2016) 926-938. [4] E. M. Sánchez-Fernández, R. Rísquez-Cuadro, M. Aguilar-Moncayo, M. I. García-Moreno, C. Ortiz Mellet, J. M. García Fernández, Generalized anomeric effect in gem-diamines: stereoselective synthesis of -N-linked disaccharide mimics, Org. Lett. 11 (2009) 3306-3308. [5] E. M. Sánchez-Fernández, J. M. García Fernández, C. Ortiz Mellet, Glycomimetic-based pharmacological chaperones for lysosomal storage disorders: lessons from Gaucher, GM1- gangliosidosis and Fabry diseases, Chem. Commun. 52 (2016) 5497-5515. [6] M. I. García-Moreno, F. Ortega-Caballero, R. Rísquez-Cuadro, C. Ortiz Mellet, J. M. García Fernández, The Impact of heteromultivalency in lectin recognition and glycosidase inhibition: an integrated mechanistic study, Chem. Eur. J. 23 (2017) 6295-6304. [7] E. M. Sánchez-Fernández, C. D. Navo, N. Martínez-Sez, R. Gonalves-Pereira, V. J. Somovilla, A. Avenoza, A., J. H. Busto, G. J. L. Bernardes, G. Jimnez-Oss, F. Corzana, J. M. García Fernández, C. Ortiz Mellet, J. M. Peregrina, Tn Antigen mimics based on sp2- iminosugars with affinity for an anti-MUC1 antibody, Org. Lett. 18 (2016) 3890-3893. [8] V. F. Vartabedian, P. B. Savage, L. Teyton, The processing and presentation of lipids and glycolipids to the immune system, Immunol. Rev. 272 (2016) 109-119.
24 [9] E. M. Sánchez-Fernández, R. Rísquez-Cuadro, M. Chasseraud, A. Ahidouch, C. Ortiz Mellet, H. Ouadid-Ahidouch, J. M. García Fernández, Synthesis of N-, S-, and C-glycoside castanospermine analogues with selective neutral -glucosidase inhibitory activity as antitumor agents, Chem. Commun. 46 (2010) 5328-5330. [10] G. Allan, H. Qualid-Ahidouch, E. M. Sánchez-Fernández, R. Rísquez-Cuadro, J. M. García Fernández, C. Ortiz Mellet, A. Ahidouch, New castanospermine glycoside analogues inhibit breast cancer cell proliferation and induce apoptosis without affecting normal cells, PLOS ONE 8 (2013) e76411. [11] N. Gueder, G. Allan, M. S. Telliez, F. Hague, J. M. García Fernández, E. M. Sánchez- Fernández, C. Ortiz Mellet, A. Ahidouch, H. Ouadid-Ahidouch, sp2-Iminosugar -glucosidase inhibitor 1-C-octyl-2-oxa-3-oxocastanospermine specifically affected breast cancer cell migration through Stim1, beta1-integrin, and FAK signaling pathways, J. Cell Physiol. 232 (2017) 3631-3640. [12] E. M. Sánchez-Fernández, V. Gómez-Pérez, R. García-Hernández, J. M. García Fernandez, G. B. Plata, J. M. Padrón, C. Ortiz Mellet, S. Castanys, F. Gamarro, Antileishmanial activity of sp2-iminosugar derivatives, RSC Adv. 5 (2015) 21812-21822. [13] A. I. Arroba, E. Alcalde-Estévez, M. García-Ramírez, D. Cazzoni, P. de la Villa, E. M. Sánchez-Fernández, C. Ortiz Mellet, J. M. García Fernández, C. Hernández, R. Simó, A. M. Valverde, Modulation of microglia polarization dynamics during diabetic retinopathy in db/db mice, Biochim. Biophys. Acta 1862 (2016) 1663-1674. [14] E. Alcalde-Estévez, A. I. Arroba, E. M. Sánchez-Fernández, C. Ortiz Mellet, J. M. García Fernández, L. Masgrau, A. M. Valverde, The sp2-iminosugar glycolipid 1-dodecylsulfonyl- 5N,6O-oxomethylidenenojirimycin (DSO2-ONJ) as selective anti-inflammatory agent by modulation of hemeoxygenase-1 in Bv.2 microglial cells and retinal explants, Food Chem. Toxicol. 111 (2018) 454-466. [15] P. Bogdanov, L. Corraliza, J. A. Villena, A. R. Carvalho, J. García-Arumi, D. Ramos, J. Ruberte, R. Simó, C. Hernández, The db/db mouse: a useful model for the study of diabetic retinal neurodegeneration, PLOS ONE 9 (2014) e97302. [16] E. M. Sánchez-Fernández, R. Gonçalves-Pereira, R. Rísquez Cuadro, G. B. Plata, J. M. Padrón, J. M. García Fernández, C. Ortiz Mellet, Influence of the configurational pattern of sp2- iminosugar pseudo N-, S-, O- and C-glycosides on their glycoside inhibitory and antitumor properties, Carbohydr. Res. 429 (2016) 113-122.
25 [17] D. Bell, J. W. Young J. Banchereau, Dendritic cells, Adv. Immunol. 72 (1999) 255-324. [18] R. L. Danner, R. J. Elin, J. M. Hosseini, R. A. Wesley, J. M. Reilly, J. E. Parillo, Endotoxemia in human septic shock, Chest 99 (1991) 169-175 [19] M. Kobayashi, S. Saitoh, N. Tanimura, K. Takahashi, K. Kawasaki, M. Nishijima, Y. Fujimoto, K. Fukase, S. Akashi-Takamura, K. Miyake, Regulatory roles for MD-2 and TLR-4 in ligand-induced receptor clustering, J. Immunol. 176 (2006) 6211-6218. [20] W. Y. C. Lu, C. Yeh, P. S. Ohashi, LPS/TLR-4 signal transduction pathway, Cytokine 42 (2008) 145-151. [21] F. Peri, V. Calabrese, Toll-like receptor 4 (TLR-4) modulation by synthetic and natural compounds: an update, J. Med. Chem. 57 (2014) 3612-3622. [22] V. Flacher, P. Neuberg, F. Point, F. Daubeuf, Q. Muller, D. Sigwalt, J. D. Fauny, J. S. Remy, N. Frossard, A. Wagner, C. G. Mueller, E. Schaeffer, Mannoside glycolipid conjugates display anti-inflammatory activity by inhibition of Toll-like receptor-4 mediated cell activation, ACS Chem. Biol. 10 (2015) 2697-2705. [23] L. Dehuyser, E. Schaeffer, O. Chaloin, C. G. Mueller, R. Baati, A. Wagner, Synthesis of novel mannoside glycolipid conjugates for inhibition of HIV-1 trans-infection, Bioconjug. Chem. 23 (2012) 1731-1739. [24] E. Schaeffer, L. Dehuyser, D. Sigwalt, V. Flacher, S. Bernacchi, O. Chaloin, J. S. Remy, C. G. Mueller, R. Baati, A. Wagner, Dynamic micelles of mannoside glycolipids are more efficient than polymers for inhibiting HIV-1 trans-infection, Bioconjug. Chem. 24 (2013) 1813- 1823. [25] J. J. Gills, S. S. Castillo, C. Zhang, P. A. Petukhov, R. M. Memmott, M. Hollingshead, N. Warfel, J. Han, A. P. Kozikowski, P. A. Dennis, Phosphatidylinositol ether lipid analogues that inhibit AKT also independently activate the stress kinase, p38alpha, through MKK3/6- independent and -dependent mechanisms, J. Biol. Chem. 282 (2007) 27020-27029. [26] C. A. Sanhueza, A. C. Arias, R. L. Dorta and J. T. Vázquez, Absolute configuration of glycosyl sulfoxides, Tetrahedron: Asymmetry, 21 (2010) 1830-1832. [27] N. Khiar, Determination of the absolute configuration of sulfinyl glycosides: the role of the exo-anomeric effect, Tetrahedron Lett. 41 (2000) 9059-9063. [28] C. Aerts-Toegaert, C. Heirman, S. Tuyaerts, J. Corthals, J. L. Aerts, A. Bonehill, K. Thielemans, K. Breckpot, CD83 expression on dendritic cells and T cells: correlation with effective immune responses, Eur. J. Immunol. 37 (2007) 686-695.