| Literature DB >> 29797675 |
Sybrin P Schröder1, Wouter W Kallemeijn2, Marjoke F Debets1, Thomas Hansen1, Lukasz F Sobala3, Zalihe Hakki4, Spencer J Williams4, Thomas J M Beenakker1, Johannes M F G Aerts2, Gijsbert A van der Marel1, Jeroen D C Codée1, Gideon J Davies3, Herman S Overkleeft1.
Abstract
N-Glycans direct protein function, stability, folding and targeting, and influence immunogenicity. While most glycosidases that process N-glycans cleave a single sugar residue at a time, enzymes from glycoside hydrolase family 99 are endo-acting enzymes that cleave within complex N-glycans. Eukaryotic Golgi endo-1,2-α-mannosidase cleaves glucose-substituted mannose within immature glucosylated high-mannose N-glycans in the secretory pathway. Certain bacteria within the human gut microbiota produce endo-1,2-α-mannanase, which cleaves related structures within fungal mannan, as part of nutrient acquisition. An unconventional mechanism of catalysis was proposed for enzymes of this family, hinted at by crystal structures of imino/azasugars complexed within the active site. Based on this mechanism, we developed the synthesis of two glycosides bearing a spiro-epoxide at C-2 as electrophilic trap, to covalently bind a mechanistically important, conserved GH99 catalytic residue. The spiro-epoxyglycosides are equipped with a fluorescent tag, and following incubation with recombinant enzyme, allow concentration, time and pH dependent visualization of the bound enzyme using gel electrophoresis.Entities:
Keywords: GH99; activity-based probes; endomannosidase; glycosidase; inhibitors
Mesh:
Substances:
Year: 2018 PMID: 29797675 PMCID: PMC6055899 DOI: 10.1002/chem.201801902
Source DB: PubMed Journal: Chemistry ISSN: 0947-6539 Impact factor: 5.236
Figure 1(A) Known GH99 endo‐1,2‐α‐mannosidase inhibitors (1–3) and fluorescent spiro‐epoxyglycosides 4 and 5 subject of this study. K D values are for B. thetaiotaomicron endo‐1,2‐α‐mannosidase (BtGH99). (B) The proposed catalytic mechanism for GH99 enzyme (amino acid numbering for B. xylanisolvens endo‐1,2‐α‐mannosidase (BxGH99)). (C) Anticipated covalent inhibition mechanism of GH99 enzymes.
Scheme 1Synthesis of fluorescent spiro‐epoxyglycosides 4 and 5. Reagents and conditions: a) tBu2Si(OTf)2, 2,6‐lutidine, DMF, −50 °C; b) PhC(OMe)3, CSA, 2 h, then AcOH, H2O, 16 h, 57 % over 2 steps; c) TBSOTf, DMAP, pyridine, 60 °C, 16 h, yield 9: 85 % over 2 steps; yield 11: 81 % over 2 steps; d) donor 9 or 11, NIS, TMSOTf, DCM, 4 Å MS, −40 °C, 1 h, yield 12 g: 92 %; yield 12 m: 88 %; e) NaOMe, MeOH, DCM, yield 13 g: 95 %; yield 13 m: 86 %; f) DMP, DCM, yield 14 g: 98 %; yield 14 m: 96 %; g) TBAF, THF, 5 days, yield 17: 97 %; yield 18: 74 %; h) Cy5‐alkyne,47 CuSO4⋅5 H2O, sodium ascorbate, DMF, rt, 16 h, yield 4: 32 %; yield 5: 34 %.
Transformation of ketones 14 g and 14 m into their corresponding spiro‐epoxides.
| Entry | s.m. | Conditions |
| Yield [%][a] |
|---|---|---|---|---|
| 1 |
| CH2N2, EtOH, 0 °C | 1:1 | 78 |
| 2 |
| 3:1 | 97 | |
| 3 |
| SOMe3I, | 1:5 | 83 |
| 4 |
| 1:2 | 85 | |
| 5 |
| SMe3I, NaH, DMSO, THF, −10 °C | 0:1 | 50 |
| 6 |
| 0:1 | 53 |
[a] Combined yield after column chromatography. s.m.=starting material.
Figure 2Fluorescent labeling of GH99 endomannanases. (A) Detection limit of Bt and Bx GH99 endomannanases (left and right, respectively), labeled with various concentrations of fluorescent spiro‐epoxyglycosides 4 or 5. (B) Effect of pH on labeling of Bt and Bx GH99 enzymes with 4 or 5. (C) Labeling of wild‐type and mutant BxGH99 with 4 or 5 (left or right, respectively) for 5, 15 or 30 minutes. (D) Effect of denaturation with 1 % (w/v) SDS and boiling on labeling of Bt and Bx GH99 enzymes (left and right, respectively) with 4 or 5. The marker is annotated with an asterisk (*).
Figure 3(A) Schematic representation of processing of human α‐galactosidase GLA by GH99 endomannosidase. GLA is pre‐labeled by fluorescent TB340, and contains high‐mannose N‐glycans which can be truncated by endomannosidase, resulting in a decrease in GLA molecular weight. Activity‐based labeling of endomannosidase by spiro‐epoxyglycosides 4 or 5 (prior to incubation with GLA) blocks its activity, and is therefore unable to process GLA. (B) BtGH99 wild‐type demannosylates GLA, causing a shift in molecular weight for the protein bands. Pre‐labeling BtGH99 wild‐type with 4 or 5 abrogates GLA demannosylation. Endo‐H cleaves high‐mannose structures, PNGase‐F cleaves full N‐linked glycan (leaving Asp‐GlcNAc). (C) BxGH99 wild‐type demannosylates GLA, while BxGH99 pre‐labeled with 4 or 5 is unable to do so. BxGH99 active‐site mutants E333Q and E3336Q are unable to process GLA. (D) Fluorescent labeling of BtGH99 (top) and BxGH99 (bottom) by 4 or 5 competed by different concentrations of 17, 18, ManIFG (2) and yeast mannan. The marker is annotated with an asterisk (*).