Ingrid 't Hart

31 Chemoenzyma�c synthesis of DSGb5 2 Glycosyla�on of disaccharide 7a and trisaccharide 8a in the presence of TMSOTf in DCM at -30 °C resulted in the forma�on of pentasaccharide 16a in a yield of 59% as a separable mixture of α/β anomers (β/α= 1.7). Op�miza�on of the reac�on condi�ons revealed that the overall yield of the glycosyla�on could be increased by lowering the reac�on temperature, however, this did not affect the poor anomeric selec�vity (SI, Table S1). Fortunately, the use of glycosyl donor 7b , having acetyl esters instead of a benzylidene acetal at the 4,6-diol of GalNTroc, 19 gave in a TMSOTf mediated glycosyla�on with acceptor 8a at -50 °C, pentasaccharide 16b as only the β-anomer in an isolated yield of 52%. Pentasaccharide 16a was deprotected by the six-step procedure to give Gb5 ( 4a ). Thus, the silyl protec�ng group was cleaved by treatment with HF·pyridine which was followed by hydrolysis of the acetyl esters and Troc protec�ng group with aqueous NaOH in THF with hea�ng (80 °C). The resul�ng compound was acetylated with ace�c anhydride in pyridine and then the anomeric methoxyphenyl (MP) protec�ng group was oxida�vely removed by cerium ammonium nitrate (CAN) in a mixture of acetonitrile and H 2 O. Finally, deacetyla�onunderZempléncondi�ons( cat. NaOMeinMeOH)followedbyhydrogena�on over Pd(OH) 2 /C in a mixture of MeOH/H 2 O/HOAc afforded Gb5 ( 4a ) in an overall yield of 53% a�er purifica�on by Bio-Gel P-2 size exclusion chromatography followed by semi- prepara�ve HPLC using a HILIC column (XBridge® Amide 5 µm, 4.6mmx 250mm, Waters). Enzyma�c synthesis of MSGb5 and DSGb5 Next, a�en�on was focused on the enzyma�c sialyla�on of 4a to give the oligosaccharide moiety of DSGb5. Thus, 4a was treated with ST3Gal1 in the presence of CMP-Neu5Ac (1.5 eq.) in sodium cacodylate buffer (pH = 7.5, 50 mM) containing MgCl 2 (20 mM) at 37 °C. The reac�on was performed in the presence of CIAP to hydrolyse CMP which may cause product inhibi�on. 20 Analysis of the reac�on mixture by TLC and MALDI-TOF MS indicated that a�er an incuba�on �me of 4 days, all star�ng material had been converted into product. Interes�ngly, Gb5 proved to be a rather poor substrate for the microbial α2,3-sialyltransferase (PmST1) and even a�er prolonged incuba�on, only par�al conversion was observed (~30%). Next, compound 5a was treated with recombinant ST6GalNAc6 and surprisingly, no product forma�on was detected. Gra�fyingly, the use of ST6GalNAc5 could readily add the second sialoside to provide DSGb5 ( 6a ). We found this enzyme requires an α2,3-linked sialoside for ac�vity and the use of Gal(β1,3) GalNAc did not give product whereas Neu5Ac(α2,3)Gal(β1,3)GalNAc was readily modified by ST6GalNAc5. These results indicate that the biosynthesis of DSGb5 involves an orchestrated a�achment of the sialosides in which the 2,3-linked Neu5Ac is first installed, followed by the introduc�on of the 2,6-sialoside. MSGb5 ( 5a ) and DSGb5 ( 6a ) were purified by Bio-Gel P-2 size exclusion column chromatography followed HPLC using a HILIC column and the resul�ng compounds were fully characterized by high resolu�on mass spectrometry and mul�-dimensional NMR. A ROESY experiment showed close proximity of H-4 and H-6 of GalNAc with H-3ax of branching Neu5Ac confirming proper connec�vity of the α2,6-sialoside of DSGb5. We also prepared the oligosaccharides of Gb5, MSGb5 and DSGb5 modified by an amino pentyl linker ( 4b , 5b and 6b , respec�vely) in a similar fashion by using acceptor 12b instead of 12a . These compounds were printed as a microarray to explore ligand requirements of Siglec-7.