Ingrid 't Hart

77 Chemical synthesis of sulfoglycolipid analog SM1a-core3 4 The HAE3 an�body binds mucin-type O -glycans, which are mainly found in the gut. In an effort to find sulfated oligosaccharide sequences on O -glycans in the colon that are similar to SM1a, Sd(a)/Cad-an�gen-like epitopes (GalNAcβ1,4(NeuAcα2,3)Gal) were found, which all contain a core3 O -glycan sequence (GlcNAcβ1,3GalNAcαSer/Thr). 20 From these findings a hybrid structure of SM1a and a core3 disaccharide was proposed as a poten�al ligand for HAE3, here named SM1-core3 (Figure 1B). 14 SM1-core 3 has not been iden�fied or isolated from mucin or other sources. Isola�on of intact sulfoglycolipids and its analogs is tedious due to the presence of both charged and lipophilic residues. Furthermore, the isola�on process o�en results in heterogenous mixtures. Therefore, a synthe�c approach is required to obtain SM1-core 3. Chemical synthesis of SM1a with two different sphingosine composi�ons has been described. 21,22 However, no sulfoglycolipid O -glycan hybrids have been synthesized. A principle difference between O -glycans and SGLs are their characteris�c reducing glycan moie�es. The reducing moiety of O -glycans consists of GalNAc α-linked to serine (Ser) or threonine (Thr) of pep�des, whereas complex glycolipids contain lactosyl residues β-linked to ceramide. 23,24 Therefore, a different chemical synthe�c strategy is required to synthesize a SGL O-glycan hybrid compared to SGL synthesis. First of all, the core3 reducing moiety require chemical glycosyla�on between GlcNAc and GalNAc, whereas SGLs can be synthesized from commercially available lactose. Furthermore, terminal 1,2- cis (in this case α-) anomeric selec�vity is required to mimic the natural O -glycan linkage. Chemical synthesis of 1,2- cis linkages is more challenging than 1,2- trans anomers (β in the case of lactosylceramide), where neighboring group par�cipa�on can be used to direct the stereoselec�ve outcome. 25 Although HAE3 is known to bind tumor glycan markers, further elucida�on on the specificity to various glycan structures is required. Here we report the first chemical synthesis of a protected formof the SM1-core 3 glycan; an poten�al ligand formonoclonal an�body HAE3. Global deprotec�on and 3- O -sulfa�on will provide the desired synthe�c target. An α-anomeric aminopentyl spacer has been incorporated at the reducing end to enable facile immobiliza�on for microarray binding studies. The results of such a binding study will provide insight in the binding specificity of HAE3. Furthermore, it will encourage the search for more relevant sulfoglycolipid (anolog)s as possible cancer- associated glycan epitopes. Highly defined cancer-associated an�gens will provide more insight in cancer cell mechanisms and serve as handles for cancer specific therapies. Results and Discussion Chemical synthesis The chemical synthesis of protected pentasaccharide 2 was designed in such way that the posi�on of the 3- O -sulfate is protected with an O -2-naphthylmethylether (Nap) as an orthogonally cleavable group (Scheme 1). All glycosidic linkages are in the β-configura�on, except the reducing aminopentyl linker, resembling the α-linkage (to Ser/ Thr) found in O -glycans. Appropriate C-2 ester protec�ng groups ( O Bz, N Phth, N HTroc) were chosen to install β-anomeric linkages by neighboring group par�cipa�on during glycosyla�on. A block synthesis approach was used, in which coupling of disaccharide 3