Synthesis and Quantitative Evaluation of Glycero‐D‐manno‐heptose Binding to Concanavalin A by Fluorous‐Tag Assistance

tion was reported earlier in the first demonstration of a microarray fabrication strategy based on noncovalent fluorous interactions. Mannose and other monosaccharides tagged by single C8F17 chains were noncovalently immobilized to fluorocarbon-coated glass slide surfaces and shown to tolerate conditions necessary for identifying binding partners such as conA. More recently, this fluorous-based microarray approach has proven valuable for the probing of other classes of small molecules. In the case of histone deacetylase inhibitors with dissocation constants of less than 0.1 s , the hits found by fluorous microarrays were comparable to those found by techniques such as surface plasmon resonance (SPR) and solution-based biochemical assays. Ideally, of course, the relative quantification of these binding interactions could also be carried out within the same fluorous microarray screening format. ConA is a plant lectin that is widely used like antibodies as research tools and diagnostics to identify the presence of specific sugars, such as mannose, on cells; however, in reality the sugar specificities of lectins have not been tested broadly, especially against less readily available carbohydrates. ConA is the most-studied lectin and is usually considered to bind terminal alpha-linked mannose, glucose, and N-acetylglucosamine. Earlier inhibition data suggest that modifications at the C-3, C-4, and C-6 positions of the d-mannopyranose deter binding to conA. In particular, the loss of the hydroxy group in the C-6 position as in 6-deoxy-d-mannose and 1,6-anhydrob-d-manno-pyranose result in complete loss of activity. To date, no studies have been realized on the addition of another hydroxymethyl group at this same position. In fact, surprisingly little is known about how seven-carbon sugars might mimic six-carbon sugars or interact with proteins. The seven-carbon sugars l-glycero-d-mannoand dglycero-d-manno-heptose (Figure 1) are common constituents in pathogenic bacteria of lipopolysaccharides (LPS) that mediate numerous responses to bacterial infections. The increasing resistance of many bacterial strains against conventional antibiotics is prompting the need for a detailed understanding of the immunological responses against these outer carbohydrate coats to inform the development of new therapeutic agents such as bacterial cell wall biosynthesis inhibitors and vaccines. Although humans lack mannoheptoses, the success of a vaccine against this bacterial sugar would presume the lack of cross-reactivity of antibodies generated against this distinct heptose antigen with the many mannose residues found in humans. As lectins are often used as antibody replacements in cellular carbohydrate-binding experiments, we reasoned that a study of the feasibility of using manno-heptose motifs in vaccine design should begin with assessing the ability of a well-characterized mannosebinding protein, conA, to cross-react with heptopyranoside motifs. To test the ability of conA to bind to l-glycero-d-mannoand d-glycero-d-manno-heptoses, we first needed a route to readily access both diastereomers. Several methods have been reported. To start from a less expensive six-carbon sugar, we required an approach to differentiate the alcohol at C-2 from Figure 1. Structure of mannose and heptomannose, and the strategy for their incorporation into a microarray by noncovalent fluorous interactions for screening with concanavalin A.