“…Therefore, although the requirement for hydroxyl-groups of the mannose-or galactose-type within the binding-site on the carbohydrate molecules is determined by the E-N/Q-D pair of residues, other differences seen in the sequences of the lectin domains, and of the extensive loop carrying the binding sites in particular, seem to be also involved in ligand selection in the in vivo interaction of C-type lectins with complex carbohydrate structures. Domains identified by sequence comparison and predicted to have an overall folding similar to that seen in the crystal structure of the C-type lectin domain of MBP (Weis et al, 1991b), but lacking some of the characteristic residues involved in calcium/carbohydrate binding, are seen in a number of type I1 membrane proteins ( QPD-E-WND gal Tanaka et al, 1988 QPD-E-WND gal Zimmerman & Ruoslahti, 1989 QPD-E-WND gal Drickamer, 1981Spiess & Lodish, 1985Drickamer et al, 1984Ii et al, 1990Hoyle & Hill, 1991Sat0 et al, 1992Curtis et al, 1992Suter et al, 1987Yokoyama et al, 1989Chan & Takei, 1989Giorda et al, 1990Van Hoegen et al, 1990Houchins et al, 1991Wong et al, 1991Yoshimatsu et al, 1992Lee et al, 1991Taylor et al, 1989Drickamer et al, 1986White et al, 1985Benson et al, 1985Rust et al, 1991Holmskov et al, 1993bHolmskov et al, 1993aLasky et al, 1989Bevilacqua et al, 1991Johnson et al, 1989Christa et al, 1994Ng & Hew, 1992Ewart et al, 1992Rouquier et al, 1991Taylor et al, 1990 EPN-E-WND man (Drickamer, 1992). Not all of the carbohydrate specificities have been determined (*), and certain proteins, although a Proteins containing C-type lectin domains or regions w...…”