Large Sialidases

The crystal structures of two large sialidases have been determined the 83 kDa Vibrio cholerae sialidase [17] and the 68 kDa M. viridifaciens sialidase [19], both shown in Figure 2. The V. cholerae enzyme revealed a central catalytic P-propeller domain, flanked by two additional domains with similar topologies, resembling that of the lectins. The first lectin domain is at the N-terminus, and the second is inserted between the second and third sheets of the P-propeller. The lectin domains bind on as yet unidentified carbohydrate, but their structure suggests a role for these domains in the small intestine where the secreted sialidase may need to grasp the cell surface to perform its catalytic function without being swept away. 

The RIP and Asp-box motifs allow the prediction of the location of the catalytic domain in other large non-viral sialidases (Figure 4). Lack of sequence identity with other domains of known function does not allow prediction of the roles of the additional domains in many other large sialidases, however they too are likely to have carbohydrate-binding functionality. 

Traving, C., Schauer, R., and Roggentin, P., 1993, The primary structure of the large sialidase isoenzyme of Clostridium perfringens A99 and its comparison with further sialidases, Gly-coconj. J. 10 238-239. 

More is known as to the pH optimum (usually, pH 4.5-5.5), the requirement for ions, the substrate specificity, and kinetic data on siali-dases, as such results, extensively reviewed,55,"<1, H7 313 can also be obtained by use of crude or partially purified enzymes. The data obtained with impure enzymes must he interpreted with care, as the presence of various sialidases, of other glycosidases modifying the substrate during incubation, or of hypothetical activators and inhibitors may lead to large errors. Correspondingly, such studies require well characterized substrates, and identification of the enzyme reaction-produets. 

Apart from the 4-O-acetylated sialic acids, another sialidase-resist-ant sialic acid exists in Nature, namely, the internal, Gal-bound, sialic acid residue of GM,. In contrast to this side-positioned sialic acid, the sialyl residues bound to the peripheral Gal of gangliosides, or to both the peripheral and the internal sialic acid residues forming oligo-sialvl chains in several members of the large ganglioside family,la can be readily removed by sialidases, as was tested with viral, bacterial, and mammalian enzymes.35 -35 1... 

Although a large number of candidates were synthesized, none of the resulting compounds demonstrated significantly increased activity against influenza virus sialidase. Moreover, the interactions of individual substituents on the benzene ring with the active site were not found to be additive. The overall interaction of the molecules with the active site of the enzyme was dependent upon the electronic and steric interaction of each unique substituent, which made the design of inhibitors difficult.111 No compound of this family has proceeded to clinical trials. 

Sialidase is a homotetramer with C4 symmetry composed of identical disulfide-linked subunits (Fig. 17.6a [62]). Each monomer is a glycosylated polypeptide with six p-sheets assuming the appearance of six blades of a propeller [6, 64] with a right-handed twist. The catalytic site is observed to be at the center of the sixfold pseudosymmetry axis, which passes through the center of each monomer and relates the six p-sheets to each other (Fig. 17.6b [63]) [64, 65], The active site contains a large number of conserved amino acid residues, which are involved in binding to sialic acid in the substrate sialoglycoconjugate [64],... 

In the challenge to develop potent influenza virus sialidase inhibitors, a large amount of research has been dedicated to the manipulation of every position on 11 except C3. Structure-activity relationship (SAR) studies carried out on compounds derived from 11 before and during the development of zanamivir (reviewed in [101-103]) revealed structural requirements to conserve the main interactions between the substrate inhibitor and the active site of NA, particularly with regards to the carboxylate, C4-guanidino, and C5-acetamido moieties. 

Ajisaka et al. examined different sialidase somces and found that Newcastle disease virus (NDV) sialidase afforded predominantly the a-2,3 regioisomers, while Arthrobacter ureafaciens and Clostridium perfringens sialidases, in addition to the Vibrio cholerae sialidase examined by Thiem, favored the a-2,6-linked products [53]. Unfortunately, the reaction yields did not improve for the new enzymes, varying from 0.8% to 3.6% isolated yield. In the case of NDV sialidase, the high selectivity for a-2,3-sialosides stemmed from a large a-2,6/a-2,3 hydrolysis ratio. Hydrolysis of the a-2,6 products was found to be 28 times faster than the a-2,3 isomers. Inter-... 

The Trypanosoma cruzi trani-sialidase catalyzes the reversible transfer of NeuAc from a NeuAc-a-2,3-Gal-)8-OR sequence to an acceptor hearing the Gal-)8-OR motif (Scheme 42) [54], The enzyme is a particularly useful sialidase because it has very little hydrolytic activity and tends to almost exclusively catalyze transsialylation to a galactose. However a major drawback to this method is that to drive the gly-cosylation to completion, there is a need for large quantities of complex a-2,3-linked sialyl donors, which are generally difficult to obtain from natural sources. Other natural donors with a-2,6- or a-2,8-linked sialic acids have been examined but were discovered to be poor sialyl donors for a-2,3-sialylations catalyzed by T. cruzi trans-sialidase [55]. Simple aryl a-sialosides, such as the 4-nitrophenyl glycoside 122 and methylumbelliferone glycoside 130 (Scheme 43), have been found to be excellent... 

---