Figure 5 A typical reaction catalyzed by mycobacterial arabinosyltransferases (AraT's).
has been cloned and expressed [52-54]. In addition, an assay for their activity that uses mycobacterial membrane preparations as the enzyme source has been developed [49,50]. The assay measures the incorporation of radiolabeled arabinose from 10 into arabinan, and it has been used to screen both potential substrates and inhibitors [46,47,55,56].
Although the natural acceptor substrates for these enzymes are lipid-bound intermediates, these AraT's also recognize small arabinofuranosyl oligosaccharides. A few of the oligoarabinosides that have been shown to be substrates for these enzymes are illustrated in Figure 6 [49,55]. Arabinofuranose disaccharides appear to be the minimum epitopes. Furthermore, investigation of the AraT recognition of the disaccharide a-D-Ara/-(1^5)-a-D-Ara/-R (13-15) and the trisaccharide a-D-Ara/-(1^5)-a-D-Ara/-(1^5)-a-D-Ara/-R (16, 17) revealed that glycosides with R = octyl (14, 17) are better substrates and those with R = methyl (13, 16). Surprisingly, the dodecyl glycoside 15 was a poorer substrate than 13. In the investigation of 14 and 17, the products of the enzymatic reactions were isolated and their structures elucidated . In both cases, mixtures of two products, corresponding to addition of an Ara/ residue to either the 2- or the 5-position of the terminal residue in the oligo-saccharide acceptor, were obtained. From this study it was concluded that 10 was the source of the 0-(1^2)-, and a-(1^5)-linked Ara/ residues.
Less is known about the formation of the a-(1^3) branch points than about the 0-(1^2) and a-(1^5) linkages. Only 0-(1^2) and a-(1^5) linkages are formed by the incubation of small oligosaccharide substrates with 10 in the presence of a mycobacterial membrane preparations . However, additional work  demonstrated that when radioactive 10 is incubated with endogenous acceptors and a mycobacterial membrane preparation, a polymer essentially identical to native AG can be obtained. Additionally the radioactivity in the polymer was equally distributed
throughout, which suggests that 10 is the major, and possibly sole, source of Ara/ residues. The lack of a-(1^3) linkages formed in the initial study  could be attributed to a number of factors: (1) the instability or absence of a-(1^3) AraT activity in the membrane preparation, (2) the possibility that this enzyme recognizes oligosaccharide substrates larger than those investigated, or (3) the possibility that another activated donor (e.g., a sugar nucleotide) is used by this AraT. Although the presence of UDP-Ara/ in mycobacteria has been reported , the incorporation of arabinose from this source into arabinan has not been demonstrated.
The range of donor substrates that are tolerated by these enzymes has also been explored. A series of analogs was prepared, each one differing from 10 only in the identity of the lipid chain, and the set was screened for biological activity . It was discovered that conjugates of C50 and C55 prenols were substrates, while those compounds with shorter (<C10) prenol derivatives (geraniol, nerol, and citronellol) were inactive.
Was this article helpful?