Islet-cell antigen 512 (IA-2) and phogrin (IA-2beta) are atypical members of the receptor protein tyrosine phosphatase (PTP) family that are characterized by a lack of activity against conventional PTP substrates. The physiological role(s) of these proteins remain poorly defined, although recent studies indicate that IA-2 may be involved in granule trafficking and exocytosis. To further understand their function, we have embarked upon developing low-molecular-mass inhibitors of IA-2 and IA-2beta. Previously, we have shown that a general PTP inhibitor, 2-(oxalylamino)benzoic acid (OBA), can be developed into highly selective and potent inhibitors of PTP1B. However, since wild-type IA-2 and IA-2beta lack conventional PTP activity, a novel strategy was designed whereby catalytically active species were generated by 'back-mutating' key non-consensus catalytic region residues to those of PTP1B. These mutants were then used as tools with which to test the potency and selectivity of OBA and a variety of its derivatives. Catalytically competent IA-2 and IA-2beta species were generated by 'back-mutation' of only three key residues (equivalent to Tyr(46), Asp(181) and Ala(217) using the human PTP1B numbering) to those of PTP1B. Importantly, enzyme kinetic analyses indicated that the overall fold of both mutant and wild-type IA-2 and IA-2beta was similar to that of classic PTPs. In particular, one derivative of OBA, namely 7-(1,1-dioxo-1 H -benzo[ d ]isothiazol-3-yloxymethyl)-2-(oxalylamino)-4,7-dihydro-5 H -thieno[2,3- c ]pyran-3 -carboxylic acid ('Compound 6 ' shown in the main paper), which inhibited IA-2beta((S762Y/Y898P/D933A)) (IA-2beta in which Ser(762) has been mutated to tyrosine, Tyr(898) to proline, and Asp(933) to alanine) with a K (i) value of approximately 8 microM, appeared ideal for future lead optimization. Thus molecular modelling of this classical, competitive inhibitor in the catalytic site of wild-type IA-2beta identified two residues (Ser(762) and Asp(933)) that offer the possibility for unique interaction with an appropriately modified 'Compound 6 '. Such a compound has the potential to be a highly selective and potent active-site inhibitor of wild-type IA-2beta.