Glutamate carboxypeptidase II (GCPII, NAALADase, or NAAG peptidase) is a catalytic zinc metallopeptidase. Its extracellular domain hydrolyzes the abundant neuropeptide, N-acetyl-L-aspartyl-L-glutamate (NAAG), to produce N-acetylaspartate and glutamate following the synaptic release of this transmitter. Thus, GCPII influences the extracellular concentrations of both glutamate and NAAG. NAAG activates group II metabotropic glutamate receptors, and activation of this receptor has been found to protect against anoxia-induced excitotoxic nerve cell death. In contrast, high levels of glutamate can be neurotoxic. Thus, GCPII is a potential therapeutic target for the reduction of excitotoxic levels of glutamate and enhancement of extracellular NAAG. To explore the structural basis of the interaction between GCPII and its inhibitors, we modeled the three-dimensional structure of the GCPII extracellular domain using a homology modeling approach. On the basis of the GCPII model, the structures of GCPII in complex with its potent inhibitors 2-(phosphonomethyl)pentanedioic acid (PMPA) and 4,4'-phosphinicobis(butane-1,3-dicarboxylic acid) (PBDA) were built by a computational docking method. The model of GCPII mainly consists of two alpha/beta/alpha sandwiches, between which two zinc ions are quadrivalently coordinated by the His379-Asp389-Asp455-H(2)O and the Asp389-Glu427-His555-H(2)O clusters, respectively. The ligand binding pocket is situated between these two sandwiches and is comprised of two subpockets: one is a surface-exposed highly positively charged subpocket; the other is a buried hydrophobic subpocket. The positively charged subpocket can accommodate the pharmacophore groups of inhibitor molecules (PMPA and PBDA) through the coordination of Zn(2+) with their phosphorus functionality and hydrogen-bonding interactions with Arg536, Arg538, and Ser456 (or Asn521), while the hydrophobic subpocket is engaged in hydrophobic and hydrogen-bonding interactions with the nonpharmacophore groups of PBDA. The predicted binding mode is consistent with the experimental data obtained from site-directed mutagenesis. On the basis of the predicted interaction mode, our structure-based design has led to a series of highly potent GCPII inhibitors.

The Ras guanyl releasing protein RasGRP belongs to the CDC25 class of guanyl nucleotide exchange factors that regulate Ras-related GTPases. These GTPases serve as switches for the propagation and divergence of signaling pathways. One interesting feature of RasGRP is the presence of a C-terminal C1 domain, which has high homology to the PKC C1 domain and binds to diacylglycerol (DAG) and phorbol esters. RasGRP thus represents a novel, non-kinase phorbol ester receptor. In this paper, we investigate the binding of indolactam(V) (ILV), 7-(n-octyl)-ILV, 8-(1-decynyl)benzolactam(V) (benzolactam), and 7-methoxy-8-(1-decynyl)benzolactam(V) (methoxylated benzolactam) to RasGRP through both experimental binding assays and molecular modeling studies. The binding affinities of these lactams to RasGRP are within the nanomolar range. Homology modeling was used to model the structure of the RasGRP C1 domain (C1-RasGRP), which was subsequently used to model the structures of C1-RasGRP in complex with these ligands and phorbol 13-acetate using a computational docking method. The structural model of C1-RasGRP exhibits a folding pattern that is nearly identical to that of C1b-PKCdelta and is comprised of three antiparallel-strand beta-sheets capped against a C-terminal alpha-helix. Two loops A and B comprising residues 8-12 and 21-27 form a binding pocket that has some positive charge character. The ligands phorbol 13-acetate, benzolactam, and ILV are recognized by C1-RasGRP through a number of hydrogen bonds with loops A and B. In the models of C1-RasGRP in complex with phorbol 13-acetate, benzolactam, and ILV, common hydrogen bonds are formed with two residues Thr12 and Leu21, whereas other hydrogen bond interactions are unique for each ligand. Furthermore, our modeling results suggest that the shallower insertion of ligands into the binding pocket of C1-RasGRP compared to C1b-PKCdelta may be due to the presence of Phe rather than Leu at position 20 in C1-RasGRP. Taken together, our experimental and modeling studies provide us with a better understanding of the structural basis of the binding of PKC ligands to the novel phorbol ester receptor RasGRP.