Purine nucleoside fosforilase from Schistosoma Mansoni: crystal structure, knetics, studies and ligands search

Author(s)

Humberto D'Muniz Pereira

• Document type: Doctoral Thesis
• Press: São Carlos.
• Institution: Universidade de São Paulo (USP). Instituto de Física de São Carlos (IFSC/BT)
• Defense date: 2003-12-22

Examining board members

Richard Charles Garratt; Walter Filgueira de Azevedo Junior; Glória Regina Franco; Francisco Javier Medrano Martin; Otavio Henrique Thiemann

• Advisor: Richard Charles Garratt

Abstract

The parasite Schistosoma mansoni, unlike its mammalian hosts, lacks the \"de novo\" pathway for purine biosynthesis and depends on the salvage pathways for its purine requirements. One component of this pathway is Purine Nucleoside Phosphorylase (PNP) (E.C. 2.4.2.1). PNP catalyzes the reversible phosphorolysis of purine nucleosides to generate the corresponding purine base and ribose 1-phosphate. In the Schistosoma mansoni Genome Project the gene for this enzyme was isolated. The SmPNP cDNA was sequenced, corresponding to 1055 bases that code for a 287 amino acid protein with 49% sequence identity to its human erythrocyte homologue. The SmPNP gene was cloned into the pMAL C2G expression vector and the recombinant fusion protein was produced and purified using an amylose affinity column (approx. 80mg/mL). The fusion protein is composed of a Maltose Binding Protein adjoined to the SmPNP. After cleavage with Factor Xa, the cleavage product was purified using a cation exchange column. The KM and kcat of SmPNP for inosine phosphorolisis was determined to be 3?M and 222 s-1 respectively. This corresponds to the lowest value for KM yet described for a low molecular mass PNP. SmPNP crystals were obtained by the hanging drop method, using 18-24% PEG 1500, 20% glycerol in the presence of 32mM sodium acetate buffer (pH 4.9-5.0). When NDSB195 was used as an additive in the SmPNP crystallization the solution used was 28-30% PEG 1500, 20% glycerol and 32mM sodium acetate buffer (pH 4.9-5.O).Five datasets were obtained in the presence and absence of ligands, varying in resolution from 2,75 to 1,75?. All structures were solved by the molecular replacement method. The first structure (at 2,75?) was solved using the bovine PNP as the search model and in the remaining structures the search model was the SmPNP structure itself. The 1,75? structure was obtained from a crystal grow in the presence of the additive NDSB195, and after its determination NDSB195 was observed bound to the SmPNP active site via its phosphate binding site. Structures were also obtained for the apo SmPNP at 1,9? and its complex with phosphate at 2,0? resolution. All structures were used in the comparison with other low molecular mass PNPs. The SmPNP structure at 1,75A resolution was used in the search small molecule ligands, which potentially bind to SmPNP via virtual screening. For this purpose the program Gold together with a database of 36000 compounds with MW lower than 280Da was used. As a result 22 compounds were selected using the docking score and the H-bonding interaction with the key residues of the SmPNP active site. Using the bovine PNP and same docking approach 19 compounds were selected. These 41 compounds were assayed against SmPNP enzyme and 12 compounds were active against the SmPNP. One complex between SmPNP and one of these compound (AT2 169) was obtained and refined. This virtual screening approach was validated using known ligands of PNPs including inosine, guanosine, immucilins, etc. (AU)