Kalkunte S. Srivenugopal

Nuclear magnetic resonance studies of polyamine binding to a defined DNA sequence.

The binding of spermine to the self-complementary DNA sequence d(C-G-C-G-A-A-T-T-C-G-C-G) has been studied by nuclear magnetic resonance spectroscopy. Free spermine gives narrow resonance lines and positive nuclear overhauser effects are observed between the spermine protons, as expected for a small molecule rotating freely in solution. In the spermine-DNA complex, there was no broadening of the spermine spectrum and very weak positive nuclear overhauser effects were observed, indicating that the spermine still has a remarkably short rotational correlation time. Spermine induced no changes in the DNA spectrum beyond those found upon addition of other salts. Although spermine interacts with DNA with a binding constant of approximately 10(6) at the low ionic strength under which these experiments were performed, it appears that the nature of the complex and the lifetime of the ligand on the DNA are such that the mobility of the spermine molecule is effectively independent of that of the DNA molecule.

DNA repair protein O6-alkylguanine-DNA alkyltransferase is phosphorylated by two distinct and novel protein kinases in human brain tumour cells.

We showed recently that human O(6)-alkylguanine-DNA alkyltransferase (AGT), an important target for improving cancer chemotherapy, is a phosphoprotein and that phosphorylation inhibits its activity [Srivenugopal, Mullapudi, Shou, Hazra and Ali-Osman (2000) Cancer Res. 60, 282-287]. In the present study we characterized the cellular kinases that phosphorylate AGT in the human medulloblastoma cell line HBT228. Crude cell extracts used Mg(2+) more efficiently than Mn(2+) for phosphorylating human recombinant AGT (rAGT) protein. Both [gamma-(32)P]ATP and [gamma-(32)P]GTP served as phosphate donors, with the former being twice as efficient. Specific components known to activate protein kinase A, protein kinase C and calmodulin-dependent kinases did not stimulate the phosphorylation of rAGT. Phosphoaminoacid analysis after reaction in vitro with ATP or GTP showed that AGT was modified at the same amino acids (serine, threonine and tyrosine) as in intact HBT228 cells. Although some of these properties pointed to casein kinase II as a candidate enzyme, known inhibitors and activators of casein kinase II did not affect rAGT phosphorylation. Fractionation of the cell extracts on poly(Glu/Tyr)-Sepharose resulted in the adsorption of an AGT kinase that modified the tyrosine residues and the exclusion of a fraction that phosphorylated AGT on serine and threonine residues. In-gel kinase assays after SDS/PAGE and non-denaturing PAGE revealed the presence of two AGT kinases of 75 and 130 kDa in HBT228 cells. The partly purified tyrosine kinase, identified as the 130 kDa enzyme by the same assays, was strongly inhibited by tyrphostin 25 but not by genestein. The tyrosine kinase used ATP or GTP to phosphorylate the AGT protein; this reaction inhibited the DNA repair activity of AGT. Evidence that the kinases might physically associate with AGT in cells was also provided. These results demonstrate that two novel cellular protein kinases, a tyrosine kinase and a serine/threonine kinase, both capable of using GTP as a donor, phosphorylate the AGT protein and affect its function. The new kinases might serve as potential targets for strengthening the biochemical modulation of AGT in human tumours.

Phosphorylation of O6-alkylguanine-DNA alkyltransferase: experience with a GST-fusion protein and a new pull-down assay

We showed recently that human O6-alkylguanine-DNA alkyltransferase (AGT), a key target for enhancing the efficacy of anticancer alkylating agents, is regulated by phosphorylation in brain tumor cells. This report describes the problems we encountered in using a glutathione S-transferase (GST)-tagged AGT as the substrate in our search for cellular AGT kinases, validation of a new pull-down assay for AGT phosphorylation, and its wide applicability for quantitating protein kinases in crude extracts and purified fractions. The GST-tag present in the fusion protein, by itself, was found to undergo significant phosphorylation by tumor cell extracts and contribute to spurious results. Instead, we used a histidine-tagged AGT protein, and its micro-scale purification with Talon resin as the basis for a quantitative pull-down assay, and applied it for measuring AGT phosphorylation by protein kinase C (PKC) and other cellular kinases. The pull-down procedure can be easily adopted for quantitating protein kinases in a variety of settings, as it overcomes the need for substrate immunoprecipitation when whole cell extracts are used, and eliminates the autophosphorylated kinase proteins, when purified kinases are used. Our observations call for caution in interpreting the results with GST-fusion proteins in phosphorylation studies.

Activity and distribution of the cysteine prodrug activating enzyme, 5-oxo-L-prolinase, in human normal and tumor tissues

5-Oxo-L-prolinase (OPase), a key enzyme of the gamma-glutamyl cycle, has the ability to metabolize L-2-oxothiazolidine-4-carboxylic acid (OTC) to cysteine, and thereby increase intracellular glutathione (GSH) levels. This strategy of GSH elevation can be potentially exploited to reduce normal tissue toxicity of anticancer agents, provided that sufficient differences exist in OPase levels between normal and malignant tissues. In this study, therefore, we quantitated OPase activity in primary specimens of matched and unmatched human normal and tumor (lung, breast, kidney, colon and ovary) tissues using a newly developed non-radioactive OPase assay, based on the production of cysteine from OTC. The rank order of OPase activity in extracts of 24 normal tissues examined was kidney > lung, breast and colon > ovary. OPase activity was present in all 37 tumor samples, but at variable levels. Tumor OPase levels were generally equivalent to those in their normal tissue counterparts, with the notable exception of Wilms tumors, which had markedly lower levels than normal kidney (P < 0.02). However, when 14 matched tumor and adjacent normal tissues were compared, OPase levels were significantly higher in normal specimens than tumors for individual patients (P < 0.005). These higher normal tissue/tumor OPase ratios suggest that OTC may be useful in decreasing normal tissue toxicity, at least, for some tissues during cancer therapy.

Modulation of the relaxing activity of Escherichia coli topoisomerase I by single-stranded DNA binding proteins

Removal of negative superhelical turns in ColE1 plasmid DNA by Escherichia coli topoisomerase I was markedly enhanced by the presence of single-stranded DNA binding protein from E. coli. A lack of species specificity makes unlikely the possibility of physical association between topoisomerase I and single-stranded DNA binding proteins. Stabilization of single-stranded regions in supercoiled DNA by single-stranded DNA binding protein would appear to be the basis of the enhancement of topoisomerase activity.

Stimulation and inhibition of 1,3-BIS(2-chloroethyl)-1-nitrosourea-induced strand breaks and interstrand cross-linking in col e1 plasmid deoxyribonucleic acid by polyamines and inorganic cations

The influence of various polyamines and metallic cations on 1,3-bis(2-chloroethyl)-1-nitrosourea (BCNU)-induced DNA single-strand breaks and DNA interstrand cross-linking was in Col E1 plasmid using electrophoretic techniques. Spermidine and spermine (0.4 to 1.5 mM concentration range) markedly stimulated BCNU-induced DNA nicking, whereas putrescine had no effect on the nicking process. In contrast to the polyamines, BCNU-induced DNA nicking was decreased by the three inorganic cations, Na+ (100 and 200 mM), Mg2+ (0.5 and 1.5 mM), and Co3+ (NH3)6 (0.2 and 0.4 mM), with the trivalent hexamminecobalt ions being most inhibitory. When the monofunctional N-methyl-N-nitrosourea (MNU) was used (instead of the bifunctionally active BCNU) to alkylate Col E1 DNA, nicking of the DNA was inhibited by spermidine. Furthermore, the ability of chloroethylated Col E1 DNA to form interstrand cross-links after treatment with BCU was inhibited by 0.5 mM spermidine and 0.5 mM spermine, both concentrations within the intracellular range. Putrescine at 3-6 mM only marginally stimulated DNA cross-linking. In comparison, the inorganic cations all enhanced Col E1 DNA cross-linking by BCNU, with the rank order of cross-link stimulation being Mg2+, Na+, and Co3+ (NH3)6. These results provide evidence that polyamines can interact with DNA to modulate chloroethylnitrosourea-induced DNA damage and that the interaction is not only a function of the charge on the polyamine molecule but also of the chemical structure of the polyamine.