Department ofBiomedicinal Chemistry, School ofPharmacy and Maryland ... Studies with wild-type and recombinant strains of S. lividans support the lack of.
Vol. 57, No. 2
APPLIED AND ENVIRONMENTAL MICROBIOLOGY, Feb. 1991, p. 610-611
0099-2240/91/020610-02$02.00/0 Copyright C 1991, American Society for Microbiology
Genetic and Biochemical Evidence for the Lack of Significant Hydrolysis of Soman by a Flavobacterium Parathion Hydrolase BURTON M. POGELL,1* SHARON S. ROWLAND,' KATHLEEN E. STEINMANN,2 MARILYN K. SPEEDIE, AND FRANCIS C. G. HOSKIN2 Department of Biomedicinal Chemistry, School of Pharmacy and Maryland Biotechnology Institute, University of Maryland, Baltimore, Maryland 21201,1 and Department of Biology, Illinois Institute of Technology, Chicago, Illinois 606162 Received 6 August 1990/Accepted 28 November 1990
Pure recombinant Flavobacterium parathion hydrolase (an organophosphorus acid anhydrase) from Streptomyces lividans was found to hydrolyze the toxic nerve agent soman at only 0. 1% of the rate observed with parathion as substrate. Studies with wild-type and recombinant strains of S. lividans support the lack of significant soman breakdown by the hydrolase and also indicate the presence in S. lividans of other significant hydrolytic enzymatic activity towards soman.
It has been suggested in the literature (1) that an organophosphorus acid (opa) anhydrase from Pseudomonas diminuta parathion hydrolase, also can catalyze the hydrolysis of the phosphonofluoridate bond of nerve gases such as 1,2,2-trimethylpropyl methylphosphonofluoridate (soman) and isopropyl methyl phosphonofluoridate (sarin). One obstacle in confirming this point has been the difficulty in obtaining purified enzyme from either Pseudomonas or Flavobacterium species, in which the enzyme is found associated with membrane fractions. The structural genes coding for these enzymes, both of which are found on plasmids, have identical nucleotide composition in both species (9, 12). The availability of the gene for this enzyme (opd), has made possible its expression in Streptomyces lividans and purification to homogeneity (13). The hydrolase was found to be excreted in soluble form in this recombinant species and to be identical with pure native Flavobacterium enzyme. Studies with different plasmid clones in S. lividans and with homogeneous enzyme have now clearly established that this enzyme does not hydrolyze soman at a significant rate, although it is very active on a chemical analog presumed to mimic the structure of soman, p-nitrophenylethylphenylphosphinate (NPEPP). In addition, these studies indicate that one or more enzymes are excreted by wild-type S. lividans 66, which can hydrolyze soman at significant rates. When the enzymatic hydrolysis of parathion and NPEPP at 0.14 mM substrate concentrations and pH 8.5 were compared by using the secreted recombinant Flavobacterium parathion hydrolase from S. lividans (13), NPEPP was found to be hydrolyzed at a rate 1.2 to 1.3 times that of parathion under the assay conditions described. Identical rates were found with both a crude concentrated broth and partially purified enzyme from a fermentation in Luria broth. These results prompted us to test this material for its effect on soman breakdown. Assays of both soman and diisopropylphosphorofluoridate hydrolysis were carried out under conditions described by Hoskin (7). A higher substrate concentration (3 mM instead of 0.14 mM) and a lower pH (7.2 instead of 8.5) were necessary because of measurement of F- release instead of p-nitrophenol formation. In all *
cases, enzyme units are expressed as micromoles of product formed per minute and have been corrected for any spontaneous hydrolysis. With dialyzed concentrated crude broth (0.72 U of parathion hydrolysis per ml of broth), soman was hydrolyzed at a rate of 0.23 U/ml of broth. Undialyzed material degraded soman at a lower rate (0.07 U/ml). However, when a high-activity broth (8.2 U/ml) from a fed-batch fermentation in tryptone-glucose (3, 10) and a preparation of pure parathion hydrolase (8.5 U/ml) (11) were analyzed, the rate of soman hydrolysis in the former had not increased (0.28 U/ml) and was almost zero with pure enzyme (0.016 U/ml). These results suggested that the soman hydrolysis was due to some other excreted enzyme. This conclusion was confirmed by analysis of the relative rates of parathion and soman hydrolysis by broth from three strains: S. lividans 66, 66(pRYE6), and 66(pRYE1). The first two strains excreted no parathion hydrolase activity (pRYE6 is a plasmid derivative of pRYE2 [13] with the opd gene deleted), and the latter had parathion hydrolytic activity of 4.3 U/ml of broth. Measurements of soman hydrolysis indicated the presence of very low activity in all three broths, 0.135, 0.125, and 0.11 U/ml, respectively. Thus, whatever enzyme was responsible for soman hydrolysis was excreted independently of expression and secretion of the parathion hydrolase gene product. Mn ion increased the rate of soman hydrolysis in crude fermentation broth, but had no effect when pure enzyme was used. This observation also is consistent with the presence of an enzyme in S. lividans different from the Flavobacterium hydrolase. Diisopropylphosphorofluoridate was hydrolyzed by S. lividans 66 and 66(pRYE1) fermentation broths at rates about equal to those with soman. The results of published kinetic measurements on three pure preparations of parathion hydrolase, our recombinant form from S. lividans (11), the native enzyme isolated from Flavobacterium sp. by Mulbry and Karns (8), and the recombinant Pseudomonas gene product isolated from insect tissue culture cells by Dumas et al. (4, 5), are summarized in Table 1. It is clear that the rate of hydrolysis of soman by both recombinant preparations is negligible, suggesting that NPEPP is not a useful analog for predicting either soman or sarin hydrolysis. Small differences in results with the recombinant preparation from insect cells may be because of a failure to remove the signal sequence. A slightly
Corresponding author. 610
NOTES
VOL. 57, 1991 TABLE 1. Comparison of kinetic constants of purified parathion hydrolasesa Substrate
Parathion NPEPP Soman
Km (,uM) for given enzyme B C A
68 357 ND
91 ND ND
240 1,100 500
Kcat (s-1) for given enzyme B A
926 2,210 1.2
936-1,760 ND ND
C
621 670 4.8
a A, Flavobacterium gene expressed in S. lividans (from reference 11). B, native Flavobacterium enzyme (from reference 8). C, Pseudomonas gene expressed in insect cells (from reference 4 and 5). ND, Not determined.
higher molecular weight for this product and the authors' reported inability to get an N-terminal amino acid sequence support this conclusion. However, the nucleotide sequences of these two bacterial genes are identical (9, 12). One of the basic problems in using an enzyme for decontamination of pollutants is the necessity of obtaining large amounts of enzyme by a simple, low-cost process. Levels of secretion of parathion hydrolase of 30 U/ml (30 mg/liter) have been attained in fed-batch fermentations of S. lividans (10). If the gene coding for soman hydrolysis from S. lividans or a similar gene from enzymes recently reported to hydrolyze soman at rapid rates in other microorganisms (2, 6) could be cloned and expressed at equal levels, then preparations of broth from such streptomycete fermentations should be useful for decontamination of this nerve agent. This research was supported by funding from the Center for Agricultural Biotechnology and the Medical Biotechnology Center of the Maryland Biotechnology Institute. REFERENCES 1. Attaway, H., J. 0. Nelson, A. M. Baya, M. J. Voll, W. E. White, D. J. Grimes, and R. R. Colwell. 1987. Bacterial detoxification of diisopropyl fluorophosphate. Appl. Environ. Microbiol. 53: 1685-1689.
611
2. Cheng, T.-C., and J. J. DeFrank. 1990. Abstr. Annu. Meet. Am. Soc. Microbiol. 1990, K102, p. 236. 3. Coppella, S. J., N. DelaCruz, G. F. Payne, B. M. Pogell, M. K. Speedie, J. S. Karns, E. M. Sybert, and M. A. Connor. 1990. Genetic engineering approach to toxic waste management: case study for organophosphate waste treatment. Biotechnol. Progr. 6:76-81. 4. Dumas, D. P., S. R. Caldwell, J. R. Wild, and F. R. Raushel. 1989. Purification and properties of the phosphotriesterase from Pseudomonas diminuta. J. Biol. Chem. 264:19659-19665. 5. Dumas, D. P., H. D. Durst, W. G. Landis, F. M. Raushel, and J. R. Wild. 1990. Inactivation of organophosphorus nerve agents by the phosphotriesterase from Pseudomonas diminuta. Arch. Biochem. Biophys. 277:155-159. 6. Gallo, B., J. Walker, and F. C. G. Hoskin. 1990. Abstr. Annu. Meet. Am. Soc. Microbiol. 1990, 066, p. 274. 7. Hoskin, F. C. G. 1985. Inhibition of a Soman- and diisopropylphosphorofluoridate(DFP)-hydrolyzing enzyme by Mipafox. Biochem. Pharmacol. 34:2069-2072. 8. Mulbry, W. W., and J. S. Karns. 1989. Purification and characterization of three parathion hydrolases from gram-negative bacterial strains. Appl. Environ. Microbiol. 55:289-293. 9. Mulbry, W. W., and J. S. Karns. 1989. Parathion hydrolase specified by the Flavobacterium opd gene: relationship between the gene and protein. J. Bacteriol. 171:6740-6746. 10. Payne, G. F., N. DelaCruz, and S. J. Coppelia. 1990. Improved production of heterologous protein from Streptomyces lividans. Appl. Microbiol. Biotechnol. 395-400. 11. Rowland, S. S., M. K. Speedie, and B. M. Pogell. 1991. Purification and characterization of a secreted recombinant phosphotriesterase (parathion hydrolase) from Streptomyces lividans. Appl. Environ. Microbiol. 57:440-444. 12. Serdar, C. M., D. C. Murdock, and M. F. Rohde. 1989. Parathion hydrolase gene from Pseudomonas diminuta MG: subcloning, complete nucleotide sequence, and the expression of the mature portion of the enzyme in Escherichia coli. Bio/technology 7:1151-1155. 13. Steiert, J. G., B. M. Pogell, M. K. Speedie, and J. Laredo. 1989. A gene coding for a membrane-bound hydrolase is expressed as a secreted, soluble enzyme in Streptomyces lividans. Bio/ Technology 7:65-68.
