Cloning and Expression of Vitreoscilla Hemoglobin Gene in Burkholderia Sp

26 Biotechnol. Prog. 2000, 16, 26−30

Cloning and Expression of Vitreoscilla Hemoglobin Gene in Burkholderia sp. Strain DNT for Enhancement of 2,4-Dinitrotoluene Degradation

Sangeeta M. Patel,† Benjamin C. Stark,† Kwang-Woo Hwang,† Kanak L. Dikshit,‡ and Dale A. Webster*,†

Department of Biological, Chemical, and Physical Sciences, Illinois Institute of Technology, Chicago, Illinois 60616, and Institute of Microbial Technology, Sector 39, Chandigarh 160014, India

The gene (vgb) encoding the hemoglobin (VHb) of Vitreoscilla sp. was cloned into a broad host range vector and stably transformed into Burkholderia (formerly Pseudomo- nas) sp. strain DNT, which is able to degrade and metabolize 2,4-dinitrotoluene (DNT). Vgb was stably maintained and expressed in functional form in this recombinant strain (YV1). When growth of YV1, in both tryptic soy broth and minimal salts broth containing DNT and yeast extract, was compared with that of the untransformed strain, YV1 grew significantly better on a cell mass basis (A600) and reached slightly higher maximum viable cell numbers. YV1 also had roughly twice the respiration as strain DNT on a cell mass basis, and in DNT-containing medium, YV1 degraded DNT faster than the untransformed strain. YV1 cells pregrown in medium containing DNT plus succinate showed the fastest degradation: 100% of the initial 200 ppm DNT was removed from the medium within 3 days.

Introduction DNT degradation both indirectly, by enhancing respiration and thus growth, and directly, by enhancing the The prokaryotic hemoglobin (VHb) from the Gram- three oxygen-dependent oxidations in the DNT degrada- negative, aerobic bacterium Vitreoscilla is well-character- tion pathway. Thus, VHb technology may be useful for ized. Its in vivo role is suggested to be an oxygen trap to bioremediation in the field, especially when oxygen-poor allow Vitreoscilla to survive and grow in the microaerobic conditions exist in contaminated sites, for example, in environments that it normally inhabits (1, 2). The gene soil or water below the surface. (vgb) encoding VHb has been isolated, sequenced, and To this end we report here the engineering of Burkhold- characterized (3, 4). When this gene is transformed into eria strain DNT to contain vgb and produce VHb. We a heterologous host (e.g., Escherichia coli, fungi), it can show that the resulting strain (YV1) does, in fact, have increase the growth of and product production by the host enhanced growth and DNT-degrading abilities compared - (5 10). Presumably the VHb produced in these cells with strain DNT. enhances respiration at low oxygen concentrations as it is thought to do in Vitreoscilla. In Pseudomonads, it has Materials and Methods been shown to have potential for enhancing bioremediation as well (11). Bacterial Strains and Plasmids. Burkholderia (orig- inally Pseudomonas) sp. strain DNT (R34) was provided In the work reported here we have extended this by Jim C. Spain (Tyndall Air Force Base, FL). The technology to Burkholderia sp. strain DNT (formerly organism was maintained on a minimal salts medium Pseudomonas sp. strain DNT, also known as R34), which supplemented with 100 ppm DNT and 200 ppm yeast can degrade 2,4-dinitrotoluene (DNT). DNT, a byproduct extract (MSB-DNT-YE) at 35 °C (13). Plasmid pSC160 of trinitrotoluene production and usage, is considered (16.0 kb) was constructed as described previously (8)by toxic at levels greater than 0.13 ppm. Recent work from inserting plasmid pUC8:16 (3) into the EcoRI site of several laboratories suggests that nitrotoluenes can be autonomous vector pKT230 (11.9 kb) (15), which inacti- biodegraded by an oxidative pathway and that mineral- vated the Strr (streptomycin resistance) but left the Kmr ization of TNT and DNT by bacteria can occur (12, 13). (kanomycin resistance). Plasmid pUC8:16 (4.1 kb), in Suen and Spain (14), by biochemical and genetic dissec- turn, contains vgb, including its native promoter cloned tions, have characterized the pathway and genes of into vector pUC8 which confers ampicillin resistance (16). Burkholderia sp. strain DNT responsible for the complete Plasmid pSC160 thus confers resistance to both ampi- mineralization of DNT by an oxidative pathway. Because cillin and kanamycin, which was the initial method of oxygen is required at three early steps in this process, selection. engineering strain DNT so that it produces VHb may lead Transformation. Burkholderia sp. strain DNT was to improvements in its ability to bioremediate DNT; transformed with pSC160 by the method of Liu et al. (8). increased intracellular oxygen via VHb may enhance Plasmid pSC160 was isolated from P. aeruginosa (rather than E. coli) by a scale-up of the alkaline lysis method † Illinois Institute of Technology. described by Birnboim and Doly (17). This transformant ‡ Institute of Microbial Technology. is denoted as YV1.

Growth, Respiration, and Hemoglobin Measure- ment. Cells were inoculated from freezer stocks into 25 mL of fresh MSB-DNT-YE medium for the untransformed strain (DNT). For the transformed strain (YV1), 100 µg/mL ampicillin and 40 µg/mL kanamycin were added to the same medium. Cells were grown overnight at 35 °C and 150 rpm, harvested by centrifugation, washed with 0.01 M sodium phosphate buffer, pH 7.2, and inoculated into 200 mL of fresh MSB-DNT-YE medium without antibiotics for both strains DNT and YV1. Growth was continued at 35 °C and 150 rpm for 120 h. Inocula were prepared in the same way in MSB- DNT-YE, and cells were grown in the same way for growth comparisons (without antibiotics) in TSB (tryptic soy broth, Difco) or MSB-DNT-YE plus 0.02 M succinate for both strains. Samples were taken at intervals for determination of viable cell counts by plating on TSB agar. For determination of plasmid stability, 100 colonies were randomly transferred with sterile toothpicks from each plate to TSB agar plates containing 100 µg/mL ampicillin and 40 µg/mL kanamycin. Samples were also taken for determination of absorbance at 600 nm, diluted as necessary with the appropriate medium to keep the measured A600 below 0.3. Measurement of respiration rates (8) and hemoglobin levels by CO-difference spectra were as previously described (3, 18) using the extinction -1 -1 coefficient, E419 -436 nm ) 274 mM cm . Other Analytical Methods. Nitrite release was measured as described by Smibert and Krieg (19). DNT degradation and formation of intermediates in the degradation pathway were assayed in the cell-free culture medium using high performance liquid chromatography (HPLC) with a Spherosorb C18 column (Alltech, Deer- field, IL) and acetonitrile/13.5 mM trifluoroacetic acid in water, 1/1, (v/v) as the mobile phase. The flow rate was 1.5 mL/min, and the UV detector was set at 230 nm. Culture medium samples were clarified by centrifugation at 5000 × g for 15 min before chromatography. DNA Isolation and Southern and Western Blots. For the Southern blot, whole DNA was purified from the cells using a modification of the procedure of Marmur (20). The DNA samples were cut with EcoRI, and fragments were separated by electrophoresis on a 1% agarose gel, transferred to a Hybond nitrocellulose Figure 1. Presence of pSC160 and vgb in YV1. (A) Total DNA membrane, and processed using an ECL kit (Amersham) was isolated from strains DNT and YV1, digested with EcoRI, following the manufacturer’s instructions; plasmid pUC8: and separated on a 1% agarose gel as described in Materials 16 was used as the probe. For the Western blot analysis, and Methods. Lanes: 1, λ DNA digested with HindIII; 2, pUC8: 16 digested with EcoRI and HindIII; 3, pSC160 digested with purified VHb was used to immunize rabbits to generate EcoRI; 4 and 5, strain DNT DNA digested with EcoRI; 6-8, YV1 polyclonal VHb-specific antibodies. Burkholderia sp. DNA (from three separate isolates) digested with EcoRI. (B) strains DNT and YV1 were grown at 35 °C at 100 rpm Southern blot of gel in A. Transfer and hybridization procedures overnight, and total cellular proteins were separated on were as described in Materials and Methods. The probe used - an SDS-PAGE (15%) gel and transferred to a nitrocel- was pUC8:16, which contains the entire vgb gene. Lanes 1 8, same as for A. The position of the 4.1 kb pUC8:16 fragment lulose membrane (Hybond, Amersham) by the procedure produced by EcoRI digestion of pSC160 is indicated by the arrow of Towbin et al. (21) using a transblot cell (BioRad). The in the right margin. The 11.9 kb EcoRI fragment of pSC160 is membrane was treated with 2% skim milk for 2 h, obscured by the chromosomal DNA band. washed twice with PBS (140 mM NaCl, 10 mM Na2HPO4, 2.7 mM KCl, 1.8 mM KH2PO4, pH 7.4) containing 0.1% Results Tween 20, and incubated with polyclonal VHb antibodies Total DNA (chromosomal plus plasmid) isolated from for approximately 2 h. The membrane was washed twice strain YV1 and subjected to restriction analysis using with PBS and incubated with the secondary antibody EcoRI showed the presence in this strain of a DNA (peroxidase-conjugated goat anti-rabbit IgG) for another fragment of the same size (4.1 kb) as that produced from 2 h. Unreacted secondary antibodies were removed by pSC160; this band was absent from strain DNT DNA washing with PBS, and the membrane was treated with (Figure 1A). Southern blots using pUC8:16 as the probe 3,3-diaminobenzidineimidazole in 0.1 M phosphate buffer, confirmed that this 4.1 kb EcoRI fragment of pSC160 pH 7.0. The membrane was developed by the addition of contained pUC8:16 (Figure 1B). The intensity of the 4.1 20 mL of 30% (w/v) H2O2 to visualize the VHb cross- kb pSC160 bands in lanes 6-8 in Figure 1A indicated reacting bands. that the plasmid exists autonomously in multiple copies 28 Biotechnol. Prog., 2000, Vol. 16, No. 1

Figure 3. Growth curves of strains YV1 and DNT in TSB medium measured by viable cell number. Each point is the average of three experiments. Error bars indicate standard deviations.

Figure 4. Growth curves of strains YV1 and DNT in TSB medium measured by absorbance at 600 nm. Each point is the average of three experiments. Error bars indicate standard deviations.

VHb. Using these spectra, the amount of VHb present in YV1 cells harvested at early stationary phase was estimated to be 8-10 nmol/g wet weight, considerably less than is found in Vitreoscilla or E. coli (3) but similar to that found by us in various Pseudomonads trans- Figure 2. Presence of VHb in YV1. (A) Western blot performed formed with vgb (8). as described in Materials and Methods. Lanes: 1, strain DNT; 2, YV1; 3, E. coli bearing pUC8:16. The arrow indicates the When growth of strains DNT and YV1 were compared polypeptide which reacted with the anti-VHb antibody. (B) CO- in TSB, the maximum growth rates monitored by cell difference spectra of intact cells of strains DNT and YV1. Cells number (Figure 3) of both strains were almost the same, were grown in TSB to early stationary phase before harvesting. but when monitored by A600 the maximum growth rate Cell concentration was 25 mg wet weight/mL in 0.01 M sodium and final A600 were both higher for YV1 (Figure 4). When phosphate buffer, pH 7.2. The cell suspension in the sample growth in MSB-DNT-YE was compared, YV1 again cuvette was bubbled with CO for 2 min. Top spectrum is YV1, middle is strain DNT, bottom (BL) is baseline. VHb is charac- achieved a higher final maximum yield whether moni- terized by a peak at 419 nm and a trough at 436 nm in such tored using cell number or A600 (data not shown). The spectra. stability of plasmid pSC160, determined by retention of resistance to both ampicillin and kanamycin, was 100% per cell; densitometry of these lanes in the agarose gel in YV1 for 72 h in both antibiotic-free TSB and MSB- yielded an estimated copy number of 15. The presence DNT-YE media. of a polypeptide of the size of VHb, which reacted with The specific respiration of both strains decreased anti-VHb antibody, was observed in YV1 and E. coli during growth in TSB with only slight differences be- transformed with vgb but not in untransformed strain tween them when compared on a per cell basis, but when DNT (Figure 2A), evidence that vgb was being expressed they were compared on an A600 basis, YV1 respired in YV1. Confirmation that VHb was produced in this roughly twice as much as strain DNT (Figure 5). When transformant was provided by CO-difference spectral washed suspensions of cells grown in DNT, cells grown analysis of intact cells: the sharp peak at 419 nm and in succinate, and cells grown in DNT plus succinate were trough at 436 nm present in the YV1 spectrum but absent transferred to fresh MSB-DNT-YE medium without in the DNT spectrum (Figure 2B) are characteristic of succinate, the DNT disappeared rapidly with little or no Biotechnol. Prog., 2000, Vol. 16, No. 1 29

Figure 5. Oxygen consumption per A600 unit of cells of strains Figure 7. Disappearance of DNT and release of nitrite into YV1 and DNT at various times in culture in TSB. Each point is the culture medium by cells grown in MSB-DNT-YE medium. the average of three experiments; error bars indicate standard Each point is the average of three experiments. O, DNT deviations. degradation by strain DNT; 9, DNT degradation by YV1; 4, nitrite release by strain DNT; X, nitrite release by YV1.

is one-third and one-twentieth of that seen in Vitreoscilla and vgb-bearing E. coli, respectively, but almost the same as other Pseudomonads that have been transformed with vgb (11). The stability of pSC160 in strain YV1 was 100% even after 72 h in nonselective medium; the estimated copy number of 15 may contribute to this stability. VHb was measured in this study in early stationary phase cells, at which time the culture would be low in oxygen; induction of vgb transcription in other bacterial species, including the native organism Vitreoscilla is known to occur under these conditions (24). Our results suggest, however, that while basal expression from the native vgb promoter occurs in YV1, as has been observed for other vgb-bearing Pseudomonads (11), this strain may not possess the machinery for induction of this gene under conditions of low oxygen. Detailed examination of Figure 6. Degradation of DNT by cells of strains DNT and YV1 the response to oxygen levels was not made in this study, pregrown in DNT and succinate. Cells grown for 72 h in MSB- so it is not yet known if VHb production can be increased DNT-YE, MSB-YE-succinate, or MSB-DNT-YE-succinate me- in this way. dium were transferred to fresh MSB-DNT-YE medium contain- The enhancement of growth due to the presence of vgb ing 200 ppm DNT. Each point is the average of three in the transformant strain YV1 relative to untransformed experiments. strain DNT was seen in a higher maximum cell mass and a slightly higher maximum viable cell number. It is accumulation of intermediate metabolites (Figure 6). In possible that this factor alone could account for the every case strain YV1 cells degraded DNT faster and increased DNT degradation by YV1 observed later in the more completely than the corresponding control strain growth cycle. The stimulations of both growth and DNT DNT cells. For both strains, growth in succinate stimu- degradation were not large, but this may be a result of lated subsequent DNT degradation more than growth in the relatively low expression of vgb in YV1. The effects DNT, and growth in succinate plus DNT was the most observed in this study parallel those observed in a stimulatory. YV1 grown in succinate plus DNT degraded previous study on the stimulation of R-amylase produc- 100% of the DNT in 3 days. As expected from the tion by VHb (6): the 3- to 4-fold increase was the result mechanism of DNT degradation nitrite appearance par- of an increase in both cell number and enzyme production alleled DNT disappearance (Figure 7). per cell and per mL of culture. A later study (10) showed that expressing production on a per cell basis can be Discussion misleading, however. Strains of Serratia marcescens Genetic engineering using vgb to express VHb in bearing vgb had about one-fifth fewer viable cells in late heterologous hosts has been used to increase the produc- log and stationary phase than control strains, but the tion of products such as alcohols, amino acids, antibiotics, cells were 5-10 times larger, so it is probably better to R-amylase, and cell protein in both bacteria and fungi express production on a cell mass (A600) basis. For DNT (5, 6, 7, 10, 22, 23) and to increase bioremediation degradation it is apparent that the approximately 15% potential (11). The work reported here is an attempt to increase in cell mass (Figure 4) alone cannot account for extend this technology to a bacterial strain (Burkholderia the nearly 2-fold increase in the rate of DNT degradation sp. strain DNT) capable of degrading DNT by transform- (Figure 6). This same figure also shows that control strain ing it with plasmid pSC160, which carries vgb. The DNT is also capable of degrading 100% of the DNT but combination of Western blot and spectral data are requires approximately twice as long. evidence that VHb is produced in YV1. The VHb level of Despite this relatively low VHb expression YV1 did about 10 nmol/g wet weight (20 nmol heme/g wet weight) show a roughly 2-fold increase in respiration relative to 30 Biotechnol. Prog., 2000, Vol. 16, No. 1 strain DNT. VHb has been reported to increase oxygen (9) Kallio, P. T.; Bailey, J. E. Intracellular expression of uptake, presumably because of its ability to trap and feed Vitreoscilla hemoglobin VHb enhances total protein secretion oxygen to the terminal oxidase (24), which may be and improves the production of R-amylase and neutral - happening with YV1, but it has also been reported to protease in Bacillus subtilis. Biotechnol. Prog. 1996, 12,31 decrease oxygen uptake, which has been observed with 39. E. coli and other bacterial species bearing vgb (6, 8, 10). (10) Wei, M. L.; Webster, D. A.; Stark, B. C. Genetic Engineer- ing of Serratia marcescens with bacterial hemoglobin gene: An explanation for the latter observation is that the Effects on growth, oxygen utilization, and cell size. Biotechnol. presence of VHb significantly increases the P/O ratio, Bioeng.. 1998, 57, 477-483. perhaps by enhancing cytochrome bo activity or by (11) Liu, S. C.; Webster, D. A.; Wei, M. L.; Stark, B. C. Genetic preventing replacement of cytochrome bo by the less engineering to contain the Vitreoscilla hemoglobin gene efficient cytochrome bd (25). The probable mechanism by enhances degradation of benzoic acid by Xanthomonas mal- which respiration would be affected in both cases would tophilia. Biotechnol. Bioeng. 1996, 49, 101-105. be by facilitated diffusion (26). This mechanism requires (12) Fernando, T.; Bumpus, J. A.; Aust, S. D. Biodegradation relatively high concentrations of hemoglobin to be effec- of TNT (2,4,6,-trinitrotoluene) by Phanerochaete chrysospo- tive. Thus, increasing the intracellular concentration of rium. Appl. Environ. Microbiol. 1990, 56, 1666-1671. VHb in DNT-degrading transformants by an order of (13) Spanggord, R. J.; Spain, J. C.; Nishino, S. F.; Mortelmans, magnitude may result in an even larger enhancement of K. E. Biodegradation of 2,4-dinitrotoluene by a Pseudomonas DNT degradation. This would be particularly beneficial sp. Appl. Environ. Microbiol. 1991, 57, 3200-3205. in the bioremediation of DNT in situ where oxygen is (14) Suen, W. C.; Spain, J. C. Cloning and characterization of limiting. Pseudomonas sp. strain DNT genes for 2,4-dinitrotoluene Nevertheless, YV1, even with its relatively low expres- degradation. J. Bacteriol. 1993, 175, 1831-1837. sion of vgb, still showed significant advantages over (15) Bagdasarian, M.; Timmis, K. N. Host:vector system for strain DNT. Under some conditions, i.e., when the cells gene cloning in Pseudomonas.InCurrent Topics in Microbi- were grown first in medium containing DNT plus succi- ology and Immunology; Hofschneider, P. H., Goebel, W., Eds.; Springer: New York, 1981; pp 47-67. nate, YV1 was capable of degrading 100% of the DNT within 3 days. Since yeast extract could replace the (16) Messing, J. New M13 vectors for cloning. Methods Enzymol. 1983, 101,20-78. succinate (data not shown) this may be simply the result of additional carbon source. The reason for this stimu- (17) Birboim, H. C.; Doly, J. A rapid alkaline extraction method for screening recombinant plasmid DNA. Nucleic Acids Res. latory effect on DNT degradation, which was also exhib- 1979, 7, 1513-1521. ited by strain DNT, will be one focus of our future work. (18) Liu, C. Y.; Webster, D. A. Spectral characteristics and Another will be attempts to increase the expression of interconversions of the reduced, oxidized, and oxygenated vgb in a DNT-degrading transformant. forms of purified cytochrome o. J. Biol. Chem. 1974, 249, 4261-4266. Acknowledgment (19) Smibert, R. M.; Krieg, N. R. General characterization. In This work was supported by NSF grants BES-9309759 Manual of Methods for General Bacteriology; Gerhardt, P., and INT-9503400 and by AFOSR grant F49620-95-1- Murray, R. G. E., Costilow, R. N., Nester, E. W., Wood, W. 0325. We thank Jim C. Spain for generously providing A., Krieg, N. R., Philips, G. B., Eds.; American Society for Microbiology: Washington, D.C., 1981; pp 409-443. strain DNT (R34) and Tom Irving for assistance with the densitometry. (20) Marmur, J. A procedure for isolation of deoxyribonucleic acid from microorganisms. J. Mol. Biol. 1961, 3, 208-218. References and Notes (21) Towbin, H. T.; Staehelin, T.; Gordon, G. Electrophoretic transfer of proteins from polyacrylamide gels to nitrocellulose (1) Costerton, J. W. F.; Murray, R. G. E.; Robinow, C. F. sheets: Procedure and some applications. Proc. Natl. Acad. Observations on the mobility and structure of Vitreoscilla. Sci. U.S.A. 1979, 76, 4350-4353. - Can. J. Microbiol. 1961, 7, 329 339. (22) Chen, W.; Hughes, D. E.; Bailey, J. E. Intracellular (2) Pringsheim, E. G.; The Vitreoscillaceae: A family of color- expression of Vitreoscilla hemoglobin alters the aerobic less, gliding, filamentous organisms. J. Gen. Microbiol. 1951, metabolism of Saccharomyces cerevisae. Biotechnol. Prog. - 5, 124 149. 1994, 10, 308-313. (3) Dikshit, K. L.; Webster, D. A. Cloning, characterization and (23) Magnolo, S. K.; Leenutaphong, D. L.; DeModena, J. A.; expression of the bacterial globin gene from Vitreoscilla in - Curtis, J. E.; Bailey, J. E.; Galazzo, J. L.; Hughes, D. E. Escherichia coli. Gene 1988, 70, 377 386. Actinorhodin production by Streptomyces coelicolor and growth (4) Khosla, C.; Bailey, J. E. The Vitreoscilla hemoglobin gene: of Streptomyces lividans are improved by the expression of a molecular cloning, nucleotide sequence, and genetic expres- bacterial hemoglobin. Biotechnology 1991, 9, 473-476. sion in Escherichia coli. Mol. Gen. Genet. 1988, 214, 158- 161. (24) Dikshit, K. L.; Spaulding, D.; Braun, A.; Webster, D. A. (5) Khosla, C.; Bailey, J. E. Heterologous expression of a Oxygen inhibition of globin gene transcription and bacterial haemoglobin synthesis in Vitreoscilla. J. Gen. Microbiol. 1989, bacterial hemoglobin improves the growth properties of - recombinant Escherichia coli. Nature 1988, 331, 633-635. 135, 2601 2609. (6) Khosravi, M.; Webster, D. A.; Stark, B. C. Presence of the (25) Kallio, P. T.; Kim, D. J.; Tsai, P. S.; Bailey, J. E. Intrac- bacterial hemoglobin gene improves R-amylase production of ellular expression of Vitreoscilla hemoglobin alters Escheri- a recombinant E. coli strain. Plasmid 1990, 24, 190-194. chia coli energy metabolism under oxygen-limited conditions. (7) DeModena, J. A.; Gutierrez, S.; Velasco, J.; Fernandez, F. Eur. J. Biochem. 1994, 219, 201-208. J.; Fachini, R. A.; Galazzo, J. L.; Hughes, D. E.; Martin, J. F. (26) Wittenberg, J. A.; Bergersen, F. J.; Appleby, C. A.; Turner, The production of cephalosporin C by Acremonium chrysoge- G. L. Facilitated oxygen diffusion. The role of leghemoglobin num is improved by the intracellular expression of a bacterial in nitrogen fixation by bacteroids isolated from soybean root hemoglobin. Biotechnology 1993, 11, 926-929. nodules. J. Biol. Chem. 1974, 249, 4057-4066. (8) Liu, S. C.; Webster, D. A.; Stark, B. C. Cloning and expression of the Vitreoscilla hemoglobin gene in Pseudomo- Accepted November 30, 1999. nas: Effects on cell growth. Appl. Microbiol. Biotechnol. 1995, 44, 419-424. BP9901421