Nocardioform arsenic resistance plasmids and construction of Rhodococcus cloning vectors

Nocardioform arsenic resistance plasmids and construction of Rhodococcus cloning vectors

PLASMID 23,242-247 ( 1990) Nocardioform Arsenic Resistance Plasmids and Construction of Rhodococcus Cloning Vectors ERIC R. DABBS, BHAVNA GOWAN, ...

512KB Sizes 5 Downloads 75 Views

PLASMID

23,242-247

( 1990)

Nocardioform

Arsenic Resistance Plasmids and Construction of Rhodococcus Cloning Vectors

ERIC R. DABBS, BHAVNA GOWAN, AND SUSANJ. ANDERSEN Department of Genetics, University of the Witwatersrand, Johannesburg, P.O. Wits, 2050, South Africa Received March 6, 1990; revised April 18, 1990 One of a number of large nocardioform plasmids previously obtained by a primarily genetic approach was reduced in size to about - 11 kb. This smaller plasmid possesseddeterminants for resistance to sodium arsenate and sodium arsenite, as well as immunity to nocardiophage Q4. It was joined to an Escherichia co&positive selection vector constructed by M. Zabeau and colleagues, which had the EcoRl endonuclease gene placed under the control of the PR promoter of X as well as a bla determinant. The resulting shuttle vector of about 14.6 kb was maintained in E. coli and in several strains of Rhodococcus. The vector was efficient in cloning DNA without prior alkaline phosphatase treatment, as a result of the presence of the positive selection function. This function was not significantly expressed in Rhodococcus, and the presence of the nocardiofonn resistance determinants led to no increase in arsenate or arsenite resistance in E. co/i. The presence of the b/a gene resulted in an increase of about threefold in ampicillin resistance in Rhodococcus strains. 0 1990 Academic PIES.. I~C.

Nocardioform bacteria of the genus Rhodococcus are notable for their ability to degrade a wide range of compounds such as phenol (Haider et al., 198 l), lignin (Rast et al., 1980) insecticides (Ferguson and Korte, 198 1), surfactants (MacDonald et al., 198 l), and acrylamide (Arai et al., 198 1). They are also able to convert cholesterol into steroids suitable as precursors to pharmacologically important substances (Ferreira et al., 1984) and some species can synthesize certain antibiotics (Wakisuki et al., 1980; Cross, 1982). Furthermore, the causative agents of tuberculosis Mycobucterium tuberculosis (Farer, 1979) leprosy M. leprue (Bullock, 1979), and nocardiosis (Gonzalez-Ochoa, 1976) are members of this group. Because of the medical and possible commercial importance of nocardioforms, cloning vectors for this group would be desirable. Several approaches for obtaining vectors have been made. Kasweck et al. (1982) described plasmids in several nocardioforms, although no phenotype could be attributed to them. Brownell et al. (1982) adopted an alternative approach, studying the nocardiophage 4EC in an attempt to develop this into a cloning vec0147-619X/90 $3.00 Copyright 0 I990 by Academic All rights of reproduction

Press, Inc. in any form reserved.

242

tor. Recently, an Escherichiu coli-Rhodococcus shuttle vector has been constructed by cloning a fragment from a cryptic Rhodococcus plasmid into pIJ30, a pBR322 derivative containing an E. coli origin of replication and an ampicillin resistance determinant together with a Streptomyces gene for thiostrepton resistance (Vogt Singer and Finnerty, 1988). As an alternative way of developing Rhodococcus cloning vectors, we previously described a primarily genetic approach used to obtain a set of resistance plasmids in Rhodococcus strain CW22 (Dabbs and Sole, 1988; Table 1). An unstable genetic element conferring resistance to sodium arsenate and arsenite (As’)’ cadmium chloride, and chloramphenico1 was identified. As’ was transduced back into a cured strain using the generalized transducing nocardiophage Q4 (Dabbs, 1987) and mutants of these transductants with increased resistance to arsenite were selected. Such mutants possessed large plasmids, migrating ’ Abbreviations used: As’, arsenate and arsenite resistance; Cd’, cadmium chloride resistance; Cm’, chloramphenicol resistance; moi, multiplicity of infection; PEG, polyethylene glycol4000.

243

SHORT COMMUNICATIONS TABLE 1 STRAINS, PHAGE, AND PLASMIDS USED IN THIS WORK Characteristics Strains (background) ATCCI 2674 (R. erythropolis) cw22 td2- 1 ATCC4277 (R. erythropolis) ATCC4277- 1 ATCC 14887 (R. equi) KDI ATCC25594 (R. rubropertinctus) A448 (R. australis) A554 (R. australis) MM294 (E. co/i)

Origin/reference N. Ferreira

Arsenic sensitive derivative of ATCC12674 CW22 carrying pDA20 Streptomycin resistant mutant of ATCC4277 Mutant of ATCC14887 unable to utilize sterols

endA thi-I hsdR17

Dabbs and Sole ( 1988) Dabbs and Sole ( 1988) N. Ferreira This work N. Ferreira This work N. Ferreira N. Ferreira N. Ferreira B. Bachmann

Phages Dabbs (1987)

Q4

Plasmids pDA20 pDA2 1 pDA30 pEcoR25 1

AKd’ AS'

As’ EcoRl bla

slower than chromosomal DNA in agarose gels. The evidence suggested (Dabbs and Sole, 1988) that these plasmids were chimeras of part of the unstable genetic element together with a portion of the phage 44 genome. The plasmids coded for As’, together with either Cd’ or Cm’. Here, we describe the development of one of these large plasmids into Rhodococcus cloning vectors. A preliminary analysis of the large plasmids indicated that they were 60-90 kb in size and had no unique restriction sites (J. Lapidos, S. Dawes, and E. Dabbs, unpublished). To reduce one of these plasmids in size, advantage was taken of three observations previously described (Dabbs and Sole, 1988): phage 44 can plaque on strain ATCC12674 and its “cured” derivative CW22. The virus can mediate generalized transduction of markers into, but not lyse, several other Rhodococcus species, including R. equi strain ATCC14887; DNA

Dabbs and Sole (1988) This work This work F. Robb

from strain ATCC 12674 and its derivatives is apparently restricted by strain ATCC14887. DNA from one large plasmid, pDA20 (Table l), was therefore introduced into strain ATCC14887 using phage 44 [lysate preparation and transduction procedures with this virus were performed as described in Dabbs ( 1987)]. The presence of Q4 genes in the plasmid suggested it might be packageable by Q4. This virus is unable to plaque on strain CW22 derivatives carrying resistance plasmids such as pDA20, presumably because that portion of the phage DNA present in the plasmid includes a Q4 immunity gene. However, at a multiplicity of infection (moi) of 10 or more, such derivatives are killed and the cells lyse. When a lysate made by infecting a pDA20containing strain with Q4 at a moi of 10 was tested for ability to transduce AS’, it was found to transduce this marker at over lo6 times the frequency which is found when 44 is mediating generalized transduction. Such results

244

SHORT COMMUNICATIONS

a BamKp”l

BglII Cla I

liiL-./

“am&-W8

EcoR I

EcoR I BqmH I

lik%b dig”:2Fl ligate die;;:;: ligate

I

L

FIG. I. (a) Partial restriction map of plasmid pDA21. (b) Schematic representation of construction of shuttle vector pDA37. (c) Partial restriction map of plasmid pDA37. In b and c, DNA of nocardioform origin is shown by a thin line, and that of E. coli origin by a heavy line.

supported an interpretation in which wild-type Q4 acted as helper in the packaging of plasmid pDA20 to produce a specialized transducing phage. A lysate, prepared as described above, of pDA20-carrying strain td2-1 (Table 1) would therefore be a mixture of residual excess wild-type Q4, together with specialized transducing phage and perhaps fresh Q4 resulting from infection and lysis of the plasmid-containing strain. When strain CW22 was recipient and a lysate of strain td2- 1 was used as a donor of As’, plasmid screens of As’ transductants revealed that all had acquired an additional band migrating similarly to pDA20 DNA in agarose gels. As expected, the lysates also transduced Cd’, the other resistance determinant on this plasmid, at high frequency. However, with re-

spect to markers not on pDA20, the lysate mediated transduction at the much lower frequency seen in QCmediated generalized transduction. A 44 lysate of strain td2-1 was used to transduce As’ into strain ATCC14887 (Fig. lb). Transduction frequency was about 0.5% of that observed when strain CW22 was recipient. The DNA of 20 transductants was analyzed on agarose gels. Eight possessed a band migrating at a rate similar to that of pDA20. The remaining 12 had each acquired a plasmid of different size, but in every case smaller than that of pDA20. This variety of plasmid sizes was consistent with strain ATCC14887 restricting DNA derived from an ATCC12674 background, as previous evidence (Dabbs and Sole, 1988) suggested. The smallest plasmid

SHORT

COMMUNICATIONS

obtained from this selection was pDA21, which was purified by cesium chloride gradient centrifugation and was shown to transform strain CW22 to As’. Transformations into this strain were done based on the method developed for Streptomyces (Bibb et al., 1978). Strain CW22 was grown at 25°C in Luria broth supplemented with 5% glycine. Five to 10 aliquots were removed every 12 h and frozen. One aliquot of cells from each sampling was thawed, washed in buffer P (Okanishi et al., 1974; Hopwood and Wright, 1978), and resuspended in the same buffer containing 5 mg/ml lysozyme. After 60 min of incubation at 37°C cells were washed in buffer P and resuspended in this buffer at twice the original concentration. Up to 20 ~1 of DNA was added to 60-~1 aliquots of cell suspension. A volume of buffer containing 50% polyethylene glycol 4000 (PEG) equal to the total of the protoplast DNA mixture was gently mixed in. After standing 3-5 min at room temperature, the protoplasts were very gently spread on chilled regeneration plates. These plates were made by dissolving 35 g sucrose in 280 ml of water. Agar (4.5 g), tryptone (3 g), yeast extract (1.5 g), and sodium chloride (0.9 g) were added. Ten milliliters of 0.25 M iV-tris(hydroxymethyl)methyl-2-aminoethanesulfonic acid, pH 7.2, and 6 ml of 1 M calcium chloride were added to the medium once it had been autoclaved. After protoplasts were spread, plates were incubated at 25°C for 8-12 h then 0.5 ml of a 3 M sodium arsenate, 0.5 M sodium arsenite solution was introduced underneath the agar using a sterile spatula. Since the volume of medium in a plate was 25 ml, the concentrations after diffusion were 60 mM arsenate, 10 IIIM arsenite. This combination of concentrations was chosen because they were the minimum ones at which no spontaneous mutants grew up on plates spread with protoplasts alone (for maintenance of plasmids, sodium arsenate alone was added to media to a final concentration of 40 mM). Plates were incubated for 10 days at 25 “C. Frozen cells from the growth stage which showed highest transformation

245

frequency were used in subsequent experiments. A restriction map was constructed of pDA21 (Fig. la), which indicated that the plasmid was about 15.3 kb. In contrast to pDA20, the Cd’ determinant was no longer present. However, this plasmid still possessed the Q4 immunity function. Purified plasmid pDA2 1 was used to attempt to transform protoplasts of a number of bacterial strains to As’ so as to obtain an idea of how wide a range of species this replicon could be maintained in. In addition to the strains listed below, R. australis strain A448 and R. rubropertinctus strain ATCC25593 were also investigated. However, they proved to already possess an As’ determinant and hence were unsuitable as recipients. Of four nocardioforms tested (Table l), three were successfully transformed with the plasmid. With strain ATCC4277, 2.1 X lo4 transformants//lg DNA were obtained; with strain CW22, 1.5 X lo4 transformants/pg DNA were obtained; with strain KD 1,5 X lo3 transformants/clg DNA were obtained. (These figures were for plasmid prepared from the corresponding strain; when DNA from another strain was used, transformation efficiency was roughly lo-fold lower). For strain A554, no successful transformation of this plasmid could be demonstrated. For every strain, several parameters were varied. A range of glycine concentrations was used during growth. Cells from various stages of growth were tested. The duration of lysozyme treatment, the PEG concentration, and the incubation time of plates before underlay were all investigated. For highest transformation efficiency (given above), conditions were generally similar for all strains. The exception was plate incubation time before addition of arsenate and arsenite. This was 1216 h for strain ATCC4277 and 18-24 h for strain KD 1, paralleling the slower growth rate of these strains. Plasmid pDA2 1 was reduced in size (Fig. lb). It was digested with any of a number of restriction endonucleases. In each case, after ligation the DNA was transformed into strain CW22 and selection made for ar-

246

SHORT

COMMUNICATIONS

senic resistance. Such experiments showed the two smaller EcoRI fragments were dispensable, as was the smallest BamHl fragment. The resulting plasmid, pDA30, was about 11.2 kb in size. Attempts to further reduce size led either to reduction in As’ or to narrowing of range of strains in which the plasmid could be maintained to only CW22. The only unique restriction site in pDA2 1 was for BglII. Removal of the smallest Kpnl fragment eliminated this site without interfering with plasmid properties. Heterologous DNA digested with BglII, BumHl, or Bell could be cloned into the BglII site of pDA30 after the plasmid had been treated with alkaline phosphatase to prevent recircularization. To make more effective Rhodococcus vectors, pDA30 was joined to the E. coli-positive selection (suicide) vector pEcoR251 via this BglII site. Plasmid pEcoR25 1 was constructed by Zabeau and colleagues based on earlier work of Cheng and Modrich (1983). It is about 3.3 kb in size. They made various E. coli plasmids in which genes are placed under the control of the Pa promoter of phage X (Zabeau and Stanley, 1982). In the case of pEcoR25 1, the EcoR 1 endonuclease gene is placed under the control of the PR promoter. The plasmid may therefore be maintained in a X lysogen using the bla gene for ampicillin resistance. Transformation into a nonlysogen results in expression of the endonuclease, digestion of the DNA, and death of the cell. If DNA is cloned into the unique Hind111 or BglII sites in the EcoRl gene, then endonuclease activity is abolished and transformation no longer leads to death of the transformed cell. Plasmid pEcoR25 1 has a unique BamHl site. Nocardioform plasmid pDA30 was cut with BglII and ligated to the E. coli plasmid cut with BamH 1. DNA was transformed into E. coli strain MM294 (Table 1) by standard methods (Maniatis et al., 1982) selecting on Luria broth plates containing 100 pg/ml ampicillin. Constructs with both possible orientations, designated pDA37 and pDA38, were shown to be unimpaired in properties compared with pEcoR25 1. Plasmids with DNA

cloned into the endonuclease gene of these constructs to inactivate it were purified from E. coli and transformed into strains CW22 and ATCC4277-1 where they were stably maintained. Such plasmids were purified in turn from the nocardioform strains and could be transformed back in E. coli. pDA37 and pDA38 are therefore shuttle vectors. Shuttle vector with endonuclease activity abolished showed a nocardioform transformation efficiency when purified from E. coli of about 1% of that seen when this DNA was obtained from the same strain of Rhodococcus, reflecting restriction by the nocardioform. Vector pDA37 itself was subsequently shown to transform nocardioforms at a similar efficiency to that seen with derivatives in which endonuclease activity was abolished. DNA of the vector purified from nocardioforms was shown by transformation back into h lysogenic and nonlysogenic E. coli strains to still have the suicide function unimpaired. This suggested that the PR promoter was not significantly expressed in these Rhodococcus strains. The presence of the blu gene resulted in a roughly 3-fold increase in ampicillin resistance of Rhodococcus strains, versus 500-fold in E. coli. The presence of the plasmid-borne As’ determinant led to no detectable increase in resistance to either arsenate or arsenite in E. coli. Plasmid pDA37 digested with BglII, but not subjected to alkaline phosphatase treatment, was ligated with appropriately digested chromosomal DNA and transformed into strain MM294. In a series of experiments, 60-75% of transformants had a detectable insert, with average insert size of 4.3 kb. REFERENCES ARAI, T., KURODA, S., AND WATANABE, I. (1981). Biodegradation of acrylamide monomer by a Rhodococcus strain. In “Actinomycetes 261 Bakt Suppl II” (K. P. Schaal and G. Pulverer, Eds.), pp. 297-308. Gustav Fischer, Stuttgart. BIBB, M. J., WARD, J. M., AND HOPWOOD, D. A. (1978). Transformation of plaamid DNA into Sfreptomyces at high frequency. Nature (London) 214398-400. BROWNELL, G. H., SABA, J. A., DENNISTON, K., AND ENQLJIST,L. W. (1982). Development of a Rhodococcus

247

SHORT COMMUNICATIONS actinophage gene cloning system. Dev. Ind. Microbial. 23,287-298.

BULLOCK, W. E. (1979). Mycobacterium leprae. In “Infectious Diseases.” (G. L. Mandell, R. G. Douglas, and J. E. Bennett, Eds.), pp. 1943-1953. Wiley, New York. CHENG, S. C., AND MODRICH, P. (1983). Positive-selection cloning vehicle useful for overproduction of hybrid proteins. J Bacterial. 154, 1005-1008. CROSS,T. (1982). Actinomycetes: A continuing source of new metabolites. Dev. Ind. Microbial. 23, l-18. DABBS, E. R. (1987). A general&d transducing phage of Rhodococcus erythropolis. Mol. Gen. Genet. 206, 116120. DABBS, E. R., AND SOLE, G. J. (1988). Plasmid-borne resistance to arsenate, arsenite, cadmium, and chloramphenicol in a Rhodococcus species. Mol. Gen. Genet. 211, 148-154. FARER, L. S. (1979). Mycobacterium tuberculosis. In “Infectious Diseases.” (G. L. Mandell, R. G. Douglas, and J. E. Bennett, Eds.), pp. 1905-1924. Wiley, New York. FERGUSON,J. A., AND KORTE, F. (198 1). Transformation of aldrin by soil microorganisms. Appl. Environ. Microbiol. 34, 7- 15. FERREIRA,N. P., ROBSON, P. M., BULL, J. R., AND VAN DER WALT, W. H. (1984). The microbial production of a 3aa-H-4or-(3’-propionic acid)&hydroxy-7&methylhexahydro-indan-1-one-&lactone from cholesterol. Biotechnol. Lett. 5 17-522. GONZALEZ-OCHOA, A. (1976). Nocardiae and chemotherapy. In “The Biology ofthe Nocardiae.” (M. Goodfellow, G. M. Brownell, and J. A. Serrano, Eds.), pp. 429-450. Academic Press, London. HAIDER, K., JUGNOW, G., KOHNEN, R., AND LIM, S. V. ( 198 1). Degradation of chlorinated benzenes, phenols, and cyclohexane derivatives by soil bacteria under aerobic conditions. In “Decomposition of Toxic and Nontoxic Organic Compounds in Soil.” pp. 207-226. Ann Arbor Science, Ann Arbor.

HOPWOOD, D. A., AND WRIGHT, H. M. (1978). Bacterial protoplast fusion: Recombination in fused protoplasts of Streptomyces coelicolor. Mol. Gen. Genet. 162, 307317.

KASWECK, K. L., LITTLE, M. L., AND BRADLEY, S. G. (1982). Plasmids in mating strains of Nocardia asteroides. Dev. Ind. Microbial.

23,279-286.

MACDONALD, C. R., COOPER,D. G., AND ZAJIC,J. ( 198 1). Surface active lipids from Nocardia erythropolis grown on hydrocarbons. Appl. Environ. Microbial. 41, 117123. MANIATIS, T., FRITSCH,E. F., AND SAMBROOK, J. (1982). Transformation by calcium chloride procedure. In “Molecular Cloning: A Laboratory Manual.” pp. 25025 1. Cold Spring Harbor Laboratory, New York. OKANISHI,

M.,

SUZUKI,

K., AND UMEZAWA,

H. (1974).

Formation and reversion of streptomycete protoplasts: Cultural conditions and morphological study. J. Gen. Microbial.

80, 389-400.

RAST, H. G., ENGELHARDT, G., ZIEGLER, W., AND WALLN~FER, P. R. (1980). Bacterial degradation of model compounds for lignin and chlorophenol derived lignin bound residues. FEMS Microbial. Lett. 8, 259263. VOGT SINGER,

M. E., AND FINNERTY, W. R. (1988). Construction of an Escherichia coli-Rhodococcus shuttle vector and plasmid transformation in Rhodococcus species. J. Bacterial. 170,638-645. WAKISUKI, Y., KOIZUMA, K., NISHIMOTO, Y., KOBAYASHI, M., AND TSUJI, N. (1980). Hygromycin and epihygromycin from a bacterium Corynebacterium equi. J. Antibiot.

33,695-704.

ZABEAU, M., AND STANLEY, K. K. (1982). Enhanced expression of cm-@ galactosidase fusion proteins under the control of the Pa promoter of bacteriophage. EMBO J. 1, 1217-1244. Communicated

by David

A. Hopwood