Hydrocarbon Utilization by Pseudomonas putida R5-3

4 downloads 0 Views 2MB Size Report
Oct 21, 1988 - spectively, were obtained from S. Haramaya and K. Tim- mis, University of ...... Rigby, P. W. J., M. Dieckmann, C. Rhodes, and P. Berg. 1977.
Vol. 55. No. 6

APPLIED AND ENVIRONMENTAL MICROBIOLOGY, June 1989, p. 1523-1530 0099-2240/89/061523-08$02.00/0 Copyright © 1989, American Society for Microbiology

Novel Alterations in Plasmid DNA Associated with Aromatic Hydrocarbon Utilization by Pseudomonas putida R5-3 BRIAN F. CARNEYt Department of Plant Pathology, University

J. V. LEARY* California, Riverside, California 92521

AND of

Received 21 October 1988/Accepted 6 March 1989

Subcultures of Pseudomonas putida R5-3 altered their plasmid DNA content in specific ways depending on the particular aromatic hydrocarbon utilized as the sole carbon source. Two indigenous plasmids, 115 and 95 kilobases (kb) in size, were observed in R5-3A, which was derived from R5-3 by growth on minimal medium containing p-methylbenzoate as the sole carbon source. When R5-3A was transferred to medium containing m-xylene or toluene, derivative strains were obtained in which the 95-kb plasmid was lost and a new plasmid of 50 or 60 kb appeared. Reversion to the original plasmid profile of R5-3A was observed when xylene- or toluene-grown cells were returned to medium containing p-methylbenzoate. Restriction enzyme analysis and Southern blot hybridizations of total plasmid DNA indicated deletions and rearrangements of DNA restriction fragments in the derivatives maintained on m-xylene and toluene when compared with the original R5-3A. In the derivatives which retrieved the original plasmid profile, the restriction enzyme fragment pattern was identical to that in the original R5-3A, in that the fragments which were missing after growth on m-xylene or toluene were again present. Southern blot hybridizations revealed that part of the plasmid DNA lost from the original plasmid profile was integrated into the chromosomal DNA of xylene-grown R5-3B and that these plasmid fragments were associated with aromatic hydrocarbon metabolism. Hybridization with pathwayspecific DNA fragments from the TOL plasmid pWWO indicated that this 95-kb plasmid contains DNA homologous to the meta-fission pathway genes.

sludge of a P. piitida strain, R5-3, which can grow on benzoate, toluene, xylenes, and methylbenzoates as the sole carbon source (B. F. Carney, L. Krockel, D. D. Focht, and J. V. Leary, Abstr. Annu. Meet. Am. Soc. Microbiol. 1987, K-180, p. 232). Plasmid analysis of the strain derived from R5-3 after growth on p-methylbenzoate (p-MB), R5-3A, revealed the presence of two large plasmids of 115 and 95 kilobases (kb) in size. Growing R5-3A in different aromatic substrates revealed alterations in plasmid content without concomitant loss in the Tol+ phenotype. This work examines the nature of these alterations. In this report we describe unique biological features of the plasmids and their association with chromosomal DNA in the maintenance of the Tol+ phenotype in P. piutida R5-3.

The ability of bacteria and fungi to utilize an array of hydrocarbons as growth substrates has been widely documented (1, 5). Pseiudoinonias piutida is well known for its ability to utilize aromatic hydrocarbons such as toluene, xylenes, and methylbenzoates as the sole carbon source (5). The genetic information for such degradative pathways has in general been shown to be plasmid encoded (2, 18, 20), with one report of these genes being chromosomal (15). The biodegradation of aromatic hydrocarbons occurs by two pathways; the chromosomally encoded ortho-fission pathway occurs when the substrates are benzoates or phenols and involves enzymes of high specificity, whereas the TOL plasmid-encoded ,neta-fission pathway occurs when methylated benzoates, xylenes, toluene, and benzene are substrates and involves enzymes of low specificity (18, 20). Continued growth on benzoate frequently results in loss of the ,neta-fission pathway owing to curing of the TOL plasmid (8). The degradation of toluene and xylene by P. pitida has been shown to be due to an inducible enzyme pathway specified by the archetypal TOL plasmid pWWO (18). Molecular and functional analysis of this plasmid has shown that all the genes coding for the degradative enzymes are located in clusters and contain regulatory elements corresponding to operons (2, 3, 7). These plasmids are autonomous and self-replicating, with little evidence for plasmid integration into the chromosome. The origin of such catabolic plasmids has often been described as analogous to that of antibiotic resistance plasmids. We have previously reported the isolation from sewage *

MATERIALS AND METHODS

Bacterial strains. P. pitida R5-3 was isolated from activated sewage sludge by enrichment culture with toluene as the sole carbon source. This isolate had the ability to grow on a range of aromatic hydrocarbons and has been described previously. The derivatives of R5-3 (listed in Table 1) were obtained by first transferring R5-3 to minimal medium containing p-MB as the sole carbon source (R5-3A). R5-3A was then transferred to minimal medium containing ,n-xylene (mn-Xyl), toluene, or benzoate to examine alterations in plasmid profiles. Pure cultures of R5-3A derivatives were obtained by transferring single colonies of R5-3A from p-MB agar medium to minimal medium agar plates containing either itn-Xyl, toluene, or benzoate to produce single colonies, which were then picked and restreaked twice on the appropriate selective substrate to obtain a final pure culture. The derivatives were obtained by subculturing several times on the selective substrate. The subcultures on the different carbon sources were examined weekly for plasmid content.

Corresponding author.

t Present address: Department of Plant Pathology. University of Georgia, Athens, GA 30602.

1523

1524

Am- ENVIRON. MICROBIOL.

CARNEY AND LEARY

TABLE 1. Source and phenotype of P. pittida R5-3 derivatives Selective substrate

Derivative

R5-3A

R5-3

p-MB

R5-3B

R5-3A

m-Xyl

RS-3C

R5-3A

Toluene

R5-3D R5-3E R5-3F

R5-3C

RS-3C RS-3C

Toluene Toluene p-MB

RS-3G

RS-3B

p-MB

R5-3H

RS-3A

Benzoate

Plasmid

Sole carbon source utilized

(kb)

p-MB, mn-MB, p-Xyl, mi-Xyl, toluene, benzoate p-MB, in-MB, p-Xyl, im-Xyl, toluene, benzoate p-MB, in-MB, p-Xyl, in-Xyl, toluene, benzoate Toluene, benzoate Toluene, benzoate p-MB, m-MB, p-Xyl,

in-Xyl, toluene, benzoate p-MB, m-MB, p-Xyl, m-Xyl, toluene, benzoate p-MB, in-MB, p-Xyl, m-Xyl, toluene, benzoate

115, 95

115, 60, 50 115, 60, 50 50

115, 60 115, 95 115, 95 115, 95

Derivatives, once established, were maintained, stored, and routinely streaked on minimal medium agar plates with the appropriate aromatic hydrocarbon as the sole carbon source. To determine the phenotypes of P. plitida R5-3 derivatives, 100 ,ul of an overnight late-log-phase 50-ml liquid culture in medium containing the appropriate aromatic hydrocarbon was plated on minimal medium agar plates containing p-MB, m-MB, p-Xyl, m-Xyl, toluene, or benzoate as the sole carbon source and scored for growth after 48 to 96 h.

The P. putida strain containing the well-characterized Tol plasmid pWWO was provided by D. Ohman, University of California, Berkeley. This strain was maintained on minimal medium with p-MB as the growth substrate and supplemented with 20 mg of methionine per ml. Escherichia coli LE392 and ED2196 containing pED3306 and pPL392, respectively, were obtained from S. Haramaya and K. Timmis, University of Geneva, Switzerland. These plasmid constructs consisted of pBR322 with cloned fragments of the TOL plasmid that code for enzymes involved in the upper pathway (pED3306) and meta-pathway (pPL392) of aromatic hydrocarbon metabolism (6, 13). For plasmid isolation, these strains were grown in Luria broth (12) containing ampicillin

(50 mg/liter). Chemicals. Toluene was obtained from Mallinckrodt, Inc., St. Louis, Mo. Sodium benzoate, m-MB, p-MB, m-Xyl, and p-Xyl were purchased from Aldrich Chemical Co., Inc., Milwaukee, Wis. Media and culture conditions. The minimal medium used consisted of phosphate buffer (10 mM KH2PO4, 10 mM Na2HPO4 [pH 7.21), 0.5 g of (NH4)2SO4, and 0.2 g of

MgSO4. 7H20 per liter of deionized water and

was

supple-

mented with 10 ml of trace-element stock solution per liter; this solution contained 530 mg of CaCl,, 200 mg of FeSO4- 7H20, 40 mg of CuS04 * 5HO, 18.5 mg of MnSO4 4H20, 20 mg of ZnSO4 7H20, 3 mg of H3BO, 4 mg of CoCl,, and 4 mg of NaMoO4 .2H20 per liter of deionized water. The stock solution of trace elements was acidified with 1 ml of concentrated sulfuric acid per liter.

Nonvolatile substrates were added at a final concentration of (wt/vol) prior to autoclaving. Toluene, p-Xyl, and

0.1%

in-Xyl were supplied to agar plates as saturated atmospheres in desiccators or were added directly to liquid cultures at a final concentration of 0.1% (vol/vol). Agar plates were incubated at 27°C or in desiccators at ambient temperature. Liquid cultures were incubated in 2.5-liter Fernbach flasks and agitated at 240 rpm at 27°C to provide suitable aeration. Analysis of 2,3-dioxygenase activity. One-liter cultures of R5-3A and the variants R5-3B and R5-3D were grown for 48 h in minimal medium containing p-MB, m-Xyl, and toluene, respectively, as the sole carbon source. Crude cell extracts were prepared by centrifugation at 20,000 rpm (48,000 x g) for 20 min and washing in phosphate buffer followed by centrifugation. The cell pellet was suspended in phosphate buffer containing 10% (vol/vol) acetone and disrupted by two passes through a French press. The cell debris were removed by centrifugation at 20,000 rpm (48,000 x g) for 20 min, and the 2,3-dioxygenase activity was assayed by the method of Gibson (4). The activity was monitored as an increase in A375 by using a spectrophotometer (Uvikon; Kontron Instruments Inc., Everett, Mass.) in the kinetics mode so that measurements were taken every 3 s for 1 min. DNA isolation. Plasmid DNA of E. coli LE392 and ED2196 was isolated by the alkaline lysis method as described by Maniatis et al. (12). P. putida plasmid DNA was isolated by a sodium dodecyl sulfate lysis procedure (12) with the following modifications. An overnight late-log-phase 500-ml culture (optical density at 600 nm, 0.8 to 1.0) was harvested and washed in 10 mM Tris-0.1 mM EDTA buffer (pH 8.0) (TE) (12) prior to lysis. Incubation times for lysozyme and sodium dodecyl sulfate steps were increased by 5 and 20 min, respectively, at room temperature. Cell debris were removed by centrifugation, and DNA was recovered (12) and dissolved in 3 ml of TE. Plasmid DNA was then purified by ultracentrifugation on cesium chloride-ethidium bromide gradients (12). The number of plasmids in each of the strains was determined by agarose gel electrophoresis (0.7% [wt/ vol] agarose) for 75 min at 100 V (10 V/cm) in Trisacetate-EDTA (TAE) buffer (12). DNA was visualized by staining gels in 1 pLg of ethidium bromide per ml and then illuminating them with a transilluminator (model 3-300; Fotodyne Inc., New Berlin, Wis.). Plasmid sizes were determined by comparison with known plasmid size standards. Chromosomal DNA was isolated by the procedure of Staskawicz et al. (17) and subsequently purified by cesium chloride-ethidium bromide ultracentrifugation (12), except when otherwise stated. Restriction enzyme analysis. Restriction enzymes were purchased from Pharmacia, Inc., Piscataway, N.J., and Bethesda Research Laboratories, Gaithersburg, Md., and were used as specified by the manufacturers, except that a digestion time of 2 h was routinely used. Digested plasmid DNA fragments were electrophoresed on a 1.2% (wt/vol) agarose gel for 2 h at 50 mA (5 mA/cm) in TAE buffer and visualized as described above. Recovery of DNA from agarose gels. Digested plasmid DNAs were electrophoresed in 1% low-melting-point agarose (FMC Corp., Rockland, Maine) gels, briefly stained with ethidium bromide (1 pLg/ml), and visualized by UV light. Gel pieces containing the desired DNA fragments were excised and separated from the agarose by using a Geneclean Kit (Bio 101 Inc., La Jolla, Calif.) as specified by the manufacturer. The eluted DNA was further purified by phenol extraction, passage through a Sephadex G50-80 spun column (12), and precipitation with absolute ethanol. The recovered DNA was then nick translated and used in Southern blot hybridizations.

VOL. 55, 1989

4_.*M0,y-';L>Xs^fv.0w-,

PLASMID DNA ALTERATIONS IN P. PUTIDA

1

2

3

4

5

6

7

I 1 15K

95K

6OKb.,,-50K

FIG. 1. Agarose gel electrophoresis of purified plasmid DNA of P. putida R5-3A and its derivatives after selective growth on aromatic hydrocarbons. Lanes: 1, R5-3A grown on p-MB; 2, R5-3B grown on m-Xyl; 3, R5-3C grown on toluene; 4 and 5, R5-3D and R5-3E, respectively, after prolonged growth on toluene; 6, R5-3G reisolated on p-MB after transfer from m-Xyl; 7, R5-3H after prolonged growth on benzoate.

Nucleic acid hybridization. DNA fragments were labeled in vitro by nick translation (14) with [32P]dCTP and hybridized to electrophoretically fractionated DNA transferred to nylon membrane filters (16). DNA-DNA hybridization, washes, and autoradiography were performed as described by Maniatis et al. (12). All hybridizations were carried out under high-stringency conditions. RESULTS AND DISCUSSION Phenotype and plasmid DNA of R5-3 derivatives. The phenotypes of P. putida R5-3 derivatives were determined by their ability to grow on a range of aromatic hydrocarbons as the sole carbon source and are summarized in Table 1. P. putida R5-3A was able to utilize p-MB, m-MB, p-Xyl, m-Xyl, toluene, and benzoate as sole carbon sources, thus showing similarity to P. putida strains carrying the TOL

1525

plasmid pWWO (19). Plasmid DNA was observed in all the R5-3 derivatives when the gentle sodium dodecyl sulfate lysis procedure was used (12). Derivative R5-3A (R5-3, p-MB grown) contained two large plasmids of 115 and 95 kb (Fig. 1, lane 1). When R5-3A was transferred to minimal medium containing m-Xyl (derivative R5-3B) or toluene (derivative R5-3C), the 95-kb plasmid was absent and two new plasmids, of 60 and 50 kb (Fig. 1, lanes 2 and 3, respectively), were observed. This alteration in plasmid profile associated with the substrate utilized was observed within 1 week after obtaining pure cultures by the procedures described above. There was no loss in phenotype in the derivatives R5-3B or R5-3C (Table 1). Growth of R5-3C on toluene with routine subculturing every 3 weeks resulted in two derivatives, R5-3D (11) and R5-3E, at 3 and 6 months, respectively, which could grow only on toluene and benzoate (Table 1). These toluenegrown derivatives contained either only one smaller plasmid, of 50 kb (R5-3D), or two plasmids, of 115 and 60 kb (R5-3E) (Fig. 1, lanes 4 and 5, respectively). The loss of the 115- and the 95-kb plasmids in R5-3D, along with the change in phenotype, suggested that these larger plasmids are involved in the meta-fission pathway metabolism of xylenes and methylbenzoates. Catechol oxygenase activity was confirmed by the catechol spray test (19) for all R5-3 derivatives except R5-3D and R5-3E, indicating that these derivatives had lost one of the main enzyme classes involved in aromatic hydrocarbon metabolism by the meta-fission pathway. The more sensitive enzyme activity assay revealed that R5-3B (m-Xyl grown) retained approximately 50% of the specific 2,3-dioxygenase activity, whereas R5-3D (toluene grown) had only 4% of the activity of R5-3A (p-MB grown). Continued growth of R5-3B on m-Xyl maintained the altered plasmid profile, and no loss in phenotype was observed in this derivative. Attempts to cure the plasmid DNA by transferring R5-3A to benzoate (R5-3H) proved unsuccessful, with no alteration in plasmid profile (Fig. 1, lane 7) or loss in phenotype after 3 months on benzoate with routine subculturing every 3 weeks (Table 1). In addition, it was not X, M

hi

t.

f::

;000:

C

f:

D; A:

00.

f--:

FIG. 2. Line drawing (A) and agarose gel electrophoresis (B) of BamHI, Hindlll, and XhoI restriction enzyme digests of plasmid DNA from R5-3A (lane 1) and pWWO (lane 2). Restriction enzyme fragments present in pWWO but absent in R5-3A plasmid DNA are indicated by arrowheads in panel A. Different line thicknesses are relative to the amount of DNA visualized on ethidium bromide-stained agarose electrophoresis gels. (C) Autoradiogram of a Southern blot hybridization of the DNA shown in panel B, with 32P-labeled pWWO.

1526

APPL. ENVIRON. MICROBIOL.

CARNEY AND LEARY

TABLE 2. Southern blot hybridization of pED3306 and pPL392 to intact plasmid DNA of R5-3 and pWWO

pWWO P A(15 kb)'

pED3306 (upper pathway) pPL392 (meta-pathway) "

+ +

P. puitida R5-3 plasmid DNA" 115 kb 95 kb 60 kb 50 kb + -

+ +

+

+

-

-

+, Homology; -. no homology.

possible to cure the large plasmids of R5-3A by growth in rich medium at elevated incubation temperatures with or without the presence of ethidium bromide. When R5-3B (m-Xyl grown) and R5-3C (toluene grown), which contained three plasmids, of 115, 60, and 50 kb, were transferred back to p-MB, the original R5-3A profile of two plasmids (115 and 95 kb) was observed (Fig. 1, lane 6). These derivatives, which recovered the original plasmid profiles (R5-3F and R5-3G [Table 1]), did not exhibit a change in substrate utilization. It was not possible to examine the recovery of the original plasmid profile in R5-3D or R5-3E, since these derivatives could no longer grow on p-MB as the sole carbon source (Table 1). Comparison of R5-3A plasmid DNA with pWWO. As stated above, P. putida R5-3A has similar phenotype to other P. putida strains carrying a TOL plasmid. However, R5-3A has two large plasmids as opposed to the single large plasmid pWWO in the archetypal P. putida TOL strains. Agarose gel electrophoresis of intact plasmid DNA revealed that pWWO is similar in size to the larger (115-kb) plasmid of R5-3A. Restriction enzyme analysis with XhoI, BamHI, and Hindlll was carried out to compare pWWO with the plasmids of R5-3A. If the 115-kb plasmids of both strains were identical, all restriction fragments of the known TOL plasmid should be present in R5-3A plasmid DNA. However, depending on the restriction endonuclease used, various fragments present in pWWO were not observed in the plasmid DNA of R5-3A (Fig. 2A and B). Specifically, 1 XhoI fragment (12 kb), 4 HindlIl fragments, and 10 BamHI fragments (Fig. 2A, arrows) were absent. Thus, the 115-kb plasmid from R5-3A is not identical to pWWO. It is important to note, however, that plasmid restriction fragments containing regions that code for enzymes involved in aromatic metabolism in

A ')

pWWO (3) are present in the plasmid DNA of R5-3A. Fragments known to encode enzymes involved in the metapathway were identified in the 95-kb plasmid only, thus indicating further variation between the two 115-kb plasmids (Table 2). Southern blot hybridizations with pWWO as probe DNA to different restriction enzyme digests of pWWO and plasmid DNA of R5-3A revealed many fragments of R5-3A plasmid DNA which are not homologous to pWWO (Fig. 2C). In the reciprocal hybridization in which total R5-3A plasmid was used as the probe, all regions of pWWO were hybridized (data not shown). In addition, when a 2-kb BamHI fragment from R5-3A plasmid DNA not observed in pWWO digests was hybridized to pWWO, there was no homology, thus identifying a particular nonhomologous region present in R5-3A. Present evidence suggests that this 2-kb BamHI fragment in R5-3A plasmid DNA may contain an origin of replication not previously identified in TOL plasmids (J. V. Leary et al., manuscript in preparation). Overall, these differences in restriction endonuclease sites and the presence of nonhomologous DNA regions indicate that the R5-3A plasmid and pWWO are similar but not identical. Relationship of plasmids from R5-3 derivatives. Southern blot hybridizations were performed to determine the relationship of the 115- and 95-kb plasmids of R5-3A to the smaller plasmids in the R5-3A derivatives. Sequence similarity was observed between plasmids of all sizes (115, 95, 60, and 50 kb) when either the larger (115-kb) plasmid of R5-3A or the smaller (50-kb) plasmid of R5-3D was used as probe DNA (data not shown). In addition, restriction endonuclease digests of total plasmid DNA from all R5-3 derivatives contained several common fragments as well as unique fragments specific to the derivative phenotypes. Three EcoRI-BglII double-digest fragments (4 and 12 kb from R5-3A and R5-3B; 15 kb unique to R5-3A) were used as probe DNA and hybridized to EcoRI-BgII digests of R5-3A, R5-3B, R5-3C, and R5-3D plasmid DNA. When the 4-kb fragment from R5-3B was used as the probe, hybridization with a homologous 4-kb restriction fragment in each of the derivatives was observed (Fig. 3B). The unique 15-kb fragment from R5-3A showed homology to itself and several smaller fragments in R5-3B, R5-3C, and

CD 2 13 41 2

8 1

4

2

1

A

3

4 :; ::

1

-~~ _

X30

04

L

w;;

t S

.

*f;

*f

D

~0

_

FIG. 3. (A) Agarose gel electrophoresis of EcoRI-Bg/II-digested plasmid DNA. Lanes: 1, R5-3A; 2, R5-3B; 3, R5-3C; 4, R5-3D on an ethidium bromide-stained gel. (B) Autoradiograms of Southern blot hybridization with the lower 4-kb EcoRI-BgIl fragment. (C) Autoradiogram of Southern blot hybridization with the 15-kb EcoRl-BgII fragment. (D) Autoradiogram of Southern blot hybridization with the 12-kb EcoRI-BglII fragment. Restriction endonuclease plasmid fragment rearrangements are indicated by nf in R5-3B, R5-3C, and R5-3D in panel A when compared with R5-3A.

VOL. 55, 1989

PLASMID DNA ALTERATIONS IN P. PUTIDA

A s 1 2 3 A 5

B s

1

2

c Aa

I

2 3

4

5

S

FIG. 4. Plasmid DNA from R5-3 derivatives digested with the restriction enzymes XhoI (A), HpaI (B), and BamHI (C). Lanes: 1, R5-3A; 2, R5-3B; 3, R5-3D; 4, R5-3E; 5, R5-3G; S, lambda/Hindlll size standard.

R5-3D (Fig. 3C), suggesting that rearrangement of the 15-kb fragment to form homologous sequences in the derived plasmids of the m-Xyl- and toluene-grown derivatives had occurred. Plasmid DNA rearrangements in these derived plasmids are further supported by the appearance of novel fragments in the digests of R5-3B, R5-3C, and R5-3D when compared with R5-3A (Fig. 3A). In the hybridization with the 12-kb fragment, homology was observed to itself, a conserved restriction fragment from each of the derivatives, and several smaller fragments in R5-3B, R5-3C, and R5-3D (Fig. 3D). This result suggests the presence of repeated DNA sequences in the plasmid DNA of the R5-3A derivatives, possibly owing to multiple rearrangements during plasmid derivation. Restriction endonuclease analysis of R5-3 derivatives. Analysis of R5-3 derivatives by using restriction endonucleases revealed several characteristics of their plasmid DNA. In all cases, relatedness between the plasmid DNAs was indicated by the presence of common restriction enzyme fragments when digested with XhoI, HpaI, and BamHI (Fig. 4). Additional fragments not present in the xylene or toluene derivatives were observed in the plasmid DNA of R5-3A. Comparison of R5-3A plasmid DNA with plasmid DNA from the strains derived after prolonged growth on toluene (R5-3D and R5-3E) consistently revealed loss of several restriction fragments in these derivatives (Fig. 4, lanes 3 and 4), which was associated with loss in phenotype (Table 1). XhoI digests of R5-3A (Fig. 4A, lane 1) contained restriction fragments of approximately 6.0, 5.0, 4.0, and 2.0 kb, which are absent from similar digests of plasmid DNA from R5-3B (Fig. 4A, lane 2), without loss of phenotype. Comparison of R5-3A plasmid DNA with pWWO (see above) indicated that these fragment sizes are similar to regions of pWWO (i.e., XhoI fragments D, F, G, and I) which code for both the upper- and meta-pathway enzymes involved in aromatic hydrocarbon metabolism (3). Subsequent transfer of R5-3B from m-Xyl back to the original p-MB substrate resulted in the precise retrieval of these lost fragments in R5-3G, apparently without rearrangement (Fig. 4A, lane 5). Similar loss and retrieval of restriction enzyme fragments in these derivatives were observed when plasmid DNA was digested with HpaI or BamHI (Fig. 4B and 4C, lanes 1, 2, and 5). This phenomenon of retrieval of lost fragments by switching substrate was also observed in the toluene-grown derivative, R5-3C. These alterations in plasmid DNA content suggest that the retrieved fragments may exist in the plasmid DNA as conserved rearranged fragments or may be located in the chromosomal DNA to allow for their precise excision and

1527

religation into the plasmid when the selection substrate is changed. Physical evidence for the integration of plasmid DNA into chromosomal DNA in R5-3 and its association with phenotype. The loss of restriction enzyme fragments from the plasmid DNA in the m-Xyl-grown (R5-3B) and toluene-grown (R53C) derivatives without concomitant loss in phenotype and the subsequent return of these fragments to the plasmid after transfer back to p-MB (R5-3F and R5-3G [Fig. 4]) suggested that these lost fragments may be integrated into the chromosomal DNA of R5-3B and R5-3C. To investigate this hypothesis, it was decided to use the XhoI digests of plasmid DNA of R5-3B and R5-3C, which showed a loss of the 6.0-, 5.0-, 4.0-, and 2.0-kb fragments when compared with R5-3A (Fig. 4A). In a preliminary Southern Blot hybridization, R5-3A plasmid DNA (115 and 95 kb) was probed to XhoI-digested plasmid and chromosomal DNA from all the R5-3 derivatives listed in Table 1. The results indicated that R5-3A plasmid DNA hybridized to three fragments (5.0, 4.0, and 2.0 kb) in the chromosomal DNA of R5-3B (Fig. 5, lane 7B). These fragments could not be due to plasmid contamination of the chromosomal DNA, since they are not present in the plasmid DNA of R5-3B (Fig. 5, lane 3A versus lane 2A). Failure to detect these fragments in the chromosomal DNA of R5-3C (Fig. 5, lane 8B) appears to be correlated with the loss of phenotype after prolonged growth on toluene (Table 1). This result indicated that plasmid fragments missing after growth on 3-Xyl had been integrated into the chromosomal DNA without apparent alteration in size or physical characteristics and suggested that these fragments contain functional genes involved in the biodegradation of aromatic hydrocarbons. This observation was based upon the maintenance of phenotype in R5-3B, the retrieval of these fragments when substrate selection was altered, and the fact that these fragments are of similar size to pWWO fragments which carry the functional genes for methylbenzoate and xylene degradation. It is important to note that the fragments detected in the chromosomal DNA of R5-3B correspond to XhoI fragments D, F, G, and I of pWWO, which code for both upper- and meta-pathway enzymes (3). Fragment G contains the genes coding for the upper-pathway enzymes xylene oxygenase (xylA) and benzaldehyde dehydrogenase (xylC) involved in the xylene-toluene degradation (2, 7). The promoter region for these genes has been localized on the adjoining fragment F (3, 10). The expression of this operon is positively regulated by the xylR gene product, located on fragment D (7). The genes coding for the meta-pathway enzymes dihydroxycyclohexadiene carboxylate dehydrogenase (xylL) and catechol-2,3-dioxygenase (xylE) are located on fragment I, whereas fragment D contains the gene coding for 4-oxalocrotonate isomerase (xylH) and the regulatory region xylS

(6).

To confirm that the fragments sequestered in the chromosomal DNA of R5-3B were associated with phenotype, we performed hybridizations with specific clones of the genes for the upper-pathway enzymes (pED3306) and metapathway enzymes (pPL392) from pWWO. Probe DNA consisted of internal XhoI fragments of pED3306 and BglII fragments of pPL392 (6, 13). The XhoI fragments of pED3306 correspond to fragments F and G of pWWO, which code for the upper-pathway enzymes (3), and the BglII fragments contain xylD, xylL, xylE, xylG, xylF, xylJ, xylI, and xylH, which code for the meta-pathway enzymes (6). We first used these probes to determine whether plasmid DNA of R5-3 and its derivatives contained homologous

1528

APPL. ENVIRON. MICROBIOL.

CARNEY AND LEARY _-

1A 1B

2A 2B

3A 3B 4A 4B

5A 5B

-

6A 6B 7A

7 r"

W lb 8A or

M

I

urti uo I

23.1 Kb~ 9.4 K b 6.6Kb4.4Kb:~~~ 40

2.3K b t9Kb

10.5KbFIG. 5. Agarose gel electrophoresis (lanes A) of X/iol-digested plasmid and chromosomal DNA of R5-3 derivatives and pWWO and Southern blot hybridization (lanes B) of the same DNA probed with R5-3A plasmid DNA (lane 2-A). Blotted filters were preblocked with unlabeled X/hol-digested plasmid DNA of R5-3B (lane 3A) prior to hybridization with labeled DNA. Lanes: 1, lambda DNA digested with HindllI as a size standard; 2, plasmid DNA of R5-3A; 3. plasmid DNA of R5-3B; 4. plasmid DNA of R5-3D; 5, pWWO; 6, chromosomal DNA of R5-3A; 7, chromosomal DNA of R5-3B; 8, chromosomal DNA of R5-3C; 9. chromosomal DNA of R5-3D; 10, chromosomal DNA of the P. piutia strain with pWWO. Arrows in lane 7B indicate Xliol fragments in the chromosomal DNA of R5-3B.

sequences. Both pathways were identified in R5-3 plasmid DNA (Table 2). However, the ,netai-fission pathway genes were present only in the 95-kb plasmid, whereas the upperpathway genes were found in all plasmids of R5-3 and its derivatives, indicating that loss of the 95-kb plasmid in R5-3B, R5-3C, R5-3D, and R5-3E derivatives (Fig. 1) should also correspond to loss in the function of the ,neta-fission pathway. However, the derivatives R5-3B and R5-3C did not show the expected loss in phenotype (Table 1), which suggests that the genes could have been integrated into the chromosomal DNA. X/uol-digested chromosomal DNA from R5-3A, R5-3B, R5-3D, and the P. plitida strain with pWWO were probed with pED3306 and pPL392. Chromosomal DNA was purified by cesium chloride-bisbenzimide gradient centrifugation (9)

A 1 2 34 A

to diminish possible plasmid contamination of the chromosomal DNA. No hybridization was observed when either pED3306 or pPL392 was used to probe chromosomal DNA

of the P. pi,tida strain containing pWWO, suggesting that the method of purification was successful (Fig. 6B and C, lanes 6). Hybridization was observed in Xhol-digested pWWO (Fig. 6B and C, lanes 2). Fragments coding for the upper pathway were identified by hybridization with pED3306 in the chromosomal DNA of R5-3A and R5-3B and had the same migration pattern as the fragments in pWWO (Fig. 6B, lanes 3 and 4). Rearrangement in the chromosomal DNA of R5-3D (toluene grown) was suggested, since pED3306 hybridized to a different-sized fragment (10 to 12 kb) when compared with pWWO or the chromosomal digests of R5-3A and R5-3B (Fig. 6B, lane 5). Hybridization with pPL392

C

B 6

1

S 6

3 4

2 _

_

W

FIG. 6. Agarose gel electrophoresis (A) and Southern blot hybridization of pWWO (lanes 2) and chromosomal DNA from R5-3A (lanes 3), R5-3B (lanes 4), R5-3D (lanes 5), and the P. pittiua strain containing pWWO (lanes 6). Lanes 1 contain lambda DNA digested with Hindill as a size standard. All other samples were digested with Xliol. The probes were X/iol-digested internal fragments of pED3306 (B) and Bg/11-digested internal fragments of pPL392 (C). The plasmids pED3306 and pPL392 contain regions coding for the upper pathway and mneta-pathway (respectively) involved in aromatic metabolism (see text for details).

VOL. 55, 1989

showed that the meta-fission pathway genes were present in the chromosomal DNA of R5-3A and R5-3B (Fig. 6C, lanes 3 and 4) but absent in R5-3D (Fig. 6C, lane 5). The hybridization signal in R5-3B cannot be due to plasmid contamination of the chromosomal DNA, since meta-fission pathway genes were not found in the plasmid DNA of this derivative. These results support the hypothesis that plasmid DNA fragments associated with aromatic hydrocarbon metabolism are integrated into the chromosomal DNA of R5-3B after transfer from p-MB to m-Xyl and that gene function is maintained. The purpose of the work presented here was to analyze the alteration in plasmid profile in the derivatives of P. putida R5-3 and the association of those plasmids with aromatic hydrocarbon metabolism. Hybridization studies with the specific probes for the genes in the upper pathway (pED3306) and for the genes in the meta-fission pathway (pPL392) confirmed the presence of fragments in the plasmid DNA of R5-3 which code for the enzymes in those pathways (Table 2). Comparison of R5-3A plasmid DNA with the archetypal pWWO revealed relatedness between these plasmids but also revealed numerous physical differences in the distribution of their restriction enzyme sites, indicating that these two plasmids, although similar, are not identical (Fig. 2). Other differences include the inability to cure the plasmid DNA of R5-3 by growth on benzoate (Fig. 1, lane 7), while prolonged growth of R5-3 on toluene resulted in the loss of the ability to utilize methylbenzoate and xylenes (Table 1). In these derivatives (R5-3D and R5-3E), which have lost the meta-fission pathway function as assayed by phenotype determination and the enzyme activity assay (Table 1; Fig. 6C), the question arises of what degradative pathway is used in the metabolism of toluene. It is possible that the continued ability of these derivatives to degrade toluene proceeds through a modified ortho-pathway. The most surprising result of these studies is the observation that specific plasmid DNA sequences associated with aromatic hydrocarbon metabolism are integrated into the chromosomal DNA and continue to be expressed in the m-Xyl-grown derivative R5-3B. It is also significant that these sequences are apparently excised from the chromosome and returned intact to the plasmid when R5-3B is grown on p-MB (Fig. 4). Recently, the integration of plasmid DNA into the host chromosome was observed in large plasmids of Acetobacter xylinum, although in this case the phenotype was not positively associated with the integrated

plasmid fragments (21). The transfer of R5-3 plasmid DNA to the chromosome and return to the plasmid were apparently induced by altering the substrate from p-MB to m-Xyl and were reversed on transfer back to p-MB. When R5-3 was grown on toluene for an extended period, however, the phenotype was lost, indicating that the genes for the meta-fission pathway are not stably maintained in the chromosomal DNA. In conclusion, the data presented here indicate that maintenance of phenotype in P. puitida R5-3 is achieved through a complex system involving an interaction between plasmid and chromosomal DNA. The particular aromatic hydrocarbon substrate utilized apparently induced the alterations in the biodegradative plasmids. Although no evidence is given about why these alterations may take place, this work may be significant in further understanding the origin of catabolic

plasmids.

PLASMID DNA ALTERATIONS IN P. PUTIDA

1529

ACKNOWLEDGMENTS We thank D. Cooksey for providing the plasmid size standards, D. Ohman for providing the P. putida strain with pWWO, and K. N. Timmis and S. Harayama for providing the pBR322 constructs, pED3306 and pPL392, used in this study. We especially thank Frank Higson, Department of Soils and Environmental Sciences, for performing the enzyme activity assay. We appreciate the helpful discussions and comments from B. Hyman and acknowledge N. Keen, D. Cooksey, and D. D. Focht for reviewing the manuscript. This work was supported by a grant from the Univcrsity of California Toxic Substances Research and Training Program. LITERATURE CITED 1. Atlas, R. M. (ed.). 1984. Petroleum microbiology. Macmillan Inc., New York. 2. Franklin, F. C. H., M. Bagdasarian, M. M. Bagdasarian, and K. N. Timmis. 1981. Molecular and functional analysis of the TOL plasmid pWWO from Pseudomonas putida and cloning of genes for the entire regulated aromatic ring meta cleavage pathway. Proc. Natl. Acad. Sci. USA 78:7458-7462. 3. Franklin, F. C. H., P. R. Lehrbach, R. Lurz, B. Rueckert, M. Bagdasarian, and K. N. Timmis. 1983. Localization and functional analysis of transposon mutations in the regulatory genes of TOL catabolic pathway. J. Bacteriol. 154:676-685. 4. Gibson, D. T. 1971. Assay of enzymes of aromatic metabolism, p. 463-478. In J. R. Norris and D. W. Ribbons (ed.), Methods in microbiology, vol. 6A. Academic Press, Inc., New York. 5. Gibson, D. T. (ed.), 1984. Microbial degradation of organic compounds. Marcel Dekker, Inc. New York. 6. Harayama, S., P. R. Lehrbach, and K. N. Timmis. 1984. Transposon mutagenesis analysis of meta-cleavage pathway operon genes of the TOL plasmid of Pseudomonas putida mt-2. J. Bacteriol. 160:251-255. 7. Harayama, S., R. A. Leppik, M. Rekik, N. Mermod, P. R. Lehrbach, W. Reinke, and K. N. Timmis. 1986. Gene order of the TOL catabolic plasmid upper pathway operon and oxidation of both toluene and benzyl alcohol by the xylA product. J. Bacteriol. 167:455-461. 8. Heinrich, K., and P. A. Williams. 1985. A new class of TOL plasmid deletion mutants in Pseudomonas putida MT15 and their reversion by tandem gene amplification. J. Gen. Microbiol. 131:1023-1033. 9. Hudspeth, M. E., D. S. Shumard, K. M. Tatti, and L. I. Grossman. 1980. Rapid purification of yeast mitochondrial DNA in high yield. Biochim. Biophys. Acta 610:221-225. 10. Inouye, S., Y. Ebina, A. Nakazawa, and T. Nakazawa. 1984. Nucleotide sequence surrounding transcription initiation site of Xyl ABC operon on TOL plasmid of Pseqdomonas putida. Proc. NatI. Acad. Sci. USA 81:1688-1691. 11. Krockel, L., and D. D. Focht. 1987. Construction of chlorobenzene-utilizing recombinants by progenitive manifestation of a rare event. Appl. Environ. Microbiol. 53:2470-2475. 12. Maniatis, T., E. F. Fritsch, and J. Sambrook. 1982. Molecular cloning-a laboratory manual. Cold Spring Harbor Laboratory, Cold Spring Harbor, N.Y. 13. Mermod, N., S. Harayama, and K. N. Timmis. 1986. New route to bacterial production of indigo. Bio/Technology 4:321-324. 14. Rigby, P. W. J., M. Dieckmann, C. Rhodes, and P. Berg. 1977. Labeling deoxyribonucleic acid to high specific activity in vitro by nick translation with DNA polymerase I. J. Mol. Biol. 113:237-251. 15. Sinclair, M. I., P. C. Maxwell, B. R. Lyon, and B. W. Holloway. 1986. Chromosomal location of TOL plasmid DNA in Pseiudomonas putida. J. Bacteriol. 168:1302-1308. 16. Soqthern, E. M. 1975. Detection of specific sequences among DNA fragments separated by gel electrophoresis. J. Mol. Biol. 98:503-517. 17. Staskawicz, B. J., D. Dahlbeck, and N. T. Keen. 1984. Cloned avirulence gene of Pseudomonas syringae pv. glycinea determines race-specific incompatibility on Glycine max (L.) Merr. Proc. Natl. Acad. Sci. USA 81:6024-6028. 18. Williams, P. A., and K. Murray. 1974. Metabolism of benzoate and the methylbenzoates by Pseudomonas putida (Arvilla)

1530

CARNEY AND LEARY

mt-2: evidence for the existence of a TOL plasmid. J. Bacteriol. 120:416-423. 19. Worsey, M. J., F. C. Franklin, and P. A. Williams. 1978. Regulation of the degradative pathway enzymes coded for by the TOL plasmid (pWWO) from Pseuidomontis piutid mt-2. J. Bacteriol. 134:757-764.

APPL. ENVIRON. MICROBIOL.

20. Worsey, M. J., and P. A. Williams. 1975. Metabolism of toluene and xylenes by Pseiudotinotiais pi(ticdt (Arvilla) mt-2: evidence for a new function of the TOL plasmid. J. Bacteriol. 124:7-13. 21. Valla, S., D. H. Coucheron, and J. Kjosbakken. 1987. The plasmids of Acetobacterx .ylintin and their interaction with the host chromosome. Mol. Gen. Genet. 208:76-83.