Vol. 171, No. 12JOURNAL OF BACTERIOLOGY, Dec. 1989, p. 6782-67900021-9193/89/126782-09$02.00/0Copyright © 1989, American Society for Microbiology

Regulator and Enzyme Specificities of the TOL Plasmid-EncodedUpper Pathway for Degradation of Aromatic Hydrocarbons and

Expansion of the Substrate Range of the PathwayMARIA-ANGELES ABRIL,1 CARMEN MICHAN,1 KENNETH N. TIMMIS,2'3 AND JUAN L. RAMOS'.3*

Estacion Experimental del Zaidin, Consejo Superior de Investigaciones Cientificas, Apto 419, 18080 Granada, Spain';GBF, D-3300 Braunschweig, Federal Republic of Germany2; and Department ofMedical Biochemistry, CMU,

University of Geneva, Geneva, Switzerland3

Received 6 April 1989/Accepted 21 September 1989

The TOL plasmid upper pathway operon encodes enzymes involved in the catabolism of aromatichydrocarbons such as toluene and xylenes. The regulator of the gene pathway, the XylR protein, exhibits a verybroad effector specificity, being able to recognize as effectors not only pathway substrates but also a widevariety of mono- and disubstituted methyl-, ethyl-, and chlorotoluenes, benzyl alcohols, and p-chlorobenzal-dehyde. Benzyl alcohol dehydrogenase and benzaldehyde dehydrogenase, two upper pathway enzymes, exhibitvery broad substrate specificities and transform unsubstituted substrates and m- and p-methyl-, m- andp-ethyl-, and m- and p-chloro-substituted benzyl alcohols and benzaldehydes, respectively, at a high rate. Incontrast, toluene oxidase only oxidizes toluene, m- and p-xylene, m-ethyltoluene, and 1,3,4-trimethylbenzene,also at a high rate. A biological test showed that toluene oxidase attacks m- and p-chlorotoluene, albeit at a lowrate. No evidence for the transformation of p-ethyltoluene by toluene oxidase has been found. Hence, tolueneoxidase acts as the bottleneck step for the catabolism ofp-ethyl- and m- and p-chlorotoluene through the TOLupper pathway. A mutant toluene oxidase able to transform p-ethyltoluene was isolated, and a mutant straincapable of fully degrading p-ethyltoluene was constructed with a modified TOL plasmid meta-cleavage pathwayable to mineralize p-ethylbenzoate. By transfer of a TOL plasmid into Pseudomonas sp. strain B13, a clone ableto slowly degrade m-chlorotoluene was also obtained.

Chemicals containing one or more benzene rings can bemineralized by soil and sediment bacteria (9). In manyinstances the genetic information for the biodegradativeprocess is harbored by plasmids which act as efficientvehicles for the spread of such information. The TOLplasmid of Pseudomonas putida is the most extensivelycharacterized catabolic plasmid; it encodes enzymes for themineralization of toluene, m- and p-xylene, m-ethyltoluene,and 1,3,4-trimethylbenzene (18, 31). In the degradation ofthese compounds, the methyl group at carbon 1 in thearomatic ring is sequentially oxidized to yield the corre-sponding carboxylic acid (upper pathway). The carboxylicacid is then oxidized to its corresponding catechol, whichundergoes meta fission to produce a semialdehyde which isfurther transformed into products (pyruvate plus aldehydes)that are finally transformed by chromosomally encodedenzymes into Krebs cycle intermediates (Fig. 1).

Studies of transposon and nitrosoguanidine mutagenesisand gene cloning have shown the genes for the catabolicpathways to be organized in two separate operons, oneencoding enzymes for the upper pathway (the upper operon)and the other encoding enzymes for the meta-cleavagepathway (the meta operon) (Fig. 1) (see reference 12 for areview). Expression of the TOL plasmid catabolic genes isregulated at the transcriptional level. The upper pathwayoperon is induced by toluene, xylenes, and their alcoholderivatives, and this induction is mediated by the xylR geneproduct (7, 8, 14) together with the sigma factor NtrA (3).Expression of the meta-cleavage pathway operon is inducedby benzoates, such induction being mediated by the xylSgene product (6, 13, 22, 30). The meta-cleavage pathway

* Corresponding author.

operon is also switched on by upper pathway substratesthrough a cascade regulatory system that involves stimula-tion of transcription from the xylS gene promoter (Ps) byXylR with NtrA (22). Hyperproduction of the XylS proteinin turn leads to stimulation of transcription from the TOLmeta-cleavage pathway operon promoter (Pm) in the ab-sence of meta-cleavage pathway substrates (15, 22).We have chosen the archetypal TOL plasmid pWWO to

study the evolution of catabolic pathways and are directingour efforts to the manipulation of such pathways to allow themetabolism of certain alkyltoluenes and alkylbenzoates. Thelaboratory evolution of TOL plasmid catabolic pathwaysinvolves the acquisition of new enzymatic and/or regulatoryspecificities (for substrates and effectors, respectively)through mutational alteration of existing proteins. This evo-lution is made possibly by the considerable plasticity exhib-ited by TOL-encoded regulators and enzymes in broadeningtheir specificities without a loss of function (24).

In the present study, we analyzed the effector profile ofthe TOL plasmid upper pathway regulator, the XylR protein,and the substrate profile of the TOL upper pathway en-zymes, namely, toluene oxidase (TO), benzyl alcohol dehy-drogenase (BADH), and benzaldehyde dehydrogenase(BZDH) (Fig. 1). Our analyses showed that TO activity is themain bottleneck for the transformation of certain substitutedtoluenes, such as p-ethyltoluene and m-chlorotoluene. Iso-lation of one TO mutation in plasmid pWWO-EB62 which, inaddition, exhibits a modified meta-cleavage pathway for themetabolism of alkylbenzoates, allowed the mineralization ofp-ethyltoluene. Furthermore, transfer of the TOL plasmidinto Pseudomonas sp. strain B13 led to the isolation ofm-chlorotoluene degraders.

6782

on January 22, 2020 by guesthttp://jb.asm

.org/D

ownloaded from

on January 22, 2020 by guest

http://jb.asm.org/

Dow

nloaded from

on January 22, 2020 by guesthttp://jb.asm

.org/D

ownloaded from

http://jb.asm.org/http://jb.asm.org/http://jb.asm.org/

LABORATORY EVOLUTION OF TOL CATABOLIC PATHWAYS

i-UPPER PATHWAY -META PATHWAY I

CH3 c2O cmo coW A A C D*L

MR10- -NNONONKreb's° E , F...H rbL OI-.'Qj s .- cycleRI RI"2 1intermediates

PU C M A B |Pm X Y Z L E G F J K I H iSPs Pr RIFIG. 1. TOL catabolic pathways encoded by plasmid pWWO and genetic organization of the genes. The xyl genes of the two operons or

their gene products are as follows: A, xylene oxidase (the protein is composed of two different subunits encoded by the xyIMA genes [12]);B, BADH; C, BZDH; D, toluate 1,2-dioxygenase (the protein is composed of three different subunits encoded by the xylXYZ genes [11]); L,dihydroxycyclohexadiene carboxylate dehydrogenase; E, C2,3-O; F to H, subsequent enzymes of the meta-cleavage pathway (10) and genesencoding the corresponding enzymes. S encodes the regulator of the meta-cleavage pathway, and R encodes the regulator of the upper operonand the xylS gene. Dashed lines indicate TOL DNA between the upper and meta-cleavage operons.

MATERIALS AND METHODS

Bacterial strains and plasmids. Bacterial strains used inthis work were Escherichia coli 5K (res thr leu thi tonAsupE) (R. L. Robson), E. coli JM101 (lac pro thi supE F'traD36 proAB lacIZA15) (29), P. putida 2440 (hsdM hsdR)(6), P. putida PaW340 (Nalr trp) (S. Harayama), P. putidaEEZ1 and EEZ3 (Rif' and Rifr Smr derivatives of P. putida2440, respectively), Pseudomonas sp. strain B13 (3-chlo-robenzoate positive and 3-chlorotoluene negative) (4), Pseu-domonas sp. strain EEZ4 (Rif' derivative of B13), andPseudomonas sp. strain EE5 (3-chlorotoluene positive).Plasmids used or constructed in the present study are listedin Table 1.Media and culture conditions. Bacteria were grown at 30°C

in LB broth (19) and in modified M9 minimal medium. Themodified medium contained the following (per liter):NaH2PO4. 7H20, 60 g; K2HPO4, 30 g; NH4Cl, 10 g; NaCl,5 g; MgSO4, 0.52 g; and 2.5 ml of A9 solution. A9 solutionwas composed of the following (in milligrams per liter):HBO3, 300; ZnCl2, 50; MnCl2 * 4H20, 30; CoCl2, 200;CuCl2 - 2H20, 10; NiCl2 - 6H20, 20; and NaMoO4 * 2H20,30. Carbon sources were glucose (0.5% [wt/vol]), acetate (10mM), or aromatic hydrocarbons that were supplied in the gas

TABLE 1. Plasmids used in this study

Plasmid Relevant characteristics Source

pERD4 Kmr IncP xylS4 25pERD99 Apr pBR replicon Pr::lacZ This studypERD100 Tcr IncQ Pm::lacZ C. MichanpKT570 Smr IncP xyIR xylS 6pMD1405 Apr pBR replicon transcriptional M. Drummond

lacZ promoter probepNM185 Kmr IncP xylS 21pRD579 Apr pBR replicon Pu::lacZ 3pRL38 Apr pBR replicon xylS xyIR S. HarayamapTS174 Cmr P15A replicon xyIR 14pWA21 Apr Smr IncP lac promoter:: M. Wubbolts

xylMApWWO Archetypal TOL plasmid Tol+ 30

pEB-pWWO-EB6 Tol+ pEB+ 25pWWO-EB61 Tol+ pEB+ pET- This studypWWO-EB62 Tol+ pET+ This study

a Abbreviations: Km, kanamycin; Ap, ampicillin; Tc, tetracycline; Sm,streptomycin; Cm, chloramphenicol; Tol, toluene; pEB, p-ethylbenzoate;pET, p-ethyltoluene.

phase. For E. coli strains, 1 mg of thiamine per liter wasadded. When appropriate, antibiotics were added at thefollowing concentrations (in micrograms per milliliter): ka-namycin, 25; ampicillin, 100; tetracycline, 10; chloramphen-icol, 30; and streptomycin, 50.

Transfer of TOL plasmids between Pseudomonas strains.About 108 cells ofP. putida bearing one of the TOL plasmidswere mixed with about 108 cells of a distinguishable Pseu-domonas recipient. The mixture was filtered through a0.45-,um-pore-diameter filter (Sartorius) and placed on LBsolid medium. After 16 h at 30°C, the mating mixture wassuspended in 5 ml of M9 minimal medium without a C sourceand serially diluted in the same medium; cells were finallyspread on selective medium. Donor and recipient strainsincubated under the same conditions were used as controls.Enzyme assays. 3-Galactosidase was determined in perme-

abilized whole cells as described previously (23). BZDH,BADH, and catechol 2,3-dioxygenase (C2,3-O) activitieswere determined in cell extracts as described previously(31). Units are expressed as micromoles of product formedper minute.TO activity was determined in whole cells either polaro-

graphically or by measuring the product resulting from theoxidation of toluene or substituted toluenes with cloned TOgenes. For the polarographic assay, P. putida 2440(pWWO)and P. putida(pWWO-EB62) cells growing exponentially onminimal medium with p-xylene were collected by centrifu-gation (5,000 x g for 5 min), washed in minimal mediumwithout a C source, and resuspended in the same medium.Two milliliters of the culture (A600, 0.46 to 0.50) was placedin the chamber of an oxygen electrode (Hansatech) at 30°C,and oxygen consumption due to endogenous respiration wasdetermined. Toluene or substituted toluenes were added inN,N'-dimethylformamide to a final concentration of 0.5 mM,and the increase in oxygen consumption in response to thearomatic hydrocarbon was determined. This increase wasconsidered to be due to TO activity, although it representsnot only the oxidation of toluenes to their correspondingbenzyl alcohols but also oxygen consumption due to metab-olism of the latter.

Plasmid pWA21 contains the TOL plasmid TO genes(xylMA) cloned downstream from the lac promoter. E. coliJM1O1(pWA21) was used to measure directly the transfor-mation of toluene and related hydrocarbons into their corre-sponding benzyl alcohols, which accumulated in the culturemedium.

6783VOL. 171, 1989

on January 22, 2020 by guesthttp://jb.asm

.org/D

ownloaded from

http://jb.asm.org/

6784 ABRIL ET AL.

Benzyl alcohol determination. Benzyl alcohols were ex-tracted twofold from the culture medium with ether, driedover anhydrous MgSO4, and concentrated by evaporation.The resulting powder was dissolved in water, and theconcentrations of benzyl alcohols were determined by gaschromatography with a Hewlett-Packard model 5730 gaschromatograph. Detection of benzyl alcohols was done byflame ionization at 300°C after the samples were run througha 100/200-mesh Chromasorb-W-AW-OMCS column in anoven set 20°C higher than the alcohol evaporation point.DNA manipulations. Procedures for plasmid isolation,

transformation, cleavage by restriction enzymes, analysis byagarose gel electrophoresis, and gene cloning have beenpreviously reported (19).Other methods. Protein levels in cell extracts were deter-

mined by the method of Bradford (2) with bovine serumalbumin as a standard.

RESULTS

Stimulation of transcription from XyIR-regulated promot-ers and effector specificity of the XyIR protein. P. putida2440(pWWO) exhibited low levels of upper and meta-cleavage pathway enzymes when grown on sugar or dicar-boxylic acids, e.g., acetate or succinate (23, 30). The addi-tion of toluene or m-xylene led to coordinated but slowinduction of upper and meta-cleavage pathway enzymeactivities, with maximal levels being reached after 6 h ofincubation, after which a decrease in enzyme activity levelswas observed (Fig. 2A).

Plasmids pRD579 and pERD100 carry transcriptional fu-sions of the TOL upper pathway operon promoter (Pu) (3)and Pm to a promoterless lacZ gene, respectively. Thekinetics of induction from Pu in pRD579 and Pm in pERD100were studied in E. coli with xylR and xylS genes in thecompatible plasmid pKT570. Stimulation of transcriptionfrom these promoters in E. coli was also slow, with inductionkinetics similar to those seen with the native TOL plasmid.To confirm the possible usefulness of the Pu::lacZ fusion

in the analysis of XylR effector specificity, we examined3-galactosidase synthesis in cells containing this construc-

tion in response to known effectors and noneffectors. In E.coli(pRD579, pTS174), the upper pathway substrates tolueneand m- and p-xylene and the pathway intermediates benzylalcohol and m- and p-methylbenzyl alcohol induced highlevels of ,-galactosidase synthesis (Tables 2 and 3). Apathway intermediate known to be a noneffector, e.g.,benzaldehyde (31), and aromatic compounds such as phenol,aniline, and benzene, which are not pathway substrates, ledto no increase in P-galactosidase synthesis. These resultsconfirm that the Pu: :lacZ fusion is effective in the analysis ofxylR gene product specificity and corroborate the fact thatthe TOL upper pathway operon is indeed expressed inresponse to hydrocarbon substrates and their alcohol cata-bolic products.The influence of an alkyl group in the aromatic ring on the

XylR-mediated stimulation of transcription from Pu wasanalyzed with toluene, ethylbenzene, isopropylbenzene, andnonylbenzene. In addition to toluene, which behaved as apotent effector, ethylbenzene was also able to promote lowlevels of transcription from Pu. However, longer chainswere unproductive.A number of substituted toluenes were tested as effectors

for the XylR protein. In addition to the TOL upper pathwaysubstrates m- and p-xylene, m-ethyltoluene, and 1,3,4-tri-methylbenzene (Table 2), other hydrocarbons that are not

.-f

E0E4E

-00-

50-

00 8

TIME (h)

.-f

c

In

;-

O,CwEDa,6Wt0

50B

25

vr~~I0 4 8

TIME Ch)FIG. 2. (A) Coordinate induction of TOL upper and meta-

cleavage pathway enzymes with toluene. A P. putida 2440(pWWO)culture growing on minimal medium with acetate was divided intotwo halves (arrow). One-half was incubated with m-xylene (O andA), while the other was kept as a control (S and A). At the indicatedtimes, BADH (O and 0) and C2,3-O (A and A) activities weredetermined. One-hundred percent values for BADH and C2,3-Owere 1.4 and 1 U, respectively. (B) 13-Galactosidase activity ex-pressed from Pu and Pm in response to m-xylene. E. coli5K(pRD579, pKT570) (O and 0) and E. coli 5K(pERD100, pKT570)(A and A) were incubated in LB medium with appropriate antibiot-ics in the absence (0 and A) or the presence (O and A) of m-xylene.At the indicated times, P-galactosidase activity was determined inpermeabilized cells.

metabolized through such a pathway, e.g., o-xylene, o- andp-ethyltoluene, and o-, m-, and p-chlorotoluene, were foundto be effectors for XylR (Table 2). In fact, the chlorinatedtoluenes were the most efficient effectors. The influence ofseveral trisubstitutions on the aromatic ring was analyzedwith the following compounds: p-chloro-o-xylene, o-chloro-m-xylene, o-chloro-p-xylene, 1,2,3-trimethylbenzene, 1,3,5-trimethylbenzene, 2,5-dichlorotoluene, and 2,6-dichlorotol-uene. All of the above-listed compounds behaved aseffectors for XylR (Table 2), suggesting that the XylR proteinexhibits a very broad effector specificity.

--o -----,A -A- &l .6

J. BACTERIOL.

on January 22, 2020 by guesthttp://jb.asm

.org/D

ownloaded from

http://jb.asm.org/

LABORATORY EVOLUTION OF TOL CATABOLIC PATHWAYS

TABLE 2. Activation of the XyIR protein by hydrocarbonsa

,3-galactosidase activity (U) with:Effector

Pu::lacZ Pm::lacZ

None 201 140Toluene 2,050 1,860Ethylbenzene 670 bIsopropylbenzene 210Nonylbenzene 200o-Xylene 1,850m-Xylene 2,960 1,650p-Xylene 2,070o-Ethyltoluene 520m-Ethyltoluene 1,510p-Ethyltoluene 500 160o-Chlorotoluene 3,660 810m-Chlorotoluene 3,320p-Chlorotoluene 3,280 1,020o-Chloro-m-xylene 1,030o-Chloro-p-xylene 1,810p-Chloro-o-xylene 1,7801,2,3-Trimethylbenzene 1,220 6401,3,4-Trimethylbenzene 1,600 1,6401,3,5-Trimethylbenzene 9102,5-Dichlorotoluene 1,710 5002,6-Dichlorotoluene 820 570

a E. coli 5K(pRD579, pTS174) and E. coli(pERD100, pKT570) were grownin LB medium with ampicillin and chloramphenicol or tetracycline andstreptomycin, respectively. Overnight cultures were diluted 100-fold withfresh medium supplemented with the appropriate antibiotics. The hydrocar-bons were supplied in the vapor phase. The relative copy number of theplasmids used did not change in the presence of the hydrocarbons. ,B-Galactosidase activity levels were determined after 5 to 6 h in the presence ofthe effector. Data given are the averages of two to six independent determi-nations. Standard deviations were in the range of 10 to 40%o of the givenvalues.b-, Not tested.

Induction from Pm with toluene involves a cascade regu-latory system, with the XylR protein promoting transcriptionfrom Ps and the XylS protein stimulating transcription fromPm in the absence of meta-cleavage substrates. In E.coli(pERD100, pKT570), ,-galactosidase synthesis from Pmin response to some of the above-tested toluenes was exam-ined. The pattern of induction from Pm with substitutedtoluenes was similar to that from Pu (Table 2).As mentioned above, benzyl alcohol and m- and p-methyl-

benzyl alcohol behaved as XylR effectors (Table 3). Thelevel of induction with those alcohols was dependent onalcohol concentration, with maximal levels of inductionbeing measured at 5 to 10 mM. The m- and p-chlorobenzylalcohol were strong effectors for the stimulation of transcrip-tion from Pu, while p-ethylbenzyl alcohol was not an ef-fector. In contrast, the stimulation of transcription from Pmwith the chlorinated benzyl alcohols was surprisingly low.Although benzaldehyde and m- and p-methylbenzalde-

hyde at 1 mM (higher concentrations were toxic for the cellsand significantly reduced cell growth) were not effectors forXylR, we tested other substituted benzaldehydes as putativeeffectors. Among those tested (Table 3), we found that XylRwith p-chlorobenzaldehyde was able to stimulate transcrip-tion from Pu. Induction with p-chlorobenzaldehyde wasobserved at concentrations as low as 100 ,uM, and maximallevels were measured at concentrations of about 1 mM.Stimulation of transcription from Pm in E. coli(pERD100,pKT570) was also observed with low concentrations ofp-chlorobenzaldehyde (Table 3).XylR synthesis autoregulation. Toluene-activated XylR

protein is the ultimate element controlling both the expres-

TABLE 3. Activation of the XylR protein by aromaticalcohols and benzaldehydesa

P-galactosidase activityEffector and concn (U) with:

(mM)Pu::lacZ Pm::lacZ

None 300 150

Benzyl alcohol1S10

m-Methylbenzyl alcohol1S10

p-Methylbenzyl alcohol1S10

p-Ethylbenzyl alcohol12.5

m-Chlorobenzyl alcohol1S

p-Chlorobenzyl alcohol1

Benzaldehyde, 1

o-Methylbenzaldehyde, 1

m-Methylbenzaldehyde, 1

p-Methylbenzaldehyde, 1

m-Chlorobenzaldehyde, 1

b

350500

720

180560

110100

490800900

1,6902,2503,040

330800

1,350

350400

1,1202,540

5503,740

300

110170

460

400

430

320

400

p-Chlorobenzaldehyde0.1 5600.5 8301 1,750 770

a Conditions were as described in Table 2, except that the benzyl alcoholsand benzaldehydes were used at the concentrations shown. Chlorobenzylalcohols and p-ethylbenzyl alcohol at concentrations higher than 2.5 to 5 mMwere toxic for the cells. The relative copy number of the plasmids used did notchange in the presence of the benzyl alcohols or benzaldehydes.b-, Not tested.

sion from Pu and the cascade regulatory system involved inthe expression of the TOL meta-cleavage pathway operon.To study expression from the xylR gene promoter (Pr), wefused Pr to a promoterless lacZ gene in pERD99. pERD99was constructed as follows. A 0.6-kilobase BglII fragmentfrom plasmid pRL38 carrying Pr was cloned at the onlyBamHI site in the promoter probe plasmid pMD1405. Theresulting plasmid was called pERD98. In pERD98, the BglIIfragment was inserted such that Pr was in front of the lacZgene; however the lacZ gene was not in the correct readingframe. pERD98 was then linearized with PstI, and thecohesive ends were made blunt with deoxynucleosidetriphosphates and the Klenow fragment of DNA polymer-ase, followed by ligation with T4 DNA ligase. In the result-

6785VOL. 171, 1989

on January 22, 2020 by guesthttp://jb.asm

.org/D

ownloaded from

http://jb.asm.org/

6786 ABRIL ET AL.

FIG. 3. Construction of pERD99. The experimental steps aredescribed in the text. Abbreviations for restriction endonucleasecleavage sites: B, BamHI; Bg, BglII; Bs, BstEII; E, EcoRI; H,HindIII; P, PstI. dNTP's, Deoxynucleoside triphosphates.

ing plasmid, pERD99, reading from Pr of the lacZ gene wasthen in the correct frame (Fig. 3).

P-Galactosidase activity measured in E. coli carryingpERD99 (2,300 to 2,800 U) was always significantly higherthan in E. coli concomitantly carrying a plasmid containingthe xylR gene, e.g., pTS174; in fact, P-galactosidase levels inE. coli(pERD99, pTS174) were about 1,200 to 1,500 U. Thisresult suggests that the XylR protein controls its own syn-thesis.

E. coli(pERD99, pTS174) was also used to study expres-sion from Pr in the presence of XylR effectors. XylRactivated by hydrocarbons, benzyl alcohols, and p-chlo-robenzaldehyde reduced the level of expression from Prfrom 30 to 50% (Table 4). Compounds that are noneffectorsfor the XylR protein failed to reduce the level of expressionfrom Pr.

Substrate specificities of TOL upper pathway enzymes. Thesubstrate specificities of the TOL upper pathway enzymesBADH and BZDH were determined in cell extracts preparedfrom P. putida 2440(pWWO) grown on toluene or m-xylene

TABLE 4. Autoregulation of XylRa

Effector -galactosidaseactivity (U)None ....................................... 1,390Toluene ....................................... 890o-Xylene ....................................... 570m-Xylene ....................................... 810p-Xyleiie ....................................... 626o-Ethyltoluene ....................................... 889o-Chlorotoluene....................................... 754p-Chlorotoluene....................................... 7742,5-Dichlorotoluene ....................................... 1,0202,6-Dichlorotoluene ....................................... 1,024Benzyl alcohol .................... ................... 1,209m-Methylbenzyl alcohol ....................................... 1,154m-Chlorobenzyl alcohol ....................................... 1,041p-Chlorobenzaldehyde ....................................... 843Benzaldehyde ....................................... 1,376Benzene ....................................... 1,430

a E. coli 5K(pERD99, pTS174) was grown overnight in LB mediumsupplemented with ampicillin and chloramphenicol. Cultures were diluted100-fold in the same medium and incubated for 6 h at 30°C in the presence ofthe XyIR effectors. The hydrocarbons were supplied in the gas phase. Benzylalcohols and benzaldehydes were added to the culture medium to a finalconcentration of 1 mM. P-Galactosidase activity was determined in perme-abilized cells. Data given are the averages of two to four independentdeterminations. Standard deviations were in the range of 10 to 25% of thegiven values.

as the inducer. In contrast, TO activity was determined ininduced whole cells with P. putida 2440(pWW0) grown onm-xylene or with cloned TO genes in plasmid pWA21. TOactivity was determined by measuring increases in oxygenconsumption in response to the addition of different substi-tuted toluenes. Marked increases in oxygen consumptionwere observed in response to the addition of hydrocarbonsthat support the growth of P. putida 2440(pWWO), e.g.,toluene and m- and p-xylene (Table 5), while no significantincreases were observed with hydrocarbons that do notsupport the growth of P. putida 2440(pWWO), e.g., o-xylene,p-ethyltoluene, and 3- and 4-chlorotoluene (Table 5). How-ever, when P. putida 2440(pWWO) was grown on glucose inthe presence of m- and p-chlorotoluene, m-chlorocatecholand the semialdehyde derived from the meta-cleavage ringfission of p-chlorocatechol accumulated in the culture me-dium as a result of the metabolism of m- and p-chlorotolu-ene, respectively. This result suggests that these chlorotol-uenes were attacked by the TOL-encoded TO, albeit at lowrates.

In plasmid pWA21, TO is expressed from the lac pro-moter. E. coli JM101(pWA21) grown on minimal mediumwith 0.4% glucose in the presence of 100 ,uM isopropyl-P-D-thiogalactopyranoside exhibited high levels of TO, andbenzyl alcohol (2.68 ,umol/ml in 16 h) accumulated in theculture medium as a result of the oxidation of toluene. Incontrast, no accumulation of p-ethylbenzyl alcohol fromp-ethyltoluene was detected. These results parallel TO ac-tivity measured as the increase in oxygen consumption inresponse to hydrocarbons.BADH exhibited a broad substrate specificity. Unsubsti-

tuted benzyl alcohol was transformed at the highest rate (1.4U/mg of protein); methyl-, ethyl-, and chloro-substitutedbenzyl alcohols were also oxidized at high rates, rangingfrom 69 to 28% of those recorded with the unsubstitutedform (Table 5). BADH exhibited a high affinity for benzylalcohols. Kms for benzyl alcohol, p-ethylbenzyl alcohol, andp-chlorobenzyl alcohol were 10 + 2, 14 + 4, and 28 + 6 ,M,respectively, as calculated from double-reciprocal (Line-

J. BACTERIOL.

on January 22, 2020 by guesthttp://jb.asm

.org/D

ownloaded from

http://jb.asm.org/

LABORATORY EVOLUTION OF TOL CATABOLIC PATHWAYS

TABLE 5. Substrate specificities of wild-type plasmid upperpathway enzymes

Enzyme Substrate Activity

TOa Toluene 15o-Xylene -0.1m-Xylene 9p-Xylene 10.2p-Ethyltoluene -O. 1m-Chlorotoluene O.1p-Chlorotoluene s0.l3,5-Dichlorotoluene NDd

BADHb Benzyl alcohol 1,443o-Methylbenzyl alcohol 132m-Methylbenzyl alcohol 1,007p-Methylbenzyl alcohol 728p-Ethylbenzyl alcohol 414m-Chlorobenzyl alcohol 771p-Chlorobenzyl alcohol 7113,5-Dichlorobenzyl alcohol 400

BZDHC Benzaldehyde 5,058o-Methylbenzaldehyde 16o-Methylbenzaldehyde 3,033p-Methylbenzaldehyde NDp-Ethylbenzaldehyde 5,016m-Chlorobenzaldehyde NDp-Chlorobenzaldehyde 3,2503,5-Dichlorobenzaldehyde ND

a P. putida 2440(pWWO) cells were grown in minimal medium with tolueneas a carbon source. Enzyme activities were assayed as described in Materialsand Methods. Given values are the averages of at least three independentdeterminations. Standard deviations were in the range of5 to 30%o of the givenvalues. Hydrocarbons were supplied at a concentration of 0.5 mM. Activity isgiven in nanomoles of 02 per minute per unit of A6w.

b The concentration of aromatic alcohols was 2.5 mM. Activity is given inmilliunits per milligram of protein.

c The concentration of aromatic aldehydes was 0.4 mM. Activity is given inmilliunits per milligram of protein.

d ND, Not determined.

weaver-Burk) plots (1/Vo versus 1/S), which gave straightlines (correlation coefficient, 0.9 to 0.95), and from Eddie-Scatchard plots (V/S versus V).BZDH also exhibited a broad substrate specificity and was

able to transform methyl-, ethyl-, and chloro-substitutedbenzaldehydes. Activities with benzaldehyde and substi-tuted benzaldehydes were kinetically complex. The reduc-tion of NAD+ to NADH followed a curve with two slopes.During the first 10 s, activity was linear and maximal;afterwards, it became linear and submaximal. Figure 4shows the kinetics obtained with different concentrations ofbenzaldehydes. V0 with substituted benzaldehydes was be-tween 99 and 64% of that determined with unsubstitutedbenzaldehyde. The direct representation (V/S versus V) orthe representation of double-reciprocal plots (1/Vo versus1/S) did not fit a straight line, so that calculation of Kms withthe Eddie-Scatchard or Lineweaver-Burk equations was notfeasible. The concentration of benzaldehydes at which VOwas 50% of the maximum was inferred from the kinetic dataobtained at different concentrations of benzaldehydes. Forbenzaldehyde, p-ethylbenzaldehyde, and p-chlorobenzalde-hyde, the concentrations of substrate at which VO was 50%of the maximum were approximately 130, 45, and 88 p,M,respectively.

Construction of a TOL plasmid carrying a double mutationof the TOL meta-cleavage pathway for the metabolism ofp-ethylcatechol and further isolation of TO mutants that

0ECla

0

z

.288 pM,230 pM

PM

57 uM

50TIME (sec)

FIG. 4. Time course of BZDH activity. The incubation mixture'j ontained 200 ,mol of glycine (pH 9.4) per ml, 7.5 ,umol of NAD+per ml, the indicated concentrations of benzaldehyde, and 112 ,ug ofprotein of a cell extract prepared from P. putida(pWWO) grown onm-xylene.

metabolize p-ethyltoluene. Our results suggest that TO is themain bottleneck for the metabolism of certain substitutedtoluenes through the TOL upper pathway. Furthermore, theeventual synthesis of p-ethyl- and m-chlorobenzoate fromtheir corresponding toluenes would lead, if they are metab-olized through the TOL meta-cleavage pathway, to dead-endproducts, namely, p-ethyl- and m-chlorocatechol, respec-tively. Hence, alternative and/or modified pathways for themetabolism of these catechols are required for the mineral-ization of p-ethyl- and m-chlorotoluene.We previously constructed a modified TOL meta-cleavage

pathway that allowed Pseudomonas strains to grow onp-ethylbenzoate (25). This pathway contains two mutations.One of them, on the xyIE6 gene on the TOL plasmidpWWO-EB6, encodes a C2,3-O resistant to inactivation byits substrate, p-ethylcatechol, in contrast to the wild-typeenzyme. The second mutation is in the xylS gene, such thatthe mutant regulator encoded by the xylS4 allele recognizesas an effector p-ethylbenzoate, a benzoate analog that is notan effector for XylS. The xylS4 allele 'is on the broad-host-range Kmr plasmid pERD4. pWW0-EB6 is a 'self-transmissible plasmid whose frequency of transfer from onePseudomonas strain to another is about 10-1. It can mobilizepERD4 at a frequency of about 10-6 (25). Since P. putida2440 is a recA+ strain, we would expect a certain frequencyof xylS4 recombination with wild-type xylS on the TOLplasmid, so that both mutations come to lie on the TOLplasmid. Such a plasmid would allow Pseudomonas strainsto grow on p-ethylbenzoate while retaining kanamycin sen-sitivity.We mated P. putida PaW340 with a culture of P. putida

2440 which had contained pWWO-EB6 and pERD4 for morethan 1 year. P. putida PaW340 Kms transconjugants able togrow on p-ethylbenzoate were found at a frequency of about2 x 10-5. Ten transconjugants were examined to check forthe presence or absence of pERD4. pERD4 was found innone of them, suggesting that in the transconjugants xylS4was on the TOL plasmid. One of the clones carrying TOLplasmid pWWO-EB61 was thought to confirm that the plas-mid conferred the ability to grow on p-ethylbenzoate: P.putida PaW340(pWWO-EB61) was mated with P. putidaEEZ1, and 100% of the EEZ1 transconjugants carrying theTOL plasmid (selected for their ability to grow on 3-meth-ylbenzoate) grew on p-ethylbenzoate.

VOL. 171, 1989 6787

on January 22, 2020 by guesthttp://jb.asm

.org/D

ownloaded from

http://jb.asm.org/

6788 ABRIL ET AL.

P. putida PaW340(pWWO-EB61), as expected, could notgrow on p-ethyltoluene as the sole source of carbon andenergy. The strain was mutagenized with ethyl methanesul-fonate, and clones able to grow on minimal medium withp-ethyltoluene were selected. Mutants appeared with a fre-quency of about 5 x 10-8 per cell per generation. One suchmutant was chosen for further study. To determine whetherthe mutation was on the TOL plasmid or on the chromo-some, we performed matings between the mutant P. putidaPaW340 strain and the Rif' Smr P. putida EEZ3 strain. P.putida EEZ3 transconjugants able to grow on p-ethyltoluenewere isolated at a frequency of about 102, a frequencysimilar to that of transfer of the TOL plasmid, thus suggest-ing that the mutation was located on the TOL plasmid. Thenew TOL plasmid was called pWWO-EB62.TO encoded by pWWO-EB62 exhibited maximal activity

with m-xylene (0.5 nmol of 02 per min per unit of A6.),whereas with p-xylene, m-ethyltoluene, p-ethyltoluene, andtoluene, activities were 6.0, 3.1, 0.7, and 1.4 nmol of 02 permin per unit of A6., respectively. These activity rates,determined as increases in oxygen consumption by wholecells in response to the addition of the hydrocarbons, corre-lated with the growth rates of the strain on these hydrocar-bons. In fact, P. putida EEZ3(pWWO-EB62) exhibited dou-bling times of 3, 3.5, 10, 12, and 18 h on m-xylene, p-xylene,toluene, m-ethyltoluene, and p-ethyltoluene, respectively.

Isolation of B13 derivatives able to degrade m-chlorotolu-ene. Pseudomonas sp. strain B13 and its Rif derivativeEEZ4 are able to grow on m-chlorobenzoate but cannotdegrade m-chlorotoluene (4). We reasoned that acquisitionof the TOL plasmid by these strains could eventually lead tothe isolation of m-chlorotoluene degraders, although wewould expect low growth rates because of the low activity ofTOL-encoded TO on m-chlorotoluene.

Matings between Pseudomonas sp. strain EEZ4 and P.putida 2440(pWWO) were set up, and EEZ4 transconjugantsable to produce microcolonies on m-chlorotoluene wereisolated. Nineteen of 20 microcolonies produced nonuniformgrowth when streaked on minimal medium with m-chloro-toluene as the sole source of carbon. However, one micro-colony produced homogeneous growth on minimal mediumwith m-chlorotoluene, and this clone was called Pseudomo-nas sp. strain EEZ5. In liquid minimal medium EEZ5 grewslowly on m-chlorotoluene but failed to grow on toluene orm-xylene. m-Chlorotoluene-grown EEZ5 cells exhibitednegligible oxygen consumption in response to m-chlorotolu-ene, confirming that TO is the limiting step for the degrada-tion of m-chlorotoluene by EEZ5. BADH and BZDH activ-ities on chlorinated substrates were, however, confirmed.

DISCUSSION

In the present study we analyzed the effector specificity ofXylR, the regulator of the upper pathway, and the substratespecificity of the upper pathway enzymes. This informationwas then used in the design of a laboratory strategy todevelop TOL plasmid pathways which would allow themetabolism of certain alkyltoluenes.The effector specificity of the xylR gene product, the XylR

protein, was analyzed by measuring expression from Pu andPm fused transcriptionally to a promoterless lacZ gene. Thedata obtained suggest that the XylR protein exhibits a verybroad spectrum of effectors, as it is able to recognize a widevariety of toluenes, benzyl alcohols, and benzaldehydes.The data support the following conclusions regarding XylR-effector interactions. (i) A -CH3 or -CH2OH group at

carbon 1 of the aromatic ring is essential for productivecontacts to be established between effectors and the XylRprotein, leading to activation of the latter. (ii) Alkyl- andchloro-substitutions are permissible at other ring carbons.(iii) The XylR protein effector pocket seems to be symmet-rical in the sense that it tolerates toluenes which are disub-stituted at carbon 2 or 3 and carbon 5 or 6. (iv) A -CHOgroup at carbon 1 of the aromatic ring is permissible if, andonly if, a chlorine atom is present at carbon 4.Hydrocarbon- and benzyl alcohol-activated XylR protein

likewise stimulated expression from Pu, leading to an ap-proximately 15-fold induction with m-xylene and m-chloro-toluene versus a 10-fold induction with 5 mM m-methyl-benzyl alcohol and 5 mM m-chlorobenzyl alcohol. Incontrast, levels of induction from Pm were higher when thehydrocarbon rather than the corresponding benzyl alcoholwas the effector: 13-, 12-, and 7-fold increases in inductionwere obtained with toluene, m-xylene, and p-chlorotoluene,respectively, versus 2-fold, 2.7-fold, and no increases with 5mM benzyl alcohol, m-methylbenzyl alcohol, and p-chlo-robenzyl alcohol, respectively. These results might reflectdifferent affinities of benzyl alcohol-activated XylR proteinfor its binding site(s) in the Pu region and the Ps region. Infact, we have identified activator sequences for XylR in Puand Ps. While these sequences are 70% homologous, thevariations may be responsible for the differences observed inthe stimulation of transcription (M.-A. Abril, A. Holtel, S.Marques, and J. L. Ramos, unpublished data). The system-atic analysis of XylR effectors allowed us to record for thefirst time activation of the protein by means of an aromaticbenzaldehyde, p-chlorobenzaldehyde. Further studies willbe required to establish the nature of the interactions be-tween XylR and its different effectors.The relatively extensive body of data on the organization

and regulation of TOL catabolic pathways has yet to be fullyexplained, however, in terms of pathway biochemistry,owing in many cases to the paucity of substrates or theinstability of the enzymes. Hence, the substrate specificitiesof the TOL upper pathway enzymes were also investigated.BADH and BZDH exhibited very broad substrate specifici-ties and were able to transform alkyl- and chloro-substitutedsubstrates. The specific maximal rates for BZDH were 4 to10 times higher than were those for BADH. In contrast,BADH exhibited 5- to 10-fold higher affinities for its sub-strates than did BZDH for its corresponding ones. TheKm/Vma. ratios for both enzymes with substituted substrateswere equivalent, suggesting that fluxes of substituted benzylalcohols to substituted benzoates may well occur without theaccumulation of pathway intermediates.

Regarding the complex kinetics observed with BZDH, wecannot rule out the presence of more than one enzyme in ourcell extracts. However, it is worth noting that P. putida 2440lacking the TOL plasmid did not exhibit BZDH activitywhen grown on glucose in the presence of toluene. Thisresult suggests that what we were measuring in the cellextracts of P. putida 2440(pWWO) grown on toluene wasprobably the TOL-encoded enzyme(s) only. Further analy-ses with purified enzymes will be required to clarify thesekinetics. Pairs of isofunctional BZDHs associated with themandalate pathway of P. putida and Aspergillus niger havebeen described, although it is not clear whether the separateenzymes have different functions (5).Although TO exhibited a relatively relaxed substrate spec-

ificity (18, 20, 31), it clearly acted as the limiting step for thefluxes of aromatic compounds through the TOL upper path-way, given that it showed the narrowest substrate specificity

J. BACTERIOL.

on January 22, 2020 by guesthttp://jb.asm

.org/D

ownloaded from

http://jb.asm.org/

LABORATORY EVOLUTION OF TOL CATABOLIC PATHWAYS

of all the pathway enzymes. The catabolism of p-ethyl- andm- and p-chlorotoluene through the TOL upper pathway islimited by TO. However, it should be noted that if thesetoluenes were metabolized through the upper pathway, theywould not allow cell growth because of the buildup of thedead-end products. Thus, it seems that in vivo TOL upperand meta-cleavage pathways coevolve to determine whichsubstrates they metabolize. In contrast, catabolic pathwayscan be independently evolved in the laboratory either byexpansion of the substrate-effector specificities of the bot-tleneck steps or by transfer or cloning of specific pathwaysegments to new hosts in which existing pathways arehorizontally or vertically evolved (24).Both genetic and biochemical characterizations of the

TOL upper pathway were considered in designing laboratorystrategies for the evolution of this pathway. When this workwas initiated we had previously manipulated the TOL plas-mid meta-cleavage pathway to allow the metabolism ofp-ethylbenzoate (25). We have now used this modifiedpathway to expand the range of alkyltoluenes mineralizedthrough the TOL catabolic pathways after isolation of oneTO mutant able to oxidize p-ethyltoluene to p-ethylbenzylalcohol, which is further transformed into p-ethylbenzoatethrough the action of TOL plasmid-encoded BADH andBZDH. The mutant TO enzyme, in addition to gaining theability to transform p-ethyltoluene, retained the ability,although to a somewhat diminished extent, to oxidize hydro-carbon substrates that the wild-type TO enzyme attacked.The constructed p-ethyltoluene-degrading strain had three

mutations on TOL plasmid pWWO-EB62. xylE6 encodes aC2,3-0 resistant to inactivation by its substrate p-ethylcate-chol, in contrast to the wild-type enzyme. xyIS4 is a mutantregulator whose gene product, in contrast to the wild-typexylS gene product, switches on the meta-cleavage pathwayin the presence of concentrations of p-ethylbenzoate as lowas 5 ,uM (J. L. Ramos, C. Michau, F. Rojo, D. Dwyer, andK. N. Timmis, J. Mol. Biol., in press). The introduction ofthis mutation on the TOL plasmid is required, as p-ethyltol-uene elicits a low induction of the TOL meta-cleavagepathway through the cascade regulatory system. These twomutations permit the metabolism of p-ethylbenzoate and,together with a third mutation in the genes encoding TO,enable Pseudomonas strains to mineralize p-ethyltoluenethrough the TOL plasmid catabolic pathways. The threeindependent mutations introduced in the TOL plasmid aroseat frequencies of 10-8 and 10-9 per cell per generation,making it unlikely that a plasmid similar to pWWO-EB62(appearing at a frequency of about 10-25) could be isolatedfrom nature.On the other hand, once a pathway segment has been

genetically and biochemically characterized, it can be ex-ploited to construct new pathways by combining it withother pathway segments. The strategy of combining criticalgenes from separate pathways to produce a hybrid pathwaywas originally tested by transfer of the TOL plasmid pWWOinto Pseudomonas sp. strain B13 (28). B13 grew on m-chlorobenzoate but not on p-chlorobenzoate, as shown bythe narrow substrate specificity of the first pathway enzyme(27). In contrast, the toluate 1,2-dioxygenase encoded by theTOL plasmid was able to oxidize m- and p-methyl- and m-and p-chlorobenzoate (27). Pseudomonas sp. strainB13(pWWO) able to grow on p-chlorobenzoate was eventu-ally isolated. Because of rearrangements in the TOL plas-mid, the B13 derivatives able to degrade p-chlorobenzoatehad lost TOL upper pathway genes and could not grow ontoluene (17, 26). In the present study, we reasoned that it

should be possible to isolate m-chlorotoluene degraders bytransferring the TOL plasmid into B13. These clones wereexpected to grow slowly on such a substrate because of thelow rates of oxidation of m-chlorotoluene by TO. One suchm-chlorotoluene degrader was in fact isolated in the courseof this study, and we are currently searching for mutantsable to degrade m-chlorotoluene more rapidly. The muta-tions are expected to appear in the TO genes.

Directed evolution of existing pathways requires a de-tailed knowledge of the biochemistry and genetics of thepathways to design appropriate laboratory strategies. TheTOL upper pathway had a regulator which was very flexiblefor effectors, together with enzymes with very broad sub-strate specificities. In light of the fact that the pathway genesare born on a transposon contained in a self-transmissibleplasmid (1, 16), this pathway is a key element in the rapidevolution of catabolic routes not only in Pseudomonasstrains, the natural hosts of the TOL plasmid, but also inmany other bacteria.

ACKNOWLEDGMENTSWe are most grateful to Marcel Wubbolts for providing plasmid

pWA21 before its publication. We also thank Ray Dixon and ShigeHarayama for strains and plasmids. Juan-Pedro Donaire kindlyprovided facilities and advice for the determination of benzylalcohols. Estrella Duque and Manuel Martinez did the artwork.M.-A.A. and C.M. were supported by fellowships from the

Spanish Ministry of Education and the Andalusian Regional Gov-ernment, respectively. Work in Spain was supported by grantB023/87 from the CICYT and by a Biotechnology Action Pro-gramme Grant (BAP.0411.E) from the European Economic Com-munity.

LITERATURE CITED1. Benson, S., and J. Shapiro. 1978. TOL is a broad-host-range

plasmid. J. Bacteriol. 135:278-280.2. Bradford, M. M. 1976. A rapid and sensitive method for the

quantitation of microgram quantities of protein utilizing theprinciple of protein-dye binding. Anal. Biochem. 72:248-254.

3. Dixon, R. 1986. The xylABC promoter from the Pseudomonasputida TOL plasmid is activated by nitrogen regulatory genes inEscherichia coli. Mol. Gen. Genet. 206:129-136.

4. Dorn, E., M. Hellwig, W. Reineke, and H.-J. Knackmuss. 1974.Isolation and characterization of a 3-chlorobenzoate degradingpseudomonad. Arch. Microbiol. 99:61-70.

5. Fewson, C. A. 1988. Microbial metabolism of mandalate: amicrocosm of diversity. FEMS Microbiol. Rev. 54:85-110.

6. Franklin, F. C. H., M. Bagdasarian, M. M. Bagdasarian, andK. N. Timmis. 1981. Molecular and functional analysis of theTOL plasmid pWWO from Pseudomonas putida and cloning ofgenes for the entire regulated aromatic ring meta-cleavagepathway. Proc. Natl. Acad. Sci. USA 78:7458-7462.

7. Franklin, F. C. H., P. R. Lehrbach, R. Lurz, B. Rueckert, M.Bagdasarian, and K. N. Timmis. 1983. Localization and func-tional analysis of transposon mutations in regulatory genes ofthe TOL catabolic plasmids. J. Bacteriol. 154:676-685.

8. Franklin, F. C. H., and P. A. Williams. 1980. Construction of apartial diploid for the degradative pathway encoded by the TOLplasmid pWWO from Pseudomonas putida mt-2: evidence forthe positive nature of the regulation by the xylR gene. Mol. Gen.Genet. 177:321-328.

9. Gibson, D. T., and V. Subramanian. 1984. Microbial degrada-tion of aromatic hydrocarbons, p. 181-252. In D. T. Gibson(ed.), Microbial degradation of organic compounds. MarcelDekker, Inc., New York.

10. Harayama, S., P. R. Lehrbach, and K. N. Timmis. 1984.Transposon mutagenesis analysis of the meta-cleavage pathwayoperon genes of the TOL plasmid of Pseudomonas putida mt-2.J. Bacteriol. 160:251-255.

11. Harayama, S., M. Rekik, and K. N. Timmis. 1986. Genetic

6789VOL. 171, 1989

on January 22, 2020 by guesthttp://jb.asm

.org/D

ownloaded from

http://jb.asm.org/

6790 ABRIL ET AL.

analysis of a relaxed substrate specificity aromatic ring dioxy-genase, toluate 1,2-dioxygenase, encoded by TOL plasmidpWWO of Pseudomonas putida. Mol. Gen. Genet. 202:226-234.

12. Harayama, S., and K. N. TimmIs. 1989. Catabolism of aromatichydrocarbons by Pseudomonas, p. 151-174. In D. A. Hopwoodand K. I. Chater (ed.), Genetics ofbacterial diversity. AcademicPress, Inc. (London), Ltd., London.

13. Inouye, S., A. Nakazawa, and T. Nakazawa. 1981. Molecularcloning of xylS gene of the TOL plasmid: evidence for positiveregulation of the xylDEGF operon by xylS. J. Bacteriol. 148:413-418.

14. Inouye, S., A. Nakazawa, and T. Nakazawa. 1983. Molecularcloning of regulatory gene xylR and operator-promoter regionsof the xylABC and xylDEGF operons of the TOL plasmid. J.Bacteriol. 155:1192-1199.

15. Inouye, S., A. Nakazawa, and T. Nakazawa. 1987. Expression ofthe regulatory gene xylS on the TOL plasmid is positivelycontrolled by the xylR gene product. Proc. Natl. Acad. Sci.USA 84:5182-5186.

16. Jacoby, G. A., J. E. Rogers, A. E. Jacob, and R. W. Hedges.1979. Transposition of Pseudomonas toluene degrading genesand expression in E. coli. Nature (London) 234:220-221.

17. Jeenes, D. J., W. Reineke, H.-J. Knackmuss, and P. A. Williams.1982. TOL plasmid pWWO in constructed halobenzoate-de-grading Pseudomonas strains: enzyme regulation and DNAstructure. J. Bacteriol. 150:180-187.

18. Kunz, D. A., and P. J. Chapman. 1981. Catabolism of pseudo-cumene and 3-ethyltoluene by Pseudomonas putida (arvilla)mt-2: evidence for new functions of the TOL (pWW0) plasmid.J. Bacteriol. 146:179-191.

19. Maniatis, T., E. F. Fritsch, and J. Sambrook. 1982. Molecularcloning: a laboratory manual. Cold Spring Harbor Laboratory,Cold Spring Harbor, N.Y.

20. Mermod, N., S. Harayama, and K. N. Timmis. 1986. Novelroute for bacterial production of indigo. Bio/Technology 4:321-324.

21. Mermod, N., J. L. Ramos, P. R. Lehrbach, and K. N. Timmis.

1986. Vector for regulated expression of cloned genes in a widerange of gram-negative bacteria. J. Bacteriol. 167:447-454.

22. Ramos, J. L., N. Mermod, and K. N. Timmis. 1987. Regulatorycircuits controlling transcription of TOL plasmid operon encod-ing meta-cleavage. Mol. Microbiol. 1:297-300.

23. Ramos, J. L., A. Stolz, W. Reineke, and K. N. Timmis. 1986.Altered effector specificities in regulators of gene expression:TOL plasmid xylS mutants and their use to engineer expansionof the range of aromatics degraded by bacteria. Proc. Natl.Acad. Sci. USA 83:8467-8471.

24. Ramos, J. L., and K. N. Timmis. 1987. Experimental evolutionof catabolic pathways of bacteria. Microbiol. Sci. 4:228-237.

25. Ramos, J. L., A. Wasserfailen, K. Rose, and K. N. Timmis. 1987.Redesigning metabolic routes: manipulation of TOL plasmidpathway for catabolism of alkylbenzoates. Science 235:593-596.

26. Reineke, W., D. J. Jeenes, P. A. Williams, and H.-J. Knackmuss.1982. TOL plasmid pWWO in constructed halobenzoate-de-grading Pseudomonas strains: prevention of meta pathway. J.Bacteriol. 150:195-201.

27. Reineke, W., and H.-J. Knackmuss. 1978. Chemical structureand biodegradability of halogenated aromatic compounds. Sub-stituent effect on 1,2-dioxygenation of benzoic acid. Biochim.Biophys. Acta 542:412-423.

28. Reineke, W., and H.-J. Knackmuss. 1979. Construction ofhaloaromatic-utilizing bacteria. Nature (London) 277:385-386.

29. Vieira, J., and J. Messing. 1982. The pUC plasmids, an M13mp7-derived system for insertion mutagenesis and sequencingwith synthetic universal primers. Gene 19:259-268.

30. Worsey, M. J., F. C. H. Franklin, and P. A. Williams. 1978.Regulation of the degradative pathway enzymes encoded for bythe TOL plasmid (pWW0) from Pseudomonas putida mt-2. J.Bacteriol. 134:757-764.

31. Worsey, M. J., and P. A. Williams. 1975. Metabolism of tolueneand xylenes by Pseudomonas ([sic]putida (arvilla) mt-2: evi-dence for a new function of the TOL plasmid. J. Bacteriol.124:7-13.

J. BACTERIOL.

on January 22, 2020 by guesthttp://jb.asm

.org/D

ownloaded from

http://jb.asm.org/

ERRATAMutagenesis of the Gene Encoding Cytochrome c550 of Paracoccusdenitrificans and Analysis of the Resultant Physiological Effects

ROB J. M. VAN SPANNING, CORRY WANSELL, NELLIE HARMS, L. FRED OLTMANN, ANDADRIAN H. STOUTHAMER

Department of Microbiology, Biological Laboratory, Vrije Universiteit, P.O. Box 7161,1007 MC Amsterdam, The Netherlands

Volume 172, no. 2, p. 990, Fig. 2: The amino acids under nucleotides 388 to 423 were transposed with those undernucleotides 424 to 474; the correct amino acid sequence of cytochrome c550 is shown below.

240 255 270 285ATG AAG ATC AGC ATC TAT GCC ACT CTC GCC GCC ATC ACC CTC GCC CTG CCCMet Lys Ile Ser Ile Tyr Ala Thr Leu Ala Ala Ile Thr Leu Ala Leu Pro

300 315 330GCT GCG GCC CAG GAT GGC GAC GCC GCC AAA GGC GAG AAA GAA TTC AAC AAGAla Ala Ala*Gln Asp Gly Asp Ala Ala Lys Gly Glu Lys Glu Phe Asn Lys

345 360 375TGC AAG GCT TGC CAC ATG ATC CAG GCG CCG GAC GGC ACC GAC ATC ATC AAGCys Lys Ala Cys His Met Ile Gln Ala Pro Asp Gly Thr Asp Ile Ile Lys

390 405 420 435GGC GGC AAG ACC GGG CCC AAC CTT TAC GGC GTC GTC GGC CGC AAG ATC GCCGly Gly Lys Thr Gly Pro Asn Leu Tyr Gly Val Val Gly Arg Lys Ile Ala

450 465 XhoI 480TCG GAG GAG GGC TTC AAA TAC GGC GAA GGC ATC CTC GAG GTC GCC GAA AAGSer Glu Glu Gly Phe Lys Tyr Gly Glu Gly Ile Leu Glu Val Ala Glu Lys

495 510 525 540AAC CCC GAC CTG ACC TGG ACC GAG GCC GAC CTG ATC GAA TAC GTC ACC GACAsn Pro Asp Leu Thr Trp Thr Glu Ala Asp Leu Ile Glu Tyr Val Thr Asp

555 570 585CCC AAG CCC TGG CTG GTC AAG ATG ACC GAC GAC AAG GGC GCC AAG ACC AAGPro Lys Pro Trp Leu Val Lys Met Thr Asp Asp Lys Gly Ala Lys Thr Lys

600 615 630ATG ACC TTC AAG ATG GGC AAG AAC CAG GCC GAC GTG GTG GCC TTC CTG GCCMet Thr Phe Lys Met Gly Lys Asn Gln Ala Asp Val Val Ala Phe Leu Ala

645 660 675 690CAG AAC TCG CCC GAT GCG GGC GGC GAC GGC GAG GCT GCG GCC GAG GGC GAAGln Asn Ser Pro Asp Ala Gly Gly Asp Gly Glu Ala Ala Ala Glu Gly Glu

699TCG AACSer Asn

Regulator and Enzyme Specificities of the TOL Plasmid-EncodedUpper Pathway for Degradation of Aromatic Hydrocarbons and

Expansion of the Substrate Range of the PathwayMARIA-ANGELES ABRIL, CARMEN MICHAN, KENNETH N. TIMMIS, AND JUAN L. RAMOS

Estacion Experimental del Zaidin, Consejo Superior de Investigaciones Cientificas, Apto 419, 18080 Granada, Spain;GBF, D-3300 Braunschweig, Federal Republic of Germany; and Department of Medical Biochemistry,

CMU, University of Geneva, Geneva, Switzerland

Volume 171, no. 12: p. 6782, abstract, line 8, and column 1, line 9; p. 6784, column 1, lines 45 and 46 of Results; and p. 6785,Table 2, line 19: "1,3,4-trimethylbenzene" should read "1,2,4-trimethylbenzene."Page 6788, column 1, line 17: "0.5 nmol" should read "10.5 nmol."

3534