assessing the role of pseudomonas aeruginosa surface ... · extracted from pg201 by using the...

APPLIED AND ENVIRONMENTAL MICROBIOLOGY, May 2002, p. 2509–2518 Vol. 68, No. 50099-2240/02/$04.00�0 DOI: 10.1128/AEM.68.5.2509–2518.2002Copyright © 2002, American Society for Microbiology. All Rights Reserved.

Assessing the Role of Pseudomonas aeruginosa Surface-Active GeneExpression in Hexadecane Biodegradation in Sand

P. A. Holden,1* M. G. LaMontagne,1 A. K. Bruce,1 W. G. Miller,2 and S. E. Lindow3

Donald Bren School of Environmental Science & Management, University of California, Santa Barbara,1 USDA ResearchAgricultural Service, Albany,2 and Department of Plant and Microbial Biology, University of California, Berkeley,3 California

Received 19 October 2001/Accepted 24 January 2002

Low pollutant substrate bioavailability limits hydrocarbon biodegradation in soils. Bacterially producedsurface-active compounds, such as rhamnolipid biosurfactant and the PA bioemulsifying protein produced byPseudomonas aeruginosa, can improve bioavailability and biodegradation in liquid culture, but their productionand roles in soils are unknown. In this study, we asked if the genes for surface-active compounds are expressedin unsaturated porous media contaminated with hexadecane. Furthermore, if expression does occur, is bio-degradation enhanced? To detect expression of genes for surface-active compounds, we fused the gfp reportergene either to the promoter region of pra, which encodes for the emulsifying PA protein, or to the promoter ofthe transcriptional activator rhlR. We assessed green fluorescent protein (GFP) production conferred by thesegene fusions in P. aeruginosa PG201. GFP was produced in sand culture, indicating that the rhlR and pra genesare both transcribed in unsaturated porous media. Confocal laser scanning microscopy of liquid dropsrevealed that gfp expression was localized at the hexadecane-water interface. Wild-type PG201 and its mutantsthat are deficient in either PA protein, rhamnolipid synthesis, or both were studied to determine if the geneticpotential to make surface-active compounds confers an advantage to P. aeruginosa biodegrading hexadecane insand. Hexadecane depletion rates and carbon utilization efficiency in sand culture were the same for wild-typeand mutant strains, i.e., whether PG201 was proficient or deficient in surfactant or emulsifier production.Environmental scanning electron microscopy revealed that colonization of sand grains was sparse, with cellsin small monolayer clusters instead of multilayered biofilms. Our findings suggest that P. aeruginosa likelyproduces surface-active compounds in sand culture. However, the ability to produce surface-active compoundsdid not enhance biodegradation in sand culture because well-distributed cells and well-distributed hexadecanefavored direct contact to hexadecane for most cells. In contrast, surface-active compounds enable bacteria inliquid culture to adhere to the hexadecane-water interface when they otherwise would not, and thus productionof surface-active compounds is an advantage for hexadecane biodegradation in well-dispersed liquid systems.

Natural biological attenuation is the present strategy forremediating greater than 25% of soil sites contaminated withpetroleum hydrocarbons (33). Because of insufficient knowl-edge about in situ controls on bioavailability and biodegrada-tion of low-solubility hydrocarbon pollutants in soil, it is diffi-cult to predict cleanup end points (33) and to selectappropriate remediation strategies. Reported mechanisms forbacterial metabolism of sparingly soluble hydrocarbons are (i)dissolution and diffusion of dissolved hydrocarbons to cellswith uptake via active or passive transmembrane transport, (ii)invagination of hydrocarbon non-aqueous-phase liquid(NAPL) into cells (9) with subsequent intracellular metabo-lism of the hydrocarbon inclusions, and (iii) bacterial produc-tion of surface-active compounds such as surfactants (biosur-factants) and emulsifiers (bioemulsifiers) that increase thelocal pseudosolubility of hydrocarbons and thus improve masstransfer to biodegrading bacteria (3, 20, 29). Compared withsynthetic surfactants, which can be either stimulatory below thecritical micelle concentration (23) or inhibitory above the crit-ical micelle concentration (28), biosurfactants reduce the in-terfacial tension in two-phase NAPL-water mixtures, while be-

ing nontoxic and fully biodegradable (40). Biosurfactants havebeen studied largely in liquid culture with a common soil bac-terium, Pseudomonas aeruginosa. Many other soil bacteria pro-duce surface-active compounds, but little is known regardingthe production of biosurfactants in soils (29). More knowledgeregarding in situ production of surface-active compoundswould enhance our understanding of natural bioattenuation.Furthermore, if surface-active compounds are produced in soiland could be stimulated in situ, engineered bioremediationcould be enhanced.

P. aeruginosa produces rhamnolipids that improve biodeg-radation of various hydrocarbons in liquid culture (17, 27,50–52) when either N (11, 12, 31), P (11, 32), or Fe (11) islimiting and with a variety of carbon substrates, including glu-cose (11, 12) and hexadecane (17, 22, 44). Rhamnolipid syn-thesis in P. aeruginosa is encoded by genes in the rhl operon(35) which are positively regulated by rhlR, a luxR homolog(36). P. aeruginosa also produces a bioemulsifying protein, PA(18), which is encoded by the pra gene (14); PA is producedunder similar conditions of nutrient limitation.

Rhamnolipids purified from liquid culture, when amendedto soil, improve residual hydrocarbon recovery (5, 6, 15, 48)and biodegradation (16, 24). However, if soil bacteria producerhamnolipids in situ, the costs associated with laboratory pro-duction and soil amendment could be deferred. To date, thereis little direct evidence that either biosurfactants or bioemul-

* Corresponding author. Mailing address: Donald Bren School ofEnvironmental Science & Management, 4670 Physical Sciences Build-ing North, University of California, Santa Barbara, CA 93106. Phone:(805) 893-3195. Fax: (805) 893-7612. E-mail: [email protected].

2509

on August 5, 2020 by guest

http://aem.asm

.org/D

ownloaded from

http://aem.asm.org/

sifying proteins are produced in situ in hydrocarbon-contami-nated porous media, e.g., soil. In fact, it has been suggestedthat microbial surfactants in liquid culture may be simply cellsurface components that are dispersed into solution followingcell lysis or excreted during unbalanced growth (13, 20, 40, 41).The dispersive conditions of well-mixed liquid cultures do notoccur in soil, so these compounds may play a different, andperhaps cell surface-limited, role in interactions of cells withinsoluble hydrocarbons. In this study, rather than infer fromliquid culture experiments the possible production of biosur-factants in soil, we used expression vectors to determinewhether P. aeruginosa transcribes genes for surface-active com-pounds when cultured in unsaturated porous media. Further-more, we asked whether the ability to produce surface-activecompounds at native levels in unsaturated porous media en-hances hexadecane biodegradation. Our results provide directevidence that genes for surface-active compounds are tran-scribed by P. aeruginosa in porous media. However, whilegenes that encode or regulate surface-active compounds wereexpressed in P. aeruginosa in unsaturated porous media, geneexpression was not tied to measurably improved hexadecanebiodegradation rates or end points.

MATERIALS AND METHODS

Chemicals, media, strains, and plasmids. The strains and plasmids used in thisstudy are summarized in Table 1. All strains were maintained at �80°C andcultured on Luria-Bertani agar (LA) for 24 h prior to use. When used, kanamy-cin, ampicillin, and tetracycline (TET) were added at final concentrations of 50,150, and 30 �g/ml, respectively.

Chemicals were purchased from Sigma Chemical Co. (St. Louis, Mo.). Thebasal medium was either half-strength 21C (19) or a variation that replacedammonium N with equimolar nitrate N that we named CW21C. Filter-sterilized(0.2-�m-pore-size filter) hexadecane was provided as the sole carbon source. Allother hydrocarbons used in green fluorescent protein (GFP) induction studieswere filter sterilized. The stock solutions for the liquid-culture GFP inductionstudies were filter-sterilized glucose (0.33 g/ml), NH4Cl (50 mg/ml), and NaNO3

(79 mg/ml). The bacterial suspension used in drop slides was prepared in 1 mM

phosphate buffer (P buffer) consisting of 4 ml of 0.2 M K2HPO4 and 1 ml of 0.2M KH2PO4 per liter (pH 6.8). Restriction enzymes were purchased from Pro-mega (Madison, Wis.), Roche Molecular Biochemicals (Indianapolis, Ind.), orStratagene (La Jolla, Calif.). DNA sequencing as well as synthesis of the inversePCR and pra probe oligonucleotides were performed at the University of Cali-fornia, Berkeley (http://mcb.berkeley.edu/barker/dnaseq). All other oligonucle-otides were synthesized by Operon Technologies (Alameda, Calif.).

Acquisition of the pra promoter fragment by inverse PCR. Although a shortregion upstream of the structural gene for the PA protein (pra) in PG201 hadbeen characterized (14), the complete promoter element for pra, including allputative regulatory elements, was not available. Therefore, inverse PCR (34) wasused to generate a complete pra promoter-containing fragment. The pra genecontains several unique restriction sites, including an NaeI site. To determine ifthis NaeI site would be suitable for inverse PCR, total chromosomal DNA wasextracted from PG201 by using the QIAamp system (Qiagen, Santa Clarita,Calif.) and digested with NaeI. The digest was Southern blotted and probed witha digoxigenin-labeled PCR product. The probe, labeled with the Genius kit(Roche Molecular Biochemicals), contains a portion of the pra gene amplifiedfrom upstream of the NaeI site with the oligonucleotide primers PROBE-5�(5�-CGG-ACG-GGA-AGG-CAC-AAG-GTG-GTC-G-3�) and PROBE-3� (5�-CCG-GAT-TAC-GGG-TTG-ACT-ACG-3�). We confirmed by positive hybrid-ization that a 500-bp NaeI fragment contained part of the pra coding region.

NaeI-digested PG201chromosomal DNA was circularized in a 20-�l ligationreaction mixture (12 h, 18°C) containing 18 �l of 10� ligation buffer, 1 �l ofDNA, and 1 �l of ligase (34). The circularized DNA was purified and concen-trated by ethanol-sodium acetate precipitation prior to use as template DNA inthe inverse PCR (34). The pra 5� flanking region, containing the putative com-plete pra promoter, was amplified using the oligonucleotide primers INVERSE-5�(5�-GCG-GAT-CCT-GCC-TGA-GCG-TTT-CGT-CC-3�; bp �49 to �63) andINVERSE-3� (5�-CCG-GAT-CCG-ATT-TCA-TTG-TTT-TGT-ATC-C-3�; bp�12 to �8), where bases in boldface are identical to sequences in the pra regionand numbers are relative to the start (�1) of pra. Inverse PCR was conducted(1.5 mM final MgCl2 concentration) in a DNA Thermal Cycler (Perkin-ElmerCetus) using 35 cycles of 94°C for 1 min, 50°C for 2 min, and 72°C for 3 min. ThePCR product was cloned into pCR2.1 (Invitrogen, Carlsbad, Calif.) and trans-formed into DH5�. Plasmid DNA was extracted from colonies containing theinsert (as determined by blue-white screening) and digested with EcoRI toestablish which clones contained the ca. 500-bp NaeI pra fragment. A 500-bpfragment was recovered from one selected clone and sequenced (Fig. 1) toconfirm that it contained the 5� end of pra in addition to a significant regionupstream of the structural gene. The sequence of the upstream region indicatedthat the INVERSE-5� primer hybridized to a secondary location, resulting in asmaller-than-expected PCR product. Thus, because the INVERSE-5� primerhybridized downstream of the NaeI site and the INVERSE-3� hybridized up-stream, the amplified product contains no NaeI site. The EcoRI-ended genefragment containing the putative pra promoter was cloned into the EcoRI site ofpUC1813 (42). In one orientation (pUC1813PRA), the promoter fragment con-tains upstream and downstream restriction sites for HindIII and KpnI, respec-tively.

Plasmid construction and transformation. The regulatory element of eitherpra or rhlR was cloned upstream of gfp in the promoter probe vectorpPROBE-TT (30); these fusions were transformed into DH5�. To construct thePpra-GFP fusion, pPROBE-TT was digested with HindIII and KpnI and theputative pra promoter fragment was ligated into the multicloning site by usingstandard protocols; the ligation product, plasmid pHX1, was transformed intocompetent DH5� (Gibco, Rockville, Md.). Transformants were selected fromcolonies that developed on LA amended with TET (LA-TET). The rhlR pro-moter region (1.1 kb) was amplified from PG201 DNA using primers FORRHL(5�-CAG-AAG-CTT-TTC-CTG-CAA-TCC-GAC-3�) and REVRHL (5�-CCG-GTA-CCT-GGT-CGA-TGT-GAA-A-3�), which contain restriction sites forHindIII (forward) and KpnI (reverse), according to standard protocols. The rhlRpromoter-containing fragment was cloned into pPROBE-TT as before to pro-duce pRhlPGB. Based on prior work with Pseudomonas species (7), gfp-contain-ing plasmids were not expected to significantly burden cells.

Late-exponential-phase, Luria-Bertani broth (LB)-grown PG201 cells weremade competent for transformation by treatment with 100 mM MgCl2 (30 min,4°C) followed by resuspension in 100 mM CaCl2. Competent cells were com-bined with purified plasmid DNA (pHX1, pRhlPGB, or pPROBE-TT) and thenincubated (4°C, 30 min) and heat shocked (2 min, 42°C) prior to addition of freshLB and a second incubation (60 min, 30°C). Cells were spread at several dilutionson LA-TET, and the plates were incubated (32°C, 48 h). Transformants wereconfirmed to harbor the gfp-containing plasmid by colony PCR of LA-TET-grown cells with previously published gfp primers (47). The stabilities of the

TABLE 1. Strains and plasmids used in this study

Strain or plasmid Descriptiona Reference(s)

StrainsP. aeruginosa

PG201 Formerly known as DSM2659 11, 14, 35PG201pra� PG201 pra 14PG201rhlA� Rhamnolipid-negative PG201

affected in rhlA gene; alsoknown as UO299

14, 35

PG201pra�rhlA� UO299 pra 14

E. coliDH5� 42NT DH5�(pUC1813PRA) This studyKtKAN DH5�(pKAN) 25

PlasmidspUC1813 26pUC1813PRA pUC1813 carrying putative pra

promoter as an EcoRI frag-ment flanked by HindIII andKpnI sites

This study

pKAN Kmr; 131-bp Tn5 nptII promoterfused to gfp

25

pPROBE-TT Tcr; gfp 30pHX1 Tcr; pra promoter fused to gfp This studypRhlPGB Tcr; rhlR promoter fused to gfp This study

a Km, kanamycin; Tc, tetracycline.

2510 HOLDEN ET AL. APPL. ENVIRON. MICROBIOL.


http://aem.asm

.org/D

ownloaded from

http://aem.asm.org/

pHX1 and pRhlPGB plasmids in PG201 were determined by cultivating theengineered strains overnight (150 rpm, 30°C) in LB-TET, resuspending cells infresh LB, then subsampling over several days and spread plating samples ontoLA-TET and LA.

Liquid and sand culturing. All culturing was at 30°C. Liquid culturing wasperformed in shaking, gas-tight, 500-ml triple-baffled Nephelo flasks (Bellco,Vineland, N.J.) with Teflon-lined top caps and 18-mm Mininert (Valco Instru-ments, Baton Rouge, La.) side caps for gas-tight sampling. Liquid cultures wereprepared with 90 ml of basal medium (either half-strength 21C or CW21C), 10ml of Hutner’s mineral solution (19), 200 �l of hexadecane, and 100 �l of inoculaprepared by suspending cells (from overnight solid-medium culture) to similaroptical densities (ODs) in basal medium. Liquid cultures were inoculated with100 �l of cell suspension. At various times during cultivation, 1 ml was sampledfor liquid tensiometry. Population density was monitored by culture absorbanceat 660 nm.

Sand cultures were prepared using 100 g of autoclaved (three times, 30 min)clean quartz sand with grains that were previously classified as being between 212and 180 �m in diameter (sizes 80 and 70 U.S. standard sieves, respectively).Culture bottles were 250-ml amber glass (Boston rounds; Fisher Scientific, Tus-tin, Calif.). Sand cultures were inoculated with a dilute 3.4-ml bacterial suspen-sion, which resulted in an unsaturated porous medium. The inocula for the sandcultures were similar concentrations of washed (two times in basal medium)exponential-phase cells grown in LB. Uninoculated controls and inoculated sandcultures were mixed by rolling the reactor bottles for 30 min on a roller mill andthen incubated overnight (30°C) prior to addition of hydrocarbon substrates.Addition of hexadecane (7 �l) resulted in an initial C/N ratio of 12 (equivalentto liquid culture). Gas and sand samples were removed periodically for analysis.Prior to sampling, the bottle contents were mixed completely on a roller mill for20 min. Headspace sampling was through a 24-mm Mininert top cap. Sand wassampled using a sterile scoop after purging (1.2 liter/min for 3 min) the bottlewith sterile, moist (100% relative humidity) air in a laminar flow hood. Sandsamples (1 g) were preserved with 2 mg of dissolved NaN3, refrigerated (4°C forless than 48 h), and then hexane extracted.

gfp induction in well-mixed liquid broth cultures. LB-TET-cultivated cellswere used as inocula for studying GFP induction over time because they wereinitially nonfluorescent. Overnight LB-TET cultures were subsampled (10 ml),and the remaining culture volume (approximately 90 ml) was washed twice andresuspended in P buffer. The washed suspensions were divided into 9- or 10-mlsubsamples, with a total of eight treatments. Amendments were made to the10-ml washed subsamples, resulting in the following treatments: negative control(P buffer only), 2 �l of hexadecane per ml, 18.3 mM glucose, 36.7 mM glycerol,9.3 mM NH4Cl, and 9.3 mM NaNO3. The proportion of N provided as eitherNH4Cl or NaNO3 was the same as that provided in our growth studies of P.aeruginosa PG201 and its mutants PG201pra�, PG201rhlA� andPG201pra�rhlA� (Table 1). The remaining two treatments consisted of 9 ml ofwashed cell suspension plus 100 �l of Hutner’s mineral solution and 900 �l ofeither half-strength 21C or CW21C. The amount of nitrogen provided in the lasttwo treatments was slightly less than that provided in our studies of PG201 andits mutants. The amended suspensions were shaken (30°C, 150 rpm) in covered50-ml Erlenmeyer flasks. Over a 48-h period, 10 �l was periodically sampled(three times) from each suspension, a wet mount on a clean glass microscope

slide was prepared, and at least 10 fields of view per slide were visualized first byepifluorescence microscopy for GFP and second by phase contrast for total cells.The apparent brightness and abundance of GFP-producing cells relative to thecontrols were noted, and qualitative scores were assigned for the purposes ofcomparing treatments.

gfp expression in drop slides. Drop slides consisted of a coverslip overlying aspecial glass microscope slide with 10 wells created by a hydrophobic barrier(Precision Lab Products, Middleton, Wis.). A 0.5-�l drop of an overnight culturein LB-TET (washed twice and resuspended in P buffer) of PG201(pHX1),PG201(pRhlPGB), or PG201(pPROBE-TT) was dispensed into duplicate wells.A 0.5-�l drop of hexadecane was added adjacent to a P-buffer cell suspensiondrop prior to applying the coverslip. When the coverslip was lowered, both dropsspread within the hydrophobic barrier. To minimize dehydration, the edges ofthe coverslip were sealed to the slide with a thin film of silicone lubricant (Dow,Midland, Mich.). The slides were monitored over time for GFP fluorescence byboth epifluorescence microscopy and confocal laser scanning microscopy(CLSM).

Microscopy and fluorescence measurements. GFP faded rapidly under the UVlight source; therefore, epifluorescence microscopy was performed prior tophase-contrast imaging. GFP induction in liquid was determined using a NikonE800 epifluorescence microscope with a GFP filter set (Chroma, Brattleboro,Vt.). The positive control for GFP was PG201(KtKan), which was cultivated onLA-kanamycin and suspended in P buffer to a cell density similar to that of thesamples. The negative control was P. aeruginosa PG201 grown in LB, washed(two times), and resuspended. The relative amount of GFP produced per cell wasinferred by the brightness of GFP-producing cells compared to positive andnegative controls. The proportion of total cells expressing GFP was estimated byobserving, for each field of view, the abundance of cells by phase contrast.

For visualizing GFP produced by sand-cultured bacteria, 1 g of moist sand wasvortexed for 10 s with 12 ml of sterile H2O. The supernatant was allowed to clearfor 1 h, and 5 ml was suction filtered onto a 13-mm-diameter, 0.2-�m-pore-sizeblack Isopore membrane (Millipore, Bedford, Mass.). A second 1-ml aliquotfrom the same 12-ml suspension was stained with a nucleic acid-specific stain(SYBR-gold; Molecular Probes, Eugene, Oreg.), filtered, and visualized for totalbacteria. Images of sand-cultivated gfp-expressing bacteria and total bacteriawere captured using a DE750 color video camera (Optronics, Goleta, Calif.)mounted on an Olympus BX60 epifluorescence microscope with a fluoresceinisothiocyanate filter set.

Drop slides were screened for GFP production by epifluorescence microscopywith a Nikon E800 microscope and then imaged in the x-y and x-z planes with aTCS SP2 CLSM (Leica, Mannheim, Germany) with a 63� water immersionobjective and an Ar 488-nm excitation laser. The thicknesses of the z scans wereapproximately 0.2 �m.

Environmental scanning electron microscopy (ESEM) was performed in wetmode using a gaseous secondary electron detector in an FEI Co. XL30 ESEMFEG (Philips Electron Optics, Eindoven, The Netherlands). Images of PG201colonizing sand under the specified experimental conditions were acquired for atleast 10 subsamples of a 2-day-old culture.

Surface tension, CO2, and hexadecane measurements. The air-water interfa-cial (surface) tension of liquid cultures was measured for 1.5 ml of either wholebroth cultures or cell-free culture broth (supernatant from centrifugation at

FIG. 1. Sequence of the inverse PCR product containing the putative promoter for pra, a structural gene encoding the PA bioemulsifyingprotein. An 8-bp region homologous to the pra structural gene (GenBank accession no. L08966) is italicized. The Shine-Dalgarno region (S/D)is underlined. PCR primers as described in the text are in boldface.

VOL. 68, 2002 RHAMNOLIPID AND PA PROTEIN PROMOTER TRANSCRIPTION IN SAND 2511


http://aem.asm

.org/D

ownloaded from

http://aem.asm.org/

10,000 � g for 10 min), using a liquid tensiometer (model KT-10; Krüss USA,Charlotte, N.C.), 10-ml (concentrated-H2SO4-washed) beakers, and a small-volume plate. A reduction in surface tension was indicated by any measurementlower than that of uninoculated media (72 mN/m).

Headspace CO2 analysis of sand cultures was performed with a GC14 gaschromatograph (GC) (Shimadzu, Kyoto, Japan) using a Porapak R column(Supelco, Bellefonte, Pa.), a TCD detector, and high-purity helium (30 ml/min)as the carrier gas. Concentrated hexane extracts of sand cultures were analyzedfor hexadecane content using an HP 6890 GC with a mass spectrometer (MS)(Agilent Technologies, Palo Alto, Calif.). Unknowns were compared to a mini-mum of 15 standards spanning three calibration ranges for hexadecane in hex-ane. The GC-MS operating conditions were as follows: 1-�l sample volume,capillary column (Alltech AT-1), splitless injection, 250°C injector temperature,ultra-high-purity helium carrier gas at 30 cm s�1, initial oven temperature of40°C (4 min), temperature increased by 10°C min�1 to 270°C then held (3 min),and detector temperature of 275°C. The GC-MS signal was automatically inte-grated using HP G1034C software for MS ChemStation (Agilent).

Data analysis. Means from replicated measurements were compared using themethod of Tukey (8) with a confidence interval of 0.05. Hexadecane depletiondata from sand cultures were analyzed by curve fitting to determine the best fitto either zero-, first-, or second-order mathematical models (46). Depletion ratecoefficients were calculated by linearizing the data for the appropriate mathe-matical model and performing regression analysis with Excel 2000. For the sandculture experiments, the biomass C originating from the utilization of hexade-cane was calculated by subtracting the total CO2 carbon (mole basis) from themolar amount of hexadecane C utilized (total added hexadecane C). The molarfraction of carbon fixed was then calculated by dividing the calculated biomass C(molar basis) by the total molar addition of hexadecane C.

RESULTS

Putative pra promoter. The sequence of a region 5� to thepra gene in PG201, which includes the pra promoter, is pro-vided in Fig. 1. A 459-bp region of the putative promoter-containing sequence was 99% identical with a homologousregion in the P. aeruginosa PAO1 complete genome (GenBankaccession no. AE004091). While not containing a consensusShine-Dalgarno sequence, a region that could serve as a ribo-some binding site was recognized (Fig. 1).

Growth and surface tension decrease in liquid culture. Wecultured PG201, PG201rhlA�, PG201pra�, and PG201pra�

rhlA� with either NH4� N or NO3

� N to determine the effectsof nitrogen on growth rate and the accumulation of surface-active compounds in liquid culture. Production of rhamnolip-ids and/or PA protein (and any other surfactant) will decreaseliquid surface tension. The surface tension was lowest in late-exponential-phase cultures (Fig. 2). Surface tension was thesame for whole cultures and cell-free supernatant (data notshown). When NH4

� was the N source, the strains had similarlag times and grew to similar maximum cell densities (Fig. 2) atsimilar rates (Table 2). Also when NH4

� was the N source, theamount of surface-active compounds was significantly smallerfor cells deficient in rhlA, while deficiency in pra did not altersurface tension (Table 2).

When NO3� was the nitrogen source, lag times were longer

for all strains and PG201 grew to higher culture densities thanthe other strains (Fig. 2). The growth rates for PG201 witheither NO3

� or NH4� were similar, but the growth rates for

the other strains with NO3� were significantly lower (Table 2).

While liquid surface tension decreased for all strains (Fig. 2),the decrease in surface tension for PG201, on a per-cell basis,was similar whether NH4

� or NO3� was the N source (Table

2). With NO3� N, the decrease in surface tension per cell was

similar across strains (Table 2).

Hexadecane biodegradation in sand culture. Relative to uni-noculated controls, where no biodegradation activity was ob-served, GC analysis of hexane-extracted sand samples overtime showed that hexadecane was rapidly and completely bio-degraded in moist sand by all cultures, including strains defi-cient in pra and rhlA (Fig. 3). Lag times, hexadecane depletion,and CO2 generation patterns were similar for all strains.Among the rate models tested, hexadecane depletion best fit afirst-order process (r2 � 0.9; r2 values for zero- and second-order models were lower). First-order depletion coefficientsand the fraction of hexadecane C fixed into cellular materialwere statistically similar (n � 3; � � 0.05) across the strainsand averaged 0.093 h�1 and 0.35, respectively. ESEM imagesindicated that PG201 was dispersed on sand grains and did notappear as thick biofilms (Fig. 4).

Expression vector stability and gfp induction in liquid cul-ture. Significant GFP induction was observed in PG201 strainscarrying pra or rhlR gene fusions when washed cells were sub-jected to a range of nutritional conditions. The expressionvector was reasonably stable (95% after 2 days). StrainPG201(pPROBE-TT), harboring a promoterless gfp reportergene only, was nonfluorescent (Table 3). Strain PG201(pHX1),in which gfp was driven by the pra promoter element, fluo-resced brightest in the presence of hexadecane (Table 3).Strain PG201(pRhlPGB) fluoresced brightest and equally inthe presence of either hexadecane, glucose, or ammonium-containing P buffer in the absence of a C source (Table 3). Itis noteworthy that the fluorescent cells appeared to be mostlyadherent to oil drops.

gfp expression in sand culture. We cultured PG201(pHX1),PG201(pRhlPGB), and PG201(pPROBE-TT) in moist sand todetermine if the promoters for rhlR and pra are active inunsaturated porous media. We used nutritional conditions(hexadecane C and NO3

�) that, in liquid culture, appeared tostimulate the production of surface-active compounds (Fig. 2and Table 2). Epifluorescence microscopy of eluted cells re-vealed that PG201(pHX1) and PG201(pRhlPGB) producedGFP in moist sand culture with hexadecane as the sole Csource and with nitrate N. No visible GFP fluorescence wasobserved in PG201(pPROBE-TT), which lacked a promotersequence 5� to the gfp reporter gene. By directly counting totaleluted Sybr-Gold-stained cells, we observed that approxi-mately 3% of PG201(pHX1) and 1% of PG201(pRhlPGB)exhibited GFP fluorescence after 1 week of cultivation in sand.Similar numbers of total cells were present in each of thecultures at the time of sampling.

gfp expression in drop slides. In our liquid culture GFPinduction studies, we observed the highest fluorescence in cellsexposed to hexadecane; most of those fluorescent cells wereadherent to hexadecane drops. This led us to hypothesize thattranscription of surface-active genes occurs mostly at the hy-drophobic hexadecane-water interface due to a lack of avail-ability of the substrate in the water phase. To test our hypoth-esis, we worked with static drop slides that would allow us tobetter visualize the spatial pattern of GFP production. As insand cultures, PG201 (pPROBE-TT), the promoterless controlstrain, did not produce GFP in the drop slides (data notshown). Epifluorescence microscopy of the slides containingPG201(pHX1) and PG201(pRhlPGB), (i.e., examination fromplan view), was inadequate to demonstrate where GFP was



http://aem.asm

.org/D

ownloaded from

http://aem.asm.org/

produced, because when viewed from the top, it appeared thatgreen cells were contained both in water and in oil. This isbecause an oil-water-glass interface was formed by water wick-ing under the hydrophilic coverslip, on top of the oil, wherethe water resided as a thin film between the glass and oil.With CLSM, we observed that PG201(pHX1) and PG201(pRhlPGB) expressed gfp at the hexadecane-water and thehexadecane-water-glass interfaces (Fig. 5). An angle of approx-imately 60° was formed between the coverslip and the gfp-expressing cells, which were oriented along the hexadecane-water interface in the x-z direction by CLSM.

DISCUSSION

An objective of this study was to assess in situ production ofsurface-active compounds in porous media. Although a num-ber of soil isolates produce surface-active compounds in brothculture (see, e.g., references 1, 2, 38, and 45), direct evidence

FIG. 2. OD at 660 nm (circles), liquid surface tension (squares), and headspace parts per thousand carbon dioxide (triangles) (data for PG201and PG201pra� only) with either ammonium (closed symbols) or nitrate (open symbols) in liquid cultures of P. aeruginosa. Strains are either wildtype (PG201) or deficient in PA protein synthesis (PG201pra�), in rhamnolipid synthesis (PG201rhlA�), or in both (PG201pra�rhlA�).Hexadecane is the carbon source. Each point represents the average of three or more independent replicates.

TABLE 2. First-order growth rate coefficients and liquid culturedensity-normalized maximum liquid surface tension decrease for

P. aeruginosa with hexadecane as the sole carbon source andeither ammonium or nitrate as the nitrogen source

Strain

Growth ratecoefficient

(h�1)a with:

Surface tensiondecreasea,b (mN/mper OD unit) with:

NH4� NO3

� NH4� NO3

�

PG201 0.20 a 0.19 a 38 a, c 52 cPG201pra� 0.22 a 0.13 b 34 a, d 54 cPG201rhlA� 0.18 a 0.11 b 13 b 34 c, dPG201pra�rhlA� 0.21 a 0.11 b 14 b 53 c

a Values with the same letter are not significantly different (� � 0.05; n � 3).Statistical comparisons were either within growth rate coefficients or withinvalues of surface tension decrease.

b The surface tension decrease was calculated by dividing the total decrease inliquid surface tension by the OD when surface tension was lowest.



http://aem.asm

.org/D

ownloaded from

http://aem.asm.org/

of related gene transcriptional activity in sand or soil has beenlacking, as is an understanding of the ecology of surface-activecompound production in soil. Here we report the transcrip-tional activity in P. aeruginosa sand culture of a gene encodingPA emulsifying protein. However, despite this transcriptionalactivity of pra in sand culture, PG201 hexadecane biodegrada-tion in sand was unaffected by the absence of intact pra or rhlAgenes.

In liquid culture with NO3� as the sole N source, the dele-

tion of either pra or rhlA slows growth at the expense ofhexadecane by PG201 (Fig. 2) (14). Therefore, we expectedthat PG201 using NO3

� in sand culture would degrade hexa-decane faster than its mutants (with mutations in pra, rhlA orboth). Instead, the biodegradation patterns were not differentin sand culture whether PG201 was with or without intactgenes encoding surface-active compounds. Our explanation isthat the direct association between PG201 and hexadecane,which is facilitated by hexadecane spreading on sand surfaceswhere cells are adherent (Fig. 4), reduces the role for surface-active compounds in sand culture and favors the biodegrada-tion mechanism of direct hydrocarbon uptake by PG201. Uponconsidering the spatial pattern of gfp expression, and takinginto account prior reports, it is possible that different predom-inant uptake mechanisms govern hexadecane biodegradationin liquid versus sand.

Studies with liquid cultures of cells harboring a pra-gfp fu-sion revealed that surface-active gene expression by P. aerugi-

nosa using hexadecane is spatially limited to the hexadecane-water interface. This may be because PG201 adheres tohexadecane, where transcription of genes (e.g., pra) for sur-face-active compounds is induced. Bacterial adherence to hy-drophobic interfaces, i.e., the oil-water or the (also hydropho-bic [21]) air-water interface (39, 43, 49), occurs withhydrophobic cells (43, 44). Cells may be hydrophobic already,or, in the case of P. aeruginosa, may become hydrophobic fromtheir rhamnolipid production (50), which reduces the amountof exposed cell surface lipopolysaccharide (4). The transcrip-tional activity of rhlR in liquid media lacking alkanes wasshown to be nearly constitutive at relatively low levels, withonly a slight increase (twofold) upon entry of cells into thestationary phase with relatively high cell concentrations in cellbroth (36). rhlR, by acting as a transcriptional activator of rhlAand rhlB, which are involved in rhamnolipid biosynthesis, thusis required for synthesis of this surfactant (27, 44) and also foradherence to hexadecane drops (44). We examined rhlR tran-scription as a way to ensure that cells were sufficiently active inour sand system to enable rhamnolipid synthesis. Indeed, wefound that there was little variation in rhlR transcription underthe various conditions tested here (Table 3) and that cellsexpressed considerable rhlR activity in both liquid and sandculture. Thus, the rhlR fusion served as a positive control forgfp expression, while transcriptional activity also suggests fa-vorable conditions for cellular adhesion to hexadecane. On theother hand, the pra fusion appears to report specific transcrip-

FIG. 3. Depletion of hexadecane (closed symbols) and accumulation of CO2 (open symbols) by four strains of P. aeruginosa PG201 in sandculture. Hexadecane is the sole C source; N is provided as either NO3

� (top) or NH4� (bottom). Strains are either wild type (PG201) or deficient

in PA protein synthesis (PG201pra�), in rhamnolipid synthesis (PG201rhlA�), or in both (PG201pra�rhlA�). Points represent either the averageof three independent replicates (NO3

�) or single measurements (NH4�) at each time point.



http://aem.asm

.org/D

ownloaded from

http://aem.asm.org/

FIG. 4. ESEM images of P. aeruginosa PG201 in moist sand after 2 days of cultivation. NO3� was the N source and hexadecane was the carbon

source. Magnifications, �2,500 (A) and �3,500 (B). Arrows point to typical cells or cell clusters.



http://aem.asm

.org/D

ownloaded from

http://aem.asm.org/

FIG. 5. CLSM of drop slides containing P buffer, hexadecane, and P. aeruginosa strains PG201(pHX1) (A) and PG201(pRhlPGB) (B). Topframes are plan views; bottom frames are x-z sections. Fluorescence is from transcription of plasmid-borne gfp controlled by the putative promotersfor pra (A) and rhlR (B). Bars, 20 �m.

TABLE 3. GFP fluorescence after 48 h in liquid culture in P. aeruginosa harboring fusions of promotersfor pra and rhlR to a gfp reporter plasmid

TreatmentbGFP inductiona by strain:

PG201(pHX1) PG201(pRhlPGB) PG201(pPROBE-TT)

Controlc � �/�d �Hexadecane ��e ��e �Glucose � �� Glycerol � � �Ammoniumf � �� Nitratef � � �Half-strength 21C plus Hutner’s mineral solutionf � � �CW21C plus Hutner’s mineral solutionf � � �

a Scores are relative to those for positive (��) and negative (�) GFP control strains.b Amendments were made to noninduced cells suspended in P buffer, as described in the text.c Control medium is P buffer without an exogenous carbon source.d Cells were mostly nonfluorescent, but a low level of fluorescence was observed in some cells.e Cells were clumped and brightest around the exterior of hexadecane drops.f Contained no carbon source.



http://aem.asm

.org/D

ownloaded from

http://aem.asm.org/

tional activity at the hexadecane-water interface. Intact pra andrhlA genes make PG201 grow faster with hexadecane andNO3

� (14), indicating that these genes facilitate the use ofhexadecane C by PG201 in liquid. Additionally, pra is inducibleby hexadecane but not by glycerol or glucose (14). Takentogether with our observations that pra promoter-driven gfpfusions fluoresced only at the hexadecane-water interface,these findings indicate that hexadecane appears to be a tran-scriptional activator for pra. Because hexadecane is so spar-ingly soluble, cellular adherence to the oil is a logical prereq-uisite for hexadecane to cause such transcriptional activity.

The involvement of rhamnolipids in cellular adhesion tohexadecane and the transcriptional activity of pra at the oil-water interface have different implications in liquid versus sandculture, which may explain why we observed relatively short lagtimes in sand culture as well as similar biodegradation patternsfor strains proficient and deficient in production of surface-active compounds. We propose that cells in sand simply do notneed surface-active compounds to facilitate rapid uptake ofhexadecane because they are already in direct contact with thehydrocarbon. In sand culture, assuming a mean grain diameterof 200 �m and well-distributed hexadecane over all sand sur-faces (a condition in our preparations), the total surface areaof hexadecane is approximately equal to the sand surface areaof ca. 1.2 � 1012 �m2. In liquid culture, this same surface areawould arise from breaking the total amount of hexadecane into1-�m-diameter drops. In liquid culture, aqueous media and oilare mixed via shaking, which is mainly intended to facilitateaeration. However, we typically observe unstable oil-in-wateremulsions before the onset of bacterial growth. The emulsionis a distribution of oil drops of different sizes dispersed into theaqueous medium. The mean drop size of oil in an oil-wateremulsion correlates positively with approximately the squareroot of oil-water interfacial tension, the mixing intensity, andthe volume fraction of the dispersed phase (10, 37), which inthe case of our PG201 liquid cultures would be 0.002. Forchlorobenzene in water (interfacial tension of 0.0334 N/m),where the volume fraction of chlorobenzene was 0.005 and thestirring rate was 180 min�1, the mean oil droplet diameter was215 �m (37). Hexadecane interfacial tension is more like thatof n-dodecane, also a low-solubility alkane, which has an oil-water interfacial tension of 52.8 N/m (43). Considering that ourliquid cultures were mixed and had volume fractions of oilsimilar to those in a previous study in which a more solublehydrocarbon resulted in larger drops (37), it is inconceivablethat the conditions of liquid culture we used could result in ahexadecane drop diameter of as small as 1 �m. The result isthat the exposed surface area of hexadecane is lower in liquidculture than in sand culture. It would seem that making sur-face-active compounds, which both facilitates cellular adhesionto oil and improves oil bioavailability, is relatively advanta-geous to cells in liquid.

In sand culture, cells are distributed over the sand surfacethat is coated with hexadecane, and thus nearly every cell is indirect contact with oil. The high spatial opportunity for cellularadherence to hexadecane is advantageous to cells in that directadherence is probably a faster route for hexadecane uptake. Inliquid culture, there is a long lag time prior to hexadecane-associated growth (Fig. 2), which contrasts with the short lagtime preceding hexadecane biodegradation in sand (Fig. 3).

Also in liquid culture, the provision of nitrate as the N source,which has been shown to stimulate surface-active compoundproduction, corresponds to rapid growth (Table 2). In bothsand and liquid culture, hexadecane biodegradation is com-plete, but the patterns of biodegradation suggest differing masstransfer mechanisms. The data support the idea that cells insand were already adherent to hexadecane while cells in liquidwere dependent on their production of surface-active com-pounds to make them adherent and to increase the availabilityof hexadecane in solution. In contaminated soil, hydrocarbonspreading will enhance the contact area between bacteria andinsoluble NAPLs. Our work supports the idea that bacterium-oil contact is a relatively more important factor for acceleratinghexadecane biodegradation in unsaturated porous media thanis the indigenous production of surface-active compounds.

ACKNOWLEDGMENTS

Funding for this work was provided by National Science FoundationAward DEB-9805946 (to P. A. Holden), by National Science Founda-tion Awards BES-99772 and DMR-9724254, by U.S. EnvironmentalProtection Agency Cooperative Agreement AERL9405 (to M. K. Fire-stone and J. R. Hunt, University of California, Berkeley), and byEPA/DOE/NSF/ONR Bioavailability Program Award R827133-01 (toP. A. Holden and A. Keller).

ESEM was performed by Jose Saleta in the MicroEnvironmentalImaging and Analysis Facility at the University of California, SantaBarbara (UCSB). CLSM was performed by Mario Yasa in the labo-ratory of Cyrus Safinya, Department of Physics and Materials Re-search Laboratory, UCSB. Joshua Schimel (UCSB) provided the in-strumentation for CO2 analysis. We acknowledge James R. Hunt forhelpful criticisms and the assistance of Portia Zamos, Claire Eustace,Cindy Wu, Sam Paik, Jennifer Lau, Perrin Pellegrin, Jennifer Guigli-ano, Don Herman, and Allen Doyle.

REFERENCES

1. Abu-Ruwaida, A. S., I. M. Banat, S. Haditirto, A. Salem, and M. Kadri. 1991.Isolation of biosurfactant-producing bacteria, product characterization, andevaluation. Acta Biotechnol. 11:315–324.

2. Akit, J., D. G. Cooper, K. I. Manninen, and J. E. Zajic. 1981. Investigationof potential biosurfactant production among phytopathogenic Corynebacte-ria and related soil microbes. Curr. Microbiol. 6:145–150.

3. Alexander, M. 1999. Biodegradation and bioremediation, 2nd ed. AcademicPress, San Diego, Calif.

4. Al-Tahhan, R. A., T. R. Sandrin, A. A. Bodour, and R. M. Maier. 2000.Rhamnolipid-induced removal of lipopolysaccharide from Pseudomonasaeruginosa: effect on cell surface properties and interaction with hydrophobicsubstrates. Appl. Environ. Microbiol. 66:3262–3268.

5. Bai, G., M. L. Brusseau, and R. M. Miller. 1997. Biosurfactant-enhancedremoval of residual hydrocarbon from soil. J. Contam. Hydrol. 25:157–170.

6. Banat, I. M., N. Samarah, M. Murad, R. Horne, and S. Banerjee. 1991.Biosurfactant production and use in oil tank clean-up. World J. Microbiol.Biotechnol. 7:80–88.

7. Bloemberg, G. V., G. A. O’Toole, B. J. J. Lugtenberg, and R. Kolter. 1997.Green fluorescent protein as a marker for Pseudomonas spp. Appl. Environ,Microbiol. 63:4543–4551.

8. Daniel, W. W. 1991. Biostatistics: a foundation for analysis in the healthsciences. John Wiley & Sons, New York, N.Y.

9. Finnerty, W. R., and M. E. Singer. 1985. Membranes of hydrocarbon-utiliz-ing microorganisms, p. 1–44. In B. K. Ghosh (ed.), Organization of prokary-otic cell membranes, vol. III. CRC Press, Inc., Boca Raton, Fla.

10. Galindo, E., A. W. Pacek, and A. W. Nienow. 2000. Study of drop and bubblesizes in a simulated mycelial fermentation broth of up to four phases. Bio-technol. Bioeng. 69:213–221.

11. Guerra-Santos, L., O. Kappeli, and A. Fiechter. 1984. Pseudomonas aerugi-nosa biosurfactant production in continuous culture with glucose as carbonsource. Appl. Environ. Microbiol. 48:301–305.

12. Guerra-Santos, L. H., O. Käppeli, and A. Fiechter. 1986. Dependence ofPseudomonas aeruginosa continuous culture biosurfactant production on nu-tritional and environmental factors. Appl. Microbiol. Biotechnol. 24:443–448.

13. Haferburg, D., R. Hommel, R. Claus, and H. P. Kleber. 1986. Extracellularmicrobial lipids as biosurfactants. Adv. Biochem. Eng. Bio/Technol. 33:53–93.



http://aem.asm

.org/D

ownloaded from

http://aem.asm.org/

14. Hardegger, M., A. K. Koch, U. A. Ochsner, A. Fiechter, and J. Reiser. 1994.Cloning and heterologous expression of a gene encoding an alkane-inducedextracellular protein involved in alkane assimilation from Pseudomonasaeruginosa. Appl. Environ. Microbiol. 60:3679–3687.

15. Harvey, S., I. Elashvili, J. J. Valdes, D. Kamely, and A. M. Chakrabarty.1990. Enhanced removal of Exxon Valdez spilled oil from Alaskan gravel bya microbial surfactant. Bio/Technology 8:228–230.

16. Herman, D. C., Y. Zhang, and R. M. Miller. 1997. Rhamnolipid (biosurfac-tant) effects on cell aggregation and biodegradation of residual hexadecaneunder saturated flow conditions. Appl. Environ. Microbiol. 63:3622–3627.

17. Hisatsuka, K., T. Nakahara, N. Sano, and K. Yamada. 1971. Formation ofrhamnolipid by Pseudomonas aeruginosa and its function in hydrocarbonfermentation. Agric. Biol. Chem. 35:686–692.

18. Hisatsuka, K., T. Nakahara, and K. Yamada. 1972. Protein-like activator forn-alkane oxidation by Pseudomonas aeruginosa S7B1. Agric. Biol. Chem.36:1361–1369.

19. Holden, P. A., L. J. Halverson, and M. K. Firestone. 1997. Water stresseffects on toluene biodegradation by Pseudomonas putida. Biodegradation8:143–151.

20. Hommel, R. K. 1990. Formation and physiological role of biosurfactantsproduced by hydrocarbon-utilizing microorganisms. Biodegradation 1:107–119.

21. Israelachvili, J. N. 1992. Intermolecular and surface forces, 2nd ed. Aca-demic Press, San Diego, Calif.

22. Itoh, S., and T. Suzuki. 1972. Effect of rhamnolipids on growth of Pseudo-monas aeruginosa mutant deficient in n-paraffin-utilizing ability. Agric. Biol.Chem. 36:2233–2235.

23. Jahan, K., T. Ahmed, and W. J. Maier. 1997. Factors affecting the nonionicsurfactant-enhanced biodegradation of phenanthrene. Water Environ. Res.69:317–325.

24. Jain, D. K., H. Lee, and J. T. Trevors. 1992. Effect of addition of Pseudo-monas aeruginosa UG2 inocula or biosurfactants on biodegradation of se-lected hydrocarbons in soil. J. Indust. Microbiol. 10:87–93.

25. Joyner, D. C., and S. E. Lindow. 2000. Heterogeneity of iron bioavailabilityon plants assessed with a whole-cell GFP-based bacterial biosensor. Micro-biology 146:2435–2445.

26. Kay, R., A. Chan, M. Daly, and J. McPherson. 1987. Duplication of CaMV35S promoter sequences creates a strong enhancer for plant genes. Science236:1299–1302.

27. Koch, A. K., O. Käppeli, A. Fiechter, and J. Reiser. 1991. Hydrocarbonassimilation and biosurfactant production in Pseudomonas aeruginosa mu-tants. J. Bacteriol. 173:4212–4219.

28. Macur, R. E., and W. P. Inskeep. 1999. Effects of a nonionic surfactant onbiodegradation of phenanthrene and hexadecane in soil. Environ. Tox.Chem. 18:1927–1931.

29. Miller, R. M. 1995. Surfactant-enhanced bioavailability of slightly solubleorganic compounds, p. 33–54. In H. D. Skipper and R. F. Turco (ed.),Bioremediation: science and applications. Soil Science Society of America,Madison, Wis.

30. Miller, W. G., J. H. J. Leveau, and S. E. Lindow. 2000. Improved gfp and inaZbroad-host-range promoter-probe vectors. Mol. Plant-Microbe Interact. 13:1243–1250.

31. Mulligan, C. N., and B. F. Gibbs. 1989. Correlation of nitrogen metabolismwith biosurfactant production by Pseudomonas aeruginosa. Appl. Environ.Microbiol. 55:3016–3019.

32. Mulligan, C. N., G. Mahmourides, and B. F. Gibbs. 1989. The influence ofphosphate metabolism on biosurfactant production by Pseudomonas aerugi-nosa. J. Biotechnol. 12:199–210.

33. National Research Council. 1997. Innovations in ground water and soilcleanup. National Academy Press, Washington, D.C.

34. Ochman, H., M. M. Medhora, D. Garza, and D. L. Hartl. 1990. Amplifica-tion of flanking sequences by inverse PCR, p. 219–227. In M. A. Innis, D. H.Gelfand, J. J. Sninsky, and T. J. White (ed.), PCR protocols: a guide tomethods and applications. Academic Press, Inc., San Diego. Calif.

35. Ochsner, U. A., A. Fiechter, and J. Reiser. 1994. Isolation, characterization,and expression in Escherichia coli of the Pseudomonas aeruginosa rhlABgenes encoding a rhamnosyltransferase involved in rhamnolipid biosurfac-tant synthesis. J. Biol. Chem. 269:19787–19795.

36. Ochsner, U. A., A. K. Koch, A. Fiechter, and J. Reiser. 1994. Isolation andcharacterization of a regulatory gene affecting rhamnolipid biosurfactantsynthesis in Pseudomonas aeruginosa. J. Bacteriol. 176:2044–2054.

37. Pacek, A. W., C. C. Man, and A. W. Nienow. 1998. On the Sauter meandiameter and size distributions in turbulent liquid/liquid dispersions in astirred vessel. Chem. Eng. Sci. 53:2005–2011.

38. Phale, P. S., H. S. Savithri, N. A. Rao, and C. S. Vaidyanathan. 1995.Production of biosurfactant “Biosur-Pm” by Pseudomonas maltophilaCSV89: characterization and role in hydrocarbon uptake. Arch. Microbiol.163:424–431.

39. Powelson, D. K., and A. L. Mills. 1996. Bacterial enrichment at the gas-waterinterface of a laboratory apparatus. Appl. Environ. Microbiol. 62:2593–2597.

40. Reiser, J., A. K. Koch, K. Jenny, and O. Kappeli. 1989. Structure, properties,and production of biosurfactants, p. 85–97. In J. W. Orbringer and H. S.Tillinghast (ed.), Bio/Technology for aerospace applications, vol. 3. PortfolioPublishing Company, The Woodlands. Tex.

41. Rosenburg, E. 1986. Microbial surfactants. Crit. Rev. Biotechnol. 3:109–132.42. Sambrook, J., E. F. Fritsch, and T. Maniatis. 1989. Molecular cloning: a

laboratory manual, 2nd ed. Cold Spring Harbor Laboratory, Cold SpringHarbor, N.Y.

43. Schafer, A., H. Harms, and A. J. B. Zehnder. 1998. Bacterial accumulationat the air-water interface. Environ. Sci. Technol. 32:3704–3712.

44. Shreve, G. S., S. Inguva, and S. Gunnam. 1995. Rhamnolipid biosurfactantenhancement of hexadecane biodegradation by Pseudomonas aeruginosa.Mol. Mar. Biol. Biotechnol. 4:331–337.

45. Singer, M. E. V., and W. R. Finnerty. 1990. Physiology of biosurfactantsynthesis by Rhodococcus species H13-A. Can. J. Microbiol. 36:741–745.

46. Snoeyink, V. L., and D. Jenkins. 1980. Water chemistry. John Wiley & Sons,New York, N.Y.

47. Stretton, S., S. Techkarnjanaruk, A. M. McLennan, and A. E. Goodman.1998. Use of green fluorescent protein to tag and investigate gene expressionin marine bacteria. Appl. Environ. Microbiol. 64:2554–2559.

48. van Dyke, M. I., P. Couture, M. Brauer, H. Lee, and J. T. Trevors. 1993.Pseudomonas aeruginosa UG2 rhamnolipid biosurfactants: structural char-acterization and their use in removing hydrophobic compounds from soil.Can. J. Microbiol. 39:1071–1078.

49. Wan, J., J. L. Wilson, and T. L. Kieft. 1994. Influence of the gas-waterinterface on transport of microorganisms through unsaturated porous media.Appl. Environ. Microbiol. 60:509–516.

50. Zhang, Y., and R. M. Miller. 1994. Effect of a Pseudomonas rhamnolipidbiosurfactant on cell hydrophobicity and biodegradation of octadecane.Appl. Environ. Microbiol. 60:2101–2106.

51. Zhang, Y., and R. M. Miller. 1995. Effect of rhamnolipid (biosurfactant)structure on solubilization and biodegradation of n-alkanes. Appl. Environ.Microbiol. 61:2247–2251.

52. Zhang, Y., and R. M. Miller. 1992. Enhanced octadecane dispersion andbiodegradation by a Pseudomonas rhamnolipid surfactant (biosurfactant).Appl. Environ. Microbiol. 58:3276–3282.



http://aem.asm

.org/D

ownloaded from

http://aem.asm.org/

assessing the role of pseudomonas aeruginosa surface ... · extracted from pg201 by using the...

Documents