Effects of Pseudomonas Putida Modified to Produce Phenazine-1

Home , Siderophore

Effects of Pseudomonas putida modified to produce phenazine-1-carboxylic acid and 2,4-diacetylphloroglucinol on the microflora of field grown wheat

Peter A.H.M. Bakker1,∗, Debora C.M. Glandorf1#, Mareike Viebahn1, Theodora W.M. Ouwens1,EricSmit2, Paula Leeﬂang2, Karel Wernars2, Linda S. Thomashow3,JaneE. Thomas-Oates4 & Leendert C. van Loon1 1Institute of Biology, Utrecht University, Section of Plant Pathology, P.O. Box 80084, 3508TB Utrecht, The Nether- lands; 2National Institute of Public Health and the Environment, Bilthoven, The Netherlands; 3USDA, Washington State University, Pullman WA, USA; 4Michael Barber Center for Mass Spectrometry, UMIST, Manchester, UK (∗Author for correspondence: E-mail: [email protected])

Abstract Pseudomonas putida WCS358r, genetically modified to have improved activity against soil-borne pathogens, was released into the rhizosphere of wheat. Two genetically modified derivatives carried the phz or the phl biosynthetic gene loci and constitutively produced either the antifungal compound phenazine-1-carboxylic acid (PCA) or the antifungal and antibacterial compound 2,4-diacetylphloroglucinol (DAPG). In 1997 and 1998, effects of single introductions of PCA producing derivatives on the indigenous microflora were studied. A transient shift in the composition of the total fungal microflora, determined by amplified ribosomal DNA restiction analysis (ARDRA), was detected. Starting in 1999, effects of repeated introduction of genetically modified microorganisms (GMMs) were studied. Wheat seeds coated with the PCA producer, the DAPG producer, a mixture of the PCA and DAPG producers, or WCS358r, were sown and the densities, composition and activities of the rhizosphere microbial populations were measured. All introduced strains decreased from 107CFU per gram of rhizosphere sample to below the detection limit after harvest of the wheat plants. The phz genes were stably maintained in the PCA producers, and PCA was detected in rhizosphere extracts of plants treated with this strain or with the mixture of the PCA and DAPG producers. The phl genes were also stably maintained in the DAPG producing derivative of WCS358r. Effects of the genetically modified bacteria on the rhizosphere fungi and bacteria were analyzed by using amplified ribosomal DNA restriction analysis. Introduction of the genetically modified bacterial strains caused a transient change in the composition of the rhizosphere microflora. However, introduction of the GMMs did not affect the several soil microbial activities that were investigated in this study.

Introduction non-modified microorganisms, concern about the ecological impact of large-scale release of genetically There is increasing interest in applying microorgan- modified microorganisms (GMMs) has been raised. isms to control soil-borne plant pathogens. Inconsist- One of the issues is the impact these genetically im- ent performance of the microorganisms, however, has proved biocontrol strains may evoke on the non-target hampered commercial application. Combining sev- soil microflora, such as affecting the microbial bal- eral modes of action against plant pathogens in one ance in the soil ecosystem and soil fertility (Cook et single organism by genetic modification can improve al. 1996). To date, risk assessment studies under field the efficacy of biological control agents (Van Loon conditions have focused mainly on microorganisms 1998). Despite long-term experience with introducing genetically modified with markers, such as antibiotic resistance, lacZY or XylE, and attention has been given # Current address: National Institute of Public Health and the to the potential impact of GMMs on the indigenous Environment, Bilthoven, The Netherlands. soil microflora (De Leij et al. 1995; Robleto et al. 618

1998). Effects of GMMs have been studied mainly Raaijmakers et al. 1995). A rifampin-resistant deriv- in microcosm experiments and transient effects have ative of WCS358, WCS358r, was used as the parental been reported on the indigenous bacterial (Natsch et strain (Glandorf et al. 1992). The phzABCDEFG genes al. 1997), fungal (Short et al. 1990; Girlanda et from P. fluorescens 2-79 (Mavrodi et al. 1998), un- al. 2001), and protozoal populations (Austin et al. der control of the Ptac promoter were introduced on 1990), on carbon turnover (Wang et al. 1991), and a mini-Tn5 LacZ1 transposon into WCS358r, result- on enzyme activities in soil (Naseby & Lynch 1998). ing in two derivatives, GMM 2 and GMM 8, with However, these microcosms lack the full biotic and different levels of PCA production (Glandorf et al. abiotic components of a field environment. 2001). Likewise, WCS358r was modified to produce Most studies of non-target effects of GMMs exam- DAPG by inserting, on a kanamycin-resistant mini- ine the culturable microflora. This limitation reduces Tn5 lacZ1 transposon (De Lorenzo et al. 1990), the the value of the results, because only a small propor- phlABCDEF genes (Bangera & Thomashow 1999). tion (0.5–2%) of the fungal and bacterial microflora The gene cluster contained its own promoter, how- can be cultured using currently-available media (Tors- ever, the phlF gene, encoding a repressor of DAPG vik et al. 1998). Cultivation independent techniques, synthesis, was disrupted to promote constitutive pro- such as amplified ribosomal DNA restriction analysis duction of DAPG. The DAPG producing derivative of (ARDRA), temperature gradient gel electrophoresis WCS358r was labeled as GMM P. In vitro produc- (TGGE), or denaturing gradient gel electrophoresis tion of PCA or DAPG by these GMMs was measured (DGGE) (Muyzer & Smalla 1998) are now available by HPLC (Bonsall et al. 1997). The presence of a to more accurately monitor microbial communities. single insertion of the mini-Tn5 in the chromosome The impact of heavy metals or pesticides on micro- of WCS358r and the absence of the transposase gene bial communities has been studied using amplified were confirmed by Southern blotting. Cultures were 16S rDNA (Smit et al. 1997; Engelen et al. 1998; stored at −80 ◦C in 35% glycerol. Fantroussi et al. 1999; Ibekwe et al. 2001). Reports in which molecular methods are used to study effects In vitro growth inhibition of pathogenic and of bacteria genetically modified to produce antibiot- saprophytic soil fungi ics on microbial communities, are scarce. Robleto et In vitro inhibition of fungal growth by WCS358r and al. (1998) demonstrated a reduction in the diversity of trifolitoxin-sensitive bacteria after inoculation of field- the GMMs was studied on KB medium supplemented with 200 µM filter-sterilized FeCl , or on diluted grown Phaseolus vulgaris with Rhizobium strains dif- 3 fering in their trifolitoxin production. potato dextrose agar (Weller et al. 1988). Growth Pseudomonas putida WCS358r was modified inhibition of both plant-pathogenic and saprophytic fungal isolates was tested. Pathogenic fungi included to produce the antifungal compound phenazine-1- Gaeumannomyces graminis carboxylic acid (PCA) (Thomashow et al. 1990), the wheat pathogens var tritici. (Ggt), Rhizoctonia solani AG8 C1, and Py- or the antifungal and antibacterial compound 2,4- thium sp.. Saprophytic fungal isolates that were tested diacetylphloroglucinol(DAPG) (Bangera & Thomashow 1999). Our objective is to determine if genetically originated from a field plot near Utrecht, The Nether- lands (Smit et al. 1999). These isolates belonged to the improved biocontrol bacteria could have effects on Absidia, Acremonium Fusarium non-target members of the fungal and bacterial rhizo- following genera: , , Gliocladium, Mortierella, Mucor, Penicillium sphere microflora, using cultivation-independent 18S and Trichoderma. and 16S rDNA analysis. Bioassay for suppression of Ggt

Materials and methods To determine suppression of take-all by WCS358r and the GMMs, wheat seeds (cv. Penewawa) were treated Bacterial strains with a suspension of the bacterial strains WCS358r, GMM 8 or the PCA producing strain P. ﬂuorescens Pseudomonas putida WCS358r is a plant growth- 2-79 in 1% methyl cellulose (Sigma Chemicals Co., promoting rhizobacterial strain with disease-suppressive St. Louis, MO), or with 1% methylcellulose (control). properties, based on the production of its ﬂuorescent The bacterial treatments were tested in soil contain- siderophore (Bakker et al. 1986; Duijff et al. 1994; ing the take-all fungus, as described by Weller et al. 619

(1985). Each treatment had six replications and con- Population dynamics of bacterial inoculants sisted of nine plants per replicate. Disease severity of take-all was rated after 4 weeks on a 0–8 scale in Roots were sampled at seven to eight different time which 0 = no disease and 8 = dead plant (Thomashow points after sowing. At harvest, 105 days after sowing, & Weller 1988). wheat plants were cut off above the soil, thereby leav- ing the roots for post-harvest samples. In each plot, Seed treatment three root samples with adhering soil were pooled, resulting in six samples for each treatment. Excess soil Cells of WCS358r, GMM 2, GMM 8, or GMM P was removed from the roots and samples of 0.5–1.0 were grown on KB agar plates (King et al. 1954) g (wet weight) roots with tightly adhering soil were supplemented with rifampin (150 µg/ml) for WCS358 shaken vigorously for 30 s in test tubes containing 5 ml of 10 mm MgSO4 and glass beads (0.11 mm dia- and additional kanamycin (50 µg/ml) for the GMMs. + Single colonies were subcultured on KB agar plates meter). Appropriate dilutions were plated on KB agar (without antibiotics) for 24 h at 28 0C. Bacteria were (KB containing 13 µg chloramphenicol, 40 µgampi- harvested by scraping cells from the agar, suspend- cillin and 100 µg cycloheximide per ml), amended with rifampin (150 µg/ml) and/or kanamycin (30 ing them in MgSO4 (10 mm) and washing the suspensions twice by centrifugation. Commercial wheat µg/ml), using a Spiral Plater (Spiral System model C, seeds (Triticum aestivum cv. Baldus) were coated with Spiral Systems Inc., Cincinnati, OH). Plates were in- 0 the washed bacterial suspensions in 1% methylcellu- cubated at 28 C for 48 h, after which colony-forming lose (w/v, Sigma). For the control treatment, bacterial units (CFUs) were counted. suspensions were replaced by 10 mm MgSO4. Treated seeds were allowed to dry overnight on filter paper in Phenazine-1-carboxylic acid extraction from wheat a laminar flow cabinet and were sown the next day. rhizosphere Population densities of the applied bacteria on seeds To determine whether PCA was produced by the were about 107 cfu/seed at the time of seeding. GMMs in the rhizosphere, in 1998 duplicate samples were collected for each treatment 12 days after sow- Experimental field ing. Samples of roots with tightly adhering soil (25 g) were extracted according to Bonsall et al. (1997), In 1997 and 1998, the experiments were conducted The presence of PCA in the residues was determ- on a site located at De Uithof, Utrecht, The Nether- ined by HPLC, according to Bonsall et al. (1997) lands. A randomized block design was used with four with some modifications (Glandorf et al. 2001). Com- treatments, with six replicates each, resulting in a total mercial PCA (Maybridge Chemical Company Ltd, of 24 plots, each of 1 m2. The four treatments were Trevillett, Cornwall, UK) and PCA produced by strain seeds treated with WCS358r, GMM 2, GMM 8, and 2-79 in liquid culture were used as reference samples. non-bacterized seeds (control). In 1999 repeated intro- To confirm the presence of PCA, HPLC fractions were ductions of GMMs were started. Again a randomized collected and the fraction corresponding to the R of block design was used, with six treatments, with six t PCA was isolated and subjected to mass spectrometric replicates each. The six treatments were seeds treated analysis (Glandorf et al. 2001). with WCS358r, GMM 8, GMM P, a mixture of GMM 8 and GMM P, non-treated seeds, and a treatment Determination of the composition of the bacterial and with a crop rotation of wheat and potato. Each year fungal microflora by ARDRA following 1999 the same treatments were applied to the same plots. Per plot about 1750 wheat seeds were For the ARDRA analysis, rDNA was amplified using sown manually in 11 rows of 1 m length at a depth of primers directed at conserved sequences of bacterial 2–3 cm. Plots were separated from each other by 60- 16S rDNA (Smit et al. 1997) or fungal 18S rDNA cm bare strips. The experimental field was fenced to (Smit et al. 1999). Fungal PCR products were diges- block rabbits from entering the site and bird entry was ted with TaqI and bacterial PCR products with HinfI. prevented using nets. The fragments were separated on polyacrylamide gels. Diagrams representing percentage similarity of resulting banding patterns per sample date were con- structed by UPGMA cluster analysis with the Biogene 620 program (version V96.15; Biogene,Vilber, Loumat, three pathogens on iron-rich medium. In contrast, the France), using the algorithm of Nei and Li (2% con- GMMs that produce either PCA or DAPG inhibited fidence) (Ferguson 1980). Rhizosphere samples from mycelial growth of all three pathogens. The PCA pro- three plots per treatment were pooled, resulting in two ducing GMMs inhibited growth of seven out of eleven independent replicate samples per treatment. saprophytic fungi isolated from field soil, and the DAPG producing GMM inhibited all of them, whereas Substrate-induced respiration (SIR), soil nitrification WCS358r did not reduce growth of any of these fungi potential (NPA), and cellulose decomposition under the experimental conditions used. Also growth of bacteria, including strains of Actinomyces, Bacillus, We measured substrate-induced respiration (SIR) (An- and Rhizobium, was strongly inhibited by the DAPG derson & Domsch 1978), soil nitrification potential producer and only slightly by the PCA producers. (NPA) (Stienstra et al. 1994), and cellulose decompos- These data indicate that PCA and DAPG production ition (Harrison et al. 1988), during the growing season. by the GMMs increases their in vitro inhibitory activ- SIR represents the activity of the total metabolically ity towards plant pathogenic as well as saprophytic active soil microbial community. NPA is very sensit- soil fungi, in addition DAPG production results in ive to perturbation in the microbial community (Tiedje increased activity against bacteria. et al. 1989). Cellulose decomposition is a microbial activity of numerous soil fungi. Effect of the GMMs on symptoms of take-all in wheat

Determination of plant growth In soil inoculated with Ggt, WCS358r and also GMM 2 did not suppress symptoms of take-all in wheat. We evaluated plant growth by determining plant However, GMM 8 did suppress activity of the fungal height, plant fresh weight and dry weight of 15 plants pathogen, even to a higher extent than the naturally per plot at each sampling date. Plant dry weight was PCA-producing Ggt-suppressive strain P. ﬂuorescens determined after incubation of the plant material for 3 2-79 (data not shown). These results demonstrate that –5daysat70◦C. a sufﬁcient level of PCA production by the GMMs leads to acquisition of biocontrol activity. Results Population dynamics of WCS358r and the GMMs and GMMs stability of the phz and phl genes in the rhizosphere

PCA-producing and DAPG-producing, kanamycin- In all years populations of WCS358r and the GMMs 7 resistant derivatives of P. putida WCS358r were ob- decreased from about 10 CFU per gram of rhizo- 2 4 tained by transposition of the phz or the phl biosyn- sphere sample to 10 to 10 CFU per gram at harvest, 2 3 thetic gene locus into the chromosome of WCS358r and to near the detection limit (10 to 10 CFU/g rhizo- using the mini-Tn5 transposon delivery system. Two sphere sample) one month after harvesting (131 or PCA-producing GMMs were selected to be included 139 days after sowing). In general no indications were in the field study, GMM 8 and GMM 2. GMM 8 found that the fitness of the GMMs was affected by produced three times as much PCA as GMM 2, and the genetic modification, as numbers of CFUs of the relative to P. fluorescens 2-79 GMM 8 produces two parental strain and the GMMs were comparable. Also times as much PCA in vitro. The DAPG producing no differences were observed between numbers of the derivative selected is GMM P, and in vitro it pro- GMMs on rifampin-containing KB+ with or without duces amounts of DAPG comparable to P. fluorescens kanamycin, suggesting that the phz and phl genes Q2-87, the donor of the phlABCDEF genes. were stable in the bacterial chromosome throughout the growing season. Effects of GMMs on fungal growth in vitro Detection of PCA in rhizosphere extracts using HPLC Three pathogens of wheat, Ggt, Rhizoctonia solani and mass spectrometry AG8 C1 and Pythium sp., were tested for inhibition of growth in vitro by WCS358r and the GMMs. Rhizosphere extracts obtained in the 1998 field trial WCS358r did not significantly inhibit growth of the 12 days after sowing were fractionated using reversed 621

Figure 1. Effects of P. putida WCS358r (W) and its PCA producing derivatives GMM 2 (#2), and GMM 8 (#8) on the composition of the fungal rhizosphere community during the field trial of 1997, at 5 and 90 days after sowing, as determined by ARDRA. C – control treatment. Dendrograms represent percent similarity of fungal communities of roots of wheat plants grown from treated seed. Similarities are based on ARDRA patterns generated from Figure 2. Effects of P. putida WCS358r (W), its PCA producing TaqI-digested, amplified 18S rDNA obtained from wheat rhizo- derivative GMM 8 (#8), DAPG producing derivative GMM P (P), sphere DNA extracts. Two independent replicates per treatment and a mixture of GMM 8 and GMM P (8/P) on the composition were used. of the fungal rhizosphere community during the field trial of 1999, at 25 and 132 days after sowing, as determined by ARDRA. C – control treatment. Dendrograms represent percent similarity of phase HPLC. In extracts of control and WCS358r- fungal communities of roots of wheat plants grown from treated seed. Similarities are based on ARDRA patterns generated from treated wheat rhizosphere no PCA was present. HPLC TaqI-digested, amplified 18S rDNA obtained from wheat rhizo- chromatograms of rhizosphere extracts of wheat plants sphere DNA extracts. Two independent replicates per treatment treated with GMM 2 and GMM 8 had peaks with the were used. same retention time as standard PCA and the presence of PCA was confirmed by mass spectrometric analysis of these peaks. Comparison of the heights of the PCA per treatment are more similar to each other than to peaks suggests that PCA production in the rhizosphere other patterns. Effects on the fungal microflora as a by GMM 8 is higher than the production by GMM 2. result of bacterization with WCS358r or the GMMs seemed differential, since the ARDRA profiles from Effects on the composition of indigenous fungal and the GMM-treated samples clustered separately from bacterial microflora the WCS358r-treated samples and from the control treatment. Effects of the GMMs could be observed up Seed application of both WCS358r and the PCA- to 40 days (1997) and 89 days (1998) after sowing, producing GMMs caused a shift in the fungal popu- whereas WCS358r-induced effects were detectable up lation of wheat roots, as indicated by cluster analysis to 12 and 40 days, respectively. In both years all treat- of replicate ARDRA-generated profiles of rhizosphere ments cluster together 1 month after harvest, indicat- samples (Figure 1). Treatments are considered to be ing that the effects induced by the bacterial treatments different, if both replicate ARDRA patterns of one were transient. In 1999 and 2000, next to effects on the treatment cluster together, apart from patterns of other fungal microflora, effects on the bacterial microflora treatments. In this case the replicate ARDRA patterns were analyzed. In Figure 2 results for the fungal mi- 622 croflora, based on 18S rDNA ARDRA patterns, from in the wheat rhizosphere, indicating that the extra the 1999 experiment are shown. The DAPG producing metabolic load did not affect the ecological fitness of derivative of WCS358r caused a shift in the fungal mi- the GMMs. croflora that lasted until the end of the growing season. In the experiments of 1997 and 1998, both intro- For the bacterial microflora a transient shift up to 40 duction of WCS358r and the PCA-producing GMMs days was observed for the treatments with the DAPG resulted in a transient effect on the composition of producers. In 2000, however, no distinct clustering the rhizosphere fungal microflora, as determined by patterns could be observed, and similarity between 18S rDNA analysis. The distinct effects of WCS358r replicate samples was low, suggesting that the nat- and the GMMs were most prominent at the beginning ural heterogeneity of microbial populations exceeded of these field trials, when the numbers of introduced possible effects of the GMMs. bacteria were relatively high (Glandorf et al. 2001). The WCS358r-induced effect on the fungal microflora Effects on soil microbial activities is probably caused by the production of pseudobactin 358, the fluorescent siderophore of WCS358 (Bak- Values obtained for substrate-induced respiration ker et al. 1986). Siderophore production by WCS358r (SIR), soil nitrification potential, and cellulose decom- is a prerequisite for suppression of fusarium wilt in position fluctuated throughout the growing season for carnation and radish by this strain (Duijff et al. 1994; all treatments, and there were no significant effects of Raaijmakers et al. 1995). GMM-induced impact on the introduction of either WCS358r or the GMMs on the composition of the fungal microflora lasted longer these microbial activities. than the WCS358r-induced impact. The observation that the GMM-induced shift in the fungal microflora Effects on plant growth was longer lasting and differed qualitatively from the There was no effect of the introduction of WCS358r shift caused by the parental strain, indicates that the and the GMMs on plant height, plant fresh weight, or PCA produced by the GMMs also affected the com- plant dry weight (data not shown) in either 1997, 1998, position of the fungal microflora. The detection of or 1999. In 2000, all bacterial treatments significantly PCA in the rhizosphere of GMM-treated plants and increased plant yield by approximately 35%. not in rhizosphere samples of WCS358-treated plants and control plants supports the role of PCA in these shifts in the fungal microflora. Discussion In 1999, introduction of the DAPG producing GMM, either as a single application or in the com- The ability to produce PCA or DAPG was intro- bination with the PCA producer, had a long lasting duced into a plant growth-promoting bacterial strain, effect on the rhizosphere fungal microflora, as determ- P. putida WCS358r, using the mini-Tn5 transposon ined by 18S rDNA analysis. For the same treatments system as a delivery vector. The production of PCA a transient effect was observed on the bacterial micro- increased antifungal activity of WCS358r and DAPG flora, based on 16S rDNA analysis. It was expected production resulted in an enhanced ability to inhibit that the intensity of the effects would increase with growth of both fungi and bacteria. The PCA pro- repeated introduction of the bacterial strains in the ducing GMM 8 was suppressive toward take-all of same plot. However, in 2000 no clear effects of bac- wheat in bioassays. Thus, this genetic modification terial treatments were observed on either the fungal or indeed conferred biocontrol activity to WCS358r. For the bacterial microflora. We speculate that by plant- the DAPG producing GMM we are currently invest- ing wheat for a second time in the same plots, each igating its disease-suppressive properties. Based on plot developed its specific microflora, reflected in a the importance of DAPG production in take-all de- high diversity of the 16S and 18S rDNA ARDRA pat- cline (Raaijmakers & Weller 1998) we expect this terns. Thus in this year of the experiment effects of the GMM also to have improved biocontrol activity. The introduced GMMs did not exceed those of natural vari- environmental fitness of a genetically modified mi- ation. In the 2000 experiment we also observed that croorganisms might be affected by the modification seed treatment with bacteria resulted in increased plant (De Leij et al. 1998). In this study no significant effects growth. This plant growth promotion was independent of the genetic modification on population densities of of the ability of the bacteria to produce PCA or DAPG. the PCA and DAPG producing GMMs were observed The wheat monoculture may have led to accumula- 623 tion of a deleterious microflora, which is subsequently Ministry of Housing, Spatial Planning and the Envir- suppressed by introduction of WCS358r. onment. We thank Bas Valstar, Jeroen van Schaik en Individual species involved in the observed shifts Fred Siesling (Botanical Gardens, Utrecht University) in composition of the fungal and the bacterial mi- for construction of the experimental field and taking croflora have as yet not been identified. Our future care of the field site. research will focus on identification of fungal species that are affected by the introduced GMMs by sequencing of discriminating fungal amplified 18S rDNA fragments after their separation by TGGE or References DGGE and by analysis of cloned 18S rDNA sequences (Kowalchuk et al. 1997; Smit et al. 1999), and also Anderson JPE & Domsch KH (1978) A physiological method for on identification of bacterial species using 16S rDNA the quantitative measurement of microbial biomass in soils. Soil Biol. Biochem. 10: 215–221. fragments (Smit et al. 2001). Austin HK, Hartel PG & Coleman DC (1990 Effects of genetically There were no detectable effects of WCS358r and altered Pseudomonas solanacearum on predatory protozoa. Soil the GMMs on the measured soil microbial activities, Biol. Biochem. 22: 115–117. Bakker PAHM, Lamers JG, Bakker AW, Marugg JD, Weisbeek PJ & indicating that the observed transient perturbations in Schippers B (1986) The role of siderophores in potato tuber yield the microflora did not significantly affect these para- increase by Pseudomonas putida in a short rotation of potato. meters. The application of RT-PCR on 16S and 18S Neth. J. Plant Pathol. 92: 249–256. rRNA followed by DGGE or TGGE, is informative Bangera MG & Thomashow LS (1999) Identification and characterization of a gene cluster for synthesis of the polyketide anti- about the predominantly active microbial populations biotic 2,4-diacetylphloroglucinol from Pseudomonas fluorescens (Duineveld et al. 2001). Use of this technique will be Q2-87. J. Bacteriol. 181: 3155–3163. a valuable addition to our field studies. Bonsall RF, Weller DM & Thomashow LS (1997) Quantification of The results of these experiments show that in- 2,4-diacetylphloroglucinol produced by fluorescent Pseudomo- nas spp. in vitro and in the rhizosphere of wheat. Appl. Environ. troduction of PCA-producing GMMs can transiently Microbiol. 63: 951–955. affect the composition of the rhizosphere fungal mi- Cook J, Bruckart WL, Coulson JR, Goettel MS, Humber RA, croflora of field-grown wheat. When introduced for Lumsden RD, Maddox JV, McManus ML, Moore L, Meyer SF, the first time, the DAPG-producing GMM had a longer Quimby PC, Stack JP & Vaughn JL (1996) Safety of microorganisms intended for pest and plant disease control: a framework lasting effect on both the bacterial and fungal mi- for scientific evaluation. Biol. Control 7: 333–351. croflora, This effect was no longer observed at the De Leij FAAM, Sutton EJ, Whipps JM, Fenlon JS & Lynch beginning of the following season, and, contrary to JM (1995) Impact of field release of a genetically-modified our expectation, no enhanced effects were observed Pseudomonas fluorescens on indigenous microbial populations of wheat. Appl. Environ. Microbiol. 61: 3443–3453. by repeated introduction. De Leij FAAM, Thomas CE, Bailey MJ, Whipps JM & Lynch For future studies this field release has provided JM (1998) Effect of insertion site and metabolic load on the the development of tools to detect effects on micro- environmental fitness of a genetically-modified Pseudomonas fluorescens isolate. Appl. Environ. Microbiol. 64: 2634–2638. bial communities. Comparative studies of the impact De Lorenzo, V, Herrero M, Jakubzik U & Timmis KN (1990) Mini- of common agricultural practices, such as plowing or Tn5 transposon derivatives for insertion mutagenesis, promoter crop rotation, on microbial communities are needed probing, and chromosomal insertion of cloned DNA in Gram- to evaluate the importance of the transient shifts we negative Eubacteria. J. Bacteriol. 172: 6568–6572. Duijff BJ, Bakker PAHM & Schippers B (1994) Suppression of observed. We expect that the impact of introduced Fusarium wilt of carnation by Pseudomonas putida WCS358 (genetically modified) microorganisms is only minor at different levels of disease incidence and iron availability. as compared to these practices. Further research will Biocontrol Sci. Technol. 4: 279–288. focus on the nature of the induced effects. Duineveld BM, Kowalchuk GA, Keijzer A, Van Elsas JD & Van Veen JA (2001) Analysis of bacterial communities in the rhizosphere of chrysanthemum via denaturing gradient gel electrophoresis of PCR-amplified 16S rRNA as well as DNA fragments Acknowledgements coding for 16S rRNA. Appl. Environ. Microbiol. 67: 172–178. Engelen B, Meinken K, von Wintzintgerode F, Heuer H, Malkomes HP & Backhaus H (1998) Monitoring impact of a pesticide treat- The 1997 and 1998 field experiments were initiated by ment on bacterial soil communities by metabolic and genetic the Coordination Commission Risk Assessment Re- fingerprinting in addition to conventional testing procedures. search in the Netherlands (CCRO), which is financed Appl. Environ. Microbiol. 64: 2814–2821. Fantroussi ES, Verschuere L, Verstraete W & Top EM (1999) Effect by the Dutch Ministry of Economic Affairs. From of phenylurea herbicides on soil microbial communities estim- 1999 onwards this project is financed by the Dutch ated by analysis of 16S ribosomal-RNA gene fingerprints and 624

community-level physiological profiles. Appl. Environ. Micro- tionships in biological control of Fusarium wilt of radish by biol. 65: 982–988. Pseudomonas spp. Phytopathology 85: 1075–1081. Ferguson A (1980) Biochemical Systematics and Evolution. Robleto EA, Borneman J & Triplett EW (1998) Effects of bac- Blackie, Glasgow, UK. terial antibiotic production on rhizosphere microbial communit- Girlanda M, Perotto S, Moenne-Loccoz Y, Bergero R, Lazarri A, ies from a culture-independent perspective. Appl. Environ. Mi- Defago G, Bonfante P & Luppi AM (2001) Impact of biocon- crobiol. 64: 5020–5022. trol Pseudomonas fluorescens CHA0 and a genetically modified Short KA, Seidler RJ & Olsen RH (1990) Survival and de- derivative on the diversity of culturable fungi in the cucumber gradative capacity of Pseudomonas putida induced or con- rhizosphere. Appl. Environ. Microbiol. 67: 1851–1864. stitutively expressing plasmid-mediated degradation of 2,4- Glandorf DCM, Brand I, Bakker PAHM & Schippers B (1992) dichlorophenoxyacetate (TFD) in soil. Can. J. Microbiol. 36: Stability of rifampin resistance as a marker for root coloniza- 821–826. tion studies of Pseudomonas putida in the field. Plant Soil 147: Smit E, Leeflang P & Wernars K (1997) Detection of shifts in 135–142. microbial community structure and diversity in soil caused by Glandorf DCM, Verheggen P, Jansen T, Jorritsma JW, Smit E, Lee- copper contamination using amplified ribosomal DNA restriction flang P, Wernars K, Thomashow LS, Laureijs E, Thomas-Oates analysis. FEMS Microbiol. Ecol. 23: 249–261. JE, Bakker PAHM & LC van Loon (2001) Effect of genet- Smit E, Leeflang P, Glandorf B, van Elsas JD & Wernars K ically modified Pseudomonas putida WCS358r on the fungal (1999) Analysis of fungal diversity in soil by sequencing of rhizosphere microflora of field grown wheat. Appl. Environ. cloned PCR amplified 18S rDNA and temperature gradient gel Microbiol. 67: 3371–3378. electrophoresis. Appl. Environ. Microbiol. 65: 2614–2621. Harrison AF, Latter PM & Walton DWH (1988) ITE Symposium Smit E, Leeflang P, Gommans S, Van den Broek J, Van Mil S No. 24. Cotton strip assay: an index of decomposition in soils. & Wernars K (2001) Diversity and seasonal fluctuations of the Institute of terrestrial Ecology, Cumbria, UK. dominant members of the bacterial soil community in a wheat Ibekwi AM, Papiernik SK, Gan J, Yates SR, Yang CH & Crowley field as determined by cultivation and molecular methods. Appl. DE (2001) Impact of fumigants on soil microbial communities. Environ. Microbiol. 67: 2284–2291. Appl. Environ. Microbiol. 67: 3245–3257. Stienstra AW, Klein Gunnewiek P & Laanbroek HJ (1994) Repres- King EO, Ward MK & Raney DE (1954) Two simple media for the sion of nitrification in soils under a climax grassland vegetation. demonstration of pyocyanin and fluorescein. J. Lab. Clin. Med. FEMS Microb. Ecol. 14: 45–52. 44: 301–307. Thomashow LS & Weller DM (1988) Role of a phenazine antibiotic Kowalchuk GA, Gerards S & Woldendorp J (1997) Detection and from Pseudomonas fluorescens in biological control of Gaeu- characterization of fungal infections of Ammophila arenaria mannomyces graminis var. tritici. J. Bacteriol. 170: 3499–3508. (Marram grass) roots by denaturing gradient gel electrophoresis Thomashow LS, Weller DM, Bonsall RF & Pierson LS (1990) Pro- of specifically amplified 18S rDNA. Appl. Environ. Microbiol. duction of the antibiotic phenazine-1-carboxylic acid by fluor- 63: 3858–3865. escent Pseudomonas species in the rhizosphere of wheat. Appl. Mavrodi DV, Ksenzenko VN, Bonsall RF, Cook RJ, Boronin AM Environ. Microbiol. 56: 908–912. & Thomashow LS (1998) A seven-gene locus for synthesis of Tiedje JM, Colwell RK, Grossman YL, Hodson RE, Lenski RE, phenazine-1-carboxylic acid by Pseudomonas fluorescens 2-79. Mack RE & Regal PJ (1989) The planned introduction of J. Bacteriol. 180: 2541–2548. genetically-modified organisms: ecological considerations and Muyzer G & Smalla K (1998) Application of denaturing gradient recommendations. Ecology 70: 298–315. gel electrophoresis (DGGE) and temperature gradient gel elec- Torsvik V, Daae FL, Sandaa RA & Ovreas L (1998) Novel tech- trophoresis (TGGE) in microbial ecology. Antonie van Leeuwen- niques for analyzing microbial diversity in natural and perturbed hoek 73: 127–141. environments. J. Biotechnol. 64: 53–62. Naseby DC & Lynch JM (1998) Impact of wild type and genetically- Van Loon LC (1998) Biotechnology as a means to improve bio- modified Pseudomonas fluorescens on soil enzyme activities and logical control of plant diseases. Med. Fac. Landbouww. Univ. microbial population structure in the rhizosphere of pea. Mol. Gent. 63/4b: 1657–1666. Ecol. 7: 617–625. Wang ZM, Crawford DL, Magnuson TS, Bleakley BH & Hertel G Natsch A, Keel C, Hebecker N, Laasik E & Défago G (1997) Influ- (1991) Effects of bacterial lignin peroxidase on organic-carbon ence of the biocontrol strain Pseudomonas fluorescens CHAO mineralization in soil, using recombinant Streptomyces strains. and its antibiotic overproducing derivative on the diversity of Can. J. Microbiol. 37: 287–294. resident root colonising pseudomonads. FEMS Microb. Ecol. 23: Weller DM, Zhang BX & RJ Cook (1985) Application of a rapid 341–352. screening test for selection of bacteria suppressive to take-all of Raaijmakers JM & Weller DM (1998) Natural plant protection by wheat. Plant Dis. 69: 710–713. 2,4-diacetylphloroglucinol-producing Pseudomonas spp. in take- Weller DM, Howie WJ & Cook RJ (1988) Relationship between all decline soils. Mol. Plant-Microbe Interact. 11: 144–152. in vitro inhibition of Gauemannomyces graminis var. tritici and Raaijmakers JM, Leeman M, Van Oorschot MMP, van der Sluis suppression of take-all of wheat by fluorescent pseudomonads. I, Schippers B & Bakker PAHM (1995) Dose-response rela- Phytopathology 78: 1094–1100.