World Journal

of Microbiology

& Biotechnology

12, 589-593

Selection of rhizosphere-competent Pseudomonas strains as biocontrol in tropical soils M.A.V. Aratijo, L.C. Mendonca-Hagler,”

agents

A.N. Hagler and J.D. van Elsas

Pseudomonas strains were isolated from the rhizosphere of maize grown in yellow-red latosol from Rio de Janeiro, Brazil, to serve as a delivery system for heterologous genes and for risk assessment studies in tropical soils. Selected strains were modified by insertion of the cryZVB gene from Bacillus fhuringiensis and tested for pathogenicity gene expression against larvae of a susceptible model species, Anopheles aquasalis. Modified strains BrS and Br12 showed similar survival performance to their parental strains, and presented a viable density of 10’ c.f.u./g dry soil 30 days after release. A strain of P. fluorescens (Br12) that presented positive results for gene expression and the best survival performance, was selected for risk assessment studies in soil microcosms. Key words: GMO,

Pseudomonas fltrorescens, rhizosphere,

tropical soil.

Applications of microorganisms in agriculture include the introduction of bacteria and fungi into soil as biofertilizers, biocontrol agents (Weller & Cook 1983; Watrud et al. 1985) and plant growth stimulators (Gaskins ef al. 1985). However, introduced microorganisms have generally shown a progressive decline in population size, limiting their effectiveness (Liang et a/. 1992; Compeau et al. 1988; van Elsas et al. 1989). Molecular biology techniques have been used to develop genetically modified microorganisms (GMOs) for specific functions such as the biocontrol of insects (Yates et al. 1985; Keeler 1988; van Elsas et al. 1991). Furthermore, environmental pressure on the introduced cells in soil may affect the performance of such GMOs (Jackman et al. 1992; Smit et a/. 1992; van Elsas et al. 1993). To be effective, the GMOs must compete with the indigenous microorganisms during the period needed to complete the intended task. Many studies have investigated the survival and competitiveness of introduced microorganisms and the persistence of GMOs in natural or model

M.A.V. Aralijo, L.C. Mendonca-Hagler and A.N. Hagler are with the lnstituto de Microbiologia. Universidade Federal do Rio de Janeiro, CCS, Bloco I, llha do Fundlo - Rio de Janeiro, RJ, CEP 21941490. Brazil. J.D. van Elsas is with the Institute for Plant Protection - IPO-DLO, P.O. Box 9060, 6700 GW Wageningen, The Netherlands. *Corresponding author.

@ 1996 Rapid Science

ecosystems such as soil microcosms (Lindow & Panopoulos 1988; Orvos et al. 1990; van Elsas et al. 1991; Jackman et al. 1992; van Elsas et al. 1993). None of these reports has shown enhanced fitness of the GMOs compared with their respective parent strains and all have presented a similar population decline with rates dependent on environmental conditions. These results and significant differences between the composition and climatic conditions of tropical and temperature soils (Primavesi 1980), suggest that the selection of suitable bacteria is fundamental to the success of field applications. The present lack of knowledge about undesirable effects of GMOs on the soil ecosystem also prevents their large-scale use (Smit et al. 1992). The fate of GMOs should initially be investigated in contained microcosms mimicking natural environmental conditions to provide a basis on which regulatory agencies can decide on field releases. Our objective was to select bacteria that perform well when reintroduced into tropical soils as carriers of heterologous genes. To assess whether they could serve as carriers in maize, the percentage of fluorescent pseudomonads in natural maize rhizosphere populations in a subtropical soil was determined. Strains with adequate phenotypes were selected for insertion of the cyIVB gene and introduction into soil microcosms to determine their performance.

Publishers World ~oumal of Microbiology & Biotechnology, Vol 12, 1996

589

M.A.

V. AraLjo

Material

et al.

and Methods

Soil and Hunting of Maize We used a clay soil (yellow-red latosol) from the Pinheral farm, Pirai, Rio de Janeiro, Brazil, described previously by Araujo et al. (1993). It was originally moderately acid (pH 4.7) however following amendment with cow manure (1.5 kg/ha) and CaCO, (2.0 kg/ha) the pH was set at approximately 6.9. The waterholding capacity (WHC) was 49.5% w/w. Seeds of Zea mays (cultivar BrlOl) were sown in the field during the summer, when the mean temperatures were 28°C (minimum) and 42°C (maximum). Soil and Root Sampling Viable bacteria were counted in non-rhizosphere (bulk) soil, in rhizosphere soil and on the rhizoplane after 10 and 30 days of seed germination. Five bulk soil samples were taken 20 cm from the stem to a depth of 10 cm and the plants were dug out of the soil for rhizosphere and rhizoplane samples. Roots plus tightly adhering soil represented the rhizosphere sample. Samples were transported on ice to the laboratory for processing. The bulk soil suspension constituted 10 g of soil placed in 95 ml of 0.1% sodium pyrophosphate (NaPP) and supplemented with 0.1% Tween 80 (Wollum 1982). The cells were dislodged by blending at low speed for 3 min. For rhizosphere soil, the roots plus adhering soil were suspended in NaPP solution as for the bulk samples and shaken at 200 rev/min for 45 min. For the rhizoplane suspension, roots were picked from the rhizosphere soil suspension, rinsed in sterile distilled water, placed with 10 g glass beads (3 mm diameter) in 95 ml of NaPP and shaken (200 rev/min for 45 min). All the suspensions were diluted tenfold in 0.25 strength Ringer’s solution (Dickinson et al. 1975). To determine colony forming units (c.f.u.) of total pseudomonads (TP) and total fluorescent pseudomonads (TFP), appropriate dilutions were plated on King’s B agar (BBL) and Sl agar (Gould et al. 1985) containing 80 mg cycJoheximide/L Total heterotrophic bacteria (THB) were counted on 0.1 strength BBL trypticase soy agar (TSA). The total Gram-negative (TGN) counts were obtained on Hajna Gram-negative agar (Difco). Total counts of Bacillus (TB) were made on Luria-Bertani agar (Sambrook et al. 1989) after pasteurization (SO”C, 10 min) of the soil suspension prior to plating. The plates were incubated at 28°C for 24 to 48 h. All counts were obtained in duplicate from 3 individual plants. Genetic Modification of Pseudomonas Strains Fifty dominant colonies with typical fluorescent pseudomonad morphology on Sl plates from the maize rhizosphere and rhizoplane were randomly selected. After purification, the strains were identified using API 20 NE kits and complementary tests according to Krieg & Holt (1984). Strains identified as P. putida (25) and P. fluoresrens (15) were screened for antibiotic resistance and those sensitive to kanamicin (Km) selected. These strains were grown on LB agar supplemented with 50 fig/ml of rifampicin (Rp) to obtain spontaneous mutants. Four P. putida strains designated Brl, Br2, Br3 and Br4 and one strain of P. fluorescens designated Br5 were selected. Modified cultures carrying the Bacillus thuringiensis cryIVB gene constructed as described by van Elsas et al. (1991) and designated Br6, Br7, Br8, BrlO and Br12 respectively. A crylVB gene carrying transposon Tn5 was inserted into the chromosome of the Rp resistant mutants by a standard filter mating procedure using Escher&a co/i strain S17.1 with a suicide plasmid loaded with Tn5::crylVB as a donor (Simon et al. 1983). Transconjugants were shown to have inserts of Tn5::crylVB by colony filter

hybridization with a radioactively labelled crylVB probe (Sambrook et al. 1989). Five transconjugants showing the strongest hybridization signals and colony development similar to the parental strain were selected and maintained at - 20°C in LB medium with 20% glycerol. Survival of Genetically Modified Pseudomonas Strains Samples from the upper 30 cm of soil were sieved (4 mm mesh), air dried to a moisture content of approximately 10% and adjusted to 50% WHC with either a cell suspension or distilled water (control). Soil microcosms were inoculated with the modified or parental bacterial strains at a density of 10’ to 10’ cells/g soil. Samples of 70 g of soil were packed to a bulk density of 1.5 (based on wet weight) in plastic beakers, incubated at 28°C in a humid chamber, and sown with three zea t?mys seeds sterilized by acid hypochlorite (Araujo et al. 1995). Sterile distilled water was added during the experimental period to compensate for evaporation. Selective plating was used to count culturable bacteria from soil microcosms at 1, 7, 14, 21 and 30 days after inoculation. Microorganisms were dislodged from the soil and counted as previously described for bulk soil samples. Counts of culturable GM0 strains were made on Sl medium containing antibiotics (50 pg/ml of both Km and Rp). Cry VIB Gene Expression: Bio-Assay Cultures were grown in 200 ml of LB broth at 28°C overnight and cells washed as described above. Dilutions were made to prepare vials of 100 ml, at 109, 10’. lo7 and IO6 ceJls/ml. Ten Anopheks aquas& larvae (second instar) were placed in each vial and incubated at room temperature for 72 h to test for killing of larvae by S-endotoxin of the Pseudomonas strains. Control systems consisted of vials with and without the same densities of the parent bacterial strains. All assays were done in triplicate. Statistical Analysis Data from duplicate field soil samples and soil microcosms were compared by analysis of variance (ANOVA) and regression analysis, using GENSTAT-5 (Rothamsted Experimental Station) and differences were considered to be significant when P < 0.05.

Results The protocol used in this work showed satisfactory rescue levels, with viable cell counts being consistent with the density

of

the

inoculum

introduced

into

the

soil.

The

St

agar medium was chosen for fluorescent Pseudomonas because our soils contained a high Bacillus population and it was more selective than the King’s B medium, making colonies easier to distinguish, and there was no significant difference between fluorescent pseudomonad counts on these media. Table I shows counts of different indigenous bacterial groups: total heterotrophic bacteria (THB) was around W/g of dry bulk and rhizosphere soil, increasing with no significant difference between days 10 and 30 in the latter. THB counts in the rhizoplane were in the order W/g dry root, and the levels were stable between days 10 and 30. The total Bacillcts (TB) viable counts in the bulk, rhizosphere soil and on the rhizoplane also remained stable. The viable counts of total Gram-negative (TGN) bacteria in

A. Pseudomonas Table 1. indigenous bacterial soil planted with maize.*

populations

Bacterial population

soil

Days

Bulk

in subtropical

Rhizosphere

clay

Rhizoplane

THB

IO 30

9.1 (0.04) 9.6 (0.22)

8.7 (0.29) 9.6 (0.29)

8.2 (0.23) 8.4 (0.23)

TB

10 30

7.3 (0.03) 7.3 (0.18)

7.4 (0.05)

6.7 (0.27)

7.2 (0.11)

6.4 (0.35)

TGN

10 30

7.4 (0.32) 7.7 (0.12)

7.6 (0.12) 7.6 (0.2)

7.2 (0.11) 7.5 (0.18)

TP

10

6.2 (0.2)

7.2 (0.28)

6.7 (0.29)

30

7.0 (0.17)

7.3 (0.23)

7.4 (0.15)

IO

5.8 (0.4)

30

6.8 (0.05)

6.7 (0.41) 7.0 (0.16)

6.6 (0.21) 7.2 (0.11)

TFP

‘All

values

are

log

c.f.u./g

heterotrophic bacteria; negative bacteria; TP, cent pseudomonads. standard deviations samples

of

dry

soil

or

TB, total Bacillus; total pseudomonads;

root;

THB.

total

TGN, total Gram TFP, total fluores-

The values in parentheses represent the between the means of two independent

(P < 0.05).

2J

I 1

5

10

15 Time

Figure

1. Survival

their

parental

strains

(Ml

Br6;

Br3;

(a)

of

(A.)

Br8;

soil

(0)

25

30

(days) modified

Pseudomonas

microcosms,

at 28°C.

genetically in clay

20

Br4;

(0)

BrlO;

(0)

and (0)

Brl;

Br5;

(*I

Br12. The full symbols represent the genetically modified strains and the open symbols represent the parental strains. The dashed represent (LSD)

lines represent the the P. fluorescens = 0.12,

P. putida strain.

strains Least

and the full lines standard deviation

for rhizosphere

bioconfrol

heterotrophic bacteria, showed a significant increase on the rhizoplane in contrast to a decreased percentage in the rhizosphere (Table 2). The percentage of TFP in relation to the TGN showed an increase in both the rhizoplane and rhizosphere. Although the &&US percentages in relation to the THB counts were significantly higher than the fluorescent pseudomonad percentage, they both decreased in the rhizosphere soil and on the rhizoplane 30 days after seed germination. Of the 50 colonies isolated, 30% were identified as Pseudomonas fluorescens, 50% as P. putida and 20% presented characteristics close to both species but with some atypical physiological variation. Insertion of the Tn5::crylVB into the chromosome of the selected Rp-resistant mutant strains of P. putidu (Brl, Br2, Br3, Bra) and P. fluorescens (Br5) showed frequencies of transfer of between lop 5 and lo-‘. All transconjugant colonies which had acquired Km resistance showed a positive hybridization signal with the cyIVB probe. Control plating for spontaneous mutants resistant to Km did not show growth of colonies. All the genetically modified strains in survival experiments in clay soil microcosms behaved similarly to their respective unmodified parental strains (Figure 1). However, two modified strains of I? ptltida (Br6 and BrlO) presented a significantly lower survival capacity than the other strains tested over 30 days of incubation. The strains Br2 and Br7 are not shown because their survival curves were identical to the curves of strain Brl. The highest survival levels were obtained with P. fluorescens strain Brl2, and consequently this strain was selected for our studies on ecological performance. Counts of viable levels of strain Brl2 in the rhizosphere soil showed a slight increase in density as compared with the values obtained from the bulk soil (Figure 2). The bio-assay for expression of the toxin showed positive results only for strain Br12. Its performance on larvae of Anopheles aqtiasalis revealed a significant decrease in the number of surviving larvae at cell densities between 10’ and 10’ c.f.u./ml, as compared with the parental strain or water negative controls (Table 3). Strain Br5 showed results comparable to those of the microcosms containing only water.

P < 0.05.

Discussion all the samples remained at about 10’ c.f.u./g of dry soil or root between the two sampling days. Again, there was no effect of the rhizosphere on the TGN viable counts. Total pseudomonads (TP) and total fluorescent pseudomonads (TFP) varied from lo6 to 10’ c.f.u./g of dry soil and of dry root on the rhizoplane. There was a significant effect of the rhizosphere on the TP and TFP counts of day 10 but not of day 30. The percentage of the TFP in relation to the total

Estimation of the composition of bacterial populations in soil based on plating methods is relative and only provides information for the culturable fractions of soil populations. This was adequate for our purposes since the selection of suitable bacteria for genetic modification implies the use of culturable strains. Bacterial counts on rich nutrient media tend to overestimate the proportions of fast-growing bacteria in total populations, and result in a poor picture of the

World Jomal of Microbrology 6 Bmtecknology. I/o/ 12. 1996

591

M.A. V. Arahjo et al. Table 2. Percentage of fluorescent rhizopiane In the clay soil.

pseudomonads

Sample

and

Bacillus

in indigenous

popuiaHons

Rhizosphere Bacillus

Days Bacterial

spp.

in the

maize

rhizosphere

and

Rhizopiane Fluorescent pseudomonads

Bacillus

spp.

Fluorescent pseudomonads

10

30

10

30

10

30

10

30

5.0

0.4t

1 .o*

0.2t

3.2’

1.ot

2.5t

6.3’

-

25.17

50.2’

-

79.5’

63.17

population

THB

-

TGN TP THB, total significantly

heterotrophic bacteria; higher than corresponding

12.6t

25.2

31.67

50.1’

TGN, total Gram-negative bacteria; values marked with t (P < 0.05).

9

c c

6

spp.

7

5

10

15

20

25

Time (days) Figure 2. Survival of the selected GM0 Br12 and its strain Br5 at 28X, in clay soil planted with maize. (+) strain P. fluorescens Br12; (0) Br5 parental strain. The lines represent the bacterial population in the bulk soil full lines represent the population in the rhizosphere LSD = 0.06. P < 0.05.

Table 3. Bio-assay against Anopheles

of P. ffuorescens aquasalis larvae.

I-endotoxin

Percentage

Strain P. fluorescens P. fluorescens

109 Br12 Br5

10 * 10 80 k 10

producing

surviving

parental modified dashed and the soil.

8112

larvae

(c.f.u. Brl2/mi) 108 10’ 23 2 6 87 + 6

30

33 f6 97 k 6

106 63fll 97 f 6

The number of surviving larvae per recipient were scored over 3 days. Controls done with larvae plus water did not show significant larval death, with a value of 97 + 6. Data given are mean values of 3 replicates f standard deviation.

TP.

total

pseudomonads.

All

values

marked

with

* are

The total heterotrophic bacterial counts on dilute medium and counts of culturable fluorescent pseudomonads should have shown more realistic proportions of these two populations. Nijhuis et al. (1993) reported bacterial population studies that showed an increase in the percentage of fluorescent Pseudomonas spp. in the grass rhizosphere of temperate zone soil during the experimental period. This is in agreement with the significant increase in the percentage of fluorescent pseudomonads in the Gram-negative population on the rhizoplane shown in this study (Table 2). Our results indicated an apparent rhizoplane competence of fluorescent pseudomonads in maize over a period of 30 days indicating that they are candidates for use as biocontrol agents against root pests. The Pseudomonas strains genetically modified by the cryIVB gene insertion had similar levels of survival to their parental strains (Figure I), suggesting no dramatic effects of the insertions on ecological fitness. We selected the best survivor, Br12, to serve as a model biocontrol agent for the maize rhizosphere. Studies of the competition between GMOs and their parent strains in temperate soils have indicated a small competitive disadvantage for the modified strains which was possibly due to an increased metabolic load caused by the presence of additional proteins or to the impairment of gene function important in survival (van Elsas et al. 1991). In contrast our strain (Br12) intended for use in release studies and showing a positive bio-assay test for 6endotoxin production, presented similar behaviour to its parent, exhibiting good establishment in the rhizosphere of soil microcosms even when introduced in mixed culture with its parental strain Br5 (Aratijo etal. 1995). The results obtained with the genetically modified strain Br12 suggested that it is a suitable model for fitness studies in tropical soils.

Acknowledgements total microbial community. Our data on lower nutrient concentration medium revealed a larger variety of colony morphology than that found with higher nutrient media.

This work was supported by an EC-IC grant awarded to JDvE and LCM-H; “Fundado Universitiria JosC Bonif&io”,

A. Pseudomonas UFRJ;

Conselho

Coordena$o Superior

National de

do

~50 Oswald0

Brasil. Cruz

de

Pesquisas

Aperfeicoamento We for

thank cooperation

de Dr

Pessoal

R. Lourenco with

CNPq-RHAE; de

Ensino

from

Funda-

the bio-assay

tests.

References Aratijo, M.A.V., Mendonca-Hagler, L.C., Hagler, A.N. & van Elsas, J.D. 1993 Survival of genetically modified Pseudomonas fltlorescens introduced into subtropical soil microcosms. FEMS Microbiology Ecology 13, 205-216. Aratijo, M.A.V., Mendonca-Hagler, L.C., Hagler, A.N. & van Elsas, J.D. 1995 Competition between genetically modified Pseudomonas fluorescens introduced into subtropical soil microcosms. Revista de Microbio2ogia 26, 6-15. Compeau, G., Al-Achi, BJ., Platsouka, E. & Levy, S.B. 1988 Survival of rifampicin-resistant mutants of Pseudomonas &orestens and P. putidu in soil systems. Applied and Environmental Microbiology 54, 2432-2438. Dickinson, C.H., Austin, B. & Goodfellow, M. 1975 Quantitative and qualitative studies of phylloplane bacteria from Lolium perenne. ]ownal of General Microbiology 91, 157-166. Gaskins, M.H., Albrech, S.L. & Hubbell, D.H. 1985 Rhizosphere bacteria and their use to increase plant productivity: a review. Agriculture, Ecosystems and Environment 12, 99-116. Gould, W.D., Hagedorn, C., Bartinelli, T.R.C. & Zablotowicz, R.M. 1985 New selective media for enumeration and recovery of fluorescent pseudomonads from various habitats. Applied and Environmental Microbiology 49, 28-32. Jackman, SC., Lee H. & Trevors, J.T. 1992 Survival, detection and containment of bacteria. Microbial Releases 1, 125-154. Keeler, K.H. 1988 Can we guarantee the safety of genetically engineered organisms in the environment? CRC Critical Review in Biotechnology 8, 85-97. Krieg, R.N. & Holt, J.G. 1984 Bergey’s Mnnllal of Systematic Bacteriology, eds Krieg, R.N. & Holt, J.G. volume I, pp. 140210. Baltimore: Williams & Wilkins. Liang, L.N., Sinclair, J.L., Mallory, L.M. & Alexander, M. 1982 Fate in model ecosystems of microbial species of potential use in genetic engineering. Applied and Environmental Microbiology 44,708-714. Lindow, SE. & Panopoulos, NJ. 1988 Field test of recombinant ice- P. syringe for biological frost control in potato. In The release of genetically engineered microorganisms, eds Sussman, M., Collins, C.H., Skinner, F.A. & Stewart-Tull, D.E. pp. 121-138. London: Academic Press. Nijhuis, E.N., Maat, MJ., Zeegers, I.W.E., Waalwijk, C. & van Veen, J.A. 1993 Selection of bacteria suitable for introduction into the rhizosphere of grass. Soil Biology and Biochemistry 7, 885-895.

for

rhizosphere

biocontrol

Orvos, D.R., Lacy, G.H. & Cairn, J., Jr., 1990 Genetically engineered Erzuinia carotovora: survival, intraspecific competition and effects upon selected bacterial genera. Applied and Environmental Microbiology 56, 1689-1694. Primavesi, A. 1980 Munejo ecol&ico do solo. p. 88. Sao Paulo, Bras& Editora Nobel. Sambrook, J., Maniatis, T. & Fritsch, E.F. 1989 Molecular cloning: u laboratory manual. Cold Spring Harbor, NY: Cold Spring Harbor Laboratory Press. Simon, R., Priefer, V. & Puhler, A. 1983 A broad host range mobilization system for in vivo genetic engineering: transposon mutagenesis in Gram negative bacteria. Biol Technology 1, 784-791. Smit, E., van Elsas, J.D. & van Veen, J.A. 1992 Risks associated with the application of genetically modified microorganisms in terrestrial ecosystems. FEMS Microbiology Reviews 88, 263-278. van Elsas, J.D., Clegg, CD., Anderson, J.M., Lapin-Scott, H.M. & Wolters, A. 1993 Fitness of genetically modified Pseudomonas fluorescens in competition for soil and root colonization. FEMS Microbiology Ecology 13, 259-272. van Elsas, J.D., van Overbeek, L.S., Feldmann, A.M., Dullemans, A.M. & de Leeuw, 0. 1991 Survival of genetically engineered Pseudomonas fluorescens in soil in competition with the parent strain. FEMS Microbiology Ecology 85, 53-64. van Elsas, J.D., Trevors, J.T., van Overbeek, L.S. & Starodub, M.E. 1989 Survival of Pseudomonas fluorescens containing plasmids RP4 and pRK2501 and plasmid stability after introduction into two soils of different texture. Canadiarz ]o~rnal of Microbiology 35.951-959. Watrud, L.S., Perlak, F.J., Kusano, K., Mayer, E.J., Miller-Wideman, M.A., Obukowicz, M.G., Nelson, D.R., Kreitinger, J.P. & Kaufman, R.J. 1985 Cloning the Bacilhts thuringiensis subsp. ktlrstaki 6.endotoxin gene into Pseudomonas fluorescens: molecular biology and ecology of an engineered microbial pesticide. In Engineered Organisms in the Environment: Scientific Issues, eds Halvorson, H.O., Pramer, 0. & Rogul, M. pp. 40-46. Washington: American Society for Microbiology. Weller, D.M. & Cook, R.J. 1983 Suppression of take-all of wheat by seed treatments with fluorescent pseudomonads. Phytopathology 73,463~469. Wollun, A.G. 1982 Cultural methods for soil microorganisms. In Methods of soil analysis, eds Page, A.L., Miller, R.H. & Keneey, D.R. part II, p. 427. Madison, WI: American Society for Agronomy. Yates, J.R., Lobos, J.H. & Holmes, D.S. 1985 The use of genetic probes to detect microorganisms in biomining operations. Indian ]ouma/ of Microbiology 1, 129-135.

(Received

in revised

form

WorldJoumal

of

13 March

1996;

accepted

16 March

19961

Microbtology b Bmtechnology, Vol 12. 1996

593