Biol Fertil Soils (2008) 45:83–92 DOI 10.1007/s00374-008-0314-2

ORIGINAL PAPER

Environmental influences on Acinetobacter sp. strain BD413 transformation in soil Stephanie K. Watson & Philip E. Carter

Received: 18 February 2008 / Revised: 19 June 2008 / Accepted: 23 June 2008 / Published online: 16 July 2008 # Springer-Verlag 2008

Abstract The ability of various environmental factors (root exudate from silver tussock, blue tussock, flax, wheat, ryegrass and lupin; simulated-root exudate; moisture; temperature; soil density; salinity; sewage sludge; fertiliser; pesticide) to promote or inhibit transformation of the soil-dwelling bacterium Acinetobacter baylyi BD413 (pFG4ΔnptII) was investigated using soil microcosm studies. A marker-rescue system was used to monitor the transfer of a functional nptII gene from exogenous chromosomal DNA to A. baylyi BD413 (pFG4ΔnptII). Significant differences were detected in A. baylyi BD413 (pFG4ΔnptII) transformation rates in three sterile New Zealand agricultural soils. Addition of simulated-root exudate to the sterile soil was essential for transformation of A. baylyi BD413 (pFG4ΔnptII) in the soil types tested, but addition of plant exudates collected from a variety of New Zealand cropping and native plants did not promote transformation rates to above detectable limits. Increases in soil temperature and bulk density increased the transformation rate but this effect was not consistent across all three soil types. Application of sewage sludge to sterile soils significantly increased transformation in the sandy soil but not in the silt loam and fine sandy loam soil types. Fertiliser (superphosphate) and herbicide (glyphosate) applied at agronomic rates did not affect transformation rates; however, when used at 5× and 50× the agronomic rate respectively, transformation was significantly reduced in all three sterile soils. These results suggest that competence and transformation of the A. baylyi BD413 (pFG4ΔnptII) in soils is highly dependent on the presence of nutrients and is also influenced by the soil texture. S. K. Watson : P. E. Carter (*) Institute of Environmental Science & Research Ltd., 34 Kenepuru Drive, P.O. Box 50-348, Porirua, New Zealand e-mail: [email protected]

Keywords Horizontal gene transfer . Root exudates . Transformation . Genetically modified plants

Introduction Horizontal gene transfer (HGT) allows bacteria to sample from the total gene pool present within an environment. It is widespread among soil-dwelling bacteria (De Vries and Wackernagel 2004) and as such has been identified as a risk factor when associated with the cropping of genetically modified (GM) plants (Nielsen et al. 1997a). Despite the presence of vast tracts of GM plant crops worldwide documented proof of HGT events in situ remain elusive due to significant methodological problems (Badosa et al. 2004; Heinemann and Traavik 2004; Bae et al. 2007). Alternative microcosm-based laboratory experiments have been employed to provide some data on the likelihood of HGT occurring in the environment (Nielsen et al. 2004). A number of studies have used the highly competent bacterial host Acinetobacter sp. which can take up DNA from any species. This ability has allowed the development of a marker-based rescue system to provide a simple HGT detection method (Gebhard and Smalla 1998) which overcomes the normal barriers to horizontal gene transfer in the wild (Nielsen 1998). Tepfer et al. (2003), De Vries and Wackernagel (2004), Mercier et al. (2006) and Simpson et al. (2007) summarise both the issues and the studies conducted under laboratory conditions to investigate transfer in the environment of GM plant DNA to soil bacteria via transformation. At least eight different GM plant species have been investigated and a variety of soil bacterial species have been transformed (Davison 1999). Most soil bacteria, however, are in a non-competent state and do not become competent until there is an up-shift in

84

nutrient availability (Demanèche et al. 2001). Uren (2001) and van Elsas (1992) reported that nutrient flow in the rhizosphere, which provides a variety of carbon-rich organic compounds, may provide a ‘hotspot’ for competence development. High levels of phosphates have also been shown to promote HGT in soil microcosms presumably due to the presence of the necessary cations involved in competency development (Nielsen and van Elsas 2001). Moreover, the heterogeneity of soil can influence microbial activity and composition as well as nutrient availability and certain clay minerals are capable of binding DNA, providing protection from nucleases (Stotzky 1966, 1986; Crecchio et al. 2005; Nielsen et al. 2006; Frey 2007). These studies highlight factors that are all capable of influencing HGT in soil-dwelling bacteria. The effect of various farming practices, such as the application of pesticides and herbicides, or the addition of sewage sludge to soils, to influence the ability of soil bacteria to take up plant DNA has yet to be fully investigated. The influence of these and other environmental factors on transformation rates within a model system were investigated using three New Zealand soils. A modified soil microcosm system (Nielsen et al. 1997b) allowed us to tightly control single factors as well as optimise the system at multiple points to enable the reproducible detection of HGT events. Transformation of bacteria in soils has been shown to be stimulated by the presence of simulated or actual root exudate (Nielsen et al. 1997b; Nielsen and van Elsas 2001; Tolove et al. 2003). The effect of simulated-root exudate or exudates taken from native and cropping New Zealand plants on transformation rates in New Zealand soils was also investigated.

Materials and methods Recipient bacteria and donor DNA The bacterium A. baylyi BD413 (pFG4ΔnptII) was the recipient strain used in all transformations (Gebhard and Smalla 1998). Bacterial stocks were grown according to Nielsen et al. (1997b). Briefly, mid-exponential cells were centrifuged (3,000 ×g for 5 min) and pellets resuspended in 1/10 volume of Luria–Bertani (LB) broth/20% glycerol supplemented with rifampicin and ampicillin (Bertani 1951). One-millilitre stocks of bacterial cells were stored frozen at −80°C ready for use in subsequent transformation experiments. All antibiotics used in this study were used at a concentration of 50 μg mL−1. Purified plant or bacterial DNA containing a functional nptII gene, encoding kanamycin resistance, was used as the donor DNA in the microcosm experiments. Bacterial chromosomal DNA was purified from A. baylyi BD413::

Biol Fertil Soils (2008) 45:83–92

KTG (containing nptII) (Nielsen et al. 1997b) and A. baylyi BD413 (control wild type DNA) using the Genomic DNA kit (Qiagen, Australia) according to manufacturer’s instructions. Chromosomal DNA was extracted from young tobacco leaves (Nicotiana tabacum cv. Samsun) containing the nptII and Cr1Ac9 genes (Beuning et al. 2001) and young non-modified tobacco leaves (control DNA) using Plant DNAzol (Invitrogen, USA) according to manufacturer’s instructions. Filter transformations To ensure all genetically modified DNA was capable of transforming A. baylyi BD413 (pFG4ΔnptII), filter plate transformations were carried out according to Nielsen et al. (1997b) and prior to the soil microcosm experiments. Briefly, pre-wet (water) 0.22-µm filters were placed on top of LB plates containing rifampicin and air-dried for 5 min. Purified DNA (in a final volume of 100 µL) and 100 µL of competent A. baylyi BD413 (pFG4ΔnptII) cells (total 2.0× 107) were mixed and spread evenly over the filter. Plates were incubated for 24 h at 30°C. The following day, the filters were removed into a sterile 10-mL glass bottle and cell growth scraped into 1 mL saline. The cell suspension was serial diluted and plated onto LB agar containing rifampicin, ampicillin and kanamycin to enumerate transformed cells and onto rifampicin and ampicillin to enumerate total recipient cell numbers. Soils Three New Zealand horticultural and agricultural soils were tested for their ability to influence the rate of transformation under a variety of conditions. Soil samples were collected from three sites: a sandy soil from Otaki, North Island (Foxton); a silt loam from Waikato, North Island, (Horotiu); and a fine sandy loam from Lincoln, South Island (Templeton). Samples were collected from the top 10 cm of the soil profile, sieved, gamma-irradiated (50 rad) and stored at −20°C. Non-sterile soils were collected as described above, sieved (2 mm) and used immediately. Soil characteristics are shown in Table 1. Soil moisture content was determined after overnight drying at 105°C. Maximum water holding capacity (MWHC) was calculated using the free-draining method described by Cassel and Nielsen (1986). Soil bulk density was measured by calculating the weight of soil per cubic centimetre of microcosm tube. To investigate effects of an increase in soil density, mimicking increased animal stocking the density the soil was compacted by a factor of three. Total C and N were measured using a Leco furnace followed by a thermal conductivity measurement for N2 analysis and an infrared measurement for the CO2 accord-

Biol Fertil Soils (2008) 45:83–92

85

Table 1 Physico-chemical properties of soils described in this study Soil series

Texture

% Sand % Silt % Clay pH pH (rhizo) Ca (Ex) cmol(+)/kg Total N (%) Total C (%) CEC me./100 g

Horotiua Templetona Foxtona Flevob Edeb

Silt/loam Fine sandy/loam Sandy Silt loam Sandy/loam

12 46 79 14.3 84.8

56 30 14 59.7 12.2

20 24 7 26.0 3.0

5.8 5.9 5.8 7.5 6.2

6.6 6.6 6.5 N/Ac N/A

2.90 0.65 5.26 N/Dd N/D

0.46 0.24 0.38 0.17 0.13

8.20 4.40 0.38 N/D N/D

23.4 13.3 4.3 29.0 9.0

a

Soils used in this study are the three New Zealand agricultural soils (Horotiu, Templeton and Foxton) Soil characteristics of The Netherlands soils Flevo and Ede (van Elsas et al. 1986) c N/A denotes analyses not applicable d N/D denotes analyses not determined b

ing to Metson et al. (1979). Particle size analysis was performed on all three soils to determine the percentage of clay, silt and sand. Exchangeable calcium (Ca (Ex)) was determined according to Daly et al. (1984). Cation exchange capacity (CEC) and soil pH was determined according to Blakemore et al. (1987). Soil transformations Soil microcosms were prepared using 1.2 g dry weight equivalent sterile soil and lightly packed (1.2 g cm−3) into 5-mL air-tight sterile polypropylene vials. All microcosms were run at 66% (w/w) MWHC except where moisture was the condition being tested. The treatments used are shown in Table 2. After weighing, the microcosm was equilibrated at the designated temperature for 24 h prior to inoculation. Recipient bacteria A. baylyi BD413 (pFG4ΔnptII) cells were removed from the freezer, washed and resuspended in sterile saline. A sample of 100 μL (108 cfu total) was inoculated onto the centre of the top of the soil microcosm.

Root exudates (simulated or actual) were added after 24 h where necessary, followed by 40 μg of chromosomal bacterial DNA 1 h later. Where amounts of DNA varied the total moisture content of the microcosm was adjusted accordingly. Microcosms were incubated for a further 24 h at the designated temperature. Harvesting consisted of the removal of soil cores from each microcosm. Approximately 50% of the dry weight of the microcosm was removed using the wide end of a sterile 1-mL pipette tip and the core was resuspended in 900 μL sodium pyrophosphate (0.1%) and 50 μL of 5 mg mL−1 DNAse. Eight sterile glass beads (2.5 mm) were then added and the mixture was vortexed for 5 s. Total numbers of recipient cells were enumerated by plating onto LB media plates supplemented with rifampicin and ampicillin. The efficiency of recovery using the glass beads based on the recipient numbers added and recovered was 90%. Total numbers of transformed cells were enumerated on LB plates containing rifampicin, ampicillin and kanamycin. Plates were incubated at 30°C for 48 h. Effect of soil factors

Table 2 Outline of microcosm treatments Treatment

% WHC Temperature Notes

20°C (standard) 10°C 30°C Low moisture High moisture High density Saline Biosolids Superphosphate

66 66 66 56 86 66 66 66 66

20°C 10°C 30°C 20°C 20°C 20°C 20°C 20°C 20°C

10 5× superphosphate 66 11 10× superphosphate 66 12 Glyphosate 66

20°C 20°C 20°C

13 10× glyphosate 14 50× glyphosate

20°C 20°C

1 2 3 4 5 6 7 8 9

66 66

N/A N/A N/A 56% MWHC 86% MWHC Increased 1/3 To 1.5% 200 kg N/ha Agronomic rate (500 kg/ha) 5× agronomic rate 10× agronomic rate Agronomic rate (3.6 g/L) 10× agronomic rate 50× agronomic rate

Soil moisture impacts were investigated using a lower limit of 56% MWHC and upper limit of 86% MWHC. These values were chosen as they were the largest range of moisture that could be practically accommodated in the system and consistently across the three soils. Soil temperatures of 10 and 30°C were chosen to test the extremes of the system. The impact of a saline soil on HGT frequencies was investigated by adding sterile saline solution (1.5% NaCl) to soil after weighing. Soil density was also investigated as some New Zealand agricultural practices (e.g. mob stocking and winter root crop feeding) can result in soil compaction due to increased stock density or grazing. Original soil density was measured and increased by compressing the soil by a factor of 3. Anaerobically digested sewage sludge (20.81% solid; 5.3%N; 1.80% P; 2.03% Ca; and 0.24% Mg) is a possible future soil amendment (McLaren et al. 2007; NZWWA 2003). A sample was sterilised by autoclaving and added, at the

86

agronomic rate of 200 kg N ha−1, to soil prior to weighing (0.35 g g−1 dry weight) and mixed lightly into the top 50% of the soil. Simulated-root exudate consisted of the following: 5× M9 salts (240 mM Na2HPO4, 110 mM KH2 PO4, 45 mM NaCl, 95 mM NH4Cl), 25× P salts (1.2M Na2HPO4, 550 mM KH2PO4), 0.4% of lactate, succinate, acetate, citrate, tartrate and 0.7 μg mL−1 glutamic acid, 0.6 μg mL−1 proline, 0.6 μg mL−1 alanine, 0.4 μg mL−1 glycine, 0.2 μg mL−1 leucine, 0.3 μg mL−1 serine, 0.3 μg mL−1 arginine, 0.1 μg mL−1 glutamine and 0.1 μg mL−1 valine according to Nielsen and van Elsas (2001). Glyphosate herbicide (Roundup Ready Renew, Monsanto) was applied at the agronomical recommended rate of 3.6 g L−1 and higher, immediately prior to the addition of DNA. Superphosphate fertiliser (0-9-0-11.5-20; N-P-K-S-Ca, Ravensdown, New Zealand) was added at the agronomic rate of 500 kg ha−1 and higher, at the time of soil weighing (4.2 mg g−1 dry weight soil) and mixed lightly into the top 50% of the soil. Seed germination and root exudate collection Approximately a dozen seeds each of Poa cita (silver tussock), Poa colensoi (blue tussock), Phormium tenax (flax), Triticum aestivum (wheat), Lolium perenne (ryegrass), Lupinus polyphyllus (lupin) were surface sterilised by incubating (shaking) in 80% ethanol (30 min), sterile distilled water (2 min) and 4.5% sodium hypochlorite/0.1% Triton X-100 (30 min) followed by six sterile distilled water washes (2 min). Seeds were aseptically placed on 0.5× potato dextrose agar (PDA) (Oxoid, UK) and incubated at 20°C in the dark for approximately 7 days with the exception of P. tenax which was placed at 4°C and incubated for 42 days. Germinated seedlings were aseptically transferred to fresh 0.5× PDA agars in bottles (100 mL, wide mouth) and incubated at 20°C with a 16h/d photoperiod for 1 week. Seedlings were then carefully extracted from the agar and placed in 15-mL sterile tubes with the root portion of five seedlings submerged in 5 mL 0.05 M CaCl2. Seedlings were incubated at 20°C with a 16-h/d photoperiod and root exudates were collected over a 5-day period (Overbeek and van Elsas 1995; Spiegel et al. 2003). Exudates were then filter sterilised, freeze-dried and resuspended in 1/10 the volume. Content of total organic carbon (TOC) of this solution was determined according to Forster (1995). Bacterial transformations For each experiment, the possibility of transformations occurring on the plate was removed by using DNase in the harvesting buffer. All experiments contained a set of triplicate tubes where either plant or bacterial DNA

Biol Fertil Soils (2008) 45:83–92

containing the nptII gene was added just prior to harvesting. This test always gave negative results. Spontaneous mutations arising in kanamycin resistance were checked for using the bacterial and plant control DNA. No spontaneous mutants were detected in early experiments (total 50 colonies tested) and a selection of putative transformants were routinely plated on fresh LB plates containing rifampicin, ampicillin and kanamycin and screened for the presence of the nptII gene by PCR as described below. All components of the transformation or soil microcosms were individually screened for presence of contaminating bacteria by plating onto appropriate antibiotic-containing LB agar plates. Either plant or bacterial chromosomal DNA without the nptII gene was used as control DNA. Putative transformants were grown on fresh LB plates containing rifampicin, ampicillin and kanamycin and confirmed by PCR amplification. PCR primers amplified across the restored section of the defective nptII in the plasmid pFG4ΔnptII. The forward primer (P1) TGCTAAAGGAAGCGGAAC and reverse primer (P2) AGGTCAACAGGCGGTAAC (Gebhard and Smalla 1999) were used as described by Nielsen and van Elsas (2001). Sequencing of the 16S DNA confirmed they were A. baylyi. Statistical analysis Generalised linear models were fitted for each soil type, each treatment and then across soil types. A treatment is defined as a unique combination of water, temperature and nutrient. Tests were done at 5% level of significance i.e. any P-value <0.05 was considered as statistically significant. SAS version 9.1.3 software was used for the analysis. For statistical analyses, variations in conditions were always compared to the standard microcosm conditions for that soil texture.

Results Plant DNA and bacterial DNA as donor DNA Genetically modified tobacco plant DNA containing a functional nptII gene successfully transformed A. baylyi BD413 (pFG4ΔnptII) cells using the filter transformation method. A transfer efficiency of 1.4×10−7 was achieved using 200 µg of purified GM tobacco DNA. Transformants, however, were not detected when GM tobacco DNA (200 µg) containing the nptII gene was added to either nonsterile or sterile soil microcosms of the three New Zealand soils in the presence or absence of simulated-root exudate and containing A. baylyi BD413 (pFG4ΔnptII) cells.

Biol Fertil Soils (2008) 45:83–92

87

To enable detection of transformants in the soil microcosm system, chromosomal DNA from A. baylyi BD413:: KTG was used as an alternative to the plant DNA as it provided a greater number of nptII gene copies per microgram of DNA and thus would optimise the system. Using the filter transformation method, A. baylyi BD413:: KTG DNA successfully transformed A. baylyi BD413 (pFG4ΔnptII) cells at a frequency of 5.4×10−4 per 40 µg of DNA. However, no transformants were detected in either non-sterile or sterile soil microcosms containing A. baylyi BD413 (pFG4ΔnptII) cells. This result was consistent across of three New Zealand soils with up to 100 µg of purified A. baylyi BD413::KTG DNA being tested.

tobacco DNA (200 µg) did not produce any transformants. These results were consistent across all three sterile and non-sterile New Zealand soils. The addition of simulatedroot exudate to the soil increased the soil pH by from approximately 5.8 to 6.6 as shown in Table 1. Using an optimum amount of A. baylyi BD413::KTG DNA (40 µg), recipient A. baylyi BD413 (pFG4ΔnptII) cells were transformed successfully in all three sterile soils under the following conditions: 20°C; 66% WHC; the addition of the simulated-root exudate (standard microcosm). Transformation frequencies under these optimal conditions were 3.4×10−6 in the silt loam soil, 6.6×10−6 in the fine sandy loam soil and 3.0×10−6 in the sandy soil (Fig. 1, with P-values shown in Table 3). Variability within a soil was tested and results showed HGT frequencies to be reproducible at a significant level (data not shown).

Simulated-root exudates and optimal conditions for bacterial transformation in microcosms The addition of simulated-root exudate to soil increased HGT frequencies significantly above the limit of detection in sterile soil microcosms containing A. baylyi BD413 (pFG4ΔnptII) cells and using the bacterial donor DNA. With the addition of 15 µg of A. baylyi BD413::KTG DNA to sterile soil microcosm, transformation frequencies of 1.4×10−6 was obtained for the sandy soil, 8.3×10−5 for fine sandy loam and 1.1×10−6 for the silt loam soil. However, addition to non-sterile soil microcosms produced no transformants. The addition of simulated-root exudate to either sterile or non-sterile soil microcosms containing modified

Actual root exudates collected from a selection of New Zealand cropping plants (lupin, ryegrass and wheat) and native plants (flax, blue tussock and silver tussock) were tested for their ability to induce transformations in A. baylyi BD413 (pFG4ΔnptII) cells in the fine sandy loam soil with 40 µg of purified A. baylyi BD413::KTG DNA at 20°C and at 66% WHC. No transformation of A. baylyi BD413 (pFG4ΔnptII) cells by A. baylyi BD413::KTG DNA in sterile soil microcosms was observed in the presence of

1.4E-05

1.2E-05

1.0E-05

8.0E-06

6.0E-06

4.0E-06

2.0E-06

fine sandy loam (Templeton)

slu dg su e pe rp ho sp su ha pe te rp ho sp h su ate pe 5X rp ho sp ha te 10 X gl yp ho sa te gl yp ho sa te1 0X gl yp ho sa te 50 X

ne sa li

de ns ity

m hi gh

silt loam (Horotiu) sandy (Foxton)

oi stu re

ur e

°C

m oi st lo w

30

°C 10

°C

*

1.0E-10

20

Transformation frequency

Fig. 1 Transformation frequencies A. baylyi BD314 (pFG4ΔnptII) cells incubated in the three soils under various treatments as outlined in Table 2. Bars represent average data and vertical lines are the standard errors. Significant differences are reported in Table 3

Effect of root exudates

Soil treatments

88

Biol Fertil Soils (2008) 45:83–92

Table 3 P-values for significant differences observed between treatments and standard microcosm Treatment

Foxton (sandy)

Horotiu (silt/loam)

Templeton (fine sandy loam)

10°C 30°C Low moisture High moisture High density Saline Biosolid application Superphosphate 5× superphosphate 10× superphosphate Glyphosate 10× glyphosate 50× glyphosate

<0.0001 N/A N/A N/A <0.0001 0.0005 0.0259 <0.0001 <0.0001 <0.0001 N/A 0.0001 <0.0001

0.0011 0.0001 N/A N/A 0.0217 N/A N/A N/A 0.0242 0.0230 N/A 0.0240 0.0007

0.0039 N/A N/A N/A N/A N/A N/A N/A 0.0116 0.0096 N/A N/A 0.0034

actual root exudates. To determine if the actual root exudates were at a similar concentration to the simulatedroot exudates, the total organic content (TOC) of the six collected root exudates and the simulated-root exudate was determined. The TOC results were very similar for all exudates collected. Even when the simulated-root exudate was diluted 2-fold in sterile soil microcosms, similar transformation frequencies were observed as described above (data not shown). Effect of temperature, moisture, salinity and bulk density Microcosm experiments carried out at 10°C significantly reduced the transformation frequencies for A. baylyi BD413 (pFG4ΔnptII) cells in all three New Zealand soils when compared to the ‘standard’ microcosm frequencies for each soil (Fig. 1 and Table 3). Transformation frequencies decreased from 3.4×10−6 to 1.6×10−7 (P=0.0011) for the silt loam, from 6.6×10−6 to 1.5×10−7 (P=0.0039) for the fine sandy loam and from 3.0×10−6 to 3.8×10−7 (P< 0.0001) for the sandy soil. This also corresponded to a 10fold drop in total bacterial recipient numbers (data not shown). At 30°C, however, the only significant difference (increase) in HGT frequencies was seen in the silt loam soil (7.3×10−6 compared to the standard microcosm, 3.4×10−6). Recipient numbers remained unchanged for all soil types at this temperature. The range of soil moisture levels tested did not have any impact on the transformation frequencies in any of the soils tested when compared to the frequencies found in the standard microcosms (Fig. 1). Similarly, adding saline to the either the silt loam or fine sandy loam had no effect but there was a significant decrease from 3.0×10−6 to a frequency of 7.5×10−7 observed for microcosms with the sandy soil. Increasing the soil bulk density by a factor of 3

caused a significant increase in transformation frequencies from 3.4×10−6 to a frequency of 6.0×10−6 in the silt loam and from 3.0×10−6 to 6.6×10−6 in the sandy soil. Effect of superphosphate, glyphosate and sewage sludge The sandy soil was the only soil to exhibit changes in transformation frequencies when superphosphate was applied at the agronomic rate. Frequencies decreased from 3.0×10−6 (standard microcosm) down to 1.7×10−6 (Fig. 1). Increasing the superphosphate loading to 5× the agronomic rate caused a significant decrease in the transformation frequencies for A. baylyi BD413 (pFG4ΔnptII) in all three soils. Frequencies in the silt loam microcosms decreased from 3.4×10−6 to 3.0×10−8, in the fine sandy loam from 6.6×10−6 to 1.8×10−7 and in sandy soil from 3.0×10−6 down to below detectable levels. Total bacterial recipient numbers in the microcosms with the lower levels of superphosphate were not affected. However, an increase of superphosphate levels to 10× that of the recommended agronomic rate reduced transformation frequencies in all three soils to below detection levels. The total bacterial recipient numbers were also reduced from 1.7×108 total cfu to an average of 3.9×107 in the silt loam soil, 8.3×104 for the fine sandy loam and to an average of 1.3×104 in the sandy soil. At the agronomic rate recommended for Roundup Ready Renew (glyphosate), there was no detectable difference in transfer frequencies of A. baylyi BD413 (pFG4ΔnptII) compared to the standard microcosms for any of the three soils tested. An increase in the application rate by a factor of 10 significantly decreased the transformation frequencies from 3.4×10−6 down to 8.4×10−7 for the silt loam soil and from 3.0×10−6 down to 8.2×10−7 for the sandy soil. Total recipient cell numbers in these microcosms were not affected. Increasing the glyphosate application 50 times decreased transformation frequencies across microcosms of all three soils to just above the detection limit (10−8). Total recipient numbers for each soil were also affected. The soil loam soil microcosm recipients decreased 10-fold, fine sandy loam microcosm recipients decreased a millionfold and the sandy soil microcosm recipients decreased 100-fold. Addition of sewage sludge material to sandy soil significantly increased the transformation frequency to 4.3×10−6 compared to the frequency of 3.0×10−6 in the untreated microcosm. Transformation frequencies in the other two soils were not significantly different in the sewage sludge treated and untreated microcosms. Total recipient numbers were not affected. Sewage sludge was also added to microcosms of all three soils, without the addition of simulated-root exudate; however, these did not yield any detectable transformants.

Biol Fertil Soils (2008) 45:83–92

Discussion Rhizosphere effects The rhizosphere has been suggested as a potential ‘hotspot’ for horizontal gene transfer (van Elsas et al. 2003). This nutrient-rich environment supports a large population of diverse bacterial species by providing carbon and nutrient sources such as organic acids, sugars and polysaccharides, amino acids, fatty acids sterols and many other compounds (Sorensen 1997; Uren 2001). In this study, we showed that a simulated-root exudate comprising of M9 salts, P salts and organic and amino acids, commonly found in root exudates, was the key component in successfully promoting transformation in A. baylyi BD413 (pFG4ΔnptII) cells in New Zealand soils but only when the soil has been sterilised. In non-sterile soils, it appears that the active components of the simulated-root exudate are degraded by the natural microbial flora before they can influence transformation. The simulated-root exudate provided the greatest positive impact on transformation rates in sterile soil microcosms compared with other factors tested. Nielsen and van Elsas (2001) tested individual components of the simulated-root exudate together with phosphate salts and concluded that both carbon and inorganic salt compounds were required for successful transformation of A. baylyi BD413 (pFG4ΔnptII) cells. They also concluded that compounds promoting bacterial growth were also able to promote competence development in A. baylyi BD413 (pFG4ΔnptII) cells. In other studies, actual root exudates have also been shown to increase the microbial activity in the rhizosphere soil (Tolove et al. 2003). Surprisingly, the plant root exudates collected from lupin, flax, ryegrass, wheat and tussock grass could not provide a similar outcome when tested in parallel experiments. Total organic carbon contents of these actual root exudates were compared to the TOC of the simulated-root exudates and at least half were within a similar range potentially indicating that the total amount of carbon-based nutrient compounds was similar. Even in experiments using 2-fold diluted simulated-root exudates the transformation rates were still well above the detection limit. However, the concentration of specific components in the actual root exudates was not determined so a potential key factor might have varied significantly in concentration or been absent in these root exudates or inhibitors of transformation may be present. It is acknowledged that exudate components vary over time and with collection conditions and that these real root exudates only represented one set time point in the growth of the plants (Uren 2001). Nielsen et al. (1997b) tested maize root exudates with A. baylyi BD413 (pFG4ΔnptII) cells and soil from The Netherlands. They were not able to detect transformation

89

in Ede loamy sand soil, however, when Flevo silt loam was used transformation rates were observed above the detection limits and they concluded that nutrient availability played an important role for soil transformations. Effect of soil properties Soil texture was shown to have a significant impact on the transformation frequencies of A. baylyi BD413 (pFG4Δ nptII) cells. Overall, the conditions tested the transformation frequencies observed in the fine sandy loam soil (Templeton) were significantly different (P=0.017) to those observed in the silt loam soil (Horotiu) and sandy soil (Foxton). On average and across all treatments, transformation frequencies were higher in the fine sandy loam soil microcosms and the lowest in the sandy soil microcosms. The greatest observable difference between these three soils is the percentage clay content with the fine sandy loam soil having the highest clay content (24%), the silt loam a 20% clay content and the sandy soil the lowest clay content at 7% (Table 1). Interestingly, Nielsen et al. (1997b) also observed that transformants were consistently detectable in soil with the higher clay content. It has been previously shown that soil minerals (e.g. montmorillonite and kaolinite) are able to adsorb DNA and protect it from nuclease degradation in the soil (Khanna and Stotzky 1992; Demanèche et al. 2001). The exact nature of the protection mechanism is not fully understood (Nielsen et al. 2006); however, the bound DNA still retains its bacterial transforming abilities (Cai et al. 2007). It has also been shown that soil pH and cation concentration may influence this binding and thus influence bacterial transformation rates (Franchi et al. 2003; Cai et al. 2006). Interestingly, overall the three New Zealand soils showed lower transformation frequencies when compared to the two Dutch soils tested by Nielsen et al. (1997b), probably because the New Zealand soils are more acidic. Nielsen et al. (1997a) used the same total number of A. baylyi BD413 (pFG4ΔnptII) cells and significantly less A. baylyi BD413::KTG genomic DNA (10 µg) but were able to produce transformants without the addition of a simulated-root extract or any other nutrient addition. The New Zealand soils did not support the transformation of A. baylyi BD413 (pFG4ΔnptII) cells at the levels found in the Dutch soil. The lower pH of the New Zealand soils might play an important part in the competence development of A. baylyi BD413 (pFG4ΔnptII) cells. Palmen et al. (1993) observed that, by increasing pH from 5.5 to 7.0, the transformation rate of A. baylyi BD413 in liquid media increased. It is important to remember that increases in pH also inversely affect DNA binding and thus DNA stability (Khanna and Stotzky 1992). Simpson et al. (2007) have reported the detection of transformants in non-sterile

90

English soil when using plant nuclear DNA. This highlights the differences in transformation frequencies achieved by different research groups which may be due differences in methodological processes or the soils examined. The application of sewage sludge (sandy soil), the increased soil bulk density (silt loam and fine sandy loam soils) and soil temperature at 30°C (silt loam soil) did not all influence the transformation of A. baylyi BD413 (pFG4ΔnptII) cells in the three New Zealand soils to the same extent. Divalent cations such as Mg2+ and Ca2+ can influence the competency development and thus transformation of A. baylyi BD413 cells (Palmen et al. 1993). The addition of sewage sludge only influenced the transformation rates in the sandy soil, which also had the highest concentration of exchangeable Ca2+ (Table 1). The sewage sludge contains both Mg2+ and Ca2+; however, they may not be at a level significant enough to promote any transformation in the other two soils (McLaren et al. 2007). Increases in soil bulk density could potentially have different effects on each of the three soils due to their different textures and clay contents. The spatial proximities of the A. baylyi BD413 (pFG4ΔnptII) cells, genomic DNA and available nutrients could vary with each of the soils thus influencing the likelihood of all three components being localised at the optimal ‘competent’ time point for the bacterial cells (Hill and Top 1998). It is well established that temperature can affect bacterial activity and A. baylyi BD413 (pFG4ΔnptII) cells grow optimally at 30°C (Palmen et al. 1993). The higher transformations rates observed at this temperature in the soil microcosm experiments in this study confirm similar observations by Nielsen et al. (1997b). The fact that the 30°C temperature affected transformation rates in the silt loam soil but not in the other two soils is probably because other soil factors (e.g. clay content, cations, soil texture) are involved. Addition of glyphosate and superphosphate Superphosphate (0-9-0-11.5-20; N-P-K-S-Ca) did not affect transformation rates at the agronomic rate but at higher agronomic rates transformation rates were significantly decreased. Previous studies by Nielsen and van Elsas (2001) showed fertiliser with a range of N-P-K values positively affected the transformation rates of A. baylyi BD413 (pFG4ΔnptII). Palmen et al. (1993) reported that DNA uptake by A. baylyi in liquid culture was Ca2+ dependent. The superphosphate used in this study was not effective probably because the Ca2+ was not available or the excessive dose may have increased the salt concentration creating osmotic stress on the bacterial cells. This effect will be increased at the higher fertilisation rates (5× and 10×).

Biol Fertil Soils (2008) 45:83–92

The application of glyphosate (Roundup Ready Renew) did not affect transformation rates at the recommended agronomic rate of 3.6 g/L but was inhibitory at higher application rates in all three soils. Contradictory effects of glyphosate on soil microbial activity have been reported (Haney et al. 2002; Hart and Brookes 1996; Wardle and Parkinson 1992). The inhibition of transformation at high application rates is most likely due to the reduced recipient numbers. Busse et al. (2001) reported a toxic effect of glyphosate in soil-free media; however, on addition to soil the toxicity was not evident even when applied at 100× the recommended dose. In conclusion, the detection of transformation in New Zealand soils is dependent on the presence of simulatedroot exudate only when soils were sterile. This suggests that under natural conditions the root exudates are degraded. Further work characterising the important components of root exudate affecting transformation is required. Soil texture and composition can affect transformation rates. Different soils can influence the rate of bacterial transformation therefore an understanding of individual soil characteristics is important in assessing the risk of an HGT event. Application of fertilisers or herbicides does not increase transformation rates. However, application of sewage sludge to sterile New Zealand soils does affect transformation rates but the significance of this in a natural system is unknown. Acknowledgements We thank Natalie Redshaw and Jeremy Grey for their assistance with laboratory analyses. Thanks to Kaare Nielsen and Kornelia Smalla for the use of their bacterial strains and constructs. Special thanks also to Vadakattu Gupta, Tom Speir, Jacqui Horswell and Maureen O’Callaghan for their valuable comments and help. This work was supported by the New Zealand Foundation for Research, Science & Technology.

References Badosa E, Moreno C, Montesinos E (2004) Lack of detection of ampicillin resistance gene transfer from Bt176 transgenic corn to culturable bacteria under field conditions. FEMS Microbiol Ecol 48:169–178 doi:10.1016/j.femsec.2004.01.005 Bae TW, Lee HY, Ryu KH, Lee TH, Lim PO, Yoon PY et al (2007) Evaluation of horizontal gene transfer from genetically modified zoysiagras. J Plant Biotechnol 34:75–80 Bertani G (1951) Studies on lysogenesis. I. The mode of phage liberation by lysogenic Escherichia coli. J Bacteriol 62:293–300 Beuning LL, Mitra DS, Markwick NP, Gleave AP (2001) Minor modifications to the cry1Ac9 nucleotide sequence are sufficient to generate transgenic plants resistant to Phthorimaea operculella. Ann Appl Biol 138:281–292 doi:10.1111/j.17447348.2001.tb00113.x Blakemore LC, Searle PL, Daly BK (1987) Methods for chemical analysis of soils. New Zealand Soil Bureau Scientific Report 80. New Zealand Soil Bureau, DSIR, Lower Hutt Busse MD, Ratcliff AW, Shestak CJ, Powers RF (2001) Glyphosate toxicity and the effects of long-term vegetation control on soil

Biol Fertil Soils (2008) 45:83–92 microbial communities. Soil Biol Biochem 33:1777–1789 doi:10.1016/S0038-0717(01)00103-1 Cai P, Huang Q, Jiang D, Rong X, Liang W (2006) Microcalorimetric studies on the adsorption of DNA by soil colloidal particle. Colloid and Surfaces. B Biointerface 49:49–54 doi:10.1016/j. colsurfb.2006.02.011 Cai P, Huang Q, Chen W, Zhang D, Wang K, Jiang D et al (2007) Soil colloids-bound plasmid DNA: effect on transformation of E. coli and resistance to DNase I degradation. Soil Biol Biochem 39:1007–1013 doi:10.1016/j.soilbio.2006.11.010 Cassel DK, Nielsen DR (1986) Field capacity and available water capacity. In: Klute A (ed) Methods of soil analysis. Physical and mineralogical methods, Part 1. Soil Science Society of America, Madison, pp 901–926 Crecchio C, Ruggiero P, Curci M, Colombo C, Palumbo G, Stotzky G (2005) Binding of DNA from Bacillus subtilis on montmorillonite–humic acids–aluminum or iron hydroxypolymers: effects on transformation and protection against DNase. Soil Sci 69:834– 841 doi:10.2136/sssaj2004.0166 Daly BK, Manu VT, Halavatau SM (1984) Soil and plant analysis methods for use at the Agricultural Research Station, Vaini, Tonga. New Zealand Soil Bureau Laboratory Report AN2. Davison J (1999) Genetic exchange between bacteria in the environment. Plasmid 42:73–91 doi:10.1006/plas.1999.1421 Demanèche S, Jocteur-Monrozier L, Quiquampoix H, Simonet P (2001) Evaluation of biological and physical protection against nuclease degradation of clay-bound plasmid DNA. Appl Environ Microbiol 67:293–299 doi:10.1128/AEM.67.1.293299.2001 De Vries J, Wackernagel W (2004) Microbial horizontal gene transfer and the DNA release from transgenic crop plants. Plant Soil 266:91–104 doi:10.1007/s11104-005-4783-x Forster JC (1995) Soil physical analysis. In: Alef K, Nannipieri P (eds) Methods in applied soil microbiology and biochemistry. Academic, London, pp 105–155 Franchi M, Ferris JP, Gallori E (2003) Cations as mediators of the adsorption of nucleic acids on clay surfaces in prebiotic environments. Orig Life Evol Biosph 33:1–16 doi:10.1023/ A:1023982008714 Frey SD (2007) Spatial distribution of soil organisms. In: Paul EA (ed) Soil microbiology, ecology and biochemistry. Academic, Oxford, pp 283–300 Gebhard F, Smalla K (1998) Transformation of Acinetobacter sp. strain BD413 by transgenic sugar beet DNA. Appl Environ Microbiol 64:1550–1554 Gebhard F, Smalla K (1999) Monitoring field releases of genetically modified sugar beets for persistence of transgenic plant DNA and horizontal gene transfer. FEMS Microbiol Ecol 28:261–272 doi:10.1111/j.1574-6941.1999.tb00581.x Haney RL, Senseman SA, Hons FM (2002) Effect of Roundup Ultra on microbial activity and biomass from selected soils. J Environ Qual 31:730–735 Hart MR, Brookes PC (1996) Soil microbial biomass and mineralization of soil organic matter after 19 years of cumulative field applications of pesticides. Soil Biol Biochem 28:1641–1649 doi:10.1016/S0038-0717(96)00249-0 Heinemann JA, Traavik T (2004) Problems in monitoring horizontal gene transfer in field trials of transgenic plants. Nat Biotechnol 22:1105–1109 doi:10.1038/nbt1009 Hill KE, Top EM (1998) Gene transfer in soil systems using microcosms. FEMS Microbiol Ecol 25:319–329 doi:10.1111/ j.1574-6941.1998.tb00483.x Khanna M, Stotzky G (1992) Transformation of Bacillus subtilis by DNA bound on montmorillonite and effect of DNase on the transforming ability of bound DNA. Appl Environ Microbiol 58:1930–1939

91 McLaren RG, Clucas LM, Speir TW, van Shaik AP (2007) Distribution and movement of nutrients and metals in a Pinus radiata forest soil following applications of biosolids. Environ Pollut 147:32–40 doi:10.1016/j.envpol.2006.08.027 Mercier A, Kay E, Simonet P (2006) Horizontal gene transfer by natural transformation in soil environment. In: Nannipieri P, Smalla K (eds) Nucleic acids and proteins in soils. Springer, Berlin, pp 353–373 Metson AJ, Blakemore LC, Rhoades DA (1979) Methods for the determination of soil organic carbon: a review, and application to New Zealand soils. N Z J Sci 22:205–228 New Zealand Waste Water Association (NZWWA) (2003) Guidelines for the safe application of biosolids to land in New Zealand. NZWWA, Wellington Nielsen KM (1998) Barriers to horizontal gene transfer by natural transformation in soil bacteria. APMIS 84(106):77–84 Nielsen KM, van Elsas JD (2001) Stimulatory effects of compounds present in the rhizosphere on natural transformation of Acinetobacter sp. BD413 in soil. Soil Biol Biochem 33:345–357 doi:10.1016/S0038-0717(00)00147-4 Nielsen KM, Ray JL, van Elsas JD (2004) Natural transformation in soil: microcosm studies. In: Kowalchuk GA, de Bruijn FJ, Head IM, Akkermans AD, van Elsas JD (eds) Molecular microbial ecology manual. Springer, New York, pp 1–12 Nielsen KM, Bones AM, van Elsas JD (1997a) Induced natural transformation of Acinetobacter calcoaceticus in soil microcosms. Appl Environ Microbiol 63:3972–3977 Nielsen KM, van Weerelt MDM, Berg TN, Bones AM, Hagler AN, van Elsas JD (1997b) Natural transformation and availability of transforming DNA to Acinetobacter calcoaceticus in soil microcosms. Appl Environ Microbiol 63:1945– 1952 Nielsen KM, Calamai L, Pietramellara G (2006) Stabilization of extracellular DNA and proteins by transient binding to various soil components. In: Nannipieri P Smalla K (ed) Nucleic acid and proteins. Marcel Dekker, New York, pp 141–157 Overbeek LS, van Elsas JD (1995) Root exudate-induced promoter activity in Pseudomonas fluorescens mutants in the wheat rhizosphere. Appl Environ Microbiol 61:890–898 Palmen R, Vosman B, Buijsman P, Breek CK, Hellingwerf KJ (1993) Physiological characterization of natural transformation in Acinetobacter calcoaceticus. J Gen Microbiol 139:295– 305 Simpson DJ, Fry JC, Rogers HJ, Day MJ (2007) Transformation of Acinetobacter baylyi in non-sterile soil using recombinant plant nuclear DNA. Environ Biosafety Res 6:101–112 doi:10.1051/ ebr:2007024 Sorensen J (1997) The rhizosphere as a habitat for soil microorganisms. In: Elsas JD, Trevors JT, Wellington EMH (eds) Modern soil microbiology. Marcel Dekker, New York, pp 21– 45 Spiegel Y, Burrows PM, Bar-Eyal M (2003) A chemo attractant in onion root exudates recognised by Ditylenchus dipsaci in laboratory bioassay. Nematol 93:127–132 Stotzky G (1966) Influence of clay minerals on microorganisms. III. Effect of particle size, cation exchange capacity and surface area on bacteria. Can J Microbiol 12:1235–1246 Stotzky G (1986) Influence of soil mineral colloids on metabolic processes, growth, adhesion, and ecology of microbes and viruses. In: Huang PM, Schnitzer M (eds) Interactions of soil minerals with natural organics and microbes: Spec Publ No 17. Soil Sci Soc Am, Madison, pp 305–428 Tepfer D, Garcia-Gonzales R, Mansouri H, Seruga M, Message B, Leach F et al (2003) Homology-dependent DNA transfer from plants to a soil bacterium under laboratory conditions: implica-

92 tions in evolution and horizontal gene transfer. Transgenic Res 13:425–437 doi:10.1023/A:1024387510243 Tolove SN, Hedley M, Kirk GJD, Bolan NS, Loganathan P (2003) Progress in selected areas of rhizosphere on P acquisition. Aust J Soil Res 41:471–499 doi:10.1071/SR02130 Uren NC (2001) Types, amounts and possible functions of compounds released into the rhizosphere by soil-grown plants. In: Pinton R, Varanini Z, Nannipiero P (eds) The rhizosphere: biochemistry and organic substances at the soil–plant interface. Marcel Dekker, New York, pp 19–40 van Elsas JD, Dijkstra AF, Govaert JM, van Veen JA (1986) Survival of Pseudomonas fluorescens and Bacillus subtilis introduced into two soils of different texture in field microplots.

Biol Fertil Soils (2008) 45:83–92 FEMS Microbiol Ecol 38:151–160 doi:10.1111/j.15746968.1986.tb01724.x van Elsas JD (1992) Antibiotic resistance gene transfer in the environment: an overview. In: Wellington EMH, van Elsas JD (eds) Genetic interactions among soil microorganisms in the natural environment. Pergamon, Oxford, pp 17–39 van Elsas JD, Turner S, Bailey M (2003) Horizontal gene transfer in the phytosphere. New Phytol 157:525–537 doi:10.1046/j.14698137.2003.00697.x Wardle DA, Parkinson D (1992) Influence of the herbicides 2, 4-D and glyphosate on the soil microbial biomass and activity: a field experiment. Soil Biol Biochem 24:185–186 doi:10.1016/00380717(92)90277-5