Influence of Cry 1 Ac Toxin From Bt Cotton on the Soil Microbiota

The first record of transgenic cotton cultivation in Brazil was in 2005, of that of the cultivar MON 531, possessing the cry1Ac gene. Since then, no evaluation has been performed to understand whether the cultivation of Bt cotton has caused any interference with the soil microbiota, including bacteria. In this context, our research was aimed to assess whether the cultivation of Bt cotton negatively affects the community of soil bacteria, through quantitative and metagenomic analyses (marker gene 16S rRNA) for phylum identification. Samples of bacterial populations obtained from the soil cultivated with Bt cotton expressing the Cry1Ac toxin were compared with soil samples from the area cultivated with conventional cotton. Significant differences were not observed in the measure of colony-forming units of bacteria between the soils cultivated with Bt and non-Bt cotton; however, differences were detected only when comparing samples from different collection times of the Bt treatment. Cultivation of Bt cotton did not affect the diversity of the soil bacterial population. Overall, our study shows that, similar to most of the works that have been reported worldwide, cultivation of transgenic cotton does not seem to affect the quantity and diversity of natural soil bacteria.


Introduction
Genetically modified plants (GMPs) or transgenic plants originate from recombinant DNA technology.This involves the insertion of foreign genes in their genome, which may be derived from various sources such as microorganisms, animals, or plants (Monnerat & Bravo, 2000).In agriculture, transgenics have been employed with the aim of expressing genes in plants that can confer resistance to insect pests (Zhu et al., 2012) and pathogens (Sarowar et al., 2009), in addition to imparting herbicide tolerance (Benekos et al., 2010).
The structural diversity of bacterial communities has been studied using methods based on the investigation of a part of their DNA sequence, such as the 16S rDNA.These methods involve the amplification of the 16S rDNA by PCR and posterior characterization through cloning and sequencing, or analysis by electrophoresis, using techniques such as amplified ribosomal DNA restriction analysis (ARDRA), terminal restriction fragment length polymorphism (T-RFLP), random amplified polymorphic DNA (RAPD), ribosomal intergenic spacer analysis (RISA), denaturing gradient gel electrophoresis (DGGE)/temperature gradient gel electrophoresis (TGGE), and single-strand conformation polymorphism (SSCP).This results in obtaining data regarding the relatedness of the bacterial species that constitute the microbial community (Ranjard et al., 2000;Kozdrój & Van Elsas, 2001).
Molecular methods for the analysis of the structure and diversity of microbes have been developed in the past that utilize the genomic DNA directly extracted from environmental samples.Such developments have allowed considerable advances to be made in the study of microbial ecology (O'Donnell & Göres, 1999;Ranjard et al., 2000).These techniques enable the identification of factors employed in soil management that interfere with the multiple functions and habitats of the microbial community (Peters, Koschinsky, Schwieger, & Tebbe, 2000).These methods help in the detailed characterization of the microbial structure and their succession in agricultural soils, including that of the non-cultivable microorganisms (Hugenholtz, Pitulle, & Pace, 1998), and allow for the identification of predominant species, that may be used as indicators of soil functionality and quality (O'Donnell & Göres, 1999).This research was designed to evaluate whether the cultivation of transgenic cotton resistant to insects (Bt) affects the community of soil bacteria, either quantitatively or qualitatively.The evaluation was carried out through quantitative analysis of total bacteria and metagenomic analysis of the samples using the marker gene 16S rRNA for species identification.For this purpose, we compared samples of bacterial populations obtained from soil cultivated with Bt cotton expressing the Cry1Ac toxin with samples from areas cultivated with conventional cotton.To the best of our knowledge, this is the first report on the effects of Bt transgenic cotton on the soil bacterial community in Brazil.

Characterization of Cultivation Area
The studies were conducted in the agricultural area of the experimental farm of Federal University of Grande Dourados (UFGD), municipality of Dourados, Mato Grosso do Sul, Brazil (L 22°11′53″ S, L 54°55′59″ W, altitude: 430 m).The climate, according to data obtained from Brazilian Agricultural Research Corporation (Embrapa) was Cwa (humid mesothermic climate with warm summer and dry winter; the temperature of the coldest period, June and July, is lower than 18 °C, and the temperature of the warmest month, January, is higher than 22 °C).In addition, the rainfall in the summer months is 10-fold higher than the monthly rainfall of July (Fietz & Fisch, 2008).
According to Thornthwaite climatic classification, the region of Dourados is subhumid, with low water deficit.With respect to the thermal index, the climate is considered megathermal (Fietz & Fisch, 2008).

Sowing of the Bt and Non-Bt Cotton Cultivars
Two sample cultivation areas of 5000 m² each were prepared in the agricultural region of the Experimental Farm of UFGD, where the Bt and non-Bt cotton cultivars were sowed.The Bt cotton treatment was represented by the transgenic cultivar NuOpal (Bollgard ® I Evento 531), expressing the Cry1Ac toxin, and the conventional treatment (without the foreign gene) by its isoline, the cultivar Delta Opal ® .The aforementioned cultivars display similar agronomic characteristics.
Sowing in both fields was performed on December 31, 2009, with a density of 10 to 14 seeds per linear meter and with line spacing of 0.90 meters.The adopted cultivation methods were based on the agricultural practices recommended by Embrapa (2001), and in both the areas, there was no application of fungicide or insecticide.The no-tillage system was adopted, and millet (Pennisetum glaucum L.; Cyperales: Poaceae), BRS 1501 ® cultivar, was the antecessor crop.

Sampling Methodology
Sample collection was carried out on a monthly basis, totaling six evaluations, the first being carried out one day before sowing (zero day: O D), the second at 30 days after sowing (D.A.S.), the third at 60 D.A.S., the fourth at 90 D.A.S., the fifth at 120 D.A.S., and the sixth at 150 D.A.S.The first collection, performed one day before sowing, was aimed to determine the chemical and microbial characteristics of the soils before sowing the cotton cultivars.Collections of soil samples were performed between the months of December 2009 and May 2010.
One part of the collected samples was brought to the Laboratory of Soils of College of Agricultural Sciences of UFGD for granulometric physical analysis and measurement of pH and amount of micro and macronutrients (Appendix A: supplementary of Table 1, Online Resource 1).
A second part of the samples was brought to the Laboratory of Microbiology of College of Biological and Environmental Sciences of UFGD to perform quantitative bacterial analysis.In addition, a third part of the samples was brought to the Laboratory of Biochemistry of Microorganisms and Plants of São Paulo State University (UNESP/FCAV), Jaboticabal Campus, to accomplish the metagenomic analysis.The soil samples were collected in three replicates, with the aid of a soil auger, Dutch type, 0 to 15 cm depth.These samples were conditioned in plastic bags placed inside a styrofoam box under environmental conditions, to maintain their humidity content (Pereira, Neves, & Drozdowicz, 1996).In the laboratories, the soil samples were preserved under refrigeration for up to 24 h for posterior analysis of physical and chemical parameters of the soils and their microbiota.Thereafter, the soil samples were sieved using a 2 mm-mesh sieve (Olsen & Bakken, 1987).

Quantitative Evaluation of the Bacterial Population
To evaluate the bacterial population, 10 g of soil was added into Erlenmeyer flasks containing 90 mL of saline solution (10 -1 dilution) and this initial suspension was vigorously shaken for 5 min.From this suspension, 1 mL was transferred into an assay tube containing 9 mL of saline suspension, and the mixture was manually homogenized.This process was repeated until a 10 -7 dilution was obtained (protocol adapted from Neder, 1992).Thereafter, 0.1-mL aliquots of each dilution were transferred by the surface plating method onto Petri dishes containing bacterial culture medium.These dishes were covered by paper film to avoid contamination and drying of the culture medium (Olsen & Bakken, 1987;Sorheim, Torsvik, & Goksoyr, 1989) and incubated in a bacteriological oven at 25 °C.Bacterial counting was carried out at 48 h after inoculation, and for this counting process the Thorton (1922) medium was selected, which is specific to bacterial enumeration (Sorheim et al., 1989).
A completely randomized design was adopted, in factorial scheme with three replicates, comprised the dilution factors and days after sowing (D.A.S.), aiming to compare variance estimates in the samples from different sowing months and before sowing.

Cloning
The PCR products amplified in both the sample sets were cloned into the vector pJET1.2/bluntusing the cloning kit Clone Jet (ThermoFisher, Waltham, USA; Fermentas; catalog #K1231) and used for the transformation of the cells of Escherichia coli strain DH5α, which were cultivated in a dish containing Luria Bertani medium supplemented with ampicillin (100 µg mL -1 ).After the cultivation of the transformed clones, the plasmid DNA was isolated by miniprep (Saitou & Nei, 1987).

Sequencing
The partial sequencing of the gene 16S rRNA was carried out according to the methodology described by Pereira et al. (2006) and Silveira et al. (2006).Thus, the sequencing of each clone was done using 0.4 mL of DNA Sequencing-Big Dye Terminator reagent in the Cycle Sequencing-ready ABI Prism (version 3); 3.2 pmols M13/pUC 1211 forward initiator oligonucleotide (5′-GTAAAACGACGGCCAGT-3′); 100 ng of plasmid DNA; 4.6 mL of buffer (400 mM Tris-HCl pH 9; 10 mM MgCl 2 ); and H 2 O milli-Q (Millipore) for a volume of 10 µL.The reaction was placed in a thermocycler model PTC-200, with a cycle of two minutes at 96 ºC and then submitted to 40 cycles of 96 ºC for 10 sec, 52 ºC for 20 sec, and 60 °C for 4 min.The amplicons were sequenced using a capillary sequencer (model ABI 3700, Applied Biosystems, Foster City, USA).After sequencing the samples, the gel image was analyzed using the Sequencing Analysis 3.4 software that generated the electropherograms of the sequences.

Phylogenetic Analysis and Comparison of the Libraries
The electropherograms were submitted to the software package Phred/Phrap/Consed (Gordon, Abajian, & Green, 1998) to verify the sequence quality and the presence of chimeric DNA.The sequences were compared with sequences in the Ribosomal Database Project (RDP) (http://rdp.cme.msu.edu/) using the software Seqmath (Dunbar, Takala, Barns, Davis, & Kuske, 1999).
Phylogenetic trees were built from the sequencing of 371 sequences with the aid of the program ClustalW, for global alignment of the sequences, and then performed with the aid of the software Mega 5.0.Thereafter, the aligned file was used for phylogenetic rebuilding, using the Neighbor-joining method.The model of nucleotides replacement used was that of Jukes-cantor, and the bootstrap phylogeny test was run 1000 times.

Statistical Analyses
To check whether the bacterial population density differed between the cotton cultivars and between the collection months, the two-way analysis of variance (ANOVA) method was used, considering the interaction between the dilution factors × D.A.S. Subsequently, the mean values were compared by Tukey's test (α = 0.05).
We used the indexes of Simpson and Shannon to calculate the diversity index, and for the calculation of the dominance index, we employed Simpson's index.These indexes were calculated using the software DivEs (Rodrigues, unpublished data).

Quantitative Evaluation of the Bacterial Population
Significant differences were not observed concerning the number of colony-forming units (CFUs) of bacteria between the soils cultivated with Bt or non-Bt cotton (F = 2.12; df = 1, 34; P = 0.16; Table 1).Earlier research conducted to investigate effects of the cultivation of transgenic maize and cotton resistant to insects (Bt) on the population of several soil microorganisms did not show significant effects (Flores, Saxena, & Stotzky, 2005;Shen, Cai, & Gong, 2006;Icoz, Saxena, Andow, Zwahlen, & Stotzky, 2008;Hu et al., 2009;Li et al., 2011), corroborating our findings in the current study.Significant differences were also not found regarding the interaction between cotton cultivar and days after sowing (D.A.S.) (F = 3.57; df = 5, 12; P > 0.05; Table 1).
However, there were significant differences detected in the quantity of bacteria in the Bt soil when comparing the collections from different D.A.S. (F = 6.41; df = 5, 12; P = 0.004) (Table 1).These results are in accordance with those shown by Li et al. (2011), in which the cultivation of transgenic cotton resistant to insects (Cry1Ac) for three consecutive years, caused fluctuation in the number of CFUs in the different assessment periods (months and years), but this occurred due to seasonal factors, not being associated with the Bt toxin.Similarly, Rui et al. (2005) did not find significant variations when evaluating different concentrations of Bt toxin in the rhizosphere of genetically modified cotton and its influence on the soil bacteria.These authors had shown that the cultivation of Bt cotton did not significantly affect the number of soil bacteria throughout the study, consistent with our research, with differences noted only between the collection months.

Characterization of the Bacterial Community
To investigate the possible effects of GMPs on the soil bacterial community, this work evaluated a total of 371 clones that were partially sequenced and belonged to two metagenomic libraries: a Bt library, constituted by 181 clones, and another non-Bt library, composed of 190 clones.Libraries of both treatments were compared with sequences in the RDP.To compare the libraries, the data were normalized wherever there were differences concerning the number of clones between cultivars.From the 181 sequences of the Bt library, 11 phyla were detected within the Archaea and Bacteria domains.Of these, 8% belonged to phylum Actinobacteria; 1.1% Bacteroidetes; 0.5% Chloroflexi; 3% Crenarchaeota; 0.3% Deinococcus-Thermus; 2.7% Euryarchaeota; 3% Firmicutes; 0.3% Gemmatimonadetes; 27.5% Proteobacteria; 0.5% Spirochaetes; 0.8% Thermotogae; and 1.3% did not group to any phylum (Figure 1).In the non-Bt library, 13 phyla were registered within the Archaea and Bacteria domains.Of these, 0.3% belonged to the phylum Acidobacteria; 7.5% Actinobacteria; 0.3% Aquificae; 0.3% Bacteroidetes; 0.5% Crenarchaeota; 0.3% Euryarchaeota; 3.5% Firmicutes; 0.3% Fusobacteria; 0.3% Gemmatimonadetes; 0.3% Planctomycetes; 36.9%Proteobacteria; 0.3% Spirochaetes; 0.3 Thermotogae; and 0.3% could not be grouped to any phylum (Figure 1).In both the treatments, most of the sequences observed belonged to the Bacteria domain.A low percentage of individuals of Archaea domain were detected, even though the oligonucleotide used was specific for the Bacteria domain (Table 2).
jas.ccsenet.Among the phyla that were observed only in the soil cultivated with non-Bt cotton are Acidobacteria, Aquificae, and Fusobacteria, showing 1% occurrence (Table 2).The phylum Acidobacteria serves as an indicator of the soil nutritional condition with low growing potential (Smit, Leeflang, & Gommans, 2001).This phylum is generally abundant in most of the soils (Kuske, Barns, & Busch, 1997;Felske, Wolterink, Van Lis, De Vos, & Akkermans, 2000;Nogales et al., 2001); however, in the current work it was recorded only in the non-Bt library, where only one clone was found, which corroborates the findings of Kishimoto and Tano (1987) and Sait, Hugenholtz, and Janssen (2002).The phylum Aquificae constitutes the oldest evolutionary line in the Bacteria domain, encompassing gran-negative autotrophic organisms; such organisms are strictly thermophilic with optimum growth generally occurring above 65 °C (Deckert et al., 1998;Reysenbach et al., 2001;Huber & Eder, 2006).In terms of metabolism, most of the species of Aquificae phylum are hydrogen-oxidants and use hydrogen as the electron donor and oxygen as the electron acceptor (Reysenbach et al., 2001;Huber & Eder, 2006).Alternatively, thiosulfate or sulfur may also be utilized as energy sources owing to their thermostability.Many of the enzymes participating in their metabolism are of interest for industrial and biotechnological applications (Huber & Stetter, 1998;Van den Burg, 2003).
The presence of Planctomycetes phylum was also reported only in the non-Bt soil, with an occurrence of 0.3% (Table 2).This phylum is constituted of aerobic organisms, which reproduce by budding, and is one of the few bacterial groups that do not possess peptidoglycan in their cell wall.Although there are several strains of this phylum in culture collections, according to Griepenburg et al. (1999), this phylum was rarely found in soil samples.This justifies its occurrence in only one of the areas studied in this work.
Apart from these, other phyla were recorded only in the soil cultivated with Bt cotton, such as Chloroflexi, which displayed an occurrence of 0.5% (Table 2).This phylum is constituted of heterotrophic, aerobic, and filamentous organisms and, despite being found in soils, little is known about the role they play.They possess a high diversity of phenotypes even among the few isolates that have been cultured (Rappé & Giovannoni, 2003).
Deinococcus-Thermus displayed an occurrence of 0.3%, with presence only in the soil sown with Bt cotton (Table 2).This phylum is well known due to its resistance to the UV rays, desiccation, ionizing radiation, hydrogen peroxide, and other agents that damage the DNA.Such resistance is possible due to a high efficiency in their DNA repair system.There is a great interest in this phylum regarding their application in bioremediation of sites contaminated with radiation and toxic chemical products, and due to their production of a series of thermostable enzymes of biotechnological importance, such as the Taq polymerase (Minton, 1994;Battista, 1997;Ferreira et al., 1997;Brim et al., 2000).
Concerning the phyla found in both the treatments, namely the soil cultivated with Bt or non-Bt cotton, the phylum Proteobacteria had the highest number of clones, followed by Actinobacteria and Firmicutes.The occurrence of Proteobacteria was 27.5% in Bt field and 36.9% in non-Bt field, Actinobacteria was 7.8% in Bt and 7.5% in non-Bt, and Firmicutes was 3.0% in Bt and 3.5% in non-Bt, respectively (Table 2).The following classes of Proteobacteria were revealed in the studied fields: Alphaproteobacteria, Betaproteobacteria, Deltaproteobacteria, Gammaproteobacteria, and Epsilonproteobacteria (Figure 4).This phylum is one of the largest and most varied among the phyla of culturable bacteria, showing a great morphological and metabolic diversity.Due to such characteristics, Proteobacteria occur in several environments (Hugenholtz et al., 1998;Dunbar et al., 1999;Smit et al., 2001), and this explains their appearance in both the soils, cultivated with Bt or non-Bt cotton.The result of our research corroborates Lee et al. (2011) study that analyzed the effect of genetically modified Zoysia grass (Zoysia japonica Steud.; Cyperales: Poaceae) on the soil bacterial community and observed that Proteobacteria was one of most abundant phyla among the assessed treatments.Other studies, such as by Janssen (2006), also reported this phylum as the most abundant, and Faoro et al. (2010) concluded that Proteobacteria was the most abundant bacteria phylum in soil samples of Atlantic forest.Moreover, in comparison with Acidobacteria, which is reported as one of the most abundant phyla in several investigations, phylum Proteobacteria is more stable and does not seem to differ in response to crop rotation, harvest, and type of cultivar.This may explain the low percentage of occurrence of Acidobacteria in the current research, highlighting that this phylum does not develop well in acid soils (Val-Moraes, Valarini, Ghini, Lemos, & Carareto-Alves, 2009), a characteristic of the cultivated the area of our study.Smit et al. (2001) suggested that the ratio between the number of Proteobacteria and Acidobacteria serves as an indication of the nutritional condition of the soil, considering that Proteobacteria has a high growth potential, whereas Acidobacteria has a low growth potential.
Figure 4 The All parameters assessed in this investigation did not indicate any effect of GMPs on populations of soil bacteria.Both treatments had some phyla that were not reported in the other one.For instance, Acidobacteria, Aquificae, Fusobacteria, and Planctomycetes were reported only in the soil cultivated with non-Bt cotton, whilst Chloroflexi and Deinococcus-Thermus were found only in the soil cultivated with Bt cotton (Figure 1).Although these phyla were registered in only one of the two treatments, the representative number of each one of these phyla is not considered statistically significant, and thus, should not be attributed as an impacting or favoring factor for these species.In this sense, it is emphasized that the bacterial community in soils may also show fluctuations and successions over time (Jung, 2007).Thus, assessments of impact of transgenic crops on microorganisms may display variations that not always are associated with the cultivation of GMPs, but to environmental factors, such as temperature and humidity.Natural differences between samples were suggested as the main causes of variation in some of the earlier studies, rather than the effects of GMPs (Heuer, Kroppenstedt, Lottmann, Berg, & Smalla, 2002;Blackwood & Buyer, 2004;Fang, Kremer, Motavalli, & Davis, 2005).Widmer (2007), in a review on effects of transgenic crops on soil bacterial community, concluded that the environmental factors or type of cultivated plant species often have higher effects on the soil microbiological characteristics than the transgenic treatments, and the reported effects are restricted to the rhizosphere of the transgenic plants and remain only during the period in which they are present.
Corroborating the results obtained in our research, other studies on transgenic plants also did not detect significant effects of Bt proteins on the soil bacterial populations (Chiarini, Bevivino, Dalmastri, Nacamulli, & Tabachioni, 1998;Min, Kremer, Motavalli, & Davis, 2005).Schmalenberger and Tebbe (2002) did not find significant differences on the bacterial community between the soil cultivated with herbicide-resistant maize and its non-transgenic isoline, as studied using the technique of 16S rDNA PCR-SSCP (Single-strand conformation polymorphism).
Li et al. ( 2011) evaluated the impact of transgenic cotton resistant to insects (Cry1Ac) during three years of consecutive cultivation and found that the long-term cultivation of transgenic cotton did not show significant adverse effects on soil microorganisms.The study developed by Shen et al. (2006), did not report any significant effects of Bt cotton on the soil microbial diversity, and the study by Saxena and Stotzky (2001), did not record any alterations in the number of aerobic bacteria and fungi that could arise due to a possible negative effect of cry1Ab protein from the maize root exudates, and, during the tissue decomposition.Results from such studies reinforce the results obtained in our research here, in which the Bt cotton caused no significant negative effect on the population of soil bacteria in field conditions, when compared with its isoline non-Bt cotton.
On the contrary, different results were found in other research studies.Donegan et al. (1995) observed that two cotton hybrids, which express the B. thuringiensis endotoxin, caused an increase in the number of total bacterial and fungal populations when compared to the other treatments, with also changes detected in the diversity of the bacterial community.Wei et al. (2006) studied the influence of the cultivation of transgenic papaya resistant to virus on the microbial population and registered significant increases in the number of colony-forming units (CFUs) of bacteria, actinomycetes, and soil fungi, compared to the soil cultivated with conventional papaya.Dunfield and Germida (2001) observed differences in the functional diversity and composition of the microbial community in the rhizosphere of a canola (B.napus L.) transgenic variety (tolerant to herbicide), when compared to the conventional variety.Nevertheless, contrasting findings have also been reported.For example, Rui et al. (2005) noted higher number of functional bacteria associated with roots of conventional cotton when compared to the rhizosphere of varieties of transgenic cotton (NuCOTN99 (Bt) and SGK321 (Bt + CpTI)).
Emphasizing these findings, some authors, such as Dunfield andGermida (2003, 2004) and Hu et al. (2009), affirmed that when there is occurrence of some type of GMP effect on soil microorganisms, such effect is considered temporary and transitory, because it disappears after removing the plants from the field.
In conclusion, our research suggests that there does not seem to be a negative influence of the Bt cotton cultivar on the quantity and diversity of the soil bacterial population.However, it is necessary to perform continuous and systematic monitoring to ascertain whether transgenic cultivars could damage the soil microbiota.Such monitoring needs to be conducted through utilization of different approaches once there is a range of Bt proteins available in the market, which may show different effects depending on the locality in which the cotton is grown.This research contributes towards revealing that the cultivation of Bt cotton does not seem to affect the soil bacterial population in a Brazilian savannah region, a result that is similar to that of most of the research conducted worldwide.
Note.Number of CFUs of bacteria in given volume (0.1 mL) in days after sowing (D.A.S.), from December 2009 to May 2010, with the means of three plots each of an area of cultivation of genetically-modified cotton (Bt) and an area of cultivation of conventional cotton (non-Bt).The means followed by the same letter in the columns are not significantly different (Tukey α = 0.05).§ Time collection mean (uppercase letters are comparing Bt with non-Bt treatment; lowercase letters are comparing collection times within Bt or non-Bt treatment); ¶ Bt and non-Bt means.

Table 1 .
Number (CFU's) of total soil bacteria observed in cultivation of Bt and non-Bt cotton by replicate and mean

Table 2 .
D cotton

Table 3 .
Indexes of Shannon and Simpson of both cultivars expressed in percentages