Experimental site and crop description

Field trials were conducted in the agricultural farm of Indian Institute of Vegetable Research (I. I.V.R.), Varanasi, India (25° 08’ N latitude, 83° 03’ E longitude, 90 m from sea level, average temperature maximum 33°C and minimum 20°C). The site has been used for intensive vegetable production but not for any transgenic crop plantation prior to the present study (during 2010–2011). The soil (WHC 39.9%) is pale brown silty loam (sand 30%, silt 70%, clay 2%), Inceptisol with pH 6.7, organic C (0.73%) and, total N (0.09%) [20].

Ten- days old seedlings of VRBT-8 Bt transgenic event are selected for the study (data not shown). Genetic transformation was brought up through Agrobacterium tumefaciens LBA4404- mediated gene transfer that harbours pBinAR binary vector for neomycin phosphotransferase (npt-II) gene with neopaline synthase (NOS) promoter and a Cry1Ac gene fused to a constitutive, widely used plant promoter (CAMV35S) and octopine synthase gene (OCS) [21]. Treatments consisted of randomised blocks design in six plots of brinjal (Solanum melongena L. var. Kashi Taru), each 12 m² (3 for transgenic -VRBT-8 and its near-isogenic non-transgenic, respectively) grown in containment condition to conform to bio-safety regulations and simulated agricultural conditions. Recommended cultivation practices were adopted in which soils prior to transplantation, were added with 25–30 tonnes/ ha farm yard manure (FYM) along with NPK (100–120 kg N, 75–85 kg P and 45–50 kg K) [22]. Irrigation was done at the interval of every 10–15 days to maintain optimum moisture conditions.

Soil sampling and analyses

Soil sampling (in triplicate for each sampling stage) was done at different crop growth stages (branching, flowering and maturation) including pre-vegetation and post-harvest stage during the consecutive years (2010 and 2011). Rhizospheric soil samples were collected from the branching, flowering and maturation stage of non-Bt and Bt brinjal crop by uprooting the plants. Roots were vigorously shaken to separate the loosely bound bulk soils [23]. Soil samples at pre-vegetation and post-harvest stage, were collected from 0–10 cm depth using a 5 cm diameter soil corer [20]. To ensure the spatial homogeneity, soil samples were pooled and homogenously mixed prior to subsequent analyses. After removal of plant debris, samples were sieved through a 2-mm sieve and divided into two sub-samples. One sample was stored for 7 days (4°C) to prevent from sunlight and to reduce the microbial activity for molecular biological analyses (microbial density and diversity), and the other air dried for soil analyses.

Soil pH was determined using pH meter (Systronics-model 361). Organic carbon content was determined by wet digestion method of Walkey and Black [24]. The available Zn, Fe, and Mn in the soil samples were extracted with a diethylene triamine penta-acetic acid (DTPA) solution (0.005 M DTPA + 0.01 M CaCl₂ + 0.1 M triethanolamine, pH 7.3 [25]. The respective micro-nutrients studied were Zn²⁺, Fe²⁺ and Mn²⁺. The available sulphur was determined using the method of Comb et al. [26], and available K₂O by the method of Licina and Markovic [27].

Soil DNA extraction

Total genomic DNA (in triplicate at each sampling stage) was extracted from 0.5 g rhizosphere soil using Fast DNA® spin kit (MP Biol, USA) combined with Fast DNA prep bead beater according to manufacturer’s protocol. The genomic DNA was eluted in 50 μl DNA eluting solution (DES) and stored (-20°C) for subsequent analysis. The concentration and purity of extracted DNA was determined using Nanodrop spectrophotometer (ND 1000, Nano Drop Technologies, Inc., Wilmington, DE, USA).

Real time PCR for total actinomycetes 16S rRNAgene copy number

Real Time Quantitative PCR (qPCR) amplification was performed using Applied Biosystems 7500 Fast Real –Time PCR system containing 96-well plate (ABI 7500) to quantify the abundance of total actinomycetes specific 16S rRNA gene copy number using universal primer sets, 517 F (5’-CCA GCA GCC GCG GTA AT-3’) and Act704R (5’-TCT GCG CAT TTC ACC GCT AC-3’) [28]. The amplifications were carried out in triplicate in a final 25 μl volume containing 10X SYBR Green PCR master mix (Fermentas, USA). The reaction mixture (25 μl) comprised of 7.5 μl master mix (2X), 10 pmol each of primer (517 F and Act704R) and 45 ng genomic DNA template. The two-step Amp + Melt protocol was as follows: (i) amplification step: denaturing at 95°C for 4 min, 40 cycles of 30 s at 94°C and 30 s at 55°C, 1 min at 95°C, 1 min at 55°C, and (ii) melting curve analysis step: 81 cycles of 30s at 55°C. Plasmid DNA containing target gene (actinomycetes- specific 16S rRNA) was used as the standard DNA in real time PCR assay, was obtained by PCR-cloning using the universal actinomycetes-specfic primers [28]. Standard curves were generated by plotting the threshold cycle for each standard, calculated with ABI Prism 7900 SDS 2.2.2 software (Applied Biosystem, USA), against the gene copy number. The amplification efficiency (E) was measured from the slope of the standard curve [29]. The standard curve revealed a slope of – 2.66 corresponding to an efficiency of 137. 39% and R² of 0.994, similar to those reported in other studies [30].

PCR amplification for actinomycetes-specific 16S rRNAgene

Genomic DNA purified from soil was used as template for PCR. PCR triplicate from each sampling stages were separately amplified using universal actinomycetes-specific primers sets, ACT283F (5’-GGG TAG CCG GCC UGA GAG GG-3’) and 1360R (5’-CTG ATC TGC GAT TAC TAG CGA CTC C-3’) [12]. The PCR amplification was carried out using thermal cycler (Bio-Rad, USA) under the following conditions: (94°C, 5 min; 10 cycles of denaturation at 94°C (1 min), annealing at 65°C (30 s), extension at 72°C (2 min) and 72°C (5 min) followed by 20 cycles of denaturation at 92°C (30 s), annealing at 65°C (30 s), extension at 72°C (2.5 min) and final extension at 72°C (5 min). Reaction mixture (25 μl) contained 2.5 μl of 10 X buffer (Bangalore Genei, India), 0.5 μl of 40 mM dNTPs (Fermentas, USA), 1.25 μl each of 10 μM forward and reverse primer (Sigma), 2.5 U Taq DNA polymerase (Bangalore Genei, India.) and 1 μl template (40 ng). The remaining volume (18.5 μl) was maintained by nuclease-free water. Three PCR replicates of each samples stage were separately amplified and visualized on a 1.5% agarose gel. The resulting PCR products (1100 bp) were purified [31] through spin column using a QIAprep spin MiniPrep Kit according to manufacturer’s protocol, and combined separately for non-Bt and Bt samples.

Cloning, restrction fragment length polymorphism and phylogenetic analyses

The purified PCR products were ligated into the p-GEM®T Easy vector at 4°C (Promega, USA) as per manufacturer’s protocol, and cloned into the CaCl₂ treated E.coli DH5α competent cells. The screening of blue and white colonies was performed on ampicillin plates (100 μg ml^-1) supplemented with X-gal (0.5 mM) and IPTG. A total of 350 clones (70 clones for each sampling stage) were checked for putative positive inserts by PCR targeted with plasmid specific primer M13 forward and M13 primers. Further details regarding the positive insert verification are as reported by Vishwakarma et al., [20]. The clones with insert showed amplification of more than 1300 bp, while the PCR products with lower bands (250 bp) corresponded to the plasmid vector without any insert.

To identify the unique, amplified insert, actinomycetes-specific clones were subjected to Restriction fragment Length Polymorphism (RFLP). Two actinomycetes-specific 16S rRNAgene libraries were constructed, one for each soil actinomycetal community from the non-Bt plot and Bt brinjal plot. PCR products with inserts were used for producing RFLP pattern by digesting them with 0.4 U each of tetrameric endonuclease Hha I [30, 32] and Hae III restriction enzymes (New England Biolabs, Beverly, MA) in 1X buffer B (New England, Biolabs), bovine serum albumin (10 mg mL^-1) in the final volume of 20 μl. Reaction mixture was incubated overnight (37°C) and restriction patterns manually verified by analyzing 20 μl reaction mixture on 3% metaphor agarose gel stained with ethidium bromide in the gel documentation system (AlphaImager, NC, DE400-220, MTZ Zoom, Alpha Innotech Corp., San Leandro, CA). The zero–one matrices were prepared on the basis of RFLP pattern and operational taxonomic units (OTUs) grouped through CLUSTAL W program using the NTSYS version 2.1 software for each soil sample, and more than one representative of each group was sequenced.

The sequencing of the actinomycetal specific 16S rRNA clones as performed on both the strands in ABI PRISM® 3100 Genetic Analyzer (ABI, USA) using the Big Dye Terminator Kit (Version 3.1). Electropherograms were generated using the Chromas freeware (Version 2.01; Chromas lite Technelysium Pvt Ltd, Australia). Clones were finally checked for chimeric artifacts using CHECK-CHIMERA of the Ribosomal Database Project, and the chimeric sequences were discarded. The 16S rRNA sequences obtained, were initially recognized and aligned against the known sequences in the GenBank database using the BLAST program of the National Centre for Biotechnology Information (NCBI, http://www.ncbi.nlm.nih.gov/). The 16S rRNA clones obtained from the non-Bt and Bt planted rhizospheric soils with > 90% similarity with the NCBI data base, were used for phylogenetic analysis using MEGA software version. Further details related to sequencing analysis are given elsewhere [33].

Statistical analysis

The complete randomized design (CRD) with three replicates was used. Multivariate analysis of variance (MANOVA) was performed to determine the effect of treatments (non-Bt and Bt) at different growth stages. Multiple comparisons for difference in the means were made using Tukey’s Highest Significant Difference (HSD) test (P < 0.05), SPSS 16.0. The correlation coefficient was calculated between different parameters using the method given by Senedecor and Cochran [34]. The levels of significance (P < 0.01) and P < 0.05) are based on Pearson’s coefficients.

Soil variables

Variation in actinomycetes population size between the non-Bt and Btbinjal crop

Significant difference in the actinomycetes population between the soil of non-Bt and Bt brinjal over the entire two year period of cropping is depicted in Figure 1. Similar trend of variation in the actinomycetes population in the soil of non-Bt and Bt brinjal crop was: flowering > maturation > branching > post-harvest > pre-vegetation. Mean values for all the stages were significantly different from each other except between pre- and post-vegetation stages. MANOVA indicated significant differences due to year and crops (Table 2).

Tarafdar et al. [41] reported significantly higher actinomycetes population in non-Bt planted soil (5.25 X 10⁶ CFU g^-1) compared to Bt brinjal planted soil (4.3 × 10⁶ CFUg^-1). No significant changes were found in the studies conducted with transgenic cotton [42], corn [3], cabbage [43], and tomato [36]. Differences in the total actinomycetes population between the non-Bt and Bt crops might attributed to the release of root exudates from the transgenic brinjal into the soil that could have changed the available organic carbon and in turn, influenced the carbon turnover [38]. Tarafdar et al. [41] suggested that reductions in the actinomycetes population under Bt cotton cultivation were due to changes in the root exudates. However, other studies [3, 36, 44] supported that genetic modification of the plant had no role in changing the microbial population.

Phylogenetic analysis of 16S rRNA gene sequences from non-Bt and Btbrinjal rhizospheric soils

Thirty eight OTUs were generated from 282 positive clones for non-Bt brinjal soils. In case of Bt soils, a total of 278 positive clones clustered into 29 OTUs for pre-vegetation, branching, flowering, maturation and post-harvest stages. Different OTUs when evaluated after RFLP finger-printing analysis, showed affiliation with 14 and 11 actinomycetal groups from the respective non-Bt and Bt brinjal soils (Figure 2 and Figure 3). The percentage similarity of the clones with their reference strain as reported in the NCBI database is also shown in Table S3 and S4 (Additional file 1: Table S3 and S4).

Phylogenetic analysis indicated Micrococaceae and Nocardioidaceae as the dominant groups in the non-Bt and Bt cultivated soils as well. These respective groups were strongly represented by high relative abundance with > 40% of actinomycetes-specific 16S rRNA clones during each sampling stage (Figure 4). OTUs of these groups were affiliated with Arthrobacter globiformis (99%) and either of the Nocardioides ganghwansis (99%) or Marmicola sp. (98%), respectively (Table S3 and S4). In addition, Intrasporangiaceae, Micromonosporaceae and Microbacteriaceae were also detected in the non-Bt and Bt rhizospheric soils, but were restricted to only some of the sampling stages. Intrasporangiaceae was present at pre-vegetation stage with relative abundance of 5% (non-Bt and Bt soils), and in post-harvest stage (non-Bt soils only) raising the abundance to 8%. Micromonosporaceae was characterized by relative abundance (> 14%) at branching and flowering stage of non-Bt brinjal crop and with more than 17% abundance at the respective stages of Bt crop. However, Microbacteriaceae group was detected only at the flowering stage (relative abundance, 5%) in the rhizospheric soils of non-Bt and Bt crop (Figure 4). OTUs belonging to these respective groups showed their resemblance with Janibacter sp., Micromonospora sp., and either of Agromyces sp. or Microbium thalassium for Microbiaceaea (Table S3 and S4) that are mostly reported from soils. [8, 15]

Groups like Promicromonosporaceae, Streptosporangiaceae, Mycobacteriaceae, Geodermatophilaceae, Frankiaceae, Kineosporaceae, Actisymmetaceae and Streptomycetaceae were exclusively detected for non-Bt while Nakamurellaceae, Corynebactericeae, Thermomonosporaceae and Pseudonocardiaceae for Bt rhizospheric soils, and were restricted to only some of the crop stages (Figure 2 and Figure 3). Both Promicromonosporaceae and Mycobacteriaceaea related OTUs were detected only during branching stages with relative abundance of 5%. OTUs of Streptosporangiaceae and Frankiaceae were present only during flowering stage (4% relative abundance) while that of Geodermatophilaceae were also present during branching in addition to the flowering stage with relative abundance of 10 and 6%. Kineosporaceae and Actisymmetaceae resembling OTUs were present only during maturation stage with 7 and 23% relative abundance, respectively. Streptomycetaceae group was confined to the post-harvest stage (20% relative abundance). Nakamurellaceae was detected during branching and maturation stages with 25 and 12% relative abundance, respectively while Pseudonocardiaceae only during flowering stage (12.5% relative abundance) (Figure 4). However, Thermomonosporaceae in addition to post-harvest stage (20%), was also detected during maturation stage (18% abundance) along with Corynebacteriaceae (5%). Except Nakamurellaceae, most of the OTUs of such exclusive groups of the non-Bt and Bt crop were affiliated with the reference strains that mostly originated from the soil / rhizospheric soil of the plants (Table S3 and S4).

In the present study, Micrococaceae and Nocardioidaceae were found to be the dominant group in cultivated soils. These taxa have been selectively enriched by the increased organic input to the soil [47, 48], and also frequently detected in the manure and organic compost treated soils [49, 50]. OTUs belonging to the exclusive groups in non-Bt and Bt planted soils as discussed above, are probably due to the specific nature of root exudates whose quantity and quality are likely to change via Cry1Ac gene based modification [3]. Rengel et al. [51] suggested that the resulting variations in the root exudates could be caused by the transformation of the plants. However, these exclusive actinomycetes groups were restricted to only a few growth stages of non-Bt and Bt crop. Also, the relative abundance of these OTUs for both the crops did not exceed the dominant taxa (Arthrobacter and Nocardia) as found for both the crops. Our findings corroborate with the result of Weinert et al. [52] wherein the genetic modification effect is more prominent only at the maturation stage compared to others in transgenic potato. Thus, it could be inferred that the genetic modification of brinjal using Cry1Ac gene, will have little impact on distribution of the dominant microbial groups (Micrococaceae and Nocardiodaceae).

Under the control of constructive promoter, the transgene Cry1Ac was expressed in all parts of the transgenic brinjal plant, throughout the entire cropping period [21]. However, the transgene was detected only during the flowering stage in the rhizospheric soils of Bt brinjal (data not shown). Sims and Holden [53] reported 50% decrease in the insecticidal activity of the Cry1Ab protein during 1.6 days, and 90% decrease within 15 days. Various studies suggested rapid degradation of Cry proteins but the reports are mostly contradictory [5]. It can be argued that transgenic brinjal influences the rhizospheric community in case trans-gene product comes in contact with the rhizospheric microbial community. Furthermore, significant changes in the organic carbon can also be one of the important soil factors to cause temporal shifts in the actinomycetal community, since changes in the microbial community are correlated with organic carbon content [45]. Changes in the other soil variables (mineral-N, K₂O, S, Zn, Fe, Mn and soil pH) with respect to plant-age [54], can also have significant role in the maintenance of the rhizospheric microbial community. The present study also supports the view that the extent of genetic modification depends on the plant type, transgenes, and the conditions prevailing [23].

Irrespective of the crop type, flowering stage harbours more diverse actinomycetes compared to others. Some studies suggested that the structure and function of rhizospheric microflora was affected by physiological activities of plant [18, 55, 56]. Therefore, flowering stage may be the favourable one for microbial proliferation due to the active release of root exudates [52, 57]. Observations in the present study are in agreement with the fact that the natural factors other than genetic modification have strong bearing on temporal shifting of the microbial community including the actinomycetes [36]. We now can summarize that changes in the actinomycetal community structure are closely associated with environmental factors such as soil variables that may favour the optimal proliferation of actinomycetal community [30]. The Cry1Ac gene induced effect has the potential in shifting of the actinomycetal community although it is transient compared to the plant-age effect in the transgenic brinjal agroecosystem.