Since the introduction of commercial cultivation of genetically modified (GM) soybeans in 1996, cultivation area of genetically modified organisms (GMO) has been increasing every year. By 2013, the GMO cultivation area spanned 181.5 million hectares across 28 countries (James 2014). Due to the global expansion in bioengineering investment, a wide variety of GM crops is being developed. Despite the advantages of GMO cultivation, such as increase in productivity and the economic, environmental, and social benefits of the consequent decrease in the use of agricultural chemicals (Brookes & Barfoot 2007), the public continues to show concern about the impact of GMO cultivation on the environment (Conner et al., 2003). However, the actual users, the general public, while perceiving the benefit of bioengineering technology to humanity, remain nervous about using GM crops (Savadori et al., 2004). However, it is likely that public acceptance of GMOs may be encouraged by developing useful, non-edible GM plants; for example, flowering plants with enhanced aesthetic qualities (e.g., color), cut flowers with increased longevity, or the production of viral vaccines and biopharmaceutical products using GM plants.

The effect of GM plants on soil microbiota has been studied for various crops such as corn, cotton, rice, canola, and potato. Some examples of the traits incorporated into the GM crop include pest resistance (Baumgarte & Tebbe 2005; Shen et al., 2006; Wei et al., 2012; Sessitsch et al., 2003), herbicide resistance (Schmalenberger & Tebbe 2002; Dunfield & Germida 2003). The effects of GM plants on the soil microbiota revealed in these experiments have mostly been researched to be negligible, seasonal or spatially or temporal restricted and comparable to natural factors (Becker et al., 2008). Important changes to the soil microbiota were reported to be from unexpected characteristic changes in trait of GM plants, especially the change in number of root exudations (Sessitsch et al., 2003; Icoz & Stotzky 2008).

Grafted cactus has higher aesthetic value with short vegetative period. It is made by grafting Gymnocalycium mihanovichii onto trigonal cactus (Hylocereus trigonus Saff). Trigonal cacti cultivated through repetitive vegetative propagation can exhibit discoloration, lower grafting rate, decreased productivity, and quarantine risk due to increased virus incidence rates (Min et al., 2006). Recently, studies have been done to provide the resistance against Cactus virus X (CVX) on trigonal cactus through RNA interference by introducing the CaMV 35S promoter- regulated sense and antisense genes that encode the coat protein of CVX (Kim et al., 2011).

Several methods were reported on analyzing soil microbial community. The first, community-level physiological profiling (CLPP), which uses Biolog Ecoplates, utilizes 31 substrates as the sole carbon source. The microbial colony characteristics can be compared by measuring color change in the substrate and turbidity due to the microbes within the soil sample. The second, denaturing gradient gel electrophoresis (DGGE) can detect qualitative and quantitative changes in the entire soil microbial community at the molecular level. In this paper, we characterized microbial communities in GM trigonal cactus cultivation soil by CLPP, DGGE, and 16S rRNA gene analysis and evaluated environmental risk factors of non-edible GM plants.

MATERIALS AND METHODS

Plant Cultivation and Soil Samples

GM trigonal cacti, provided with CVX resistance through the insertion of sense and antisense genes that encode the coat protein of CVX (Kim et al., 2011), and non-GM trigonal cacti were planted in horticultural bed soils in three pots [590 mm (W) x 385 mm (D) × 150 mm (H)] with 12 GM and 12 non-GM plants in each pot. The trigonal cacti were cultivated by vegetative propagation until they grew 3 - 4 bulblets. Soil samples were collected from the trigonal cactus cultivation pot once every 2 months according to a previously described method (Kim et al., 2008). Two months after removing the trigonal cacti from the horticultural bed soil, soil samples were collected to analyze any changes in soil microbial colonies with three replications. The soil samples were stored at −20°C for further analysis. GM trigonal cactus was analyzed for the presence of CVX transcripts. Total RNA was extracted from stems and roots using RNeasy kit (Qiagen, USA) according to the instructions of the manufacturer provided. Reverse transcriptionwas performed using a RevertAid First Strand cDNA Synthesis Kit (Fermentas, USA). One microgram of total RNA was primed with oligo (dT)₁₈ and reverse transcribed in a total volume of 20 μL. Four μL of reverse transcription products were used for PCR with primers P1 (5'-ATGTCTACTACTGGAGTCCAGTCTT- 3') and P2 (5'-TCACTCAGGGCCTGG GAGAAATTGA-3'). The cDNA was denatured at 95°C for 5 min followed by 37 amplification cycles of primer annealing at 58°C, extension at 72°C and denaturation at 95°C (30 sec each). Four microliters of PCR products were analyzed on a 1% agarose gel to assay for the presence of the expected RT-PCR fragment of 650 bp.

Carbon Substrate Analysis using Ecoplate

Changes in carbon substrate due to soil microbial colonies were analyzed by using Ecoplate (Biolog Inc., USA). Ten grams of the collected soil sample was diluted in 90 ml of sterile water, which was mixed for 30 min at 200 rpm. The solution (150 μl) was then placed in the well of the Ecoplate and cultivated at 20°C. Color changes were measured at 24-hour intervals at 590-nm wave length. Each EcoPlate was divided into three identical zones, serving as replicates, and the absorbance value of each carbon source was corrected by subtractingthe absorbance value of the control well without any carbon substrate; negative values were set to zero. Average Well Color Development (AWCD) was calculated for each replicate and for each sampling time as the mean of the 31 corrected values (Garland & Millis 1991; Wei et al., 2012).

Molecular Biological Analysis using DGGE

The genomic DNA was extracted using the FastDNA Spin Kit (MP Biomedicals, USA) from the GM and non- GM trigonal cactus cultivation soil. For bacterial analysis, PCR was performed using 518R primer and 341F, which had been added with a GC-rich sequence that amplifies the 16S rRNA site of the bacteria. Fungi were analyzed by using NS7-GC and NS8 primer to amplify the internal transcribed spacer (ITS) site. The 16S rRNA site of actinomycetes was amplified using 243F and 513R-GC primer (Wei et al., 2012). PCR was done as follows: pre-denaturation for 5 min at 95°C, 30 cycles of reaction with denaturation for 1 min at 95°C, primer annealing for 1 min at 60°C and elongation for 1 min at 72°C, and final extension for 7 min at 72°C. The PCR results was electrophoresed in an 8% acrylamide gel with a formamide concentration gradient of 40-70% using Dcode Universal Mutation Detection System (Bio-Rad, Hercules, U.S.A.). The amplified DNA was stained with SYBR Green I and EtBr and observed using a UV trans-illuminator.

Clone Library Construction and Sequencing

PCR amplifications were carried out using the universal bacterial 16S rRNA-targeting primers 27F (5'-AGAGTTTGATCMTGGCTCAG- 3') and 1492R (5'-ACGGGCGGTGTGT RC-3') (Gremion et al., 2003). PCR products were comfirmed by 1% agarose gel, purified on PCR Dual Kit (BBI) following manufacturer’s instructions, ligated into the pTOPO-TA cloning vector (Invitrogen, USA) and used to transform Escherichia coli JM109 competent cells (Promega, USA). White colonies were picked randomly for plasmid isolation using a Plasmid Mini prep kit (Invitrogen, USA). Cloned inserts were sequenced using universal M13 (−20) F and M13R primers and BigDye Terminator version 3.1 cycle sequencing kit (Applied Biosystems, USA). The obtained sequences were resolved on ABI 3730 XL Genetic analyzer. Similarity comparisons were performed using the online program (BLAST) to known 16S rRNA sequences in the database.

Data analysis. The absorbance was measured using Multiskan Ascent (Thermo Labsystems, Finland), and the average well color development (AWCD) was computed using the following method. AWCD = Σ(C − R)/n, where C is the mean value of OD590 nm of each well, R is the value of control well, and n is the number of substrates 31. The principal component analysis (PCA) was based on 144-hour AWCD data and performed using BioPAT-SIMCA version 13 (Umetrics, Sweden) to evaluate the differences among groups of multivariate data. The PCA output consisted of score plots to visualize the contrast between different samples and loading plots to explain the cluster separation. The data file was scaled with unit variance scaling before all variables were subjected to the PCA. Similarities between DGGE patterns were calculated using InfoQuest FP software Ver. 4.50 (Bio-Rad, USA) via pairwise similarity of the banding patterns for the different samples, and clustering of patterns was performed with the unweighted pair group method using average linkages (UPGMA).

RESULTS AND DISCUSSION

Expression of CVX Gene in GM Trigonal Cactus

RT-PCR was used to determine qualitatively whether or not the integrated and intact CVX transgenes were transcribed. Template RNA from stems and roots of the GM trigonal cacti allowed the synthesis of CVX PCR products of the expected size, whereas no expression was found in non-GM trigonal cactus (Fig. 1). RT-PCR analysis revealed that CVX transgene is expressed routinely in all tissues from GM trigonal cactus.

Effect of Trigonal Cactus on Functional Diversity of Microbial Communities

We investigated the carbon substrate utilization of microbial colonies in the GM and non-GM trigonal cactus cultivation soil using CLPP analysis. The absorbance was measured at 24-hour intervals from the Ecoplate, and the resulting value was expressed in terms of AWCD (Fig. 2). There was no difference in AWCD curve between the GM and the non-GM trigonal cactus group in soil samples collected during the vegetative growth period. For the postharvest soil, the two groups didn’t exhibit significant difference except at 48- 96 hour. However, the difference was not detected after that time point.

PCA was conducted by using the AWCD value for 31 components. The two highest ranking components accounted for 36.6% of the total variation within the data set (Fig. 3A). The PCA results revealed a distinct separation of the two growth stages on the first principal component representing 23.9% of the total variation. In contrast, no clear separation was observed between the GM and the non-GM trigonal cactus, in which the variation was not higher than that between sample collection periods. These results indicated that the changes of functional diversity of microbial communities in the rhizosphere during growth were not affected by GM cactus. The major components responsible for the separation between the two growth periods were determined by analyzing the corresponding loading plots (Fig. 3B). This variation was mainly attributed to hydroxybenzoic acid, asparagine, N-acetylglucosamine, and a-glycerolphosphate of which the loading was positive for N-acetylglucosamine and a-glycerolphosphate and negative for asparagine, 2- and 4-hydroxybenzoic acids. The loading plot indicated that N-acetylglucosamine and a-glycerolphosphate were higher in vegetative growth period than in post-harvest period.

Effect of Trigonal Cactus on Microbial Communities’ Composition

The microbial communities compositions of the soil samples were compared using the data gathered from the characteristic amplifications of 16S rRNA for bacteria and actinomycetes and ITS site for fungi by the DGGE analysis method (Fig. 4). The DGGE profiles of the GM and non-GM groups were nearly identical, except slight differences due to the sample collection period. In the analysis of DGGE fingerprinting (Fig. 5), a consistent cluster was not observed between GM and non-GM trigonal cacti.

The clone libraries amplified from the GM, non-GM trigonal cacti, and control soils yielded 200, 200, and 100 high quality clones, respectively. All of sequenced clones were found to be unique in their cultivated site. Fig. 6 shows a Venn diagram representing the unique and overlapping sequences at the three sites, indicating that most of the members comprising each of the soil communities were unique to their sample sites. The clone sequences were attributed to 15 phyla that were divided into the three sites. The most abundant clone in the soil samples collected from the three different sites was Archaea. Gammaproteobacteria and Sphingobacteria were more abundant in the GM trigonal cactus soil than in the other two soils. The Opitutae and Mortierellomycotina were only detected in clones amplified from non-GM trigonal cactus soil.

Since Grove et al. (2004) successfully used CLPP for characterizing microbial community, the Ecoplate was used to analyze the effect of GM plants such as Bt rice, Bt cotton, and Bt maize on the diversity of microbial communities (Mulder et al., 2006; Wei et al., 2012). The use of Ecoplates to calculate substrate-based diversity index was shown to yield useful ecological information (Garland & Mills 1991; Insam & Goberna 2004). Our results indicated that changes of microbial metabolic diversity were mainly related with development period, and GM trigonal cactus had a little difference at the post-harvest period. Direct incubation of soil samples in Ecoplates produces patterns of metabolic response useful in the classification and characterization of microbial community. These results were also confirmed by PCA analysis. On the basis of PCA analysis, microbial communities associated with vegetation period showed a greater relative utilization of carbohydrates (N-acetylglucosamine and a-glycerolphosphate) and a lesser relative utilization of amino acid (asparagines) and carboxylic acid (hydroxybenzoic acids) compared with microbial community sample from post-harvest period. The temporary changes observed in the post-harvest soil were caused by environmental factors and cultivation period rather than by the differences in plant proteins due to genetic modification (Icoz et al., 2008; Xue et al., 2011). After the trigonal cacti were harvested, the pots were maintained under high-temperature, high-humidity conditions in the greenhouse. This extreme condition probably results in the dominance of specific microbes in soil microbial community in our post-harvest analysis. Our experiment shows that the bacteria isolate Rhizobium sp. JEYF14 and uncultured bacterium clone UK10.40 significantly increased in the post-harvest soil compared to the vegetation period (Fig. 5). According to previous work, rhizobial bacteria including Rhizobium sp. can be greatly affected by the climate conditions of the site. They suggest that these bacteria are adaptive to their environmental conditions (Bachar et al., 2010; Egamberdiyeva & Hoflich, 2003). Further studies will be needed to explore the mechanism behind such as high microbial presence. One possible hypothesis is that environmental condition may well be one of the important factors affecting microbial community composition in soils. The soil sample of the GM papaya cultivation area showed a difference in band patterns due to the vegetation period, resulting from the heterogeneity of the soil in the DGGE analysis (Hsieh & Pan 2006). In addition, within the soil samples of post-harvest period, 2 months since the vegetative growth period, DNA disintegration resulted in less amplified bands for bacteria and actinomycetes compared to those of the vegetative growth period. With respect to fungi, post-harvest soil sample exhibited a slightly different band pattern such that the change was attributed to the community level physiological profiling (CLPP) factors. The changes that occurred in microbial communities due to temperature, humidity, or other external environmental factors were temporary and reverted back to the normal state during the post-harvest period. Because the half-life of proteins released by root exudates or residues is less than 20 days and the same proteins disintegrate within the soil at a fast rate (Icoz et al., 2008), GM sugar beet plants have not been found to have an impact on the diversity of microbial colonies (Gebhard & Smalla 1999). In the present study, we predicted that the inserted CVX gene in GM trigonal cactus would quickly disintegrate within the soil because of its small size. A significant difference in the AWCD of the post-harvest soil was not attributed to soil chemical properties difference but to the introduction of GM trigonal cactus debris. Consequently, changes in soil microbial communities due to the cultivation of virus-resistant trigonal cactus were predicted to be minimal. Based on the overall observations, cultivating GM trigonal cactus was found to have no impact on soil microbial community. Temporal variations observed between GM and non-GM trigonal cactus were attributed to differences in environmental factors rather than gene expression. Our results suggest that growing of virus-resistant trigonal cactus may not pose ecological or environmental impact.

적 요

본 연구는 국내에서 개발된 CVX 저항성 삼각주 재배가 토 양 미생물에 미치는 영향을 분석하기 위해 수행되었다. 근권 토양 DNA에 대한 DGGE 분석 결과, 삼각주의 생육시기와 생육 이후의 근권 토양 세균, 방선균 및 곰팡이의 군집밀도는 CVX 저항성 삼각주 재배 토양과 비형질전환 삼각주의 재배 토양간에 차이를 보이지 않았으며, 토양 미생물 군집의 profile 변화도 나타나지 않았다. Biolog Ecoplate을 이용하여 토양 미 생물의 기능적 다양성을 분석한 결과, 삼각주의 재배 기간에 서는 CVX 저항성 삼각주와 비형질전환 삼각주의 토양 시료 간의 유의한 차이를 발견하지 못했으나, 재배 이후 토양에서 는 CVX 저항성 삼각주 시료의 탄소 기질 이용도가 비형질전 환 삼각주 시료보다 높게 나타났다. 이는 48, 72, 96시간에서 만 관찰되었으며, 이후 시간에서는 유의적 차이를 보이지 않 았다. 토양 미생물 군집의 profile을 이용하여 주성분 분석을 수행한 결과, CVX 저항성 삼각주와 비형질전환 삼각주간의 차이보다는 재배 기간과 재배 이후, 즉 재배 시기에 의한 미 생물 군집의 차이가 나타남을 확인하였다. 결과적으로 CVX 저항성 삼각주 재배가 토양 미생물 군집 밀도와 다양성에 영 향을 주지 않는 것으로 사료된다.