Impact of chemical- and bio-pesticides on bacterial ... - Springer Link

Ecotoxicology (2013) 22:1479–1489 DOI 10.1007/s10646-013-1134-1

Impact of chemical- and bio-pesticides on bacterial diversity in rhizosphere of Vigna radiata Sukriti Gupta • Rashi Gupta • Shilpi Sharma

Accepted: 23 September 2013 / Published online: 2 October 2013 Ó Springer Science+Business Media New York 2013

Abstract To study the effects of two chemical pesticides (chlorpyrifos and endosulfan), and a bio-pesticide (azadirachtin) on bacterial diversity in rhizospheric soil, a randomized pot experiment was conducted on mung bean (Vigna radiata) with recommended and higher doses of pesticides. Denaturing gradient gel electrophoresis was used to analyze such effects on both resident and active bacterial communities across two time points. It was observed that higher doses of azadirachtin mimicked the effects of chlorpyrifos on bacterial diversity. Both azadirachtin and chlorpyrifos showed a dose- and time-dependent effect, which was observable only at the RNA level. Endosulfan treatments showed dissimilar profiles compared to control. Most of the bands showed high sequence similarities to known bacterial groups, including many nitrogen-fixing, phosphate-solubilizing, and plant-growthpromoting bacteria. This study indicates that pesticides display non-target effects on active microbial populations that serve important ecosystem functions, thereby emphasizing the need to critically investigate and validate the use of bio-pesticides in agriculture before accepting them as safe alternatives to chemical pesticides. Keywords Azadirachtin Endosulfan Chlorpyrifos DGGE Bacterial diversity Rhizosphere

S. Gupta R. Gupta S. Sharma (&) Department of Biochemical Engineering and Biotechnology, Indian Institute of Technology Delhi, Hauz Khas, New Delhi 110016, India e-mail: [email protected]

Introduction The use of pesticides on leguminous crops in the world, especially in India, has seen a significant rise over the last few decades. Several pesticides, including organochlorines, organophosphates, fungicides, and herbicides, are used in modern agriculture to meet the need for abundant and affordable food and fiber. Although the use of pesticides has led to increased agricultural production, they have been associated with several major health and environment related concerns (Ejaz et al. 2004). Therefore, there has been an increased popularity of bio-pesticides owing to their plant origin. There is a belief that such natural insecticides are ‘‘safe’’ or less damaging to the ecosystem, however, it has been reported that the numbers of fungi and nitrifiers are reduced significantly after application of azadirachtin at recommended doses while higher doses have high biocidal effects on soil microorganisms and its activities (Gopal et al. 2007). Therefore, there is a need for studies looking at the effects of such pesticides of biological origin on the structure and function of microbial community under realistic exposure conditions. Microbial communities in soil ecosystems provide many important functions like the decomposition of organic material, recycling of nutrients, biogeochemical cycling and by constituting a major food source at the base of food webs. Additionally, microbes are also able to degrade soil-associated organic pollutants and can thus contribute to remediation of contaminated ecosystems. Thus, many microbial functions are critical to crop production, soil sustainability, and environmental quality and the effect of pesticides on the underlying structure and diversity of soil microbial communities become crucial in this regard. A number of available cultivation-dependent and cultivation-independent techniques allow monitoring and quantification of in situ

123

1480

responses of microbial populations to pesticides with increasing precision. However, due to the complexity involved in detecting the effects of environmental contaminants on natural microbial aggregations, a very few studies have addressed this topic. Community changes including both enhancement and inhibition of a variety of dominant organisms have been reported in a study done to assess the impact of the fungicide chlorothalonil on dominant bacterial and fungal populations in agricultural soils (Sigler and Turco 2002). Most of the research done to study the effect of pesticides on soil microbial diversity has been conducted to imitate massive pesticide pollution events using enormously high doses of pesticides, which are unrealistic (Widenfalk et al. 2008; Zhang et al. 2010). The response of bacterial communities to low-level exposure to pesticides that represent the most common type of soil contamination by pesticides, is still not well studied and understood. Although studying the bacterial diversity at the level of DNA would reveal the presence of bacterial members resident in the rhizosphere, it is only by analyzing the diversity at the level of RNA that the presence of the metabolically active members can be confirmed. However, very few studies have targeted the active populations in plant rhizospheres. It has previously been reported that pesticide (Ally and Marathon) addition led to major shifts in the active soil bacterial community structure, without significantly affecting the bacterial numbers or heterogeneity (Girvan et al. 2004). Chlorpyrifos and endosulfan have been detected as contaminants in human blood samples in a study performed in the villages of Punjab, India (Mathur et al. 2005). Chlorpyrifos (an organophosphorus pesticide), which is the fourth highest consumed pesticide in India, was detected in significant concentrations in 85 % of the samples tested. Endosulfan (an organochlorine insecticide), which is quite persistent in nature due to its slow decomposition rate, long half-life, and high stability in the environment, was also detected in 25 % of the blood samples tested. Vigna radiata (mung bean), which is the third most important pulse crop of India after chickpea and pigeon pea, was selected for this study. The present study aims to assess and compare the effect of two chemical pesticides (chlorpyrifos and endosulfan) and a bio-pesticide (azadirachtin) at recommended and three times higher dosage on the microbial population in rhizospheric soil. Both DNA and RNA were analyzed to target and compare the resident as well as active microbial populations. Universal bacterial primers (for 16S rRNA) were used to analyze the total bacterial community from DNA and RNA extracted from the system. Fingerprinting, using denaturing gradient gel electrophoresis (DGGE), was performed after amplifying the 16S rRNA genes and transcripts. Identification of key players was done by eluting bands from DGGE gel and sequencing them.

123

S. Gupta et al.

Materials and methods Soil characteristics The soil used in this study had the following physical properties: pH of 7.8 (in water), electrical conductivity of 0.32 dS m-1 and a water holding capacity of 20 %. Experimental set-up, seed sterilization and sowing Seeds of V. radiata variety Pusa Vishal were procured from National Seed Corporation (NSC), Indian Agricultural Research Institute (IARI), New Delhi. Seeds were surface sterilized with 70 % ethanol for 30 s followed by treatment with sodium hypochlorite (0.01 %) for 2 min. The seeds were washed with 0.01 M hydrochloric acid to remove sodium hypochlorite (Abdul-Baki 1974), and then rinsed thoroughly with sterile water several times to remove all chemicals. A completely randomized design was used for the experiment set up with cylindrical pots of approximate height and diameter of 40 cm. 45 pots [(6 treatments 9 2 time points 9 3 replicates) ? (1 bulk soil control 9 3 replicates) ? (3 plant controls 9 2 time points)] were filled with soil (with no known history of contamination) mixed well with Rhizobium strain suitable for growth of V. radiata, as recommended by IARI, New Delhi. Seeds were sown in each pot at a depth of about 4–5 cm. Three different pesticides (chlorpyrifos, endosulfan and azadirachtin) were added at two different concentrations each. For each concentration pots were set in triplicates for two sampling points. Besides, untreated plants, and unplanted and untreated bulk soil, were included as controls. The pots were kept under approximately 16/8 light/dark regime without any fertilization in IIT Delhi nursery at temperature of ranging from 24 to 29 °C. They were watered equally every alternate day to maintain constant soil moisture. The health of the plants was monitored during this period by the visual observation of the colour and state of their leaves. Pesticide treatment Plantlets were treated with pesticides 20 days after sowing of seeds. The ‘low’ concentration of each of the pesticides used was formulated in water as per recommendation of the manufacturer and the ‘high’ concentration was taken to be 3 times higher than the ‘low’ concentration based on the trends followed in previous studies (Sigler and Turco 2002; Parween et al. 2011). Soil application with 500 ml of prepared formulation was performed. Instead of using pure compounds, commercial formulations were used—Dhanvan 20 ec with recommended dose of 1 g active ingredient

Non-target effects of chemical and bio-pesticides

(chlorpyrifos) per l, Endo 35 ec with recommended dose of 1.75 g active ingredient (endosulfan) per l, and neem oil with recommended dose of 7.5 9 10-3 g active ingredient (azadirachtin) per l of treatment solution. The nomenclature used for the treatments were: untreated ‘bulk soil control (BS), untreated plant control (PC), azadirachtin (7.5 9 10-3 g active ingredient per l) treated planted soil (AL), azadirachtin (2.25 9 10-2 g active ingredient per l) treated planted soil (AH), chlorpyrifos (1 g active ingredient per l) treated planted soil (CL), chlorpyrifos (3 g active ingredient per l) treated planted soil (CH), endosulfan (1.75 g active ingredient per l) treated planted soil (EL), endosulfan (5.25 g active ingredient per l) treated planted soil (EH) (1 g/l corresponds to 0.1 g/kg dry soil). Sampling First and second samplings were performed 30 and 70 days after pesticide treatment, respectively. The plants for each treatment and control were completely uprooted. The roots were shaken vigorously to separate loosely held soil from the roots. The soil, tightly adhering to the roots, was collected using a soft brush without damaging the root and root nodules and this was termed as ‘‘rhizosphere soil’’. These rhizospheric samples were stored at -20 °C (after shock freezing in liquid nitrogen) for nucleic acid extractions. Total nucleic acid extraction Co-extraction of DNA and RNA from the rhizospheric soil samples was performed using the method described by Griffiths et al. (2000). This involved bead beating and solvent extraction of the nucleic acids. To obtain pure RNA prior to reverse transcription, DNA was removed by treatment with DNase (1 U ll-1; RNase-free; Fermentas) according to manufacturer’s instructions. PCR and RT-PCR amplification PCR and RT-PCR, targeting the 16S rRNA gene and transcript was performed using the universal bacterial primers 341F(GC) and 518R (Muyzer et al. 1993). cDNA was synthesized from template RNA using RevertAidTM First Strand cDNA Synthesis Kit (Fermentas) in a final reaction mixture of 20 ll (8 ll of total-RNA sample added to 12 ll RT mixture) as per manufacturer’s instructions. Random hexamer primer was used in the reaction. Synthesized cDNA and directly extracted DNA were further subjected to PCR using Bioline Master mix (Bioline, USA). The following thermocycler programme was used: 95 °C for 5 min initial denaturation followed by a touchdown annealing protocol (20 cycles of 95 °C for 30 s,

1481

annealing started at 65 °C for 30 s, 0.5 °C decrement in annealing temperature per cycle and elongation at 72 °C for 30 s) followed by 10 additional cycles of 95 °C for 30 s, annealing at 59 °C for 30 s, elongation at 72 °C for 30 s, and in the end a final elongation at 72 °C for 7 min. A 1.2 % agarose gel stained with ethidium bromide was used to check the size of amplification product. Denaturing gradient gel electrophoresis (DGGE) The PCR and RT-PCR products obtained were subjected to DGGE analysis using the Bio-Rad DcodeTM system (BioRad, Hercules, CA, USA). Amplicons derived from 16S rRNA gene and transcript were loaded onto a 0.75 mm thick 8 % (w/v) polyacrylamide gel with a linear denaturing gradient of 52–65 % denaturant (100 % denaturant is 7 M urea and 40 % (v/v) deinonized formamide). It was run at 60 °C in 19 TAE buffer (40 mM Tris–acetate, 1 mM Na-EDTA, pH 8.0) at a constant voltage of 70 V for 15 h. Following electrophoresis, the gels were stained in an ethidium bromide solution (1 lg ml-1) for 40 min and thereafter photographed using a Bio-Rad GelDoc station (Bio-Rad, Hercules, CA, USA). DGGE images were analyzed using Gel Compar II software version 6.6 (Applied Biomath, Sint-Martens-Latem, Belgium). Similarities between band patterns were calculated using SørensenDice coefficients (unweighted data based on presence or absence of bands). Cluster analysis based on the observed similarity values between profiles at different time points or treatments was performed to obtain dendrograms and the linkage type used was the unweighted pair group method with arithmetic mean (UPGMA). The tolerance value used was 1 %. Cophenetic correlations were calculated using the same software. Sequencing and sequence analysis Individual bands from the DGGE gels were excised using a sterile blade and stored overnight at 4 °C in 50 ll sterile water. After heating at 94 °C for 10 min, 1 ll of the sample was used as template for PCR using the 341F/518R primer pair and the PCR products were purified by Exo-I/ SAP treatment. The purified PCR products were used in sequencing reactions with the 341F primer, using a BigDye Terminator Cycle Sequencing Reaction Kit version 3.0 (Applied Biosystems, USA). Sequencing was performed on ABI 3500 sequencer (Applied Biosystems, USA) at GCC Biotech (India). The sequence data was compared with known 16S rRNA sequences stored in GenBank using the Basic Local Alignment Search Tool (BLAST). The sequences have been submitted in NCBI database with accession numbers KF386613–KF386634.

123

1482

S. Gupta et al.

Results and discussion Two chemical pesticides (chlorpyrifos and endosulfan) and a bio-pesticide (azadirachtin) were studied for their nontarget effect on rhizospheric bacterial diversity in V. radiata at two time points (30 and 70 days post treatment) and two dosages.

AH1

Effect on resident bacterial communities

CH1

Universal bacterial primer pair 341F (GC)/518R was used to amplify the hypervariable V3 region of the 16S rRNA gene. Discrete bands with the expected size of 217 bp (177 bp insert ? 40 bp GC clamp) were obtained. DGGE profiles of replicate soil samples (3 replicates per treatment) were compared to look for sampling variations. About 25–30 distinct bands could be observed in each profile (data not shown). UPGMA analysis using Sørensen-Dice coefficient was performed to create a dendrogram depicting pattern similarities (Fig. 1). When the profiling patterns of replicates were compared amongst themselves, replicates of the same treatment were greater than 96 % in all cases with DNA derived fingerprints. Thus, high reproducibility of the patterns was confirmed and only one representative sample for each pesticide treatment was used for further analysis. DGGE profiles for EL treatments formed a separate cluster from the rest of the treatments although the percentage similarity was quite high (94 %). Likewise, AL treatment profiles also formed a separate cluster but showed a 95 % similarity to other profiles, whereas AH profiles clustered along with CL, CH and EH profiles showing a 100 % similarity. This indicates that at higher doses, the effects of the bio-pesticide mimicked those of the chemical pesticides. Cophenetic correlation was used as a parameter to express the consistence of a cluster.

CH3

Effect on active bacterial communities RT-PCR of the RNA extracted from soil was performed using the same primer set as used for 16S rRNA gene amplification. About 20–30 distinct bands could be observed in the profiles generated (data not shown). Similar patterns were obtained for all the replicates showing a 94 % or above similarity level (Fig. 2), suggesting a low degree of variability caused by sampling, nucleic acid extraction, and further processing. Hence, one representative sample for each pesticide treatment was used for further analysis. The profiles of L and H treatments of endosulfan clustered separately from all other treatments showing a similarity of 81 % with the other two pesticides. While the 16S rRNA gene profiles showed 94 % similarity between EL

123

AH2 AH3

CL1 CL2 EH1 EH2 EH3 CL3 CH2 AL1 AL2 AL3 EL3 EL2 EL1 Fig. 1 Cluster analysis of DGGE patterns of 16S rRNA gene fragments amplified from pesticide treated Vigna radiata plants at first time point (30 days post-treatment). Scale represents % similarity. Cophenetic correlation values are mentioned at the branch nodes. AL azadirachtin (7.5 9 10-3 g active ingredient per L), AH azadirachtin (2.25 9 10-2 g active ingredient), CL chlorpyrifos (1 g active ingredient per L), CH chlorpyrifos (3 g active ingredient per L), EL endosulfan (1.75 g active ingredient per L), EH endosulfan (5.25 g active ingredient per L)

and other treatments, the 16S rRNA profiles showed 81 % similarity suggesting that there were relatively more differences between the ‘‘active’’ bacterial communities


1483

Comparison of resident and active bacterial communities AH1 AH2 AH3 CH3 CL2 CL3 CH1 CH2 CL1 AL1 AL2 AL3 EL1 EL2 EL3 EH2 EH3 EH1 Fig. 2 Cluster analysis of DGGE patterns of 16S rRNA transcripts amplified from pesticide treated Vigna radiata plants at first time point (30 days post-treatment). Scale represents % similarity. Cophenetic correlation values are mentioned at the branch nodes. AL azadirachtin (7.5 9 10-3 g active ingredient per L), AH azadirachtin (2.25 9 10-2 g active ingredient), CL chlorpyrifos (1 g active ingredient per L), CH chlorpyrifos (3 g active ingredient per L), EL endosulfan (1.75 g active ingredient per L), EH endosulfan (5.25 g active ingredient per L)

present in the rhizosphere as compared to the total resident bacterial community. As observed for the PCR profiles, the RT-PCR profiles of AH also showed a 94 % similarity with CL and CH profiles, while the AL profiles clustered away and were only about 85 % similar to the chlorpyrifos profiles. Thus, in this case also, the higher dose of biopesticide followed the trends of the chemical pesticide, chlorpyrifos.

DNA obtained from environmental samples could originate from dormant or dead cells or even from naked DNA. Thus, DNA-based profiling methods do not discriminate between dormant and active populations, which is possible through RNA based profiling methods. Comparative DGGE profiles of 16S rRNA gene and transcripts were obtained for L and H treatments of azadirachtin, chlorpyrifos and endosulfan (Figs 3, 4, 5a). In each of the gel images, the patterns were also compared across two time points between treated plants, untreated plants and bulk soil. Both the PCR and RT-PCR profiles showed 20-30 distinct bands indicating a high diversity in both the active and resident bacterial communities. Unlike previous report by Duineveld et al. (2001), which showed a reduced number of bands in RNA generated profiles from the rhizospheres of chrysanthemum as compared to DNA generated profiles, this study shows an equally complex RNA profile as compared to the DNA profile obtained from the rhizosphere of V. radiata. Similar to our observation, Duarte et al. (1998) also reported 20–30 bands in both PCR and RT-PCR profiles of bacterial 16S rRNA in soil. However, it should be noted that the number of bands does not exactly correlate with the number and abundance of active species within the microbial community since one organism may produce more than one DGGE band because of the presence of multiple heterogenous sequences in ribosomal genes (Cillia et al. 1996). On the other hand, it is possible that bands at similar positions may have originated from different sequences with minor variations leading to similar melting points (Vallaeys et al. 1997). Thus, the DGGE profiles generated could be an over or under-estimation of the actual diversity and might not give the real picture. Presence of large number of bands in the upper part of the gel corresponding to lower gradient indicated a prevalence of bacteria with low GC content in the studied rhizosphere. Most of the bands (about 15–18) were common for the PCR and RT-PCR profiles suggesting that the bacterial populations that were numerically dominant were the ones responsible for most activity. However there were 5-7 bands that were unique to either the PCR or RT-PCR profiles (bands 3, 5 in Fig. 3a) or were conspicuously more intense in the DNA-based or RNA-based patterns (bands 1, 2, 6, 7 in Fig. 3a). Likewise, in an earlier study performed on ryegrass sand, Corgie’ et al. (2006) reported that certain bacterial species that were dominant members of the population were only moderately active, while others that represented a minor segment of the population were highly active. Some bands that were only present in DNA generated profiles and absent from RNA generated profiles

123

1484

S. Gupta et al. b Fig. 3 a DGGE gel image of PCR and RT-PCR profiles of 16S rRNA

(a) 1 2

3

4

5

6

7

8

9 10 11 12 13 14

1

2 3 4 5 6 7 6’

(b)

AL/2/R

PC/2/R AH/1/R AL/1/R BS/0/R PC/1/R AH/2/R AH/1/D AH/2/D AL/1/D AL/2/D BS/0/D PC/1/D PC/2/D

123

amplified from azadirachtin treated Vigna radiata plants at two time points (30 and 70 days post-treatment). Lane numbers are represented above the image. Arrows mark the position of bands excised and sequenced. Lane 1 AH/2/R; lane 2 AH/2/D; lane 3 AL/2/R; lane 4 AL/2/D; lane 5 PC/2/R; lane 6 PC/2/D; lane 7 AH/1/R; lane 8 AH/1/ D; lane 9 AL/1/R; lane 10 AL/1/D; lane 11 PC/1/R; lane 12 PC/1/D; lane 13 BS/0/R; lane 14 BS/0/D. b Cluster analysis of DGGE patterns obtained in part a. Scale represents % similarity. Cophenetic correlation values are mentioned at the branch nodes. BS bulk soil, PC plant control, AL azadirachtin (7.5 9 10-3 g active ingredient per L), AH azadirachtin (2.25 9 10-2 g active ingredient), 0 start of experiment, 1 1st time point, 2 2nd time point. D and R represent PCR and RT-PCR profiles, respectively

indicate that several predominantly present bacterial groups in soil are metabolically dormant under certain conditions but would still be detected by DNA based analyses, however certain differences could also be due to inefficiency of reverse transcription (Duineveld et al. 2001). On the other hand, bands that were absent from DNA generated profiles but present in RNA based profiles suggest that bacteria which are too few in number to be detected by PCR, could be metabolically very active showing high transcription levels. From the comparitive DNA and RNA profiles for azadirachtin treated plants, it can be seen that two distinct clusters were obtained (Fig. 3b). The PCR profiles clustered together separated from the cluster formed by the RTPCR profiles and the two clusters showed a similarity of only about 64 %. Thus the UPGMA analysis clearly distinguished between the DGGE patterns derived from DNA and RNA forming two separate clusters. Among the DNA profiles, the UPGMA analysis revealed no differences across time points or between L and H treatments of azadirachtin. Among the RNA profiles, the profile for AH treatment at second time point showed only a 76 % similarity to the control RNA profile, while the profile for AL treatment showed high similarity (100 %) to the control RNA profile. This indicates that azadirachtin affected only the active microbial populations, leaving the resident population unaffected, and this effect was dose dependent since only the high dose showed a significantly different profile compared to control. A possible explanation could be that the responses at the RNA level are usually more rapid and have a greater magnitude compared to those at the DNA level (Duineveld et al. 2001). An earlier study done on propargyl bromide and 1,3-dichloropropene treated soils also showed that the DNA profiles for treated soils began to cluster away from the control profiles as the fumigant concentration was increased, indicating a dosedependent response (Dungan et al. 2003). However, in the present study, the dose-dependence was only recorded at the RNA level and not at the DNA level. The RNA profiles of AL and AH treatments also showed a time-dependent


1485 b Fig. 4 a DGGE gel image of PCR and RT-PCR profiles of 16S rRNA

(a) 1

2

3

4

5

6

7

8

9 10 11 12 13 14

9

10 11 12

14

amplified from chlorpyrifos treated Vigna radiata plants at two time points (30 and 70 days post-treatment). Lane numbers are represented above the image. Arrows mark the position of bands excised and sequenced. Lane 1 CH/2/R; lane 2 CH/2/D; lane 3 CL/2/R; lane 4 CL/ 2/D; lane 5 PC/2/R; lane 6 PC/2/D; lane 7 CH/1/R; lane 8 CH/1/D; lane 9 CL/1/R; lane 10 CL/1/D; lane 11 PC/1/R; lane 12 PC/1/D; lane 13 BS/0/R; lane 14 BS/0/D. b Cluster analysis of DGGE patterns obtained in part a. Scale represents % similarity. Cophenetic correlation values are mentioned at the branch nodes. BS bulk soil, PC plant control, CL chlorpyrifos (1 g active ingredient per L), CH chlorpyrifos (3 g active ingredient per L), 0 start of experiment, 1 1st time point, 2 2nd time point. D and R represent PCR and RT-PCR profiles, respectively

14’

(b) BS/0/D BS/0/R PC/1/R CL/2/R PC/2/R CH/1/R CL/1/R PC/2/D CH/1/D CL/1/D PC/1/D CL/2/D CH/2/D CH/2/R

response, since the similarity between the two profiles dropped from 90 % at the first time point to 76 % at the second time point. The comparative DGGE profiles for chlorpyrifos treated plants showed similar patterns compared to azadirachtin

treated plant profiles, generating a larger number of bands in the upper section of the gel with a lower gradient (Fig. 4a). Here again, the DNA and RNA profiles formed separate clusters showing only 68 % similarity level, with the exception of bulk soil DNA profile at zero time point (BS/0/D) which clustered along with bulk soil RNA profile at zero time point, indicating that at the start of the experiment, bacterial groups most active were also the ones that were numerically dominant (Fig. 4b). While the DNA profiles for CL and CH treatments were similar (about 80 %) to control at both time points, it was not the same for RNA profiles. The RNA profiles for CL and CH showed only a 70 % similarity to control at the first time point, indicating that the effect of pesticide was more pronounced at the RNA level. Moreover, at the second time point, while the CL profile was 88 % similar to control, the CH profile clustered far away showing a mere 64 % similarity to control. Thus, similar to the observation for azadirachtin profiles, chlorpyrifos also showed a dose- and timedependent effect on the active bacterial populations. Another possible explanation for the recoup observed in the case of CL with RNA-derived profiles (similarity level increased from 70 to 88 % with control at 2nd time point) is a probable state of ‘‘metabolic freezing’’ at the first time point (as suggested by the increased level of similarity between control, and profiles derived from DNA at the two time points), which was later overcome. Similar to the previous profile patterns, the DNA and RNA profiles of endosulfan treatments were clustered separately with only 65 % similarity level (Fig. 5b). However, most of the control profiles clustered separate from the treated plant profiles. Similar to our observation, Dungan et al. (2003) also reported that the DNA profiles for soils treated with propargyl bromide and 1,3-dichloropropene clustered away from the controls. Likewise, Seghers et al. (2003) reported a differentiation between control and herbicide (atrazine?metolachlor)-treated maize plants through cluster analysis although the similarity level was very high (93 %). Unlike our observation for azadirachtin and chlorpyrifos profiles, the effect of

123

1486

S. Gupta et al. b Fig. 5 a DGGE gel image of PCR and RT-PCR profiles of 16S rRNA

(a) 1

2

3

4 5

6 7

8

9 10 11 12 13 14 16 17 19 21 22 23 24 25 27 28

30

(b) PC/2/R EL/2/R EH/2/R EH/1/R EL/1/R EH/1/D EL/1/D

amplified from endosulfan treated Vigna radiata plants at two time points (30 and 70 days post-treatment). Lane numbers are represented above the image. Arrows mark the position of bands excised and sequenced. Lane 1 EH/2/R; lane 2 EH/2/D; lane 3 EL/2/R; lane 4 EL/ 2/D; lane 5 PC/2/R; lane 6 PC/2/D; lane 7 EH/1/R; lane 8 EH/1/D; lane 9 EL/1/R; lane 10 EL/1/D; lane 11 PC/1/R; lane 12 PC/1/D; lane 13 BS/0/R; lane 14 BS/0/D. b Cluster analysis of DGGE patterns obtained in part a. Scale represents % similarity. Cophenetic correlation values are mentioned at the branch nodes. BS bulk soil, PC plant control, EL endosulfan (1.75 g active ingredient per L), EH endosulfan (5.25 g active ingredient per L), 0 start of experiment, 1 1st time point, 2 2nd time point. D and R represent PCR and RT-PCR profiles, respectively

endosulfan treatment on both resident and active bacterial communities. At the second time point, the EL and EH profiles showed similar trends but became more closer to control, showing 82 and 78 % similarity at DNA and RNA level, respectively. This suggests that the system might have recovered from the impact of pesticide treatment at the second time point. Dungan et al. (2003) also reported that while the DNA profiles of propargyl-bromide and 1,3dichloropropene treated soils clustered away from controls in week 1, the treated soils began to cluster closer to the control by week 4. Thus, for endosulfan, dose-dependence could not be observed since both treatments showed similar trends, but a temporal variation was observable. The observation of strong non-target effects of endosulphan, even at the level of DNA and at the lower dosage, in addition to its persistence and other reported toxic effects (Sutherland et al. 2004), further emphasizes the need for immediate discontinuation of this toxic pesticide in the couple of countries still continuing its application.

PC/2/D

Bacterial identification EL/2/D EH/2/D PC/1/D BS/0/D BS/0/R PC/1/R

endosulfan treatment was noted at both the DNA and RNA level. At the first time point, the EL and EH profiles were quite similar to each other at the DNA (82 %) and RNA (92 %) level. But, compared to control, the treated profiles were only 66 % and 60 % similar at the DNA and RNA level respectively, indicating a substantial impact of

123

Bacterial identification was done by eluting the DGGE bands, re-amplifying them and sequencing the 177 bp region of 16S rRNA. Although the 177 bp target sequence is not sufficient to derive species information, phylogenetic affiliation up to the genus level was derived for most sequences using the NCBI database (Table 1). In this study, 25 eluted bands were sequenced out of which 2 pairs of bands (6, 60 and 14, 140 in Table 1) were excised at the same position based on mobility from two different lanes to check for consistency of gradient throughout the gel. The matching sequences obtained for the bands cut from similar positions confirmed the correctness of gradient formation in the gels. Thereafter, only one representative band from any one of the lanes was cut from a single position. The sequences obtained showed high % similarity with the known sequences in the NCBI database and gave a diverse variety of bacterial species including several uncultured bacteria, Pseudomonas sp.,

Non-target effects of chemical and bio-pesticides Table 1 BLAST search matches of 16S rRNA sequences amplified from DGGE bands

a

Numbers correspond to the gel images in Figs. 3, 4, and 5a

1487

Band (accession number)a

BLAST sequence match

Similarity (%)

1 (KF386613)

Pseudomonas aeruginosa (JX514422.1)

2 (KF386622)

Uncultured Pseudomonas sp. (FR690577.1)

100

99

3 (KF386614)

Uncultured Staphylococcus sp. (JQ820176.1)

98

4 (KF386623)

Pseudomonas aeruginosa (KC791702.1)

5 (KF386615)

Uncultured bacterium (JF150599.1)-closely related to uncultured Firmicutes (JQ433791.1)

100 99

6, 60 (KF386616)

Uncultured bacterium isolate (HQ439336.1)

99

7, 28 (KF386624) 9 (KF386625)

Uncultured bacterium clone (JF205668.1) Uncultured Alcanivoracaceae (JX523668.1)

99 82

10 (KF386626)

Uncultured bacterium (GQ032759.1)—closely related to Staphylococcus haemolyticus (KC573499.1)

93

11 (KF386617)

Uncultured beta proteobacterium (HM110954.1)

88

12 (KF386618)

Cupriavidus necator (NR_102851.1)

89

14, 140 (KF386619)

Bacillus sp. (DQ658998.1)

96

16 (KF386627)

Uncultured bacterium (HM558401.1)—closely related to Acinetobacter (KC466099.1)

97

17 (KF386628)

Acinetobacter schindleri (KC453989.1)

98

19 (KF386629)

Uncultured alpha proteobacterium (KC449409.1)

95

21 (KF386630)

Uncultured Pseudomonas sp. (KC539472.1)

97

22 (KF386631)

Uncultured Pseudomonas sp. (HE590066.1)

93

23 (KF386632)

Uncultured bacterium (JQ373071.2)—closely related to Burkholderia phenazinium (AM502994.1)

92

24 (KF386620)

Burkholderia sp. (AB011287.1)

90

25 (KF386633)

Anoxybacillus toebii (AY466701.1)

94

27 (KF386634) 30 (KF386621)

Weissella confusa (HQ711354.1) Uncultured Pseudomonas sp. (JQ013060.1)

86 90

Bacillus sp., and a- and b-proteobacteria. A major proportion (70 %) of the isolated sequences displayed relationships with a wide range of as yet uncultured bacteria. Earlier studies on plant rhizospheres have reported such high numbers of sequences related to uncultured bacteria (Gremion et al. 2003; Janssen 2006), which clearly demonstrates the significance of the molecular methods over traditional cultivation-dependent methods. About one-third of the sequences tested showed alignments with the bacterial phylum Proteobacteria. In a previous study, a 16S rRNA clone library constructed out of prairie grass soil revealed Proteobacteria to be the most abundant phylum in soil (Elshahed et al. 2008). Most of the Proteobacteria including Pseudomonas showed more intense bands in the DNA profiles compared to RNA (bands 1, 2, 4, 21, 22 in Figs. 3, 5a). However, band 11 (Fig. 4a) representing Proteobacteria showed a greater intensity in RNA profiles compared to DNA profiles. Bands 23 and 24 (Fig. 5a) were closely related to Burkholderia, which is a gamma-proteobacterium. Soil has shown to be an extremely complex ecosystem harboring several yet uncultured bacteria belonging to diverse phyla including Proteobacteria, Acidobacteria and Actinobacteria

(Janssen 2006). The Proteobacteria display an enormous level of morphological and physiological diversity, and are crucial to carbon, nitrogen and sulfur cycling (Kersters et al. 2006). Oliveira et al. (2009) isolated a number of phosphatesolubilizing Burkholderia strains from the rhizosphere of maize. Some Pseudomonas species such as Pseudomonas putida are known to be phosphate-solubilizers and the effect of fungicides (tebuconazole, hexaconazole, metalaxyl and kitazin) was studied on the plant-growth promoting activities of P. putida isolated from mustard rhizosphere (Ahemad and Khan 2012). It was observed that higher concentrations of fungicides suppressed the PGP activities of P. putida. The presence of gram-negative bacteria has been reported in an earlier cultivation dependent study on pea rhizosphere (Scott and Knudsen 1999). Band 12 (Fig. 4a), closely related to Cupriavidus necator, also showed an increased intensity in the RNA profile of chlorpyrifos treated plants. A Cupriavidus strain showing 98 % sequence similarity to Cupriavidus necator has been isolated from the rhizosphere of different agricultural plants growing in alkaline soils (Santos et al. 2012). Band 5 (Fig. 3a) was found to be affiliated to

123

1488

uncultured Firmicutes. This phylum was found to be the most abundant in a study performed to analyze the bacterial diversity in rhizospheres of two Antarctic vascular plants (Teixeira et al. 2010). Band 14 (Fig. 4a) was also identified as a Bacillus species (phylum Firmicutes), which showed the maximum intensity in RNA profiles of chlorpyrifos treated plants at the first time point, indicating that chlorpyrifos treatment initially enhanced the activity of Bacillus strains. In a recent study, a Bacillus strain isolated from the rhizosphere of common bean was shown to have plant growth promoting activities and inhibited several phytopathogens (Kumar et al. 2012). Felske et al. (1998), also found Bacillus to be the most predominantly active bacterium in grassland soils. Certain bands (16, 17, 19 in Fig. 5a) were observed only in bulk soil profiles and not in pesticide treated profiles. These showed sequence similarities to various bacterial groups including Acinetobacter and uncultured alpha-proteobacterium isolates. However, since these bands were also absent from untreated plant profiles, it indicates a ‘‘rhizospheric’’ effect, wherein a plant differentially influences the survival, growth and activity of microorganisms in the rhizosphere depending on the plant species, developmental stage etc. (Duineveld et al. 2001; Felske et al. 1998). Although the bands showing varying intensities in the RNA versus DNA profiles point towards shifts in numerical abundance as well as activity of target bacteria, such band intensities cannot be taken as strictly quantitative data and more sophisticated molecular techniques must be employed, that can provide more useful information such as gene copy number. This study indicates that both chemical- and bio-pesticides display non-target effects on active microbial populations that serve important ecosystem functions. While the recommended dosage of azadirachtin could restore the structure of microbial community faster as compared to the chemical pesticides, it is worthy to note that at higher dosage the biopesticide exerted effects similar to chemical pesticide. Hence, there is a need to educate farmers and reemphasize the significance of recommended dosages. This study also highlights the importance of studying the bacterial diversity in plant rhizosphere at the RNA level and not just at the DNA level, since pesticide treatments were seen to have significant effects on the active populations even when the numerically dominant resident populations remained unaffected. Group-specific primers should be used in future studies to analyze such effects on specific populations, especially populations with essential soil functions such as ammonifiers and denitrifiers. Moreover, while more and more research is being done to study the effects of pesticide use on microbial functional diversity, most of these studies are being performed in a controlled pot experiment, including the current study. Although the

123

S. Gupta et al.

experiments were subjected to natural environmental variations, it cannot mimic the field conditions in which such food crops actually grow. Thus, future research should be targeted towards studying the plant rhizospheres in actual field conditions. Acknowledgments The work was supported by Science and Engineering Research Board, Department of Science and Technology, Government of India. The authors wish to thank Prof. T. R. Sreekrishnan for his constant support in carrying out the work. Conflict of interest of interest.

The authors declare that they have no conflict

References Abdul-Baki AA (1974) Pitfalls in using sodium hypochlorite as a seed disinfectant in 14C incorporation studies. Plant Physiol 53:768–771 Ahemad M, Khan MS (2012) Effect of fungicides on plant growth promoting activities of phosphate solubilizing Pseudomonas putida isolated from mustard (Brassica compestris) rhizosphere. Chemosphere 86:945–950 Cillia V, Lafay B, Christen R (1996) Sequence heterogeneities among 16S ribosomal RNA sequences, and their effect on phylogenetic analyses at the species level. Mol Biol Evol 13:451–461 Corgie’ SC, Beguiristain T, Leyval C (2006) Profiling 16S bacterial DNA and RNA: difference between community structure and transcriptional activity in phenanthrene polluted sand in the vicinity of plant roots. Soil Biol Biochem 38:1545–1553 Duarte GF, Rosado AS, Seldin L, Keijzer-Wolters AC, van Elsas JD (1998) Extraction of ribosomal RNA and genomic DNA from soil for studying the diversity of the indigenous bacterial community. J Microbiol Methods 32:21–29 Duineveld BM, Kowalchuk GA, Keijzer A, van Elsas JD, van Veen JA (2001) Analysis of bacterial communities in the rhizosphere of chrysanthemum via denaturing gradient gel electrophoresis of PCR-amplified 16S rRNA as well as DNA fragments coding for 16S rRNA. Appl Environ Microbiol 67:172–178 Dungan RS, MarkIbekwe A, Yates SR (2003) Effect of propargyl bromide and 1,3-dichloropropene on microbial communities in an organically amended soil. FEMS Microbiol Ecol 43:75–87 Ejaz S, Akram W, Lim CW, Lee JJ, Hussain I (2004) Endocrine disrupting pesticides: a leading cause of cancer among rural people. Exp Oncol 26:98–105 Elshahed MS, Youssef NH, Spain AM et al (2008) Novelty and uniqueness patterns of rare members of the soil biosphere. Appl Environ Microbiol 74:5422–5428 Felske A, Wolterink A, Van Lis R, Akkermans ADL (1998) Phylogeny of the main bacterial 16S rRNA sequences in Drentse A grassland soils (The Netherlands). Appl Environ Microbiol 64:871–879 Girvan MS, Bullimore J, Ball AS, Pretty JN, Osborn AM (2004) Responses of active bacterial and fungal communities in soils under winter wheat to different fertilizer and pesticide regimens. Appl Environ Microbiol 70:2692–2701 Gopal M, Gupta A, Arunachalam V, Magu SP (2007) Impact of azadirachtin, an insecticidal allelochemical from neem on soil microflora, enzyme and respiratory activities. Bioresour Technol 98:3154–3158 Gremion F, Chatzinotas A, Harms H (2003) Comparative 16S rDNA and rRNA sequence analysis indicates that Actinobacteria might

Non-target effects of chemical and bio-pesticides be a dominant part of the metabolically active bacteria in heavy metal-contaminated bulk and rhizosphere soil. Environ Microbiol 5:896–907 Griffiths RI, Whiteley AS, O’Donnell AG, Bailey MJ (2000) Rapid method for coextraction of DNA and RNA from natural environments for analysis of ribosomal DNA and rRNA-based microbial community composition. Appl Environ Microbiol 66:5488–5491 Janssen PH (2006) Identifying the dominant soil bacterial taxa in libraries of 16S rRNA and 16S rRNA genes. Appl Environ Microbiol 72:1719–1728 Kersters K, De Vos P, Gillis M, Swings J, Vandamme P, Stackebrandt E (2006) Introduction to the Proteobacteria. In: Dwarkin M, Falkow S, Rosenberg E, Schleifer K-H, Stackebrandt E (eds) The prokaryotes, vol 5, 3rd edn. Springer, New York, pp 3–37 Kumar P, Dubey RC, Maheshwari DK (2012) Bacillus strains isolated from rhizosphere showed plant growth promoting and antagonistic activity against phytopathogens. Microbiol Res 167:493–499 Mathur HB, Agarwal HC, Johnson S, Saikia N (2005) Analysis of pesticide residues in blood samples from villages of Punjab. Centre for Science and Environment Report, New Delhi. www. cseindia.org/userfiles/Punjab_blood_report.pdf. Accessed 28 Sept 2013 Muyzer G, de Waal EC, Uitterlinden AG (1993) Profiling of complex microbial populations by denaturing gradient gel electrophoresis analysis of polymerase chain reaction-amplified genes coding for 16S rRNA. Appl Environ Microbiol 59:695–700 Oliveira CA, Alves VMC, Marriel IE, Gomes EA, Scotti MR, Carneiro NP, Guimara˜es CT, Schaffert RE, Sa NMH (2009) Phosphate solubilizing microorganisms isolated from rhizosphere of maize cultivated in an oxisol of the Brazilian Cerrado Biome. Soil Biol Biochem 41:1782–1787 Parween T, Jan S, Fatma T (2011) Assessing the impact of chlorpyrifos on growth, photosynthetic pigments and yield in

1489 Vigna radiata L. at different phenological stages. Afr J Agric Res 6:4432–4440 Santos PE, Martı´nez-Aguilar L, Lo´pez-Lara IM, Caballero-Mellado J (2012) Cupriavidus alkaliphilus sp. nov., a new species associated with agricultural plants that grow in alkaline soils. Syst Appl Microbiol 35:310–314 Scott JS, Knudsen GR (1999) Soil amendment effects of rape (Brassica napus) residues on pea rhizosphere bacteria. Soil Biol Biochem 31:1435–1441 Seghers D, Verthe K, Reheul D, Bulcke R, Siciliano SD, Verstraete W, Top EM (2003) Effect of long-term herbicide applications on the bacterial community structure and function in an agricultural soil. FEMS Microbiol Ecol 46:139–146 Sigler W, Turco R (2002) The impact of chlorothalonil application on soil bacterial and fungal populations as assessed by denaturing gradient gel electrophoresis. Appl Soil Ecol 21:107–118 Sutherland TD, Home I, Weir KM, Russell RJ, Oakeshott JG (2004) Toxicity and residues of endosulfan isomers. Rev Environ Contam Toxicol 183:99–113 Teixeira LCRS, Peixoto RS, Cury JC, Sul WJ, Pellizari VH, Tiedje J, Rosado AS (2010) Bacterial diversity in rhizosphere soil from Antarctic vascular plants of Admiralty Bay, maritime Antarctica. ISME J 4:989–1001 Vallaeys T, Topp E, Muyzer G, Macheret V, Laguerre G, Rigaud A, Soulas G (1997) Evaluation of denaturing gradient gel electrophoresis in the detection of 16S rDNA sequence variation in rhizobia and methanotrophs. FEMS Microbiol Ecol 24:279–285 Widenfalk A, Bertilsson S, Sundh I, Goedkoop W (2008) Effects of pesticides on community composition and activity of sediment microbes—responses at various levels of microbial community organization. Environ Pollut 152:576–584 Zhang C, Liu X, Dong F, Xu J, Zheng Y, Li J (2010) Soil microbial communities response to herbicide 2,4-dichlorophenoxyacetic acid butyl ester. Eur J Soil Biol 46:175–180

123