土壌中における農薬分解菌の生態と農薬が及ぼす影響評価に関する研究

Abstract

Ecological properties of microorganisms involved in aerobic degradation of 2,4-D, 2,4,5-T and salithion and in anaerobic degradation of chlorophenols were examined based on their phylogenetic classification, identification and characterization of genes and enzymes, and/or distribution, diversity and succession in the soil or sediment environment by culture- and molecular-based methods. Side effects of pesticides were evaluated on the soil microbial community by analyzing community-level physiological profiles, on duckweed proliferation considering long-term exposure, recovery potential from the damage and mixture effects of the pesticides, and on river biofilm by assembling a model river biofilm.

Introduction

The role of soil microorganisms in the degradation of pesticides is critical for environmental safety; therefore, understanding their ecological properties is important. In addition, there are some pesticide degraders that can not only degrade the pesticide but also use it as sole carbon and energy source using specific enzymes for the degradation. The microbial ability to adapt to a variety of compounds is interesting from an evolutionary point of view. On the other hand, it is important to evaluate the side effects of pesticides on soil microorganisms due to their crucial role in the soil ecosystem. Moreover, evaluation of the side effects on aquatic organisms is necessary considering the unintentional exposure of the pesticides beyond the agricultural fields. A review of our studies on the ecology of pesticide-degrading microorganisms in soil and the assessment of pesticide effects on the ecosystem is presented in this article.

1. The ecology of pesticide-degrading microorganisms in soil

1.1. 2,4-D and 2,4,5-T degraders

The 2,4-D-catabolizing bacterial strain RD5-C2 was isolated from Japanese upland soil that had no history of 2,4-D exposure. RD5-C2 was closely related to Bradyrhizobium sp. in α-Proteobacteria and also 2,4-D degraders previously isolated from pristine environments of Hawaii, Canada and Chile. The 2,4-D/α-ketoglutarate dioxygenase gene homolog (tfdAα) was cloned from the strain, and a MalE-TfdAα fusion protein exhibited approximately 10 times greater activity for phenoxyacetate than did 2,4-D. It was suggested that tfdA in β- and γ-Proteobacteria and tfdAα in α-Proteobacteria arose by divergent evolution from a common ancestor.

The tfdAα gene was found in all Bradyrhizobium strains examined including non-2,4-D-degrading root-nodulating strains. The phylogenetic tree was congruent with that of 16S rRNA genes, indicating evolution of tfdAα without horizontal gene transfer. On the other hand, cadA, another gene encoding 2,4-D-degrading proteins that are characteristic of the 2,4-D-degrading Bradyrhizobium sp., showed limited distribution in 2,4-D-degrading Bradyrhizobium sp. and in some strains of non-2,4-D-degrading B. elkanii. The cadA genes were phylogenetically separated between 2,4-D-degrading and nondegrading strains, indicating the incongruence of cadA with 16S rRNA genes and suggesting that the cadA genes of 2,4-D-degrading and nondegrading Bradyrhizobium spp. have different origins. The tfdB gene encoding chlorophenol hydroxylase and its homolog were found in RD5-C2. TfdBa showed the highest activity for 2,4-dichlorophenol, while TfdBb presented no observable activity for any chlorophenols tested. Their different codon usage patterns and GC contents suggest that tfdBa was obtained through horizontal gene transfer.

Diverse 2,4-D- and 2,4,5-T-degrading bacteria were isolated countrywide in Vietnamese soils. The major degraders phylogenetically consisted of Burkholderia spp. (43.3%), Sphingomonas spp. (40.2%) and Ralstonia spp. (15.3%). The 2,4,5-T degraders, 65% of all degraders, were Sphingomonas spp., Burkholderia spp. and Bradyrhizobium sp. The distribution of the degradative genes, tfdA (tfdAα), tfdB, tftA (cadA) and tftC, was examined for the isolates. It was found that the 2,4-D- and 2,4,5-T-degrading microbial consortia have spread countrywide and are diverse on a genetic as well as geographic basis.

Taxonomically various fungal strains isolated from the Vietnamese soils included 45, 12 and 4% degraded phenoxyacetic acid (PA), 2,4-D and 2,4,5-T, respectively. While the PA-degrading fungi were distributed among many genera, the 2,4-D degraders were found only in the order Eurotiales in the class Eurotiomycetes. All of the 2,4,5-T-degrading fungal strains were isolated from southern Vietnam and were phylogenetically close to Eupenicillium spp. It was suggested that the heavily contaminated environments enhanced the strains toward obtaining the ability to degrade 2,4,5-T.

Changes in the bacterial community in soil-water suspensions during the enrichment period of 2,4-D and 2,4,5-T degraders were examined using denaturing gradient gel electrophoresis (DGGE) analysis of the 16S rRNA gene. The nucleotide sequences of almost all major bands at the degradation corresponded to those of 2,4-D- and 2,4,5-T-degrading isolates, and successions of the degraders were demonstrated in the DGGE profile. These results suggest that 2,4-D- and 2,4,5-T-degrading isolates were responsible for the degradation.

1.2. Salithion degraders

Salithion (SLT) degraded rapidly in aerobic soils to CO₂ via cleavages of P-O-aryl and P-O-aralkyl linkages, demethylation, and oxidative desulfuration of the thiophosphoryl moiety. When each enantiomer of SLT was applied separately, (S)-SLT disappeared 1.5–1.7 times faster than did (R)-SLT. Acinetobacter sp. (B-60) with high demethylating activity and Agrobacterium sp. (B-7, B-15 and B-17) with high activity in cleaving the P-O-aryl bond were isolated. (S)-SLT was degraded faster by B-7 than was (R)-SLT by stereoselective cleavages of the P-O-aryl and P-O-aralkyl linkages. Both (R)- and (S)-SLT were degraded at similar rates by B-15. (S)-SLT was degraded faster than was (R)-SLT by B-17, in which cleavages of the P-O-aryl and P-O-aralkyl linkages were the main pathways of (S)-SLT, whereas demethylation is a main pathway of (R)-SLT. B-60 degraded both enantiomers; however, the pathway of each enantiomer was in contrast to that by B-17. The salithion degraders showed weak activity for fenitrothion, parathion and cyanophos, suggesting that the enzymes involved were fairly specific to the unique cyclic structure of SLT.

1.3. Anaerobic degradation of chlorophenols

Among mono- and dichlorophenol isomers, anaerobic dechlorination was observed in 2-chlorophenol, 3-chlorophenol and/or 2,6-dichlorophenol in the sediment of Lakes Shinji and Nakaumi. The highest activity was observed at 0.7% of sodium chloride and 6.0 mM of sodium sulfate in the Nakaumi sediment, whereas it was observed without the addition in the Shinji sediment. The chlorophenols were degraded via phenol and benzoate in both sediments under methanogenic conditions. Microbial consortia enriched with each monochlorophenol showed different substrate specificities for chlorophenols. Using the limiting dilution method with PCR-DGGE, it was suggested that Dehalobacter sp. corresponded to the dechlorination of 4-chlorophenol, and that the enriched culture transforming phenol to benzoate consisted of the homologous strain with Cryptanaerobacter phenolicus.

2. Assessment of pesticide effects on the ecosystem

2.1. Effects on the soil microbial community

Natural fluctuations in community-level physiological profiles of microorganisms in low-input and conventional rice paddy soils were monitored using Biolog GN plates for almost two years in order to establish criteria for assessing side effects of pesticides on soil microbial ecosystems. Soil microbial communities were grouped into three clusters based on the sampling season; the first cluster was further divided between the two paddy soils. Low soil temperatures and reductive soil conditions seemed to influence the establishment of the second and third clusters, respectively. The effects of the herbicide Zark D51 (ZD, daimuron and bensulfuron methyl) powder and the fungicide Fuji-One Moncut (MC, isoprothiolane and flutolanil) powder on microbial communities in a paddy soil were examined in laboratory experiments. In the case of MC applied at 50 times the recommended rate, the microbial communities changed for at least four weeks. The significance of the change was assessed by comparing the magnitude of the change with that caused by soil flooding, which had been shown to be the most influential environmental parameter acting on the microbial communities in paddy soils.

2.2. Effects on duckweed

The toxicity of several herbicides with a different mode of action was examined in short- and long-term exposure tests using Lemna gibba. The herbicides tested showed stronger toxicity with an increasing exposure period. The recovery potential of L. gibba from the damage after short- and long-term exposure was also examined. In the recovery period, L. gibba started to grow again even in plots where it did not grow during the exposure period; a longer period of exposure affected its recovery. The phytostatic and phytocidal concentrations decreased with the exposure period. The mixture effects of different types of herbicides were evaluated on the basis of the effects of long-term exposure and potential for recovery from the damage. Their combined effect often appeared stronger than the effect expected individually; therefore, their effect could not be predicted from the standard toxicity test using a single herbicide. In addition, the effect of long-term exposure and the recovery potential were also affected by their combination, indicating that further understanding of the mechanisms of the mixture effects is needed.

2.3. Effects on river biofilm

Assessment of the side effects of pesticides on river biofilm is difficult due to their high variability. To evaluate the effects of pesticides, we constructed a model biofilm consisting of two diatom strains and three bacterial strains on the glass surface. Microbial cell numbers, esterase activity, chlorophyll-a content and the community structure of the model biofilm were measured and found to be useful as biological parameters for evaluating pesticide effects. The model biofilm was reproducibly formed through the cooperative interaction of bacteria and diatoms, and a herbicide atrazine, which only inhibits diatoms, indirectly inhibited bacterial growth in the model biofilm in a dose-response manner. The results suggest that this model of biofilm could be used for testing the potential effects of pesticides on natural river biofilms at a community level with high reproducibility.

Conclusion

There are diverse pesticide-degrading microorganisms in the soil. Their behavior and activities depend on physicochemical and biological interactions with their microenvironments in the soil. Both culture-dependent and molecular-based approaches are important for understanding their ecological properties. In an assessment of the side effects of pesticides, several aspects, such as long-term exposure, recovery potential, combined pesticides effects, and biological community in the ecosystem, should be incorporated in the future.