Abstract
Polychlorinated biphenyl (PCB)-contaminated soils represent a major treat for ecosystems health. Plant biostimulation of autochthonous microbial PCB degraders is a way to restore polluted sites where traditional remediation techniques are not sustainable, though its success requires the understanding of site-specific plant–microbe interactions. In an historical PCB contaminated soil, we applied DNA stable isotope probing (SIP) using 13C-labeled 4-chlorobiphenyl (4-CB) and 16S rRNA MiSeq amplicon sequencing to determine how the structure of total and PCB-degrading bacterial populations were affected by different treatments: biostimulation with Phalaris arundinacea subjected (PhalRed) or not (Phal) to a redox cycle and the non-planted controls (Bulk and BulkRed). Phal soils hosted the most diverse community and plant biostimulation induced an enrichment of Actinobacteria. Mineralization of 4-CB in SIP microcosms varied between 10% in Bulk and 39% in PhalRed soil. The most abundant taxa deriving carbon from PCB were Betaproteobacteria and Actinobacteria. Comamonadaceae was the family most represented in Phal soils, Rhodocyclaceae and Nocardiaceae in non-planted soils. Planted soils subjected to redox cycle enriched PCB degraders affiliated to Pseudonocardiaceae, Micromonosporaceae and Nocardioidaceae. Overall, we demonstrated different responses of soil bacterial taxa to specific rhizoremediation treatments and we provided new insights into the populations active in PCB biodegradation.
Similar content being viewed by others
Introduction
Polychlorinated biphenyl (PCB)-contaminated soils represent a major environmental source of persistent organic pollutants (POPs) worldwide, with important concerns for ecosystems and public health1,2. Microbial communities providing PCB biodegradation abilities through reductive dichlorination and aerobic pathways are the main agents of natural attenuation in historically contaminated environments, hence, when properly biostimulated, constitute a natural resource for the reclamation of polluted soils3. Several studies have reported that the “rhizosphere effect” exerted by plant root exudation and root deposition on the soil microbiome has a positive influence on the degradation of POPs, including PCBs, as a consequence of the induction of microbial degradation pathways by secondary plant metabolites4,5,6,7. Therefore, plant-driven biostimulation of PCB-degrading soil microbial populations (i.e., rhizoremediation) has been identified as a way to accelerate the natural attenuation process in the attempt to restore ecosystem services in those sites where the implementation of traditional remediation techniques based on soil removal and incineration are not sustainable due to costs and environmental impact, such as extended agricultural areas8. However, root growth and exudation profile, soil microbial community, and their interdependent relationships may undergo temporal and spatial variations depending on the plant species and phenological growth stage along with soil edaphic conditions and contamination profile9,10,11.
In a previous, 18-month rhizoremediation study12 we reported how ten plant species combined with soil treatments differentially affected the structure and activity of the soil microbial community and ultimately the degradation rate of different PCB congeners in a weathered contaminated soil. In the present work we focused on the soils that during this rhizoremediation trial had been i) biostimulated with reed canary grass (Phalaris arundinacea L.) and/or ii) subjected to periodic flooding with the aim to induce a redox cycle that would allow both reductive dichlorination and aerobic degradation of PCBs by autochthonous microorganisms13. These treatments were chosen because they promoted the reduction of the concentration of tri-, tetra- and penta-chlorinated PCB in the original soil after 18 months12. We hypothesized that the combined or independent application of plant biostimulation and cyclic oxic-anoxic conditions given by the periodical flooding would shape the soil bacterial communities and would enrich different populations among the autochthonous PCB degraders present in this historically polluted soil. Therefore, we applied DNA-based stable isotope probing (SIP) incubation combined with 16S rRNA amplicon sequencing to determine how the plant growth and the redox cycles affected the structure of bacterial communities and their diversity and, in particular, the metabolically active bacteria potentially involved in PCB degradation. SIP has been successfully employed in bioremediation-oriented studies to detect bacterial taxa able to incorporate carbon from diverse organic pollutants14,15. Compared to other molecular techniques, this approach allows to directly link metabolic capability to phylogenetic information within a microbial community by tracking isotopically labeled carbon into the DNA, and therefore in phylogenetic biomarkers such as the 16S rRNA gene, only in actively duplicating cells. This technology enables the identification of only the organisms active in the utilization of a specific substrate, requiring less sequencing effort than a full metagenomic analysis of the total community15. To the best of our knowledge, whereas the non-chlorinated biphenyl backbone is generally considered a model molecule for PCB mineralization, only one study compared the results obtained from SIP incubation of sediment samples with 13C-labeled biphenyl or 13C-4-chlorobiphenyl (4-CB)16. This implies a lack of knowledge on the metabolization of chlorinated congeners in environmental matrices and especially in soils differentially biostimulated. To overcome this limitation and target more specifically the PCB biodegradation, in this work we used 13C-labeled 4-chlorobiphenyl as a metabolic tracer.
Materials and methods
Soil samples
Samples of soil were collected at the end of an 18-month rhizoremediation trial performed in greenhouse conditions. The trial was conducted on former agricultural soils from the National Priority Site for remediation (SIN) Brescia-Caffaro17 which were contaminated by PCBs but also PCDDs, PCDFs, DDT and its isomers, metalloids, and metals (mainly As and Hg). The experimental set-up, conditions, and applied treatments are described in detail in our previous publication12. In this work, we used soils collected from four different treatments: biostimulation with Phalaris arundinacea (Phal) and non-planted control (Bulk); biostimulation with Phalaris arundinacea subjected to periodic soil flooding obtaining an oxic-anoxic (redox) cycle (PhalRed) and non-planted control subjected to the same redox cycle (BulkRed). Each treatment was performed on the same homogeneous original batch of soil in separate pots in three biological replicates.
Stable isotope probing (SIP) incubation
We set up SIP microcosms for three incubation times: seven (D07), twenty-one (D21) and twenty-eight (D28) days. Triplicate microcosms were established for each time point and soil treatment, corresponding to each original biological replicate of the four rhizoremediation treatments, totaling 36 microcosms. The set-up procedure was as follows: 10 µl of 13C-labeled 4-chlorobiphenyl (4-CB, 99% 13C, Alsachim) in acetone solution (50 mg/ml) were spiked on 0.2 g of quartz sand placed at the bottom of 100-ml sterile serum bottles to have a final quantity of 0.5 mg of 4-CB per microcosm. After acetone evaporation, 2 g of soil were added to each microcosm, manually mixed with the sand, and supplemented with 200 µl of basal mineral salt solution18. The microcosms were sealed and placed at 25 °C in the dark, to avoid the proliferation of photosynthetic microorganisms, until destructive harvesting when the content was stored at -80 °C. Parallel incubation with unlabeled 4-CB was carried out for each microcosm as a control.
Isotopic analysis of headspace CO2
At each harvest time point, the % of O2 and CO2 in the headspace of each microcosm was measured using a CheckPoint II—Portable Headspace Analyzer (Dansensor). For the determination of CO2 isotopic composition, 500 μl of the headspace gas was sampled with a gas-tight syringe from labeled microcosms and transferred into helium-filled borosilicate vacutainers. The isotopic composition of headspace CO2 was analyzed by Gasbench II device equipped with a cryotrap coupled to a Delta V Advantage isotope ratio mass spectrometer (ThermoFisher Scientific, Bremen, Germany). The mineralization of 13C-labeled 4-chlorobiphenyl was calculated stoichiometrically based on the amount of 13CO2 evolved19.
DNA extraction and isopycnic separation
Basing on the results of the isotopic CO2 analysis, the metagenomic DNA was extracted from D21 and D28 SIP microcosms and the initial, non-incubated, time zero (T0) soil samples using the FastDNA™ SPIN Kit for Soil (MP Biomedicals) and purified with the Genomic DNA Clean & Concentrator™ (Zymo Research) following the manufacturer’s instructions. D07 samples were excluded from DNA extraction because no 13C-labelled CO2 was retrieved in the microcosms’ headspace. DNA concentration was measured with a Qubit™ fluorometer (Thermo Scientific) and all samples were adjusted to a final concentration of 25 ng/µl.
For isopycnic separation, 40 µl (1,000 ng) of DNA were placed into a Quick-Seal™ centrifuge tube (Beckman) filled with cesium trifluoroacetate (CsTFA) pre-diluted to a concentration of 1.62 g/ml. Ultracentrifugation was run for 72 h at 145,000 × g, 24 °C, using an Optima™ MAX-XP Ultracentrifuge with TLN 100 rotor (Beckman). After ultracentrifugation, each gradient was fractionated into fractions of 100 μL using a Beckman Fraction Recovery System (Beckman) and a Harvard Pump 11 Plus Single Syringe (Harvard Apparatus) by replacement with water at a flow rate of 400 μL/min. To ensure reproducibility and accuracy, each isopycnic centrifugation and gradient fractionation were run in parallel with two blanks (water instead of DNA), then the buoyant density (BD) of each gradient-recovered fraction was inferred by measuring the refractive index of fractionated blank samples with a Digital Hand-held Refractometer (Reichert Analytical Instruments). DNA was retrieved from CsTFA by isopropanol precipitation with glycogen18 and stored at -20 °C.
The distribution of bacterial DNA across the density gradients was assessed by quantitative PCR (qPCR) targeting the 16S rRNA gene using primers 357F (3’-CCCTACGGGAGGCAGCAG-5’) and 907R (3’-CCGTCAATTCCTTTGAGTTT-5’) as previously described20. In each gradient, fractions that were determined to contain 13C-labeled DNA based on 16S rRNA gene copies distribution as a function of buoyant density (BD) were combined into pools. Equivalent pools were prepared for controls incubated with unlabeled substrates and were subjected to the same downstream analyses as 13C-labeled DNA to identify any background contamination. Fractions corresponding to the main 16S rRNA peak containing 12C-DNA of each gradient from both labeled and unlabeled SIP incubations and from initial, non-incubated T0 soils were also pooled and analyzed.
16S rRNA gene sequencing and data analysis
We applied Illumina MiSeq sequencing to the pooled-DNA samples after amplification using primers 515F (5’-GTGYCAGCMGCCGCGGTAA-3’) and 926R (5’-CCGYCAATTYMTTTRAGTTT-3’) targeting the V4-V5 hypervariable region of the 16S rRNA gene21. Reactions were performed at the Core Facility for Nucleic Acid Analysis, University of Alaska Fairbanks (USA). The 16S rRNA gene amplicon sequence data were processed using the DADA2 pipeline22 in R software with a few modifications to the DADA2 SOP. In brief, the primer sequences were trimmed off or, if absent, the whole read was discarded. In order to manage the lower quality towards read ends, forward and reverse reads were truncated to the length of 248 and 176 bp, respectively. The values were calculated as the average positions where 75% of reads had a quality score > = 25 while ensuring a hypothetical minimum of 25 bp overlap between the paired reads. Filtering was based on the reads quality using the following parameters: maxN = 0, maxEE = 2, truncQ = 2. After dereplication and application of DADA2-based removal of sequencing errors, denoised forward and reverse reads were merged and chimeric sequences were removed using the method = "pooled". Based on the analysis of the mock community, which consisted of three bacterial strains (i.e., Rhodococcus sp. 3B12, ENA accession LT978383; Acinetobacter sp. P320 ENA accession LT838128; Bacillus sp. P28, ENA accession LT838114) that were amplified in parallel with the DNA samples, the sequences which differed by one base were clustered together and the most abundant sequence from each of the clusters was considered as the valid one. Taxonomy was assigned by using Silva v13823 to create a database of amplicon sequence variants (ASVs). The sequence reads were deposited in the NCBI SRA database under the BioProject ID: PRJNA809320.
Prior to further analysis, sequencing data were rarefied at 4,000 reads per sample to ensure the comparability between samples. A principal coordinate analysis (PCoA) was used to assess the phylogenetic β-diversity based on Bray–Curtis distance matrix on the normalized (log transformed) ASV table. Significant differences in bacterial community composition were investigated by canonical analysis of principal coordinates (CAP) and permutational multivariate analysis of variance (PERMANOVA), according to the factors ‘biostimulation’ (levels: ‘planted’, ‘no plant’), ‘redox’ (levels: ‘redox’, ‘no redox’), ‘treatment’ (levels: ‘Phal’, ‘Bulk’, ‘PhalRed’, ‘BulkRed’), ‘time’ (levels: T0, D21, D28) and ‘treatment’ and ‘time’ interaction. PERMDISP analysis was implemented prior to CAP and PERMANOVA to test the homogeneity of the data dispersion. Statistical analyses were conducted in PRIMER v. 6.1, PERMANOVA + for PRIMER routines24. Alpha-diversity indices (i.e., Shannon diversity and dominance) were calculated using the PAST software25 and their statistical difference was evaluated using the R software version 4.0.226.
To identify the bacterial taxa incorporating 13C from 13C-labeled 4-CB, we summed the number of sequences for each ASV present in each of the three 13C-labeled replicate samples and made a comparison with the respective unlabeled control. Only ASVs that were at least ten-fold more abundant in the 13C-DNA pools were considered to have incorporated 13C.
Sequencing results of pools of fractions corresponding to the main 16S rRNA peak of each control DNA gradient from T0 and D21 / D28 unlabeled SIP incubations were analyzed to characterize the community dynamics among different soil treatments and over time.
Results
Evolution of labeled CO2 and 4-chlorobiphenyl mineralization
Results showed that after seven days (D07) of incubation there was no evolution of 13CO2 in the microcosms supplemented by 13C 4-CB, while it was detected after twenty-one days (D21) and increased at the end of SIP incubation (D28) (Fig. 1, Supplementary Table S1). For this reason, D07 samples were excluded from further processing and analysis. Mineralization of 4-CB was calculated from the concentration of 13CO2 and depletion of the labeled substrate occurred in all D21 and D28 microcosms and varied between 10% in Bulk soil and 22% in BulkRed soil microcosms at D21 and reached up to 39% for the PhalRed soil at D28. Even though planted and flooded soils seemed to support a higher 13C 4-CB mineralization rate (Fig. 1), no significant differences (p > 0.05, Student t-test) were observed among treatments, possibly because of the intrinsic variability of soil samples and the limited number of biological replicates.
Soil samples from D21 and D28 were therefore selected for labeled DNA isolation and further analysis to identify 13C-labeled bacterial populations.
Isolation of labeled DNA
Following density gradient ultracentrifugation and fractionation of DNA, we performed 16S rRNA gene qPCR analyses to assess DNA distribution in the CsTFA gradients (Supplementary Table S2 A-F). Gradients from all D28 samples presented a small 16S rRNA peak corresponding to heavy fractions containing labeled DNA in addition to samples Bulk 3 and BulkRed 2 from D21. Thus, we considered only D28 samples for further analyses. A main 16S rRNA peak containing unlabeled DNA was observed at buoyant density of 1.60 g/mL, while 13C-DNA-containing fractions were identified at BDs between 1.61 and 1.64 g/mL. Minor quantities of 16S rRNA gene (Supplementary Table S2 A-F) were detected at the same BDs in samples from parallel incubations with unlabeled 4-CB, that were analyzed as controls.
Diversity of soil bacterial communities
A total of 10,921,966 high-quality paired-end reads with an average length of 352 bp were obtained from 16S rRNA amplicon sequence data spanning all samples and replicates. Taxonomy assignment resulted in a database of 21,185 unique ASVs. After rarefaction to 4,000 reads per sample, the resulting ASV table was composed of 21,167 ASVs. Samples that were below the rarefaction threshold were excluded from the dataset (Supplementary Table S3 A-B). ASVs representing less than 0.005% of the total bacterial community were cut off prior to further analyses.
Following PERMDISP validation (p = 0.179), the beta-diversity analysis on the ASV dataset showed that bacterial communities of soils subjected to different treatments grouped separately as represented by the PCoA (Fig. 2A) and differed significantly from one another as demonstrated by CAP (delta_1^2: 0.98974, p = 0.0001) and PERMANOVA main and pairwise tests, with the exception of the two non-planted soils (Supplementary Table S4A-B). Beta-diversity based on ASV composition according to the time of SIP incubation cannot be discussed due to the difference in dispersion among soil samples of T0, D21, and D28 (PERMDISP: p = 0.021), however we could observe a diversity given by this factor when considered together with the soil treatment (PERMDIS p = 0.142; PERMANOVA: F2,6 = 2.46, p = 0.0001). Estimation of the factors’ contributions to the observed variations determined by PERMANOVA showed that “treatment” alone explained 28% of the variation and an additional 23% was explained by its interaction with the factor “time” (Supplementary Table S4C). Furthermore, PCoA and PERMANOVA (Supplementary Figure S1; Supplementary Table S5) separated the communities of planted soils from the bulk soils (PERMDISP: p = 0.072), and those of soils subjected to a redox cycle from those that were not (PERMDISP: p = 0.748).
Bacterial alpha-diversity was affected by different soil treatments, with Phal soils hosting the most diverse community compared to the other treatments, as shown by the Shannon index and demonstrated by ANOVA (Fig. 2B). Soils planted with P. arundinacea and not subjected to redox cycle (Phal) were also characterized by lower dominance, even though the difference was statistically significant only compared to the non-planted Bulk control (Fig. 2C). The analysis of the alpha-diversity also showed a decrease in the Shannon index over time, while the dominance index followed an opposite trend. However, both the parameters differed significantly only between pre-incubation (T0) and SIP-incubated soils (D21, D28) according to two-way ANOVA (Supplementary Table S6).
The main bacteria phyla, having relative abundance between 25 and 5%, observed in the total community of pre-incubation samples (T0) were Proteobacteria, Actinobacteria, Acidobacteria, Chloroflexi and Planctomycetes. The different soil treatments presented a different profile in terms of phyla/class relative abundance (Supplementary Figure S2). ASVs affiliated with Actinobacteria were dominant in soil planted with P. arundinacea (Phal) (25% on average between the replicates), while they declined in non-planted soil samples (Bulk, 15.5%) and in both planted and non-planted soils subjected to redox cycle (PhalRed, 13.2% and BulkRed, 9.2%). ASVs affiliated with Proteobacteria showed similar abundance in all the T0 soil samples but were mainly represented by Alphaproteobacteria in Phal (11.9%) and Bulk (16.7%) soils, and by Betaproteobacteria in the repeatedly flooded soils (Phalred and BulkRed, 10% on average). Chloroflexi were enriched in PhalRed and BulkRed soils, representing respectively 20.5% and 19.8% of the total community. Other less abundant taxa whose relative abundance changed with soil treatments were Firmicutes, which were reduced in plant-biostimulated soils, and Desulfobacterota, which were increased in soils subjected to redox conditions.
After incubation with 4-chlorobiphenyl, ASVs affiliated with Proteobacteria became prevalent, representing the 30% of the communities of Phal and PhalRed soils at D21 and further increasing up to 40% in Phal soils at D28. Actinobacteria were more abundant in PhalRed soils at D28 compared to T0 while the abundance of Chloroflexi was decreased in all the D28 communities.
Identification of bacterial taxa deriving carbon from 4-chlorobiphenyl
Fifty-seven different ASVs corresponding to bacterial taxa incorporating 13C derived from the mineralization of 4-CB were identified in D28 soils after rarefaction (Supplementary table S7A). We observed the highest number of ASVs (n = 30) in Bulk soils, followed by PhalRed (n = 19), Phal (n = 14), and BulkRed (n = 6). Most of these ASVs (n = 46) were unique to the different soil treatments and none was shared among all of them (Fig. 3A). Only one ASV, affiliated with the genus Rhodococcus, was common in three of the considered soils (i.e., Bulk, PhalRed and BulkRed).
Within the bacteria deriving carbon from 4-CB, the most abundant class/phyla in all four treated soils were Betaproteobacteria and Actinobacteria (Fig. 3B). In particular, the bacterial ASVs detected in Phal soils mostly belonged to the family Comamonadaceae (85.7% of the 13C-incorporating bacteria), which represented only 20% in Bulk and 24.3% in PhalRed soils and were absent in BulkRed soils (Table 1). Instead, we found Rhodocyclaceae among the 13C-incorporating bacteria only in non-planted soils (4.2% in Bulk and 30.8% in BulkRed). Among Actinobacteria, Nocardiaceae were represented only by ASV3, and were particularly abundant in Bulk and BulkRed soils (28.1% and 58.4% respectively), while in planted soils they represented only 2.6% in PhalRed and were absent in Phal. PhalRed soil bacteria that derived 13C from 4-CB were affiliated with other Actinobacteria families, such as Pseudonocardiaceae (18.3%), Micromonosporaceae (8.1%), and Nocardioidaceae (4.7%). Alphaproteobacteria deriving carbon from 4-CB were also retrieved in all the soils but were more abundant in planted (7.7% in Phal and 19.7% in PhalRed) than in non-planted soils (4.3% in Bulk and 5.7% in BulkRed). The most represented families within this clade included mainly Dongiaceae (7.8%) and other unidentified taxa (6.7%) in PhalRed soils and Sphingomonadaceae in BulkRed (5.7%). Other phyla that were identified to have derived 13C from 4-CB were Chloroflexi (13.6% in PhalRed, 8.7% in Bulk and 2.4% in Phal samples) and Acidobacteria (5.0% in BulkRed, 3.9% in PhalRed and 1.9% in both Phal and Bulk samples). Considering the relative abundance of ASVs corresponding to 13C-deriving bacteria in the total community of each sample (i.e., 4,000 reads obtained after rarefaction) and comparing it with the relative abundance of the same ASVs in T0 soil bacterial communities, we could infer that the most relevant families identified herein as deriving carbon from 4-CB were enriched during the SIP incubation period (Supplementary Table S8).
Discussion
Rhizoremediation through plant biostimulation of autochthonous microbial degraders is considered an environmentally sustainable and cost-effective promising technology for the clean-up of weathered PCB-polluted soils8. However, plant species-specific rhizosphere effect together with soil treatment practices, differentially impact soil bacterial communities, which further affects pollutants’ depletion rates6,27,28. Consequently, a lack of knowledge still exists about how biodegradation of PCB congeners occurs and which are the main microbial players under different soil rhizoremediation treatments. In this work, a historically contaminated soil containing an initial microbiome naturally selected during twelve years of natural attenuation17 was biostimulated for 18 months using reed canary grass and/or redox cycle, aiming at the enrichment of microbial degraders. Subsequent soil incubation with 13C-4-CB demonstrated the presence of bacteria that were able to assimilate 4-CB.
The results of the evolution of 13CO2 in soil microcosms showed that after one week of SIP incubation there was no mineralization of the labeled substrate regardless of the original soil treatment, and, apart from two D21 samples, we were able to detect 13C labeled DNA only after twenty-eight days (D28). These results are in contrast with the only previous studies that used 4-CB as a metabolic tracer of PCB biodegradation in a stream sediment16,29, in which the substrate mineralization and the identification of bacterial degraders occurred after one to seven days of incubation. Possible explanations could be that the different used matrices are characterized by diverse native bacterial communities together with environmental conditions and pollution profiles that can differentially impact the biodegradation rate, as shown for biphenyl by Chen and co-authors30. These differences may reflect in the incorporation of 13C by bacterial cells and consequently in DNA-labelling: for instance, in other works labeled DNA was detected only after fourteen days of incubation with biphenyl31,32. As 4-CB has a slower depletion rate compared to biphenyl due to the presence of chlorine33, it is possible that the conditions of the soil samples used in this study, such as the presence of other classes of contaminants in addition to PCBs34, determined a longer time for 4-CB mineralization and for the incorporation of derived carbon into bacterial DNA.
The results of 16S rRNA amplicon sequencing from the DNA of the T0 soils before the 4-CB microcosm enrichment, confirmed that the applied treatments, i.e. biostimulation, periodic flooding, and their combination, induced the differentiation of the structure and taxonomic composition of the total bacterial community, corroborating previous information on beta-diversity obtained by DNA fingerprinting targeting the 16S rRNA- 23S rRNA intergenic spacers12. Considering the dominant taxa, plant biostimulation with P. arundinacea induced an increase of Actinobacteria, though this effect was partially attenuated by the soil flooding. The phyla Chloroflexi and Desulfobacterota (formerly included in the class Deltaproteobacteria35) were significantly enriched in soils subjected to the redox cycle, suggesting the occurrence of reductive dechlorination metabolisms, potentially able to reduce the concentration of highly chlorinated molecules. Members of dehalorespiring taxa, such as Dehalococcoides and Geobacter, have been in fact studied for their PCB dehalogenation activity36,37. Besides the effect exerted by soil treatments on the original soil microbiota, our data also report a shift towards less diverse bacterial communities over SIP incubation time in terms of alpha-diversity, with the enrichment of Proteobacteria and Actinobacteria, putatively driven by the incubation condition.
The combination of SIP and 16S rRNA gene sequencing allowed us to identify the ASVs deriving carbon from 4-CB at the end of SIP incubation (D28). The bacterial cells incorporating 13C may have acquired carbon both through direct degradation of 4-CB and via cross-feeding on catabolic intermediates or other metabolites. In a time-flow experiment with parallel incubation using biphenyl or benzoate as substrates32 it was suggested that different taxa are responsible for either the upper or lower PCB degradation pathways, even though a partial overlap in the two communities may include also bacteria performing the whole pathway. The dataset generated in this study does not allow for a clear discrimination between the two groups, however the results are consistent with previous DNA-SIP studies, where the predominant taxa detected also herein, such as Comamonadaceae, Rhodocyclaceae, Nocardiaceae and Pseudonocardiaceae, were identified as biphenyl and/or 4-CB degraders in PCB contaminated soils and sediments16,18,29,38,39. The prevalent ASVs belonged to Comamonadaceae and were classified as Hydrogenophaga or Caenimonas in Phal and Bulk soil samples, and as Methylibium in PhalRed ones. In addition to the above-mentioned DNA-SIP works, Hydrogenophaga sp. has also been characterized as a PCB degrader by Lambo and Patel40 whereas Caenimonas and Methylibium were associated with the degradation of aromatic compounds in soil and groundwater41,42,43 and may therefore be cross feeders on PCB catabolic intermediates. Among Rhodocyclaceae, Azoarcus sp., a genus that we found abundant among 13C-labeled ASVs in Bulk soils, was also identified as capable of oxidizing biphenyl in upland soil by a recent study based on a protein-SIP approach30. Considering the family Nocardiaceae, Rhodococcus strains are well known PCB degraders and have been shown to carry out different degradation pathways from biphenyl to benzoate in a metagenomic study conducted on an enrichment culture growing on biphenyl44. Interestingly, in the present study, the only ASV representing this bacterial family was common to three of the considered soils (Bulk, BulkRed and PhalRed) and was classified as Rhodococcus. Three Rhodococcus sp. strains were indeed previously isolated from this soil after three months of rhizoremediation treatment with reed canary grass and a redox cycle (PhalRed), and their capacity to utilize biphenyl as a sole carbon source and to degrade different PCB congeners present in the Delor 103 commercial mixture was demonstrated in vitro through a resting-cell assay45. Therefore, we can hypothesize that this genus was involved in the direct degradation of 4-CB in the present study as well as in the degradation of other PCB congeners in this historically polluted soil.
Although most of the ASVs derived from D28 4-CB amended microcosms were unique to each of the four soil treatments, we could not observe the same pattern comparing their relative abundance in initial T0 soils. In fact, in pre-incubation samples, the same ASVs were mostly below or close to the detection threshold, indicating that the enrichment of the detected degraders occurred over time during the incubation in microcosms supplemented with 4-CB This may imply that, likewise the total bacterial community, the alpha-diversity of active PCB-degraders was higher in the initial soils subjected to the rhizoremediation treatments, with active taxa residing in the rare biosphere. Only ASV6 within Comamonadaceae, the prevalent taxon incorporating carbon from 4-CB in Phal soil at D28, even though representing less than 1% of the bacterial communities before SIP incubation, seemed to be enriched in previously planted treatments compared to bulk soils at T0. This taxon was already observed to be positively selected by plant species within the Poaceae family, both in uncontaminated and polluted soils46,47 and to be associated with an increase in organic pollutant biodegradation48. It is possible that the predominance of ASV6 among degraders in Phal soils was favored by an efficient capacity to utilize the compounds derived from root deposition that may be implied in the co-metabolism of 4-CB or its catabolic intermediates5. Instead, among Comamonadaceae in PhalRed soils we observed the enrichment of ASV4 classified as Methylibium, a facultative aerobe previously described for PAH anaerobic degradation43, that could have been favored in this partially anoxic environment.
In a former study, we showed that rhizoremediation with P. arundinacea alone or in combination with soil flooding resulted in a significant reduction in the concentration of low chlorinated PCBs (≤ pentaCBs) by stimulating microbial activity and inducing a shift in the overall soil bacterial community structure12. In this study, we described at taxonomic level how these treatments shaped the structure of soil bacterial communities and we observed that Actinobacteria were enriched by the plant biostimulation. Using 4-CB as a metabolic tracer for DNA Stable Isotope Probing (SIP), we also provided new insights into the populations metabolically active in PCB degradation, identifying members of Actinobacteria and Betaproteobacteria as the main players involved. Altogether this information pointed out that Actinobacteria and Betaproteobacteria are the key bacterial targets to be enriched in a plant biostimulation-based approach aimed at the on-site bioremediation of the highly and historically PCB-polluted site considered in our study. Future research perspectives include in situ-SIP coupled with metagenomic analysis49 and will allow to deepen the investigation of plant-microbiome interactions and increase the knowledge of the metabolic reactions occurring during the PCB biodegradation in the field.
Data availability
All relevant data are within the paper and its Supporting Information files. The 16S rRNA gene sequence reads were deposited in the NCBI SRA database under the BioProject ID: PRJNA809320 as indicated in the Materials and Methods section.
References
Quinete, N., Schettgen, T., Bertram, J. & Kraus, T. Occurrence and distribution of PCB metabolites in blood and their potential health effects in humans: A review. Environ. Sci. Pollut. Res. 21, 11951–11972 (2014).
Desforges, J.-P. et al. Predicting global killer whale population collapse from PCB pollution. Science 1979(361), 1373–1376 (2018).
Wolf-Reiner, A., Balbina, N., Golyshin Peter, N., Pieper Dietmar, H. & Timmis, K. N. Polychlorinated biphenyl-degrading microbial communities in soils and sediments. Curr. Opin. Microbiol. 5, 246–253 (2002).
Rolli, E. et al. ‘Cry-for-help’ in contaminated soil: A dialogue among plants and soil microbiome to survive in hostile conditions. Environ. Microbiol. 23, 5690–5703 (2021).
Musilova, L., Ridl, J., Polivkova, M., Macek, T. & Uhlik, O. Effects of secondary plant metabolites on microbial populations: Changes in community structure and metabolic activity in contaminated environments. Int. J. Molecul. Sci. 17, 1205 (2016).
Leewis, M. C. et al. Differential impacts of willow and mineral fertilizer on bacterial communities and biodegradation in diesel fuel oil-contaminated soil. Front. Microbiol. 7, 837 (2016).
Lopez-Echartea, E., Macek, T., Demnerova, K. & Uhlik, O. Bacterial biotransformation of pentachlorophenol and micropollutants formed during its production process. Int. J. Environ. Res. Public Health 13, 1146 (2016).
Vergani, L. et al. Phyto-rhizoremediation of polychlorinated biphenyl contaminated soils: An outlook on plant-microbe beneficial interactions. Sci. Total Environ. 575, 1395–1406 (2017).
Chaparro, J. M., Badri, D. V. & Vivanco, J. M. Rhizosphere microbiome assemblage is affected by plant development. ISME J. 8, 790–803 (2014).
Dassen, S. et al. Differential responses of soil bacteria, fungi, archaea and protists to plant species richness and plant functional group identity. Mol. Ecol. 26, 4085–4098 (2017).
Correa-García, S., Rheault, K., Tremblay, J., Séguin, A. & Yergeau, E. Soil characteristics constrain the response of microbial communities and associated hydrocarbon degradation genes during phytoremediation. Appl. Environ. Microbiol. 87, e02170 (2021).
Terzaghi, E. et al. Rhizoremediation of weathered PCBs in a heavily contaminated agricultural soil: Results of a biostimulation trial in semi field conditions. Sci. Total Environ. 686, 484–496 (2019).
Chen, C. et al. Paddy field–a natural sequential anaerobic-aerobic bioreactor for polychlorinated biphenyls transformation. Environ. Pollut. 190, 43–50 (2014).
Cupples, A. M. Contaminant-degrading microorganisms identified using stable isotope probing. Chem. Eng. Technol. 39, 1593–1603 (2016).
Uhlik, O. et al. Stable isotope probing in the metagenomics era: A bridge towards improved bioremediation. Biotechnol. Adv. 31, 154–165 (2013).
Koubek, J., Mackova, M., Macek, T. & Uhlik, O. Diversity of chlorobiphenyl-metabolizing bacteria and their biphenyl dioxygenases in contaminated sediment. Chemosphere 93, 1548–1555 (2013).
di Guardo, A. et al. Differentiating current and past PCB and PCDD/F sources: The role of a large contaminated soil site in an industrialized city area. Environ. Pollut. 223, 367–375 (2017).
Uhlik, O. et al. Biphenyl-metabolizing bacteria in the rhizosphere of horseradish and bulk soil contaminated by polychlorinated biphenyls as revealed by stable isotope probing. Appl. Environ. Microbiol. 75, 6471–6477 (2009).
Wald, J. et al. Pseudomonads rule degradation of polyaromatic hydrocarbons in aerated sediment. Front. Microbiol. 6, 1268 (2015).
Mapelli, F., Riva, V., Vergani, L., Choukrallah, R. & Borin, S. Unveiling the microbiota diversity of the xerophyte Argania spinosa L. skeels root system and residuesphere. Microb. Ecol. 80, 822–836 (2020).
Caporaso, J. G. et al. Ultra-high-throughput microbial community analysis on the Illumina HiSeq and MiSeq platforms. ISME J. 6, 1621–1624 (2012).
Callahan, B. J. et al. DADA2: High-resolution sample inference from Illumina amplicon data. Nat. Methods 13, 581–583 (2016).
Quast, C. et al. The SILVA ribosomal RNA gene database project: Improved data processing and web-based tools. Nucleic Acids Res. 41, D590 (2013).
Anderson M. J., Gorley R. N., Clarke K. R. & Clarke R. K. PERMANOVA C for PRIMER: Guide to software and statistical methods. (2008).
Hammer, Ø., Webb, K. E. & Depreiter, D. Numerical simulation of upwelling currents in pockmarks, and data from the Inner Oslofjord Norway. Geo-Mar. Lett. 29, 269–275 (2009).
R Core Team. R: A language and environment for statistical computing. (2020).
Iffis, B., St-Arnaud, M. & Hijri, M. Petroleum contamination and plant identity influence soil and root microbial communities while AMF spores retrieved from the same plants possess markedly different communities. Front. Plant Sci. 8, 1381 (2017).
Thomas, F., Corre, E. & Cébron, A. Stable isotope probing and metagenomics highlight the effect of plants on uncultured phenanthrene-degrading bacterial consortium in polluted soil. ISME J. 13, 1814–1830 (2019).
Uhlik, O. et al. Plant secondary metabolite-induced shifts in bacterial community structure and degradative ability in contaminated soil. Appl. Microbiol. Biotechnol. 97, 9245–9256 (2013).
Chen, S. C. et al. Novel clades of soil biphenyl degraders revealed by integrating isotope probing, multi-omics, and single-cell analyses. ISME J. 15, 3508–3521 (2021).
Uhlik, O. et al. Identification of bacteria utilizing biphenyl, benzoate, and naphthalene in long-term contaminated soil. PLoS ONE 7, e40653 (2012).
Leewis, M. C., Uhlik, O. & Leigh, M. B. synergistic processing of biphenyl and benzoate: Carbon flow through the bacterial community in polychlorinated-biphenyl-contaminated soil. Sci. Rep. 6, 1 (2016).
Balley Robert, E., Gonslor, S. J. & Rhinehart, W. L. Biodegradation of the monochlorobiphenyls and biphenyl in river water. Environ. Sci. Technol. 17, 617–621 (1983).
Bouwer, E. J. & Zehnder, A. J. B. Bioremediation of organic compounds-putting microbial metabolism to work. Trend Biotechnol. 11, 360 (1993).
Waite, D. W. et al. Proposal to reclassify the proteobacterial classes deltaproteobacteria and oligoflexia, and the phylum thermodesulfobacteria into four phyla reflecting major functional capabilities. Int. J. Syst. Evol. Microbiol. 70, 5972–6016 (2020).
Wang, S. et al. Genomic characterization of three unique dehalococcoides that respire on persistent polychlorinated biphenyls. Proc. Natl. Acad. Sci. USA 111, 12103–12108 (2014).
Praveckova, M., Brennerova, M. V., Holliger, C., de Alencastro, F. & Rossi, P. Indirect evidence link PCB dehalogenation with geobacteraceae in anaerobic sediment-free microcosms. Front. Microbiol. 7, 933 (2016).
Sul, W. J. et al. DNA-stable isotope probing integrated with metagenomics for retrieval of biphenyl dioxygenase genes from polychlorinated biphenyl-contaminated river sediment. Appl. Environ. Microbiol. 75, 5501–5506 (2009).
Leigh, M. B. et al. Biphenyl-utilizing bacteria and their functional genes in a pine root zone contaminated with polychlorinated biphenyls (PCBs). ISME J. 1, 134–148 (2007).
Lambo, A. J. & Patel, T. R. Biodegradation of polychlorinated biphenyls in Aroclor 1232 and production of metabolites from 2,4,4′-trichlorobiphenyl at low temperature by psychrotolerant hydrogenophaga sp. strain IA3-A. J. Appl. Microbiol. 102, 1318–1329 (2007).
Zhang, Y. et al. Straw return promoted the simultaneous elimination of sulfamethoxazole and related antibiotic resistance genes in the paddy soil. Science of the Total Environment 806, 150525 (2022).
Zhang, S., Wang, Q. & Xie, S. Stable isotope probing identifies anthracene degraders under methanogenic conditions. Biodegradation 23, 221–230 (2012).
Kane, S. R. et al. Whole-genome analysis of the methyl tert-butyl ether-degrading beta-proteobacterium Methylibium petroleiphilum PM1. J. Bacteriol. 189, 1931–1945 (2007).
Garrido-Sanz, D., Manzano, J., Martín, M., Redondo-Nieto, M. & Rivilla, R. Metagenomic analysis of a biphenyl-degrading soil bacterial consortium reveals the metabolic roles of specific populations. Front. Microbiol. 9, 232 (2018).
Vergani, L. et al. Novel PCB-degrading rhodococcus strains able to promote plant growth for assisted rhizoremediation of historically polluted soils. PLoS ONE 14, e0221253 (2019).
Bulgarelli, D. et al. Structure and function of the bacterial root microbiota in wild and domesticated barley. Cell Host Microbe 17, 392–403 (2015).
Feng, J. et al. Assembly of root-associated microbiomes of typical rice cultivars in response to lindane pollution. Environ. Int. 131, 104975 (2019).
Li, J. et al. Diversity of the active phenanthrene degraders in PAH-polluted soil is shaped by ryegrass rhizosphere and root exudates. Soil Biol. Biochem. 128, 100–110 (2019).
Li, J. et al. The catabolic pathways of in situ rhizosphere PAH degraders and the main factors driving PAH rhizoremediation in oil-contaminated soil. Environ. Microbiol. 23, 7042–7055 (2021).
Acknowledgements
SB, FM, LV acknowledge the funding agency Ente Regionale per i Servizi all'Agricoltura e alle Foreste (ERSAF), Decreto ERSAF n. III/5426 del 09.12.2013. SB acknowledges the funding support received by the European Union’s Horizon 2020 research and innovation programme under the Marie Sklodowska-Curie grant agreement N° 841317 with the project “SENSE”. MF, JP, and OU acknowledge the support of the INTER-EXCELLENCE program of the Ministry of Education, Youth and Sports of the Czech Republic under grant no. LTAUSA19013.
Author information
Authors and Affiliations
Contributions
L.V., F.M., O.U. and S.B. conceived the experiments; L.V., O.U., M.F. and J.J. conducted the experiments; J.P. performed the bioinformatic analysis; L.V., F.M. and O.U. analyzed the data; L.V., F.M. and S.B. wrote the paper; all authors critically reviewed the manuscript; O.U. and S.B. provided reagents, materials and analysis tools.
Corresponding author
Ethics declarations
Competing interests
The authors declare no competing interests.
Additional information
Publisher's note
Springer Nature remains neutral with regard to jurisdictional claims in published maps and institutional affiliations.
Rights and permissions
Open Access This article is licensed under a Creative Commons Attribution 4.0 International License, which permits use, sharing, adaptation, distribution and reproduction in any medium or format, as long as you give appropriate credit to the original author(s) and the source, provide a link to the Creative Commons licence, and indicate if changes were made. The images or other third party material in this article are included in the article's Creative Commons licence, unless indicated otherwise in a credit line to the material. If material is not included in the article's Creative Commons licence and your intended use is not permitted by statutory regulation or exceeds the permitted use, you will need to obtain permission directly from the copyright holder. To view a copy of this licence, visit http://creativecommons.org/licenses/by/4.0/.
About this article
Cite this article
Vergani, L., Mapelli, F., Folkmanova, M. et al. DNA stable isotope probing on soil treated by plant biostimulation and flooding revealed the bacterial communities involved in PCB degradation. Sci Rep 12, 19232 (2022). https://doi.org/10.1038/s41598-022-23728-2
Received:
Accepted:
Published:
DOI: https://doi.org/10.1038/s41598-022-23728-2
Comments
By submitting a comment you agree to abide by our Terms and Community Guidelines. If you find something abusive or that does not comply with our terms or guidelines please flag it as inappropriate.