Soil Microbiome Response to Contamination with Bisphenol A, Bisphenol F and Bisphenol S

The choice of the study objective was affected by numerous controversies and concerns around bisphenol F (BPF) and bisphenol S (BPS)—analogues of bisphenol A (BPA). The study focused on the determination and comparison of the scale of the BPA, BPF, and BPS impact on the soil microbiome and its enzymatic activity. The following parameters were determined in soil uncontaminated and contaminated with BPA, BPF, and BPS: the count of eleven groups of microorganisms, colony development (CD) index, microorganism ecophysiological diversity (EP) index, genetic diversity of bacteria and activity of dehydrogenases (Deh), urease (Ure), catalase (Cat), acid phosphatase (Pac), alkaline phosphatase (Pal), arylsulphatase (Aryl) and β-glucosidase (Glu). Bisphenols A, S and F significantly disrupted the soil homeostasis. BPF is regarded as the most toxic, followed by BPS and BPA. BPF and BPS reduced the abundance of Proteobacteria and Acidobacteria and increased that of Actinobacteria. Unique types of bacteria were identified as well as the characteristics of each bisphenol: Lysobacter, Steroidobacter, Variovorax, Mycoplana, for BPA, Caldilinea, Arthrobacter, Cellulosimicrobium and Promicromonospora for BPF and Dactylosporangium Geodermatophilus, Sphingopyxis for BPS. Considering the strength of a negative impact of bisphenols on the soil biochemical activity, they can be arranged as follows: BPS > BPF > BPA. Urease and arylsulphatase proved to be the most susceptible and dehydrogenases the least susceptible to bisphenols pressure, regardless of the study duration.


Introduction
Twentieth-century chemical synthesis created conditions for the production of plasticisers, and in consequence, of their components [1]. The most hazardous of these include dioxins [2] and acrylonitrile [3], as well as bisphenol A (BPA) [4]. Currently, alternative substances are being sought to alleviate concerns associated with bisphenol. Sixteen BPA analogues have been used in the chemical industry with bisphenol S (BPS) and bisphenol F (BPF) being its chief substitutes [1,5]. According to Molina-Molina et al. [6], the toxicity of the bisphenols decreases with the increasing polarity of their molecules. Bisphenol F (BPF) is obtained in the process of phenol hydroxyalkylation and formaldehyde catalysis. The catalysis runs with a Brönsted's ionic liquid [7] or aluminosilicate catalysts MCM-41 [8]. Bisphenol F (BPF) differs from BPA by the absence of two methyl groups bound to the central carbon atom of its molecule that results in its lower polarity. On the other hand, because of the presence of two phenolic groups on each side of the sulphonic group, BPS has a similar chemical structure to BPA, but higher thermal stability [9,10]. Bisphenol A (BPA) is formed by condensation of an acetone molecule and two molecules of phenol, catalysed by hydrogen chloride or an ion-exchange resin [11].
Except for BPF, bisphenols are covered by the Community Rolling Action Plan (CoRAP) [12]. According to the data from the European Chemicals Agency (ECHA 2020), the annual production output microorganism growth. Bisphenol stimulated an increase in the count of all groups of microorganisms except for the cellulolytic bacteria and Azotobacter sp. on day 30 of the experiment ( Table 2).
The opposite trends were observed on day 15. The microorganism response to soil contamination with BPF was much more varied. Regardless of the time of exposure to BPF, the organotrophic and cellulolytic bacteria count was decreased by this bisphenol. Moreover, the nitrogen immobilising bacteria and Azotobacter sp. count was found to decrease on day 15 of the experiment and that of copiotrophic, ammonification bacteria, and Arthrobacter sp. on day 30 of the experiment. The application of BPF to soil had the most beneficial effect on fungi, Pseudomonas sp. and oligotrophic bacteria. On the other hand, BPS had a stronger inhibitory impact on the soil microbiome on day 15 of the experiment. It reduced the count of Azotobacter sp., cellulolytic, nitrogen immobilising, organotrophic and oligotrophic bacteria by 77.18%; 64.77%; 62.34%; 47.16% and 8.34%, respectively. After 30 days of the soil incubation, stimulation of the count of microorganisms exposed to BPS was observed, considering an increasing count of all the microorganism groups under study, except Arthrobacter sp., cellulolytic and copiotrophic bacteria.
on the count of the cellulolytic bacteria (67.02%) and Pseudomonas sp. (51.07%). Bisphenol A (BPA) modified the soil microorganism growth. Bisphenol stimulated an increase in the count of all groups of microorganisms except for the cellulolytic bacteria and Azotobacter sp. on day 30 of the experiment ( Table 2). The opposite trends were observed on day 15. The microorganism response to soil contamination with BPF was much more varied. Regardless of the time of exposure to BPF, the organotrophic and cellulolytic bacteria count was decreased by this bisphenol. Moreover, the nitrogen immobilising bacteria and Azotobacter sp. count was found to decrease on day 15 of the experiment and that of copiotrophic, ammonification bacteria, and Arthrobacter sp. on day 30 of the experiment. The application of BPF to soil had the most beneficial effect on fungi, Pseudomonas sp. and oligotrophic bacteria. On the other hand, BPS had a stronger inhibitory impact on the soil microbiome on day 15 of the experiment. It reduced the count of Azotobacter sp., cellulolytic, nitrogen immobilising, organotrophic and oligotrophic bacteria by 77.18%; 64.77%; 62.34%; 47.16% and 8.34%, respectively. After 30 days of the soil incubation, stimulation of the count of microorganisms exposed to BPS was observed, considering an increasing count of all the microorganism groups under study, except Arthrobacter sp., cellulolytic and copiotrophic bacteria.
The bisphenol impact factor (IFBP) confirmed the sensitivity of organotrophic and cellulolytic bacteria to pressure from BPA, BPF and BPS. It also provides grounds for the claim that BPF is the most toxic phenolic compound, followed by BPS, and the impact of BPA on the microorganism count is the weakest (Figure 2, Figure 3). 62 Figure 1. The share of independent variables in the evolution of the microorganisms activity (η 2 ): t-time, BP-kind of bisphenol; time*kind of bisphenol; Org-organotrophic bacteria, Act-Actinomycetes, F-mould fungi, Ps-Pseudomonas sp., Art-Arthrobacter sp., Cel-cellulolytic bacteria, Olig-oligotrophic bacteria, Cop-copiotrophic bacteria, Im-nitrogen immobilizing bacteria, Am-ammonification bacteria, Az-Azotobacter sp. (two-way analysis of variance, ANOVA, at p < 0.05).   Figure 2. Coefficient of impact (IFBP) of bisphenols for group of microorganisms in soil contaminated with BPA, BPF and BPS on the 15th day of research, A-soil contaminated with BPA; F-soil contaminated with BPF; S-soil contaminated with BPS; Homogeneous groups denoted with letters (a-c) were calculated separately for each group of microorganisms (for abbreviations see Figure 1).   Valuable information on the scale of disruption of the microbiological balance in soil exposed to bisphenol contamination and the complexity of the processes that initiate them is provided by the colony development index (CD) ( Figure 4) and the ecophysiological diversity index (EP) ( Figure 5). These indices were correlated with the phenolic compound type and the soil incubation time. The differences in the impact of individual bisphenols were demonstrated by multidimensional PCA analysis ( Figure 4). The first principal component (PCA1) explaining 67.81% of the total data variance generated a positive value of the primary variable vector only for fungi (0.887). The case dislocation emphasised the fungi colony development stimulation by BPA, BPF and BPS. The tendency was further made more precise by the PCA analysis. The highest CD for organotrophic bacteria suggests that this group is dominated by fast replicating microorganisms. Nevertheless, the application of bisphenols to soil was not as effective in accelerating the development of organotrophic bacteria as it was for fungi, either on day 15 or day 30 of the experiment. The PCA also revealed that Actinobacteria are slowly growing organisms, regardless of whether or not they were exposed to bisphenols. However, unlike for fungi, the ecophysiological diversity index (EP) for this group of microorganisms reached the highest values ( Figure 5).

Org
The position of all cases on the PCA plot shows that all the bisphenols reduced the ecophysiological diversity of organotrophic bacteria and fungi on day 15 of the experiment and the impact of BPS was the greatest. The negative impact of this bisphenol persisted until day 30, while the inhibitory strength of the other bisphenols diminished. The relationships revealed made more credible the corresponding standardised vectors of primary variables with positive values relative to the first principal component (PCA1) for all the microorganism groups and negative for Act (−0.974) relative to the second principal component (PCA2) describing 33.37% of the variance of variables.
The greatest number of OTU among all the identified phyla was observed in Actinobacteria and Proteobacteria ( Figure 6). Actinobacteria accounted for 37.5% of all bacteria in soil uncontaminated with bisphenols and 39.0% in soil contaminated with BPA, 46.4%-with BPF and 52.7%-with BPS. Proteobacteria accounted for 32.6% (control), 32.2% (BPA), 23.9% (BPF) and 26.2% (BPS). Bisphenol F (BPF) and bisphenol S (BPS) increased the abundance of Actinobacteria by 8.9% and 15.2%, respectively, compared to the control, whereas they reduced the abundance of Proteobacteria (BPS by 8.7% and BPF by 6.4%) and Acidobacteria (BPS by 4.6% and BPF by 5.5%). Conversely, BPA did not have a significant impact on these types. Only BPF had a positive impact on TM7.
emphasised the fungi colony development stimulation by BPA, BPF and BPS. The tendency was further made more precise by the PCA analysis. The highest CD for organotrophic bacteria suggests that this group is dominated by fast replicating microorganisms. Nevertheless, the application of bisphenols to soil was not as effective in accelerating the development of organotrophic bacteria as it was for fungi, either on day 15 or day 30 of the experiment. The PCA also revealed that Actinobacteria are slowly growing organisms, regardless of whether or not they were exposed to bisphenols. However, unlike for fungi, the ecophysiological diversity index (EP) for this group of microorganisms reached the highest values ( Figure 5).     According to the number of OTU at the class level, the most representative were Actinobacteria and Thermoleophilia of phylum Actinobacteria, and Alphaproteobacteria of phylum Proteobacteria ( Figure  7a). The highest (and the most diverse) number of OTUs for individual objects was identified within the Actinobacteria class: control-16,511 OTU, BPA-22,715 OTU, BPF-24,481 OTU, BPS-26,760 OTU. Interestingly, BPF generated a large number of OTU in classes of Saprospirae and TM7-3, and BPA: Phycisphaerae, iii1-8, Chloracidobacteria. It also induced an increase in this parameter for Gammaproteobacteria, whereas BPF and BPS reduced it. The impact of bisphenols on bacteria classes also affected the lower taxon, i.e., order ( Figure 7b). Actinomycetales of Phylum Acidobacteria was the dominant order in this study. Its abundance was increased all the bisphenols. Subsequent orders in terms of their abundance in this Phylum: Gaiellales, iii1-15, Solirubrobacterales and Acidimicrobiales. Furthermore, Phylum Proteobacteria were dominated by orders Rhizobiales and Rhodospirillales. All the bisphenols reduced the OTU number for bacteria of the orders Rhodospirillales, and the OTU number in the order Rhizobiales-only BPF and BPS. According to the number of OTU at the class level, the most representative were Actinobacteria and Thermoleophilia of phylum Actinobacteria, and Alphaproteobacteria of phylum Proteobacteria ( Figure 7a). The highest (and the most diverse) number of OTUs for individual objects was identified within the Actinobacteria class: control-16,511 OTU, BPA-22,715 OTU, BPF-24,481 OTU, BPS-26,760 OTU. Interestingly, BPF generated a large number of OTU in classes of Saprospirae and TM7-3, and BPA: Phycisphaerae, iii1-8, Chloracidobacteria. It also induced an increase in this parameter for Gammaproteobacteria, whereas BPF and BPS reduced it. The impact of bisphenols on bacteria classes also affected the lower taxon, i.e., order ( Figure 7b). Actinomycetales of Phylum Acidobacteria was the dominant order in this study. Its abundance was increased all the bisphenols. Subsequent orders in terms of their abundance in this Phylum: Gaiellales, iii1-15, Solirubrobacterales and Acidimicrobiales. Furthermore, Phylum Proteobacteria were dominated by orders Rhizobiales and Rhodospirillales. All the bisphenols reduced the OTU number for bacteria of the orders Rhodospirillales, and the OTU number in the order Rhizobiales-only BPF and BPS. The lower taxon was dominated by Gaiellaceae, Nocardioidaceae and Geodermatophilaceae of phylum Actinobacteria and Hyphomicrobiaceae and Xanthomonadaceae of phylum Proteobacteria ( Figure  8). All bisphenols reduced the abundance of OTUs of the following bacteria: Gaiellaceae, Rhodospirillaceae, Bacillaceae, Syntrophobacteraceae, Pseudomonadaceae, Polyangiaceae, Rhodobacteraceae, and increased that of Nocardioidaceae, Geodermatophilaceae, Intrasporangiaceae, Sphingomonadaceae, Microbacteriaceae and Micrococcaceae. The microbiological balance on the family level was caused to a lesser extent by BPA than by BPS and BPF.  The Venn analysis helped to identify unique bacteria types characteristic of individual study objects ( Figure 9). In the control soil, those included: Hyphomicrobium, DA101, Nitrospira, Nannocystis, Methylibium, in that contaminated with BPA: Lysobacter, Steroidobacter, Variovorax, Mycoplana, in that contaminated with BPF: Caldilinea, Arthrobacter, Cellulosimicrobium and Promicromonospora and in that contaminated with BPS: Dactylosporangium, Geodermatophilus and Sphingopyxis. Twelve genera common to all the objects were identified along with the native genera. It is notable that all the genera in the core microbiome in the soil contaminated with BPA were of Phylum Proteobacteria, whereas in the soil contaminated with BPF 3 the genera belong to Actinobacteria and 1 to Chloroflexi, and in soil contaminated with BPS-2 genera belong to Actinobacteria and 1-to Proteobacteria. The Venn analysis helped to identify unique bacteria types characteristic of individual study objects ( Figure 9). In the control soil, those included: Hyphomicrobium, DA101, Nitrospira, Nannocystis, Methylibium, in that contaminated with BPA: Lysobacter, Steroidobacter, Variovorax, Mycoplana, in that contaminated with BPF: Caldilinea, Arthrobacter, Cellulosimicrobium and Promicromonospora and in that contaminated with BPS: Dactylosporangium, Geodermatophilus and Sphingopyxis. Twelve genera common to all the objects were identified along with the native genera. It is notable that all the genera in the core microbiome in the soil contaminated with BPA were of Phylum Proteobacteria, whereas in the soil contaminated with BPF 3 the genera belong to Actinobacteria and 1 to Chloroflexi, and in soil contaminated with BPS-2 genera belong to Actinobacteria and 1-to Proteobacteria.

Enzyme Activity
The study results showed that all three bisphenols, bisphenol A, bisphenol F and bisphenol S, modified the soil biochemical activity significantly, although to a different extent. This is evidenced by the η 2 values for each enzyme, which can be put in the following order: Deh > Ure > Aryl > Pac > Glu > Pal > Cat ( Figure 10).

Enzyme Activity
The study results showed that all three bisphenols, bisphenol A, bisphenol F and bisphenol S, modified the soil biochemical activity significantly, although to a different extent. This is evidenced by the η 2 values for each enzyme, which can be put in the following order: Deh > Ure > Aryl > Pac > Glu > Pal > Cat ( Figure 10).

Enzyme Activity
The study results showed that all three bisphenols, bisphenol A, bisphenol F and bisphenol S, modified the soil biochemical activity significantly, although to a different extent. This is evidenced by the η 2 values for each enzyme, which can be put in the following order: Deh > Ure > Aryl > Pac > Glu > Pal > Cat ( Figure 10).  Time affected the activity of Pal (15.62%), Glu (12.19%) and Aryl (11.30%) to the greatest extent. Contrary to expectations, BPA was found to stimulate the activity of Deh, both on day 15 and on day 30 of the experiment ( Table 3). The enzyme activity increased 2 and 3 fold compared to the control objects. Bisphenol A (BPA) also stimulated the activity of Pal and, to a lesser extent Pac, regardless of the experiment duration, which also corresponds to the homogeneous groups identified in the study. Bisphenol A (BPA) had an inhibitory effect on the activity of Cat, Aryl, Ure and Glu on day 15, and on the activity of Aryl and Ure on day 30. The application of 100 mg of BPF kg −1 d.m. of soil significantly disrupted its homeostasis on day 15 of the experiment. Of the enzymes under analysis, Ure and Glu proved to be the most sensitive to the presence of BPF in soil, which is indicated by the 29% and 18% inhibition of these enzymes activity caused by 100 mg of BPF kg −1 d.m. of soil. Similarly, urease proved to be the most sensitive to BPF exposure. This bisphenol also proved to have a beneficial effect on Deh and Pal, whose activity increased 3-and nearly 2-fold compared to the control samples. This relationship also corresponded to the parallel objects on day 15 of the experiment. BPS proved to be the most controversial substitute of BPA. The study results revealed that it inhibited the activity of Pac, Aryl, Glu and Ure while at the same time stimulating the activity of Deh both on day 15 and day 30 of the soil incubation, and Pal on day 30. It is notable that the strength of BPS inhibition was much higher than that of BPA and BPF, which is shown by the homogeneous groups identified in the analysis. The impact factors for each bisphenol (IF BP ) confirmed the sensitivity of individual enzymes to the applied phenolic compounds (Figure 11, Figure 12). Ure, Aryl and Glu proved to be the most sensitive to pressure from BPA, and Deh the least so. This index emphasised the differences in bisphenol toxicity. Considering the strength of their negative impact on the soil biochemical activity, the phenolic compounds used in this study can be arranged as follows: BPS > BPF > BPA.  Figure 12. Coefficient of impact (IFBP) of bisphenols for enzymes activity in soil contaminated with BPA, BPF and BPS in 30th day of research, Homogeneous groups denoted with letters (a-c) were calculated separately for each enzyme, (for abbreviations see Figure 11).  Figure 12. Coefficient of impact (IFBP) of bisphenols for enzymes activity in soil contaminated with BPA, BPF and BPS in 30th day of research, Homogeneous groups denoted with letters (a-c) were calculated separately for each enzyme, (for abbreviations see Figure 11).  Figure 11).

Counts and Diversity of Bacteria
The response of microorganisms to the pressure of the phenolic compounds is associated with the activation of a wide range of mechanisms induced by soil microbiome. In the research, only BPA did not stimulate the multiplication of Pseudomonas sp. on the 15th day of the study. Consequently, on the 30th day the number of Pseudomonas sp. was more than twice as high. BPF and PBS enhanced the multiplication of Pseudomonas sp. both on the 15th day respectively by: 71.08% and 24.46% and the 30 th day of analysis respectively by: more than twice as high and 86.58%. A similarly large number of mechanisms and, most of all, an impressive pool of enzymes which catalyse bisphenol degradation, were observed in fungi, which is a probable cause of the positive response of these microorganisms to the application of BPA, BPF and BPS to the soil. The count of fungi was particularly stimulating due to BPS. After the application of bisphenol, the number of this group of microorganisms increased as much as twice on the 15th day and three times on the 30th day of analyses. One could expect the Pseudomonas sp. count to be stimulated by all the bisphenols. According to the findings of Chuanchuen et al. [78], Pseudomonas sp. respond to the substance pressure in soil by activating Resistance-Nodulation-Division (RND) efflux pumps which remove phenols from the cell. They also have a gene (HHDH) encoding halohydrin dehalogenase, which opens the epoxide ring [79]. Moreover, a number of metabolic pathways were identified, which are induced by Pseudomonas sp. and which effectively degrade phenolic compounds, including bisphenols [53,54].
Fungi are a reservoir of triphenylmethane reductase, which breaks down phenolic compounds [80], lignin peroxidase, which catalyses their single-electron oxidation [81], polyketide synthase (PKS) and cytochrome P 450, which catalyses monomeric dimerization of phenolic compounds [82]. Reduction of the Azotobacter sp. count in soil contaminated with BPA and BPF was not surprising in this study. This is because Azotobacter sp. is regarded as a reliable and sensitive indicator of soil contamination with xenobiotics [83,84], although G-negative bacteria, including Azotobacter sp., Arthrobacter sp. and Pseudomonas sp. of phylum Proteobacteria, are less susceptible to phenolic compound pressure [85]. These tendencies may be a consequence of a higher isoelectric point (pH = 4-5) for Gram-negative bacteria [86]. More controversy was raised by inhibition of cellulolytic bacteria replication by all the bisphenols under study: BPA, BPF and BPS. However, it turns out that 40% of the sequenced bacteria genomes encode one gene of cellulase, but only 4% are described as proper cellulolytic bacteria [87]. One of the reasons for the negative response of cellulolytic bacteria can be a disruption of the newly discovered strategy of the functional dependence of beneficiaries on helpers, which generates commensalistic and even mutualistic interactions, that provide an opportunity for tangible activity in this group of microorganisms [88]. Also noteworthy are the results of studies in which a beneficial effect of 100 mg BPA on the biomass of microorganisms in wheat-seeded soil was obtained [89]. Nevertheless, the plant is a deposit of microorganisms with phenol degrading genes located in plasmids. In turn, the horizontal transfer of these genetic conditions associated with the plant is favourable in relation to the degradation of volatile contaminants [90].
The CD values for Actinobacteria are nearly twice lower than for organotrophic bacteria and the high values of the EP index, ranging from 0.0584 to 0.799 show that although Actinobacteria replicate more slowly than organotrophic bacteria [91], they pretend being regarded as microorganisms which effectively biodegrade organic contaminations [92]. This is associated with a large pool of enzymes: proteases, cellulases, amylase, lectinase, catalase, chitinase and urease, which catalyse the process of degradation of complex polymers [93], and with the participation of Actinobacteria, in solubilisation of phosphates, siderophores production and nitrogen fixation [92]. The observed succession of the microorganism groups under analysis corresponds to the reports of Sarathchandra et al. [94], according to which fast replicating microorganisms generate higher values of CD regardless of the ecosystem stability and they are called r-strategists, as opposed to slowly replicating or dormant microorganisms, k-strategists. The effect of bisphenols on CD for organotrophic bacteria may be explained by the fact that such microorganisms, such as Bacillus, are capable of breaking down BPA. However, bisphenol significantly changes the membrane permeability, sporulation, amino acid and protein expression and metabolism of carbon, purines, pyrimidines and fatty acids (PLFA) [95]. Bisphenol toxicity is associated with indefinite toxicity related to the hydrophobicity of a single compound and the formation of free radicals. Phenols damage the endoplasmic reticulum and mitochondrion. Exposure of microorganisms to bisphenols leads to changes in the properties of the cell membrane, which results in inhibition of respiration and growth of microorganisms and cell lysis [96,97] Determination of microorganism biodiversity, which is the main factor affecting the soil function has attracted much interest among researchers, which is proven by over 900 papers on the soil metagenome being published before 2017 [98]. The best predictors of soil phylotype abundance include soil pH, climatic factors (temperature, precipitation seasonality) and plant productivity. The most commonly occurring microorganisms include Alphaproteobacteria, Betaproteobacteria, Actinobacteria, Acidobacteria and Planctomycestes [57]. This study showed that the dominant microorganisms in uncontaminated soil include the phyla: Actinobacteria, Proteobacteria, Acidobacteria, Chloroflexi, Firmicutes, Planctomyces. According to Bakker et al. [99], these are the most common phyla in arable lands along with Bacteroidetes. Salam and Varma [100] determined the effect of e-waste as a source of phenolic compounds and observed a change in soil bacterial composition towards phylum Actinobacteria which overtook Proteobacteria and Firmicutes. The application to the soil of BPA, BPF and BPS induced corresponding changes in the soil microbiome, significantly enriched by TM7 in objects with BPF. This was also correlated with the findings described by Siczek et al. [101]. In objects contaminated with a phenolic compound, the authors observed a high abundance of Actinobacteria and Proteobacteria, with reduced phylum Chloroflexi. Similarly, Hassen et al. [102] revealed the presence of α, β and γ Proteobacteria and Firmicutes in soil contaminated with phenolic compounds. Four bacteria types were identified in the author's research in soil exposed to all the bisphenols: Nocardioides, Agromyces, Spinghomonas and Devosia. These findings are consistent with those of other researchers. Siczek et al. [101] identified all of these types in soil contaminated with a phenolic compound, except Spinghomonas. Tian et al. [103] proposed two phenol degrading strains: Sphingomonas sp. PH20 and Sphingomonas sp. 31853. It is notable that Sphingomonas sp. MV1 was one of the first discovered strains using BPA as a source of carbon and energy, which it obtained in the process of hydroxylation of the phenolic ring and meta-cleavage [104]. Its capability for biodegradation has been confirmed in scientific research many times [101,105]. Hassen et al. [102] and Siczek et al. [101] found bacteria of the genus Bacillus, which were also identified in soil contaminated with bisphenols in this study, to be active in decomposing phenolic compounds. According to Li et al. [94], Bacillus bacteria are capable of degrading up to 85% of BPA within 24 h. Similarly, Li et al. [106] found the strain Bacillus sp. GZB to decompose BPA to p-hydroquinone (HQ) and 4-(2-propanol)-phenol.

Soil Enzyme
Soil enzymes are natural factors which accelerate several soil processes, which gave them the status of indicators of early changes of soil degradation and intensity of biological processes closely connected with the physicochemical properties of the soil environment [46,47]. The high activity of dehydrogenases was expected in line with scientific reports, regardless of the type of bisphenol applied to soil. These expectations stemmed from the participation of dehydrogenases in the conversion of ethylbenzene to 1-phenylethanol and dehydrogenation of phenolic compounds in the presence of dehydrogenases as biocatalysts [107,108]. Bilal et al. [109] report that oxidoreductases, effective in bisphenol biodegradation, are active in a wide range of pH and temperature values. They found a three times higher Deh activity in soil contaminated with BPA and BPF compared to the control sites, which does not correspond to the findings of studies by Zaborowska et al. [110] and Zaborowska et al. [111], in which the correlations were different. Deh proved to be particularly sensitive to BPF, after the application of as little as 5 mg of bisphenol kg −1 d.m. of soil. BPS had a similar toxic impact on them. This may be attributed to the fact that stimulation of dehydrogenases activity may also result from the accumulation by microorganisms of lactate dehydrogenase dependent on NAD, which is induced by the demand for a higher energy level following the oxidative stress caused by the pressure of phenolic compounds [112].
A controversy is also raised by a response of acid phosphatase to soil contamination with bisphenols, mainly inhibition of its activity by BPS, whereas the presence of hydroxyl groups in the phenolic ring implies phosphatase adsorption on soil colloids and activates these enzymes. However, given the fact that the process of adsorption may have changed the structure of the function groups, resulting in inhibiting the Pac activity as a consequence of collocation between the enzyme and the substrate, this relationship could be expected [113]. Such concerns are also dissipated by the findings described by Tang et al. [114] who suggest that quinone in soil is formed by phosphorylation of disodium phosphate and the formation of phenol, catalysed by phosphatases. Quinones, in turn, are regarded as important inhibitors of enzymatic activity. Together with the SH group in cysteinyl, transformed to SS bonds, they reduce the urease activity [115]. The toxicity of quinones is associated with arylation of the thiol group in Cys. 1,4 and covalent modification. Currently, the group of urease inhibitors has been expanded to include hydroquinone, derivatives of coumarin, phenolic aldehydes and catechol [116]. Catechol inactivates the urease metallocentre by the formation of catechol-metal complexes. It is also oxidised to form orto-benzoquinone, which is a strong urease inhibitor as a result of the modification of the functional groups in the enzyme protein [117,118]. The effectiveness of urease activity inhibition is also correlated with the presence and position of substituents in phenolic compounds [119]. They are probably caused more by methoxyl and hydroxyl groups in phenolic organic compounds, as opposed to nitro groups which exhibit low-protein inhibition of urease activity [120,121]. Due to their structure with two ortho hydroxyl groups, isoflavone polyphenols become less toxic to urease only after the c-isoflavone ring is broken [122]. It should be emphasised that binding substituents with single pairs of electrons to the phenyl ring around the thiourea core and the presence of fluorine atom in phenyl groups, regardless of its position, also results in strong urease inhibition [123,124]. The findings of Zaborowska et al. [110,111] and Siczek et al. [101] corroborate the experiment results. Both BPS and BPF proved to be significant inhibitors of both Pac and Ure. The findings related to Glu activity were more debatable. BPF and BPS had a negative effect on the Glu activity on day 30 of the study. This is probably a consequence of low kinetic parameters, such as maximum reaction rate (V max ) and Michaelis-Menten constant (K m ), which generate Glu adsorption at the site of binding with hydroxyl phenol [112]. This thesis can be partly corroborated by the findings of Zaborowska et al. [111], who found the response of Glu to BPF was similar, while the application of BPS did not generate such low enzyme sensitivity to the phenolic compound. A phenolic compound was also a significant Glu inhibitor according to Siczek et al. [101].
The most important factors shaping the microbiological response to bisphenols are: genes encoding enzymes responsible for catabolism of bisphenols located in bacterial chromosomes or degrading plasmids, functional viability of enzymes different in organic and mineral soils, desorption and adsorption of phenols in soil affecting their mobility and availability as well as organic matter and pH. In addition, the intensity of phenolic compound inhibition is significantly affected by the presence and position of selected substituents It is also important to be aware of the diverse degradative activity of bacteria, and in particular the activation of various metabolic pathways under both aerobic and anaerobic conditions.
The changes in the reaction of microorganisms and the activity of individual soil enzymes, that were observed in this study, were a response to a biotic stress caused by soil contamination with bisphenols. The diverse reaction of microorganisms and enzymes to BPA, BPF and BPS that was observed in this study is likely due to the fact that enzymes can be both extracellular and intracellularly metabolized by microbes. In the soil environment the enzymes react with minerals, are broken down by proteolytic enzymes or undergo thermal denaturation [125]. It should be emphasized that phenolic compounds are a growth medium for microorganisms, a source of carbon and energy. However, the condition for using these substrates is the presence of hydroxyl groups equivalent to the presence of molecular oxygen as a co-substrate [126].

Soil and Facilities
The experiment was conducted with proper brown soil classified as loamy sand, which had a granulometric composition that was determined following the international soil classification for soil nomenclature and creating legends for soil maps [125,127]. The soil was collected from the genetic horizon Ap, at the Teaching and Experiment Centre in Tomaszkowo, developed agriculturally for cereal cultivation, situated in the Olsztyn Lakeland. This area is dominated by Eutric Cambisol soils formed on sand and loam. The Olsztyn Lakeland (Pojezierze Olsztyńskie) is the largest mesoregion in the Mazurian Lake District in the East European Lowland in the temperate warm transitional climate zone with an average annual temperature of 7.2 • C. A full characterisation of the soil material, with its granulometric composition and selected physicochemical, biochemical and microbiological properties is provided in Table 1, which was determined using the methodology described by Borowik et al. [126,128].

Experiment Setup
The choice of the key experiment stages was based on the fact that there is scarce data on the effect of BPA, BPF and BPS, on both the biochemical and microbiological activity of soil. Therefore, variable factors which could potentially modify the expected analysis results were eliminated and the experiment was conducted in a laboratory, under strictly controlled ex situ conditions. To determine ultimately which of the bisphenols disrupts the soil homeostasis to the greatest extent and mainly to emphasise the difference in toxicity of the proposed phenolic compounds, one level of soil contamination with bisphenols was applied-100 mg BP kg −1 d.m. of soil. The assays were performed against control objects, uncontaminated with bisphenols. Since bisphenols are poorly soluble in water, adding them to air-dry soil was preceded by dissolving each bisphenol in ethanol at the ratio of 3:1 (ethanol:bisphenol). The experiment was conducted in 150 cm 3 glass beakers in three replicates. The soil samples (100 g d.m.) were prepared separately for the two experiment runs. After being thoroughly homogenised and their moisture content stabilised at 50% of the capillary water capacity, the soil samples were incubated at the constant temperature of 25 • C for 15 and 30 days. The soil moisture content was monitored throughout the experiment.

Soil and Facilities
The experiment was conducted with proper brown soil classified as loamy sand, which had a granulometric composition that was determined following the international soil classification for soil nomenclature and creating legends for soil maps [125,127]. The soil was collected from the genetic horizon Ap, at the Teaching and Experiment Centre in Tomaszkowo, developed agriculturally for cereal cultivation, situated in the Olsztyn Lakeland. This area is dominated by Eutric Cambisol soils formed on sand and loam. The Olsztyn Lakeland (Pojezierze Olsztyńskie) is the largest mesoregion in the Mazurian Lake District in the East European Lowland in the temperate warm transitional climate zone with an average annual temperature of 7.2 °C. A full characterisation of the soil material, with its granulometric composition and selected physicochemical, biochemical and microbiological properties is provided in Table 1, which was determined using the methodology described by Borowik et al. [126,128].

Experiment Setup
The choice of the key experiment stages was based on the fact that there is scarce data on the effect of BPA, BPF and BPS, on both the biochemical and microbiological activity of soil. Therefore, variable factors which could potentially modify the expected analysis results were eliminated and the experiment was conducted in a laboratory, under strictly controlled ex situ conditions. To determine ultimately which of the bisphenols disrupts the soil homeostasis to the greatest extent and mainly to emphasise the difference in toxicity of the proposed phenolic compounds, one level of soil contamination with bisphenols was applied-100 mg BP kg −1 d.m. of soil. The assays were performed against control objects, uncontaminated with bisphenols. Since bisphenols are poorly soluble in water, adding them to air-dry soil was preceded by dissolving each bisphenol in ethanol at the ratio of 3:1 (ethanol:bisphenol). The experiment was conducted in 150 cm 3 glass beakers in three replicates. The soil samples (100 g d.m.) were prepared separately for the two experiment runs. After being thoroughly homogenised and their moisture content stabilised at 50% of the capillary water capacity, the soil samples were incubated at the constant temperature of 25 o C for 15 and 30 days. The soil moisture content was monitored throughout the experiment.

Characteristics of Bisphenols
The research dealt with three bisphenols (Figure 13). Referring to the safety data sheet issued by Sigma-Aldrich, (Poznań, Poland) bisphenol A (BPA) (synonyms: 4,4′-isopropylidenediphenol; 2,2bis(4-hydroxyphenyl)-propane), bisphenol F (BPF) (synonyms: 4,4′-methylenediphenol, bis(4hydroxyphenyl) methane and bisphenol S (BPS) (synonyms: 4,4′-sulfonyldiphenol, 4-hydroxyphenyl sulfone, bis(4-hydroxyphenyl) sulfone) were used. All the bisphenols were white crystalline substances with the purity of ≥98.0% (HPLC -highperformance liquid chromatography). Selected physical and chemical properties of the bisphenols are shown in Table 4. The level of soil contamination with BPA, BPF and BPS was determined based on two guidelines. First, legal guidelines, according to which the permissible bisphenol concentration at the depth of 0-0.3 m was 0.1 mg kg −1 in soil which is classified as agricultural land, excluding built up and urban areas [129]. The second guideline was established based on the need to determine the All the bisphenols were white crystalline substances with the purity of ≥98.0% (HPLC -high-performance liquid chromatography). Selected physical and chemical properties of the bisphenols are shown in Table 4. The level of soil contamination with BPA, BPF and BPS was determined based on two guidelines. First, legal guidelines, according to which the permissible bisphenol concentration at the depth of 0-0.3 m was 0.1 mg kg −1 in soil which is classified as agricultural land, excluding built up and urban areas [129]. The second guideline was established based on the need to determine the level of potentially negative bisphenol impact (lower than that analysed in the experiments conducted so far) common to the selected phenolic compounds [110,111].

Sample Analysis
On days 15 and 30 of the experiment, the organotrophic bacteria (Org), Actinobacteria (Act) and fungi (F) count was determined with the serial dilution method, in three replications. The composition of the microbiological media was as follows: organotrophic bacteria (Bunt and Rovira medium), Actinomycetes (Parkinson medium) and fungi (Martin medium). Microbial counts were performed according the media and procedure described by Borowik et al. [128]. Colony-forming units (cfu) were counted on 10 consecutive days with a colony counter. The results provided grounds for determining the colony development index (CD) [94] and the ecophysiological diversity index (EP) [131] from the following formulas: where N 1 , N 2 , N 3 ,...N 10 -sum of ratios of the colony numbers identified on each day (1, 2, 3,...10) and the sum of all the colonies identified during the entire experiment and: where pi denotes the number of microbe colonies replicated on a specific day, divided by the number of all the colonies.
To consider the bisphenol effect on the soil microbiological activity in a broader perspective, the count of eight microorganism groups was also determined. The composition of the microbiological media was as follows: oligotrophic bacteria (Olig) and copiotrophic bacteria (Cop) (Ohta and Hattori medium), cellulolytic bacteria (Cel), nitrogen immobilising bacteria (Im) and ammonification bacteria (Am) (Winogradski medium), Pseudomonas sp. (Ps) and Arthrobacter sp. (Art) (Mulder and Antheumisse medium), Azotobacter sp. (Az) (Fenglerowa medium). The determination was performed by serial dilutions, in three replicates. The microorganisms were cultured on Petri dishes on microbiological medium whose composition and preparation procedure were described by Borowik et al. [128]. All the microorganism groups were incubated at a constant temperature of 28 • C.

DNA Extraction and Bioinformatic Analysis of Bacteria Taxons
Isolation of genomic DNA from a soil sample was based on a modified method utilising the "Genomic Mini AX Bacteria +" kit (A&A Biotechnology). Determination of bacterial DNA was preceded by mechanical lysis of the samples with zirconium balls in a FastPrep-24 device and additional purification in an Anti-Inhibitor Kit. The presence of bacterial DNA in the soil samples was determined by the. fluorometric method on a Qubit 4 Fluorometer server. Real-time PCR was performed with SYBR Green dye. Preliminary denaturation was carried out for 3 min at 95 • C and denaturation was carried out for 15 s at 95 • C. Universal starters: 1055F (5 -ATGGCTGTCGTCAGCT-3 ) and 1392R (5 -ACGGGCGGTGTGTAC-3 ) which amplify a fragment of the bacterial gene 16S RNA were bound during 30 s at 58 • C. The PCR product was extended for 30 s at 72 • C. The melting curve of the PCR product was determined by measurement fluorescence in the temperature range 65 • C->95 • C. Sequencing of the gene encoding amplicon 16S sequences was performed based on the complementarity determining region V3-V4. The bioinformatic analysis was performed on the MiSeq sequencer of the QIIME software package based on the GreenGenes v13_8 reference sequence database. Sequencing performed by Genomed SA (Warsaw, Poland) enabled the reading classification down to the genus level. The sequencing readings on the amplicon of gene 16S RNA were assigned with OTU in accordance with the taxon in the manner described by Ravel et al. [132].

Statistical Data Analysis and Methodology of Calculations
The experiment results were configured based on the statistical analyses performed in the Statistica 13.1 package [133]. The η 2 coefficient of percentile variability of the variable under study was determined by an ANOVA analysis of variance. A multidimensional analysis, PCA, was applied to determine the impact of BPA, BPF and BPS on the colony development index (CD) and the ecophysiological diversity index (EP). Tukey's test at p = 0.01 was used to determine homogeneous variances between soil enzymes and microorganism groups. The differences in their responses to soil contamination with BPA, BPF and BPS were emphasised by showing the fluctuating tendencies with the bisphenol impact factor (IF BP ), calculated from the formula: Where IF BP -the factor of the impact of increasing bisphenol (BP) soil contamination levels, (IF BP < 1-inhibition of the enzyme activity and groups of microorganisms by BPA, BPF and BPS, IF BP > 1-stimulation of the soil enzyme activity and groups of microorganisms by BPA, BPF and BPS; A BP -enzyme activity and groups of microorganisms in the soil subjected to the increasing BPA, BPF and BPS contamination pressure; A C -activity of the enzyme and groups of microorganisms in the control soil non-contaminated with BPA, BPF and BPS. Visualisation of the genome data only with the sequences exceeding 1% was performed with the following statistical analyses: gplots library [134], a bilateral test of statistical hypotheses-G-test (w/Yates') + Fisher's, with the interval confidence method Asymptotic with CC [135] configured in the STAMP 2.1.3. software, the thermal map based on the RStudio v1.2.5033 software [136], system R v3.6.2 [137]. The data were presented in a circular arrangement with Circos 0.68 software (Canada's Michael Smith Genome Sciences Center, Vancouver, British Columbia V5Z 4S6, Canada).

Conclusions
Bisphenols A, F and S interfered significantly in the soil microbiome and disrupt its homeostasis. Bisphenol F (BPF) was regarded as the most toxic, followed by BPS and BPA. Cellulolytic bacteria