1. Introduction
Polycyclic hydrocarbons (PAHs) are a class of the most toxic and persistent chemically stable compounds. Their danger to living organisms, including humans, increases with increased numbers of rings in their molecular structure [
1,
2,
3]. PAHs are widespread in the soils of industrial enterprises, due to emissions, as well as in agricultural soils in connection with irrigation with sewage [
4].
For minimizing the availability or eliminating the PAHs from terrestrial ecosystems, various chemical, physical, and biological approaches have been adopted [
5,
6]. The availability of PAHs for microbial destruction depends on the size of the soil fraction and the types of PAH [
7], as well as the composition of the soil microbial community and the presence of metabolic pathways genes and operons of PAH catabolism [
8].
The biochemical pathways of aerobic degradation of PAHs by bacteria have been described in sufficient detail. In short, the initial stage is the PAHs’ oxidation to dihydrodiol. The dihydrodiols can then be degraded via the ortho- or meta-pathway, yielding metabolites such as protocatechoates and catechols, that further undergo transformation to the tricarboxylic acid cycle’ metabolites [
9]. Therefore, PAH-degrading bacteria may utilize these molecules as their single carbon source, as well as the energy source. Representatives of such genera as Rhodococcus, Pseudomonas, Sphingomonas, and Arthrobacter are the most prevalent natural PAH destructors.
Use of PAHs as sole energy and carbon source has been repeatedly noticed in Gram-positive bacteria. Mycobacterium strains can grow on pyrene and anthracene, while Rhodococcus strains can only grow on anthracene [
10]. Although many Rhodococcus strains may utilize PAHs, particularly naphthalene, the complete biodegradation process has yet to be identified. Some Rhodococcus species use gentisate to transport naphthalene, whereas others use catechol [
11,
12]. In Rhodococcus, naphthalene and salicylate activate the genes involved in naphthalene metabolism [
13].
The breakdown of PAHs is controlled by dioxygenase-encoding genes. The narAa and narAb genes, for example, encode the two subunits of naphthalene-1,2-dioxygenase (NDO), which converts naphthalene to (+)-cis-(1R, 2S)-dihydroxy-1,2-dihydronaphthalene [
13,
14]. The nar genetic systems were also found in other actinobacteria, in particular, in representatives of the genus Gordonia [
15].
In Gram-negative bacteria, particularly Proteobacteria, the organization of naphthalene catabolism genes is well studied. These genes are mainly located on plasmids and are grouped into 2 operons. The upper operon includes the nahA-nahF genes, while the lower one contains the nahG-nahM genes. These operons control the transformation of salicylate from, and salicylate metabolism through, catechol with ring destruction and the incorporation of metabolites into the tricarboxylic acid cycle.
The study of the genetic organization of the PAH destruction metabolic pathways by individual strains makes it possible to further identify these pathways in the general pool of the soil microbial community genes.
Many researchers in their studies on the metabolization of PAHs in soils focused on the degradative abilities of individual isolated strains [
7,
16]. Nevertheless, isolated bacteria can quickly utilize PAHs in model environments, but lose their properties in soil. Isolates are not fully adapted to soil conditions, namely, to the content of carbon, nitrogen, and phosphorus, the types and bioavailability of PAHs, and interactions with native soil bacteria [
17,
18]. Currently, a lot of attention is paid to bacterial communities, since, as a rule, representatives of different taxa are involved in multi-stage degradation of the same pollutants [
19]. Soil microbial communities are complex systems, the representatives of which are in inextricable interaction with each other [
20]. The mechanisms of these interactions cannot be investigated by isolating individual members of the community. To understand the metabolic processes occurring in soils and grounds, the microbiome of such ecosystems should be considered as a single complex of organisms.
Lack of knowledge about the metabolic and adaptive potential of microbial soil communities is a serious fundamental problem of modern systemic ecology. This work aimed to study the PAH destruction genes (abundance and diversity) in pristine and technogenically transformed soils of Southern Russia using shotgun metagenomics methods. Shotgun metagenomics is an untargeted sequencing of all microbial genomes present in a sample [
21,
22]. This makes it possible to study the genetic and functional potential of microbial communities in a sample. Here, a comparison was done of the taxonomic diversity of microbial communities in contaminated and uncontaminated (control) plots, as well as a comparison of the abundance of genes (abundancy) and operons of PAH catabolism in soil samples. During the study, metagenome-assembled genomes were reconstructed and evaluated in terms of bioremediation of polyaromatics-contaminated soils.
3. Results
In the background soil of the monitoring site of protected area A, the total content of 12 priority PAHs was 262 µg kg
−1 (see
Figure 2), which does not exceed the content of these pollutants in the soils of background regions of different regions of the world [
39,
40,
41,
42]. The content of benzo(a)pyrene (BaP) is lower than the 0.02 mg kg
−1, maximum permissible concentration (MPC) (
Figure 3) [
43]. The composition of PAHs is dominated by low-molecular-weight 2- and 3-ring compounds.
In the soil of the monitoring site B, located in the zone of influence of the power plant, 4 annular PAHs prevail, which is typical when exposed to the enterprises of the fuel and energy complex [
44].
For the soils of both monitoring sites, phenanthrene predominates in the composition of 3 ring-shaped PAH compounds. For the soil of the background plot No. 1, its share of the total content reaches 31%. The composition of 4 rings PAHs is represented mainly by fluoranthene and pyrene. Benzo[b]fluoranthene dominates in the composition of 5 ring PAHs (
Figure 4). Thus, the prevailing composition of pollutants, such as PAHs, differs depending on the historical conditions of use of the studied territories, as well as on the genesis of soils.
3.1. Taxonomic and Functional Data Classification
The relative numbers of bacteria at the class level show that in both samples (uncontaminated and contaminated), the Terrabacteria group dominates (50 and 47% of all bacteria, respectively). The phyla Actinobacteria (84 and 82%, respectively) as well as Chloroflexi (5%) and Firmicutes (7 and 8%) are the most well-represented in the Terrabacteria group. Actinobacteria are well-known soil microflora members, that are known for their metabolic potential to survive in the presence of a wide range of contaminants. In an uncontaminated sample, the Actinomycetia class accounts for 73% of all actinobacteria and 31% of all bacteria, while in a contaminated sample, it accounts for 74% of all actinobacteria and 28% of all bacteria. It is represented by the orders Micrococcales, Corynebacteriales, Propionibacteriales, Streptomycetales, Pseudonocardiaceae, Micromonosporaceae (
Figure 5a,b). Firmicutes are represented by the classes Bacilli and Clostridia. The Bacilli class accounts for 50 and 51% of all Firmicutes in uncontaminated and contaminated samples, respectively, while Clostridia accounts for 40 and 39%, respectively.
3.2. Reconstruction of Metagenome-Assembled Genomes
The raw filtered readings were collected in contigs (
Table 3).
Actinobacteria have the highest abundance in the samples. Their share in communities A and B is 50.32% and 48.22%, respectively. The share of proteobacteria is 22.41% and 23.37%, respectively, in A and B (
Figure 6).
With the data obtained, we assume that the microbiome of Haplic Chernozem soils in its existence for a long time reaches a certain homeostasis. The soil rich in macro- and microelements makes it possible to maintain a taxonomically diverse community, which has a high potential for adaptive reactions. Due to these properties of the soil, the ingress of complex pollutants into it (in our case, PAHs) does not suppress the microbiome, but allows it to adapt. Based on the calculated values of the alpha-diversity parameter when comparing Shannon’s diversity test (
Table 4), we assume that in terms of the abundance of different species and their diversity, the communities of the unpolluted and polluted areas are similar.
Even though archaea are an important part of soil microbiomes, their metabolic potential in comparison with bacteria is extremely poorly understood. However, we can already see that in a contaminated sample containing a mixture of heavy PAHs, the archaeal part of the community is not inhibited. Archaea are dominated by 3 types—Euryarchaeota, Thaumarchaeota and Crenarcheota (
Figure 7). The ability to catabolize PAHs in archaea has been repeatedly noted but practically nothing is known about PAH catabolism genes in them [
45,
46].
In the taxonomic classification of reads, we found that among Euryarcheota, the most represented by reads are the Halobacteria and Methanomicrobia (Stenosarchaea group) classes. Among Thaumarchaeota and Crenarcheota, the Nitrososphaeraceae family and the Thermoprotei class (
Figure 8A,B), respectively, are most represented by reads.
3.3. Identification of Aromatic Hydrocarbon-Degrading Coding DNA Sequences
With the help of the KEGG service, a functional annotation of the collected metagenomes has been summarized (
Table 5).
In sample B, 316 coding sequences (CDS) were found, and in sample A, 291 CDS were found, that are involved in the catabolism of xenobiotics, including PAHs (
Figure 9).
The breakdown of aromatic hydrocarbons occurs through a series of reactions including oxidation, hydroxylation, dehydrogenation, and ring breakdown. To study the representation of PAH catabolism genes in the metagenome of soil communities, we selected 179 bacterial genera that can be classified as common in communities, and using the Genbank NCBI database, we compiled a set of sequences of PAH catabolism genes and operons in representatives of these genera. Since many bacteria do not have information on such genes in the Genbank NCBI, the final version of our database includes information on the genes and operons of PAH catabolism in 55 bacterial genera (
Supplementary Table S2).
Complete operons of PAH catabolism are known in such genera as Burkholderia, Pseudomonas, Rhodococcus, Marinobacter, Azoarcus, Gordonia. There is also information on gene clusters of PAH catabolism in Variovorax, Bradyrhizobium, Albidiferax, Sphingobium, Ralstonia. In our samples of Haplic Chernozem (both in unpolluted and contaminated PAHs), readings of the PAH catabolism operons were found in Azoarcus, Bradyrhizobium, Burkholderia, Gordonia, Rhodococcus, Variovorax.
The genetic system of naphthalene catabolism in actinobacteria is quite universal. It is represented by a gene cluster that includes genes encoding rubredoxin (Rub1), regulator proteins NarR1 and NarR2, Naphthalene dioxygenase subunits NarAa and NarAb, Naphthalene dihydrodiol dehydrogenase NarB and Hydratase—aldolase NarC [
15].
In the results of sequencing of soil metagenomes of both contaminated and uncontaminated samples, we found readings corresponding to the genes of the catabolism operons of naphthalene for Gordonia (128 and 192 reads, respectively) and Rhodococcus (195 and 264 reads, respectively).
The ability of representatives of Azoarcus to utilize PAHs is periodically mentioned in the literature [
47,
48]. However, not much is known about the pathway of PAH metabolism in this genus. We analyzed the genomes of known Azoarcus, placed in the Genbank database, and identified 11 genes related to the destruction of naphthalene. Among them are naphthalene dioxygenase in large (nagAa) and small (nagAb) subunits, salicylate-5-hydroxylase (nagG and nagH subunits), nagI dioxygenase gentisate, as well as two regulators-nahR1 and nahR2 and the nahY protein, which is responsible for chemotaxis towards the carbon source. The genes are organized into 2 operons, similar to the structure characteristic of pseudomonads, and the chemotaxis protein is separate.
For Variovorax, we found readings corresponding to two naphthalene dioxygenase subunits and two salicylate hydroxylase subunits. The total length of genes for naphthalene catabolism in Variovorax is about 7000 bp. (
Supplementary Table S2). However, it should be noted that these genes in Variovorax are currently poorly understood.
Normalized gene counts of the main PAH catabolic genes were identified in the samples (
Table 6). It is worth noting that the number of genes is comparable in a contaminated and uncontaminated sample. Therefore, it can be assumed that both the number of these genes and the number of organisms carrying them do not depend on the presence of the pollutant in the soil.
4. Discussion
Basically, the aerobic catabolism of PAHs by bacteria is studied using the example of a model two-ring compound, naphthalene [
49]. However, the process of destruction of PAHs with 3 or more rings has also repeatedly attracted the attention of many scientific groups [
50,
51]. When analyzing soils subjected to technogenic stress (a sample from the NPS), we found that there are only 3 times more 2-ring PAHs in this sample than in the control one. As for PAHs with 3–6 rings, their content in the contaminated sample was at least 10 times (
Figure 3) higher than the amount in the uncontaminated sample. Such compounds, denoted by the general term HMW-PAH (high molecular weight), are extremely resistant to bacterial degradation [
52,
53]. In this regard, we focused on studying the genetic potential of the microbial community, which makes it possible to utilize these compounds.
The ability to biodegrade three-ring PAHs has been found in many microorganisms [
54,
55]. We identified genes of the phenanthrene operon Novosphingobium, as well as genes for catabolism of fluorene and protocatechuic acid, in the assemblies of complete metagenomes. In Sphingomonas, genes for catabolism of fluorene and protocatechuic acid are included in the fld operon. According to Pinyakong et al., Sphingomonas can convert fluorene to fluorenone and then through phthalate and protocatechuic acid. The fld cluster of genes has been described in detail [
56]. The readings corresponding to the representatives of Sphingomonodaceae in our data (uncontaminated and contaminated samples, respectively) constitute 1 and 0.7% of reads of all bacteria, 4 and 6% of all proteobacteria.
It is also interesting that the pathway of fluorene catabolism in rhodococci is assumed to be different [
57]. According to the results of Wang et al., the destruction of fluorene begins with hydroxylation and cleavage of one of the outer rings, similar to the destruction of dibenzothiophene along the Kodama pathway [
58]. However, unlike the Kodama pathway, the putative pathway for the cleavage of fluorene by rhodococci is not a dead end but leads to the formation of salicylate. We did not find information in the literature or in the Genbank database on specific genes that control this pathway. Hence, at the moment, we assume that the degradation of fluorene by rhodococci is controlled by the nar gene system, the main reaction of which is hydroxylation of naphthalene in one of the rings.
Among the genes for anthracene catabolism, we noted in our data the specific gene Anthrone monooxygenase. The transformation of anthracene through anthrone is typical for fungi, especially for Penicillium simplicissimum. According to Jove et al., P. simplicissimum oxidizes anthracene to anthrone, which is then transformed into anthraquinone and phthalate [
59]. We found an anthrone monooxygenase belonging to Streptomyces. This gene is included in the anthracene catabolic operon, which is specific for streptomycetes.
Addition of rings and increase in molecular wight does not imply the need to use a different type of reaction for their destruction, in addition to those listed above. HMW-PAHs are cleaved by bacteria through sequential hydroxylation and further ring opening, as shown, for example, for bacilli [
60]. These reactions are carried out by the same enzyme systems as the catabolism of the lighter PAHs. We found an extensive pool of PAH catabolism genes in the microbial communities of Haplic Chernozem. Genes have different taxonomic affiliations. We got similar results for contaminated and uncontaminated samples. This allows us to assume the following:
Biodegradation of polyaromatic hydrocarbons, as a rule, does not imply highly specific reactions requiring specialized enzymatic systems or unique taxa. As described in detail above, PAHs undergo sequential multistage conversion by oxygenation, hydroxylation, modification of functional groups and cycles to facilitate steric access to them by enzymes, as well as ring opening. Moreover, in the overwhelming majority of cases, enzymes of different microorganisms are simultaneously involved in this process, and the microorganisms themselves are taxonomically quite distant from each other (for example, streptomycetes and pseudomonads). This is especially typical for HMW-PAH due to physicochemical reasons: poor-water solubility and high recalcitrance due to sorption on soil matter lead to very low rates of mobilization and transport of these molecules or their derivatives into the cells of microorganisms, rates that are not compatible with the metabolic needs of individual cells [
61]. For this reason, we must talk about the degradation of complex PAHs by the enzymatic system of a consortium of microorganisms, and not by its individual representatives [
19].
This state of enzymatic systems for the biodegradation of PAHs as a not-one-genome phenomenon, but dependent on the existence at the level of the metagenome of the entire microbial community, inevitably leads to low specificity, but at the same time multifunctionality of enzymes that catabolize the transformation of PAHs. Thus, many oxygenases have a very wide range of substrates and low specificity [
62,
63]. For example, oxygenases of the P450 family isolated from fungi have been shown to oxidize both PAHs and lignin, while naphthalene dioxeganases have been shown to be capable of enzymatic conversion of humic substances [
64]. Several laccases have been shown to oxidize PAHs, which is true for both fungi and bacteria [
65]. All this indicates the ability to transform and metabolize a wide range of complex organic substrates, especially those with a similar structure, which is especially important in the context of the Haplic Chernozem soils in this study.
Haplic Chernozems are characterized by a high content of total and dissolved organic matter, which includes supramolecular organic complexes of varying complexity—humin, humic and fulvic acids, as well as their polymer precursors at different stages of transformation—degradation and condensation products of cellulose, chitin, lignin and proteins. It seems highly probable that at least some of the enzymes involved in PAH catabolism can be used to convert other (poly) aromatic compounds in the soil organic matter of Haplic Chernozem.
This is not the only reason why the found soils exhibit a similar abundance and diversity of PAH destruction genes. Although PAHs present in ecosystems are mainly of petrogenic and pyrogenic origin, polyaromatic hydrocarbons entering soils can also be of non-anthropogenic, namely of biogenic and/or diagenetic nature [
66,
67]. Non-biogenic diagenetic processes that introduce PAHs into soils are fires, volcanic activity, kerogen diagenesis, and erosion of ancient sedimentary rocks [
68]. As for biogenic PAHs, perylene and retene are typical examples [
69,
70]; naphthalene, phenanthrene, and chrysene derivatives are less frequently mentioned [
70,
71,
72]. Such PAHs are most often formed extracellularly by diagenesis of various biogenic molecules, such as plant waxes, terpenoids, and hopanoids from prokaryotic membranes [
67,
70,
73]. The object without anthropogenic load in this study was represented by the soils of the protected area, where agricultural activities have never been carried out. Since the sorption of PAHs by soil matter increases with the time of the presence of PAHs in the soil, we can assume the accumulation of biogenic polycyclic hydrocarbons in these soils, and hence the selection pressure in favor of the preservation of the potential for PAH destruction by the microbial community.
Finally, a significant part of the soil community capable of PAH degradation does not have to be metabolically active and be constantly presented to PAHs. High concentrations of PAHs in soils are still oligotrophic concentrations, and in addition to PAHs, microorganisms capable of utilizing them have to be able to use other carbon sources. This also implies the previously discussed low specificity and multifunctionality of enzymatic systems for PAH degradation. At the same time, some, mostly oligotrophic, microorganisms may exist most of the time in a dormant state as dormant cells or viable, but non-culturable cells, or in a quasi-stationary state, when the population maintains a relatively constant number of cells [
61]. Thus, the processes and causes listed above, most likely, provide a similar representation and diversity of PAH destruction genes in Haplic Chernozem with radically different levels of anthropogenic load.