3.1. Phylogeny and Overall Genome Characteristics
The assembly consisted of 140 contigs with a total sequence length of 5,733,046 bp, an N50 value of 126,441 bp, a GC content of 68%, and coverage of 53.0×. A total of 5331 CDSs, 5209 CDSs with proteins, and 67 RNAs were found in the
R.
rhodochrous IEGM 1362 genome (
Table 1).
Using phenotypic methods (morphology, physiological tests, chemotaxonomy), the IEGM 1362 strain was identified as belonging to the
R. rhodochrous species. During the analysis of the genome and phylogenetic markers, such as the 16S rRNA gene, it was revealed that
R. rhodochrous IEGM 1362 was not a common member of this species. The scores for the key taxonomic markers (16S rRNA gene similarities, dDDH, and ANI) of
R. rhodochrous IEGM 1362 were close to threshold levels of 98.7%, 70%, and 94–95%, respectively [
26]. According to the gene relatedness of 16S rRNA, the
R. rhodochrous strain IEGM 1362 was in the same phylogenetically close group with the
R. coprophilus strain NCTC10994, the
R. gordoniae strain DND3, and various
R. pyridinivorans and
R. rhodochrous strains (
Figure 1). Only the percent of identity with the 16S rRNA gene of
R. coprophilus NCTC10994 was below the threshold of <98.7% [
26], while the levels of identity with 16S rRNA genes of other strains were ≥98.7% (
Figure 1). This indicated that
R. rhodochrous IEGM 1362 could be classified as
R. gordoniae,
R. pyridinivorans, or
R. rhodochrous.
The dDDH scores were more than 70% in comparisons of the IEGM 1362 genome with genomes of
R. rhodochrous type strains and less than 70% in comparisons with genomes of
R. gordoniae and
R. pyridinivorans type strains (
Table 2). This parameter was evidence for the IEGM 1362 strain belonging to
R. rhodochrous, which agrees with the phenotypic traits of this strain. However, ANI scores were above the threshold of 94–95% for two species,
R. pyridinivorans and
R. rhodochrous, and G+C content was less different (0.05–0.25%) between
R. rhodochrous IEGM 1362 and type strains of
R. gordoniae and
R. pyridinivorans than that between the IEGM 1362 strain and type strains of
R. rhodochrous (0.38–0.45%) (
Table 2).
According to the ANI and dDDH values, the IEGM 1362 strain did not belong to R. gordoniae. However, its difference from R. pyridinivorans was not evident. Summarizing all the scores (the 16S rRNA gene relatedness, the highest ANI score, and dDDH values within the species borders), the IEGM 1362 strain continued to belong to R. rhodochrous. Apparently, R. rhodochrous IEGM 1362 accumulated genetic traits of both closely related species R. pyridinivorans and R. rhodochrous and occupied an intermediate position between these two species. The heterogeneity of the R. rhodochrous IEGM 1362 genome, distancing this strain from common representatives of R. rhodochrous, could be a basis for its advanced catalytic properties, including its ability to transform (–)-isopulegol.
3.2. The Whole Genome Analysis
The category distribution of genes with known functions is shown in
Figure 2. In the genome of IEGM 1362, 75 CDSs encoding monooxygenases/hydroxylases, 31 CDSs encoding dioxygenases, 7 CDSs encoding peroxidases, and 351 CDSs encoding dehydrogenases were found. Among monooxygenases, the genes of two alkane 1-monooxygenases, nine cytochrome P450s, eight cyclohexanone 1,2-monooxygenases, three 3-ketosteroid-9-alpha-monooxygenases, and two steroid C27-monooxygenases were revealed. Additionally, we found several genes encoding enzymes of steroid metabolism, such as cholesterol oxidases (six CDSs), 3-oxosteroid 1-dehydrogenases (four CDSs), 3-alpha-hydroxysteroid dehydrogenase (one CDSs), steroid monooxygenases (three CDSs), and 3-ketosteroid-9-alpha-monooxygenases (three CDSs).
Based on the results of the analysis of metabolic pathways in the genome of strain IEGM 1362, a non-mevalonate pathway for C5 isoprenoid biosynthesis was detected, including the following enzymes: 1-deoxy-D-xylulose-5-phosphate synthase (EC 2.2.1.7), 1-deoxy-D-xylulose-5-phosphate reductoisomerase (EC 1.1.1.267), 2-C-methyl-D-erythritol-4-phosphatecytidyl transferase (EC 2.7.7.60), 4-diphosphocytidyl-2-C-methyl-D-erythritol kinase (EC 2.7.1.148), 2-C-methyl-D-erythritol 2,4-cyclodiphosphate synthase (EC 4.6.1.12), 1-hydroxy-2-methyl-2-(E)-butenyl-4-diphosphate synthase (EC 1.17.7.1), and 4-hydroxy-3-methylbut-2-enyldiphosphate reductase (EC 1.17.1.2) (
Supplementary Materials, Figure S1) [
28]. This pathway provides the formation of isopentenyl diphosphate and dimethylallyl diphosphate, which are necessary for the synthesis of terpenes, sterols, carotenoids, and dolichols.
It should be noted that, at present, the vast majority of processes for the biosynthesis of terpene compounds in actinomycetes have been discovered and studied mainly for representatives of the genus
Streptomyces [
29], whereas for representatives of the genus
Rhodococcus, only a few examples of the biosynthesis of zeatins, isoprenoid cytokinins with a dimethylallyl moiety attached to the N atom of adenine or adenosine [
30,
31], and valuable carotenoids such as dihydroxyneurosporine, hydroxyequinenone [
32], beta-carotene, zeaxanthin, and isorenieratin [
33], etc. [
34], have been detected. Since the strain we used in our work is characterized by the red color of the colonies, the discovered pathway is most likely responsible for the synthesis of terpene carotenoids. In addition, using the online genomic mining service AntiSMASH, the presence of three biosynthetic gene clusters for the synthesis of terpenes in the genome, as well as biosynthetic gene clusters for the synthesis of polyketides, non-ribosomal peptide synthases, ectoines, butyro-, and beta-lactones and ε-poly-L-lysine, was shown (
Table 3).
3.3. Determination of Putative Monoterpenoid Transformation Genes
To date, only a few genes and the enzymes they encode have been identified that catalyze the conversion of individual monoterpenoids [
16,
17,
18,
19,
20]. Using limonene as an example, the enzymes limonene-1,2-epoxide hydrolase, limonene-1,2-monooxygenase, and limonene-6-monooxygenase, which are responsible for its transformation through hydroxylation, oxidation, and hydrolysis in
R.
erythropolis cells, were identified [
17]. A gene encoding limonene-1,2-epoxide hydrolase was also found in the genome of IEGM 1362 (fig|6666666.993513.peg.1867), which indicates that this strain may exhibit limonene-transforming activity.
Based on the recorded transformations of (–)-isopulegol, including the stages of hydroxylation and oxidation (
Figure 3), and the analysis of the literature data [
35], we assumed that enzymes of the cytochrome P450 family can be involved in the bioconversion process. Automatic annotation of the obtained sequences of
R.
rhodochrous IEGM 1362 using the RAST service made it possible to search for genes encoding enzymes presumably involved in the bacterial oxidation of (–)-isopulegol. We discovered nine genes encoding enzymes of the family of cytochrome P450-dependent oxygenases and hydroxylases (
Table 4). Pairwise comparison of these genes within strain in the BLAST service showed the absence of significant matches in Megablast mode and no more than 33.5% similarity accordingly to the blastp (protein–protein BLAST) algorithm (
Supplementary Materials, Figure S2). This confirms the fact that all of the discovered genes are not copies, but separate functional units. It is worth noting that, based on the results of a search for genes encoding CYP450 in the NCBI database, 13 genes were identified, but the transcription products of 4 of them in the RAST system were identified more specifically as steroid C27-monooxygenases (EC 1.14.13.141) and lanosterol 14-alpha demethylase (EC 1.14.13.70); therefore, they were not used by us in further experiments.
Based on the detected sequences, for the first time, pairs of primers for individual CYP450 genes of
R.
rhodochrous IEGM 1362 were selected using the Primer-BLAST service. Optimal conditions for their amplification were selected. These guaranteed the formation of products with an expected size without the amplification of extra DNA fragments of other sizes typical for GC-rich matrices (
Table 4,
Figure 4).
Genes encoding CYP450 enzymes in
R. rhodochrous IEGM 1362 differed from their homologues in other
R. rhodochrous strains; this could be a basis for the advanced transformation capabilities of this strain. Percents of identity between the CYP450 nucleotide and amino acid sequences varied in wide ranges of 88.72–100.00% and 56.36–100.00%, respectively (
Table 5). The highest (100.00%) and the lowest (88.72% and 56.36%) similarities were not typical and were detected in a comparison of
R. rhodochrous IEGM 1362 with a few
R. rhodochrous strains. Some of
R. rhodochrous IEGM 1362 cytochrome genes were rare and unique for bacteria. Gene nos. 4, 6, and 7 (see
Table 4) were found only in 13, 59, and 46 bacterial strains. Among
R. rhodochrous strains, gene no. 6 was unique for the IEGM 1362 strain, since it was not found in other representatives of
R. rhodochrous. However, according to amino acid sequencing, an enzyme with 94.55% identity with CYP450 no. 6 of IEGM 1362 was detected in one
R. rhodochrous strain (
Table 5). This enzyme was also found in
R. pyridinivorans with high identities of 99.85–99.92% and 99.77% for DNA and amino acid sequences, respectively (
Table 5). The high similarity of
R. rhodochrous IEGM 1362 cytochromes with those in
R. pyridinivorans was also evident from the close identities of these enzymes, which varied in ranges of 82.05–100.00% and 56.11–100.00% for DNA and amino acid sequences, respectively (
Table 5). Moreover, cytochromes of
R. rhodochrous IEGM 1362 were more frequently detected in
R. pyridinivorans. Homologues of these enzymes and their genes were found in 1–27 strains of
R. pyridinivorans compared with 0–12 strains of
R. rhodochrous (
Table 5). This was not related to the over-representation of sequenced genomes for one of the species in the NCBI. There were 31 strains with sequenced genomes for both species.
The genetic surroundings of CYP450 genes were analyzed. The genes that coded for transcriptional regulators were found directly before or in close proximity to eight CYP450 genes. Only gene no. 6 had no transcription regulators nearby. However, partner proteins for cytochromes, such as ferredoxin and ferredoxin reductase, were detected only in the surrounding of CYP450 no. 6. We supposed that genes encoding CYP450 no. 6, ferredoxin, and ferredoxin reductase consisted of one gene cluster expressed as a polycistron transcript, and that these proteins were responsible for a distinct metabolic process, probably (–)-isopulegol transformation. Other cytochromes in
R. rhodochrous IEGM 1362 could be involved in complex metabolic processes requiring global regulation along with other genes. Such processes could be the biosynthesis of lipids, fatty acids, or acylated polymers (glycolipids or lipoproteins), since genes coding for corresponding enzymes were found near CYP450 genes. Other metabolic processes could include the neutralization of toxic organic compounds, since genes coding for various transporting proteins and other oxidoreductases were also found near CYP450 genes. Gene no. 8 apparently was involved in the metabolism of aromatic compounds, since genes encoding benzoate transporters, benzoate, and catechol dioxygenases, and some specific dehydrogenases, were annotated in the same gene cluster (
Table 6). A common characteristic of cytochromes is their wide substrate specificity. Consequently, any of the CYP450 enzymes found in the genome of
R. rhodochrous IEGM 1362 can theoretically enable the transformation of (–)-isopulegol, including providing co-oxidation conditions.
Additionally, a perspective of horizontal gene transfer (HGT) of genes coding for cytochrome P450 in
R. rhodochrous IEGM 1362 was estimated. Mobile elements were only found near three CYP450 genes. Just one mobile element protein was detected in each case (
Table 6). Apparently, HGT plays a minor role in the distribution of cytochrome-coding genes between
Rhodococcus species; however, it is probably responsible for the appearance of the unique CYP450 no. 6 in the genome of
R. rhodochrous IEGM 1362.
The data obtained expand the understanding of the molecular genetic basis of the transformation of terpenoids by actinomycete cells of the genus Rhodococcus and create the prerequisites for further analysis of gene expression levels in order to identify genes and enzymes for the highly efficient selective biotransformation of the plant monoterpenoid (–)-isopulegol.