Strain YC-RL4^T was isolated from a petroleum-contaminated soil sample collected from Heze City (35°16′ N, 115°28′ E), Shandong Province, China, as described by Ren et al. (2016). Strain MB418^T was isolated from a gravel soil sample collected from the Gurbantunggut desert (45°22′ N, 88°20′ E) in Xinjiang. Strain HX176^T was isolated from rhizosphere soil associated with a medicinal plant in Xinjiang (43°36′ N, 82°11′ E). About 2 g of each sample was suspended in 18 mL of 0.85% (w/v) NaCl solution. Then, 200 μL of 10⁻⁴ diluted soil suspensions were spread on humic acid agar medium, as previously described (Deng et al., 2022). After incubation for 2 weeks at 28°C, visible colonies were picked and streaked on peptone yeast glucose (PYG) medium containing peptone (3 g L⁻¹), yeast extract (5 g L⁻¹), glycerol (10 g L⁻¹), betaine hydrochloride (1.25 g L⁻¹), sodium pyruvate (1.25 g L⁻¹), and agar (15 g L⁻¹), with adjustment to pH 7.2. The cultures were again incubated at 28°C to obtain isolated colonies. Purified isolates were maintained in glycerol suspensions (20%, v/v) at −80°C.

The reference strain M. fluoranthenivorans JCM 14741^T was obtained from the Japan Collection of Microorganisms (JCM), while strains M. frederiksbergense DSM 44346^T and M. litorale DSM 45785^T were acquired from the German Collection of Microorganisms and Cell Cultures (DSMZ).

Strain growth was evaluated on glucose–yeast extract–malt extract agar (GYM; DSMZ medium 65), Middlebrook 7H10 agar (MB7H10; Difco), nutrient agar (NA; Difco), and tryptic soy agar (TSA; Difco), with incubation at 28°C for 48–72 h. Cell motility was examined by microscopic observations and inoculation on semisolid GYM medium with 0.3% agar (w/v). The growth conditions of strains YC-RL4^T, MB418^T, and HX176^T were evaluated over temperature, pH, and salt tolerance ranges. The temperature range of growth was determined in MB7H10 mediu (Difco) at 4, 15, 20, 22, 25, 28, 30, 32, 35, 37, 45, and 50°C. The growth range across pH was determined in GYM medium at pH 4.0–10.0, at intervals of 1.0 unit. NaCl tolerance was determined with cultivation at 28°C in modified GYM medium (pH 7.0, without Na⁺ and Cl⁻), and with NaCl supplemented at concentrations of 0–10.0% (w/v) at increasing increments of 1.0%. All growth condition assays were performed in triplicate.

Biochemical and physiological tests were conducted for strains YC-RL4^T, MB418^T, HX176^T and other closely related strains. Oxidase activity was investigated using the API oxidase reagent (bioMérieux) according to the manufacturer’s instructions. Catalase activity was evaluated via production of bubbles after addition of a drop of 3% (v/v) hydrogen peroxide. Voges–Proskauer, H₂S production, and starch hydrolysis tests were evaluated using a biochemical identification kit (Huankai Microbial) according to the manufacturers’ instructions. Additional biochemical characteristics including enzymatic activities, acid production from fermentation, and assimilation of carbon sources were evaluated using API ZYM, API 50CH, and Biolog GEN III MicroPlates (Biolog) according to the manufacturers’ instructions. API strip and Biolog results were recorded every 24 h after incubation at 28°C until all reactions had stabilized. The API and Biolog tests were performed in duplicate using consistent conditions.

Biomass for chemotaxonomic investigations of strains was obtained by cultivation in flasks on a rotary shaker (180 r.p.m.) using ISP2 broth as the medium and with incubation at 28°C for 3 days, except that cellular fatty acids and cell-wall mycolic acids extraction and analyses were conducted using cultures grown on Middlebrook 7H10 (MB7H10) medium. Cellular polar lipids were extracted, detected with two-dimensional TLC silica-gel 60 F₂₅₄ thin-layer plates (10 × 10 cm, Merck), and analyzed, as previously described (Minnikin et al., 1984). Menaquinones were extracted and purified using reverse-phase HPLC, also as previously described (Minnikin et al., 1984). Cellular fatty acids were extracted, methylated, and analyzed using the Sherlock Microbial Identification System (MIDI) with the standard ACTIN1 database (version 6.0) according to the manufacturer’s instructions (Sasser, 2006). Diagnostic isomers of diaminopimelic acid in whole cell hydrolysates (6 N HCl, 120°C, 30 min) of the strains were subjected to thin-layer chromatography on cellulose plates (10 × 20 cm, Merck) using a previously described solvent system (Schleifer and Kandler, 1972). The cell-wall mycolic acids were determined following saponification, extraction and derivatization, and then separated using a gradient of methaand 2-propanol via high-performance liquid chromatography (HPLC) as recommended by the Sherlock Mycobacteria Identification System Operating Manual Version 1.0 (MIDI) (Butler and Guthertz, 2001; Zimenkov et al., 2015).

16S rRNA genes of strains were amplified using PCR and the universal bacterial primers 27F (5′-AGAGTTTGATCCTGGCTCAG-3′) and 1492R (5′-GGTTACCTTGTTACGACTT-3′). Purified PCR products were cloned into the vector pMD19-T (TaKaRa) and recombinant plasmids were transformed into Escherichia coli DH5α cells, followed by sequencing in Sangon Biotech (Shanghai, China). The 16S rRNA gene sequences of the isolates were compared with publicly available sequences in the EzBioCloud platform² to determine the approximate taxonomic affiliations of the strains (Yoon et al., 2017a). Multiple sequence alignments and phylogenetic reconstructions of 16S rRNA genes were performed in MEGA (version 11) (Tamura et al., 2021). A phylogenetic tree was then inferred using neighbor-joining methods, and evolutionary distances were calculated using the Kimura 2-parameter substitution model (Shamsuzzaman et al., 2023). Maximum Parsimony and Maximum Likelihood phylogenetic methods were also used to evaluate the phylogenetic affiliations of the strains. The topologies of the resultant phylogenetic trees were evaluated using bootstrap analysis with 1,000 replicates (Deng et al., 2023).

Genome sequencing was conducted on an Illumina HiSeq 4,000 system platform at the BGI sequencing company (Shenzhen, China). To prepare sequencing libraries, genomic DNA was randomly sheared to construct three read libraries with lengths of 300 bp using a Bioruptor ultrasonicator (Diagenode, Denville, NJ, United States) and physico-chemical methods. The paired-end fragment libraries were then sequenced on the Illumina platform. Low quality reads (those with consecutive bases covered by fewer than five reads) were discarded. The sequenced reads were then assembled using the SOAPdenovo (version 1.05) assembly software program (Li et al., 2010). Estimated completeness and contamination values for the genomes were estimated using the CheckM pipeline (Parks et al., 2015). Digital DNA–DNA hybridization (dDDH) and average nucleotide identity (ANI) values between the strains and related strains were calculated using the Genome-to-Genome Distance Calculator (GGDC, version 3.0)³ (Auch et al., 2010) and with the ezbiocloud platform (Yoon et al., 2017b), respectively. Genome-based phylogeny of supermatrix approach from protein sequences of the bac120 gene set (a collection of 120 single-copy protein sequences prevalent in bacteria) was constructed by using EasyCGTree version 3.0⁴ as described previously (Zhang et al., 2020). Evolutionary distances were calculated using the IQ-Tree software program (version 1.6.1) (Nguyen et al., 2015). The genome sequences of the strains of interest were downloaded from the NCBI genome database.⁵

Protein sequences encoded by the strains were predicted and annotated using the NCBI Prokaryotic Genome Annotation Pipeline (PGAP). Gene prediction was conducted for the genome assemblies using glimmer3⁶ with Hidden Markov models. tRNA, rRNA, and sRNA identification was conducted with tRNAscan-SE (Chan et al., 2021), RNAmmer (Lagesen et al., 2007), and the Rfam database (Gardner et al., 2009), respectively. Tandem repeat annotation was conducted using the Tandem Repeat Finder⁷, with minisatellite and microsatellite DNAs identified based on the numbers and lengths of repeat units. Functional annotation was also conducted based on the best hits within BLAST analyses. In addition, general functional annotations were identified by comparisons against several databases included the Kyoto Encyclopedia of Genes and Genomes (KEGG), Clusters of Orthologous Groups (COG), Non-Redundant Protein (NR), Swiss-Prot, Gene Ontology (GO), TrEMBL, and EggNOG databases. Predictions of gene clusters involved in natural product formation were conducted using antiSMASH (Blin et al., 2019).

The bacterial pan-genome analysis (BPGA) pipeline (version 1.3) was used to assess the genomic diversity among Mycobacterium strains using default settings (Chaudhari et al., 2016). A total of 73 protein datasets were used for the pan-genome analysis, including for strains YC-RL4^T, MB418^T, HX176^T, and 70 related mycobacterial strains that were previously classified as “Mycolicibacterium.” Orthologous gene/protein clusters (homologous families) were identified using the USEARCH clustering tool (Kim et al., 2014).

Cells of strains YC-RL4^T, MB418^T, and 205372^T were Gram-positive, non-motile and rod-shaped. The three isolates grew well on GYM and MB7H10 agar, with weak growth on NA and TSA agar. Colonies of strains YC-RL4^T and HX176^T grown on GYM medium and incubated for 48 h were 1.0–2.5 mm in diameter, opaque, ivory, and convex with a smooth surface. Colonies of strain HX176^T grown on GYM medium incubated for 48 h were 1.0–3.2 mm in diameter, dry, rough, and orange-colored with undulated/scalloped edges. All strains grew over the pH range of 6.0–7.0 (optimum: 7.0). Strain YC-RL4^T grew in the presence of 0–5.0% (w/v) NaCl (optimum: 0–3.0%, w/v), and 10–37°C (optimum: 28°C), while strain MB418^T grew at 0–5.0% (w/v) NaCl (optimum 0–3.0%, w/v), and 15–45°C (optimum: 28°C), and strain HX176^T grew in the presence of 0–4.0% NaCl (w/v) (optimum: 0–2.0%), and at 15–28°C (optimum: 28°C). All strains exhibited positive catalase reactions, but negative results for oxidase, starch hydrolysis, Voges–Proskauer, and H₂S production tests. All strains exhibited positive enzymatic activities for alkaline phosphatase, esterase (C4), leucine arylamidase, lipase (C14), naphthol-AS-BI-phosphohydrolase, and valine arylamidase, based on API ZYM strip tests. None of the strains could produce acid from fermentation of amidon, amygdalin, arbutin, D-arabinose, D-ardonitol, D-cellobiose, D-fucose, D-galactose, D-lactose, D-maltose, D-mannitol, D-melezitose, D-melibiose, D-raffinose, D-sorbitol, D-trehalose, D-turanose, dulcitol, erythritol, gentiobiose, glycogen, inulin, L-arabitol, L-fucose, L-rhamnose, L-sorbose, L-xylose, mehyl-α-D-glucopyranoside, mehyl-α-D-mannopyranoside, mehyl-β-D-xylopyranoside, N-acetylglucosamine, potassium 5-ketogluconate, salicin, or xylitol. The major differentiating features among strains YC-RL4^T, MB418^T, HX176^T, and other species are shown in Table 1.

Strains YC-RL4^T, MB418^T, and HX176^T exhibited chemotaxonomic properties consistent with their classification as Mycobacterium. The strains produced whole organism hydrolysates rich in meso-diaminopimelic acid (meso-A₂pm), arabinose, galactose, glucose, mannose, and ribose. MK-9 (H₂) was the predominant menaquinone and the polar lipid profile contained diposphatidylglycerol (DPG), phosphatidylethanolamine (PE), and phosphatidylinositol (PI) (Supplementary Figure S1). The three strains nevertheless exhibited differences in several chemotaxonomic characteristics relative to other Mycobacterium strains. The fatty acid profile of strain YC-RL4^T comprised major levels of C_16:0 (15.3%) and C_17:1ω7c (15.3%), while the fatty acid profile of strain MB418^T exhibited major levels of C_17:1ω7c (47.7%) and sum in feature 9 (iso- C_17:1ɷ9c/C_16:0 10-methyl) (43.3%). Strain HX176^T exhibited higher levels of C_17:1ω7c (43.2%) and C_19:1 trans 7 (30.9%), but lower C_16:0 levels (8.8%), while also lacking sum in feature 9 (Supplementary Table S1). In addition, only the polar lipid profiles of strains MB418^T and HX176^T included glycolipids (GLs), contrasting with that of strain YC-RL4^T that lacked GLs (Supplementary Figure S1). The HPLC profile of cell-wall mycolic acids from strain YC-RL4^T, containing 60–70 carbon atoms of mycolic acids as the main peaks, was a little similar to that observed for M. fluoranthenivorans JCM 14741^T and M. frederiksbergense DSM 44346^T while quite different from that of strains MB418^T, and HX176^T. The HPLC pattern of cell-wall mycolic acids from strains MB418^T and HX176^T were more similar to that of M. litorale DSM 45785^T than others, and the chromatogram of strain MB418^T was distinct from that of strain HX176^T (Supplementary Figure S2).

Nearly complete 16S rRNA gene sequences for strain YC-RL4^T (1,527 bp, accession OQ096616), MB418^T (1,509 bp, accession OP522341), and HX176^T (1,508 bp, accession OQ096617) were obtained. BLAST searches of the 16S rRNA gene sequences against the GenBank database indicated that strains YC-RL4^T, MB418^T, and HX176^T were closely related to Mycobacterium species of the family Mycobacteriaceae even though they were previously classified as the genus Mycolicibacterium. Strain YC-RL4^T was most closely related to M. fluoranthenivorans JCM 14741^T and M. frederiksbergense DSM 44346^T, with 16S rRNA gene sequence similarities of 99.3 and 99.2%, respectively. The 16S rRNA gene of MB418^T exhibited 98.5% nucleotide similarity to 16S rRNA genes of M. celeriflavum DSM 46765^T and M. moriokaense DSM 44221^T, but with 97.2% similarity to the 16S rRNA genes of strain YC-RL4^T, and 96.2–98.3% similarity to 16S rRNA genes of other various Mycobacterium type strains. The 16S rRNA gene sequence of strain HX176^T exhibited a nucleotide similarity of 98.3% to that of M. litorale DSM 45785^T, 98.0% to that of strain YC-RL4^T, and similarities of 95.8–97.9% to those from other Mycobacterium type strains (Supplementary Table S2). Phylogenetic analysis of the 16S rRNA gene sequences of the three isolates with other mycobacterial strains that were previously classified as “Mycolicibacterium” revealed that the strains belonged to the genus Mycobacterium. The 16S rRNA gene sequence of strain YC-RL4^T formed a sub-clade with that of the type strain M. fluoranthenivorans JCM 14741^T, which were together related to the 16S rRNA gene of M. frederiksbergense DSM 44346^T, consistent with BLAST comparisons. The 16S rRNA gene of strain MB418^T formed a well-supported sub-clade with the 16S rRNA gene of strain HX176^T, which were together associated with that of M. litorale DSM 45785^T (Figure 1). The whole-genome phylogenomic analysis recapitulated the 16S rRNA gene phylogenetic analysis, with the exception of the strain YC-RL4^T genome being most closely related to that of M. frederiksbergense DSM 44346^T (Supplementary Figure S3), contrasting with the 16S rRNA gene phylogenetic analysis.

The genomes of strains YC-RL4^T, MB418^T, and HX176^T were estimated to be 99.95, 100.00, and 100.00% complete, with estimated contamination of 0.62, 0.30, and 0.45%, respectively. The draft genome sizes of strains YC-RL4^T, MB418^T, and HX176^T were 6.1, 5.6, and 5.9 Mbp, respectively, and were assembled from two contigs with an N50 length of 5,801,417 bp; 15 contigs with an N50 length of 710,882 bp; and 41 contigs with an N50 length of 302,245 bp, respectively (Supplementary Table S3). The genomic sequence for strain YC-RL4^T encoded 5,881 total genes including 47 tRNA, 6 rRNA, 4 other ncRNA, and 83 pseudo genes. The genome of strain MB418^T encoded 5,444 total genes including 47 tRNA, 3 rRNA, 3 other ncRNA, and 98 pseudo genes. The genome of strain HX176^T encoded 5,692 genes including 46 tRNA, 4 rRNA, 3 other ncRNA, and 64 pseudo genes. The genomic G + C content of all three strains ranged between 66.5 and 69.3%. ANI values calculated between strains YC-RL4^T, MB418^T, HX176^T, and other Mycobacterium species were all <86.9%, with corresponding dDDH values all <31.8% (Supplementary Table S4). The values were lower than the thresholds used to delineate bacterial species (i.e., ANI < 95–96% and dDDH <70%) (Kim et al., 2014), thereby indicating that the strains represented novel, uncharacterized species.

Polycyclic aromatic hydrocarbons ring-hydroxylating dioxygenases (PAH-RHDs) are critical for PAH degradation by aerobic bacteria, because they catalyze the initial oxidation of O₂ atoms to form cisdihydrodiol and this step controls the PAH degradation rate (Kim et al., 2007; Chemerys et al., 2014). Among PAH-RHDs, aromatic ring-hydroxylating dioxygenases are particularly important in the aerobic bacterial degradation of aromatic compounds, because they catalyze the oxidation of the compounds during degradation. They typically catalyze the addition of two hydroxy groups to vicinal carbons, thereby disrupting aromaticity and yielding dihydrodiol compounds with cis, cis stereochemistry. The enzymes are biotechnologically important because they act as biocatalysts in the stereospecific synthesis of chiral synthons and the degradation of aromatic pollutants (Kahl and Hofer, 2003). Multiple genes encoding dioxygenases were identified in the genomes of strains YC-RL4^T, MB418^T, and HX176^T. Specifically, four genes putatively encoding alpha subunits with homology to aromatic ring-hydroxylating dioxygenases were encoded by the genome of strain YC-RL4^T, while genes putatively encoding alpha subunits of aromatic ring-hydroxylating dioxygenases were identified in the genomes of strains MB418^T and HX176^T (Supplementary Table S5).

Strains of the genus Mycobacterium could degrade PAHs to central intermediates via the o-phthalate and the β-ketoadipate pathway (Kim et al., 2007; Augelletti et al., 2020). The o-phthalate degradation pathway of Gram-positive bacteria involves oxygenation to form 3,4-dihydro-3,4-dihydroxyphthalate, dehydrogenation to 3,4-dihydroxyphthalate, and finally decarboxylation to generate protocatechuate (Bhattacharyya et al., 2023). Genes (phtAa) were identified that encoded homologous proteins involved in the degradation of phthalate in the genomes of strains YC-RL4^T, MB418^T, and HX176^T (Supplementary Table S5). Thus, these three strains could likely degrade various types of phthalates, as previously suggested. In addition, genes of the β-ketoadipate pathway cluster (pcaC, pcaD, pcaG, and pcaH) were also identified in the genomes of strains YC-RL4^T, MB418^T, and HX176^T (Supplementary Table S5), suggesting that anthracene degradation by these strains could proceed through the ortho-cleavage of protocatechuate (protocatechuic acid [PCA] 3,4-dioxygenase) (Habe et al., 2005). These protocatechuate catabolic genes were required for the complete degradation of PAHs to TCA cycle intermediates. Consequently, strains YC-RL4^T, MB418^T, and HX176^T exhibited the potential for the bioremediation of PAHs and smaller ring aromatics.

Although heavy metals naturally exist in many environments (Masindi and Muedi, 2018), the ubiquitous distribution of heavy metals in natural environments is a consequence of global industrialization and urbanization that has led to negative impacts on human and environmental health. Microorganisms have adapted multiple resistance mechanisms to overcome the physiological stress from heavy metal exposure (Sherpa et al., 2020). Several genes were identified in the genomes of strains YC-RL4^T, MB418^T, and HX176^T that encoded heavy metal resistance proteins. For example, the three strains exhibited the potential to mediate copper toxicity via copper-resistant genes (copC) identified in all of their genomes. Further, the copper-resistance gene copD was also identified in the genomes of strain MB418^T and HX176^T. Free form copper is highly toxic due to its ability to produce radicals during cycling between oxidized Cu(II) and reduced Cu(I) forms. Consequently, intracellular copper must remain complexed within a tightly controlled copper homeostatic system. Importantly, Cop proteins sequester excess copper in periplasms and outer membranes (Arnesano et al., 2003). Consequently, these proteins might confer copper resistance to strains YC-RL4^T, MB418^T, and HX176^T by helping to sequester and accumulate copper in periplasms with copper binding proteins, thereby preventing toxic levels of copper from entering the cytoplasm.

In addition to the above, genes (merA and merB) that encoded mercury reductase enzymes were identified in the strain YC-RL4^T genome that enable resistance to mercury stress. Mercury toxicity arises from the strong affinity of monomethyl-Hg and Hg²⁺ to sulfur atoms in cysteine residues and, hence, interference with protein structure and function. The mercury reductase encoded by merA and merB reduced the toxicity of Hg2+ to Hg⁰ (Møller et al., 2014). Specifically, organomercurial lyase (MerB) catalyzed the protonolysis of the carbon-mercury bond, resulting in the formation of ionic mercury and reduced hydrocarbon. The ionic mercury (Hg²⁺) was subsequently reduced to less reactive elemental mercury (Hg⁰) by mercuric reductase (MerA) (Di Lello et al., 2004). The presence of these genes involved in PAH-degradation and metal resistance genes in the genomes of these strains suggested that strains YC-RL4^T, MB418^T, and HX176^T could be applied in bioremediation of soils contaminated with both PAHs and heavy metals.

The pan-genomes of strains comprise the entire gene repertoire for a given species across several populations, and their analysis can be informative for understanding the distribution of core, important genes shared by all strains of a species, in addition to strain-specific accessory genes that may be more dispensable (Reis and Cunha, 2021). A total of 401,362 protein-coding genes (Table 2) were encoded by the genomes of strains YC-RL4^T, MB418^T, HX176^T, and 70 other related strains of the genus Mycobacterium, comprising 51,712 homologous families based on cluster analysis. A total of 734 core genes were shared by the 73 strains and comprise the core genome identified in this study. The core genes accounted for approximately 13.5% of the pan-genome for the 73 strains of the genus Mycobacterium. The identification of the “open” pangenome based on consideration of increasing numbers of strains was fit with a powerlaw regression function [f(X) = 5991.36×^0.50], while the core genome identification fit an exponential regression [f(X) = 2838.85e^−0.03X; Figure 2]. Thus, the relatively extensive open pangenome of the strains indicates that they have considerable repertoires of accessory genes, possibly due to extensive horizontal gene exchanges with various microbial species (Coyte et al., 2022). Aromatic ring-hydroxylating dioxygenase alpha subunit-encoding genes were identified in all of the 73 genomes, suggesting it was an important core gene shared by all strains and plays a critical role in PAH degradation by the strains. In addition, the copy numbers of these genes varied among different strains. The lowest number of copies was observed in the genome of M. brumae DSM 44177^T, while the largest number of copies was observed in the genome of strain M. gadium DSM 44077^T.

A functional classification of core, accessory, and unique genes was performed based on comparison to the COG database, revealing clear differences between core and accessory genes. Core genes were most associated with the COG categories of R (general function prediction only; 13.0% of genes), J (translation, ribosomal structure, and biogenesis; 12.4%), E (amino acid transport and metabolism; 9.1%) and I (lipid transport and metabolism; 8.2%) (Figure 3), while accessory and unique genes were most associated with the COG categories R (general function prediction only; 16.3 and 17.0%, respectively), K (transcription; 13.0 and 12.6%), and Q (secondary metabolite biosynthesis, transport, and catabolism; 8.8 and 9.3%). Thus, core genes were mostly involved in basic physiological functions, resulting in the overall prevailing phenotypes of the strains. In contrast, accessory genes related to tachytely evolution could increase the gene and functional diversity of their respective genomes. These accessory genes may be involved in metabolic pathways that critically ensure adaptations or functioning in variable ecological niches that differ among species (Vernikos et al., 2015).

Polyketide synthesis operons and others involved in non-ribosomal peptide synthesis (NRPS) are common in the genomes of Mycobacterium and are responsible producing cell wall-associated lipids, siderophores, and other biologically active molecules (Stinear et al., 2008). AntiSMASH analysis revealed the presence of 14–21 biosynthetic gene clusters (BGCs) in the genomes of strain YC-RL4^T, MB418^T, and HX176^T that were especially enriched in genes encoding Type I polyketide synthases (PKSs) and non-ribosomal peptide synthases (NRPSs) (Supplementary Table S6). A BGC that was most similar to that encoding mycobactin was identified in all 73 genomes, consistent with the production of mycobactin being essential for Mycobacterium strains to access iron (McMahon et al., 2012). In addition, several putative BGCs exhibited low similarities to known clusters, indicating their potential for synthesizing putatively novel secondary metabolites.

(a.di.pa’tum. L. neut. adj. adipatum, fat, greasy)

Cells are Gram-positive, non-motile, and rod-shaped, producing opaque, ivory, and convex colonies on GYM and MB7H10 agar after cultivation at 48 h and 28°C. Cells are catalase-negative and oxidase-positive. Optimum growth was observed at 10–37°C, pH 6–7, and with NaCl concentrations ranging between 0 and 5%. Cells were negative for Voges–Proskauer, H₂S production, and starch hydrolysis tests. Acetic acid, D-gluconic acid, D-mannose, glucuronamide, glycerol, L-glutamic acid, L-malic acid, propionic acid, tween 40, α-D-glucose, β-hydroxy-D, and L-butyric acid can be utilized as carbon sources. 3-methyl glucose, acetoacetic acid, bromo-succinic acid, citric acid, D-arabitol, D-aspartic acid, D-cellobiose, dextrin, D-fructose, D-fucose, D-galactose, D-galacturonic acid, D-glucose-6-PO₄, D-glucuronic acid, D-lactic acid methyl ester, D-maltose, D-mannitol, D-melibiose, D-raffinose, D-saccharic acid, D-salicin, D-serine, D-sorbitol, D-trehalose, D-turanose, gelatin, gentiobiose, glycyl-L-proline, inosine, L-aspartic acid, L-fucose, L-galactonic acid lactone, L-histidine, L-pyroglutamic acid, L-rhamnose, L-serine, mucic acid, myo-inositol, N-acetyl neuraminic acid, N-acetyl-D-galactosamine, N-acetyl-D-glucosamine, N-acetyl-β-D-mannosamine, pectin, p-hydroxy-phenylacetic acid, quinic acid, stachyose, sucrose, α-D-lactose, α-keto-butyric acid, α-keto-glutaric acid, β-methyl-D-glucoside, and γ-amino-butryric acid are not used as carbon sources. D-fructose, D-glucose, D-xylose, esculin ferric citrate, glycerol, inositol, potassium 2-ketogluconate, and potassium gluconate can be assimilated and fermented to produce acid. Cells exhibit positive acid phosphatase, alkaline phosphatase, cystine arylamidase, esterase (C4), esterase lipase (C8), leucine arylamidase, lipase (C14), naphthol-AS-BI-phosphohydrolase, trypsin, valine arylamidase, and α-glucosidase activities via API ZYM strip assays. Whole-cell hydrolysates contain meso-diaminopimelic acid (meso-A₂pm), arabinose, galactose, glucose, mannose, and ribose. Cellular polar lipid profiles comprise diposphatidylglycerol (DPG), phosphatidylethanolamine (PE) and phosphatidylinositol (PI). The primary menaquinone is MK-9 (H₂). The major fatty acids were C_16:0 (15.3%) and C_17:1ω7c (15.3%). The mycolic acids primarily contain 60–70 carbon atoms. The type strain YC-RL4^T (= CPCC 205684^T = CGMCC 1.62027^T) was isolated from a petroleum-contaminated soil sample from Shandong, China. The draft genome of the type strain is 6.1 Mbp in size and exhibits a genomic G + C content of 67.4%.

(de.ser’ti. L. gen. n. deserti, from a desert)

Cells are Gram-positive, non-motile, and non-spore forming. Ivory colonies develop on GYM and MB7H10 agar after incubation for 48 h at 28°C. Optimal growth was observed at 15–45°C, pH 6–7, and NaCl concentrations ranging from 0 to 5%. Cells were oxidase-positive and catalase-negative, but exhibited negative results for Voges–Proskauer, H₂S production, and starch hydrolysis tests. Cells can utilize acetic acid, D-arabitol, D-fructose, D-galactose, D-gluconic Acid, D-mannitol, D-mannose, D-saccharic acid, D-sorbitol, myo-inositol, propionic acid, quinic acid, and α-D-glucose as sole carbon sources. API 50CH tests indicated that arbutin, D-lyxose, D-saccharose, D-tagatose, esculin ferric citrate, potassium 2-ketogluconate, and potassium gluconate are used as carbon sources. In contrast, amidon, amygdalin, D-arabinose, D-arabitol, D-ardonitol, D-cellobiose, D-fructose, D-fucose, D-galactose, D-glucose, D-lactose, D-maltose, D-mannitol, D-mannose, D-melezitose, D-melibiose, D-raffinose, D-ribose, D-sorbitol, D-trehalose, D-turanose, dulcitol, D-xylose, erythritol, gentiobiose, glycerol, glycogen, inositol, inulin, L-arabinose, L-arabitol, L-fucose, L-rhamnose, L-sorbose, L-xylose, mehyl-α-D-glucopyranoside, mehyl-α-D-mannopyranoside, Mehyl-β-D-xylopyranoside, N-acetylglucosamine, potassium 5-ketogluconate, salicin, and xylitol are not used as carbon sources. API ZYM analysis revealed the presence of positive acid phosphatase, alkaline phosphatase, cystine arylamidase, esterase (C4), leucine arylamidase, lipase (C14), naphthol-AS-BI-phosphohydrolase, trypsin, valine arylamidase, α-chymotrypsin, α-glucosidase, and β-glucosidase activities. Whole-cell hydrolysates contained meso-diaminopimelic acid (meso-A₂pm), arabinose, galactose, glucose, mannose, and ribose. Cellular polar lipid profiles comprised diposphatidylglycerol (DPG), phosphatidylethanolamine (PE), phosphatidylinositol (PI), and glycolipids (GL). The primary menaquinone was MK-9 (H₂). The major fatty acids were C_17:1ω7c (47.7%) and summed feature 9 (iso- C_17:1ɷ9c/C_16:0 10-methyl) (43.3%). The mycolic acids mainly contain 78–86 carbon atoms, with moderate amount of 76–84 carbon atoms, and minor of 62–70 carbon atoms. The type strain MB418^T (= CPCC 205710^T = KCTC 49782^T) was isolated from a gravel soil sample from the Gurbantunggut desert, China. The draft genome of the type strain is 5.6 Mbp in length and exhibits a genomic G + C content of 66.5%.

(hip.po.pha’is. L. gen. n. hippophais, from Hippophae, a plant genus whose rhizosphere the type strain was isolated from)

Cells are Gram-positive, non-motile, and rod-shaped, producing yellow colonies. Cells were positive for oxidase activity, but negative for catalase, Voges–Proskauer, H₂S production, and starch hydrolysis tests. Optimum growth was observed at 15–28°C, pH 6–7, and with NaCl concentrations ranging between 0 and 4%. Acetic acid, D-gluconic acid, D-saccharic Acid, propionic acid, and β-hydroxy-D,L-butyric acid can be used as substrates. API 50CH analysis revealed that D-fructose, D-glucose, D-ribose, esculin ferric citrate, potassium 2-ketogluconate, and potassium gluconate can be used as carbon substrates. Enzymatic activity analysis revealed the presence of alkaline phosphatase, esterase (C4), esterase lipase (C8), leucine arylamidase, lipase (C14), naphthol-AS-BI-phosphohydrolase, and valine arylamidase activities, but the lack of activity for acid phosphatase, cystine arylamidase, N-acetyl-β-glucosamimidase, trypsin, α-chymotrypsin, α-fucosidase, α-galactosidase, α-glucosidase, α-mannosidase, β-galactosidase, β-glucosidase, and β-glucuronidase. Whole-cell hydrolysates contained meso-diaminopimelic acid (meso-A₂pm), arabinose, galactose, glucose, mannose, and ribose. Cellular polar lipid profiles comprised diposphatidylglycerol (DPG), phosphatidylethanolamine (PE), phosphatidylinositol (PI), and glycolipids (GL). The primary menaquinone was MK-9 (H₂). The major fatty acids were C_17:1ω7c (43.2%), C_19:1 trans 7 (30.9%), and C_16:0 (8.8%). The mycolic acids primarily contain 68–78 carbon atoms. The type strain HX176^T (= CPCC 205372^T = KCTC 49413^T) was isolated from a rhizosphere soil sample of a medicinal plant in Xinjiang, China. The draft genome of the type strain is 5.9 Mbp in length and exhibits a genomic G + C content of 69.3%.

The 16S rRNA gene sequences of strains YC-RL4^T, MB418^T, and HX176^T were deposited in GenBank under the accession numbers OQ096616, OP522341 and OQ096617, respectively. The whole genome shotgun data for the genomes of strains YC-RL4^T, MB418^T, and HX176^T were deposited in the DDBJ, ENA, and GenBank databases under the accession numbers CP015596, JAODWD000000000, and JAPZPY000000000, respectively.