A novel eco-friendly Acinetobacter strain A1-4-2 for bioremediation of aquatic pollutants

Abstract

The increasing accumulation of hydrocarbons and aromatic compounds in aquatic ecosystems, stemming from anthropogenic activities, poses severe ecological challenges, including disrupting biodiversity and threatening human health through the food chain. This study presents Acinetobacter strain A1-4-2, isolated from a hairy crab farming base, which could represent a novel Acinetobacter species. The metagenomic analysis of approximately 12,000 publicly available datasets revealed that this novel Acinetobacter species is widely distributed across various environments, particularly in those with high organic matter content, such as sludge, feces, and wastewater. Strain A1-4-2 exhibited exceptional metabolic capabilities, effectively degrading a diverse range of substrates, including amino acids, organic acids, oils, n-alkanes, lignin, and aromatic monomers. Genomic analysis, coupled with biological experiments, revealed that strain A1-4-2 exhibited resistance to a very limited kind of antibiotics. Moreover, the strain’s biosafety, affirmed through zebrafish toxicity assays, underscores its suitability for environmental release. Additionally, the feasibility of genetic manipulation of strain A1-4-2 gives it the potential to become a chassis cell, enabling it to degrade organic pollutants more efficiently through genetic engineering. Our findings elucidate the strain’s genomic and metabolic attributes, offering insights into its biodegradation potentials and developing effective strategies for ecological restoration in face of pollution.

Supplementary Information

The online version contains supplementary material available at 10.1038/s41598-025-05431-0.

Introduction

The increasing accumulation of hydrocarbons, such as n-alkanes and plant oils, along with aromatic compounds in aquatic environments, poses a significant ecological challenge due to their anthropogenic origins, primarily from petroleum leaks, kitchen waste, and industrial outflows¹. These pollutants persist in the environment, causing profound harm to aquatic ecosystems by disrupting microbial communities, reducing soil fertility, and curtailing phytoplankton productivity². The ensuing decline in aquatic biodiversity and the potential for these toxins to accumulate in organisms, leading to reproductive failure and population declines, and disruption in the delicate balance of aquatic food webs³. Furthermore, the pollutants’ infiltration into the soil can adversely impact agricultural productivity, affecting crop yields and quality, and introducing these toxins into the human food chain⁴. The bioaccumulation and biomagnification of these compounds in the food chain ultimately pose risks to human health⁵. Hence, robust remediation strategies are needed to counteract the effects of these pollutants on aquatic ecosystems and terrestrial agriculture, thereby protecting human health and ecological balance.

Biodegradation emerges as a sustainable and eco-friendly strategy, capable of reducing pollutant levels and rejuvenating the ecological health of affected areas. Within the pantheon of microbial degraders, Acinetobacter species stand out for their metabolic versatility, enabling the utilization of a wide array of substrates, including environmental pollutants⁵. For example, Acinetobacter sp. HX09 showed a very high removal of short- and medium-chained alkanes, approximately 64% after 7 days of incubation at 30℃¹. Moreover, co-cultured Acinetobacter baumannii and Tararobacter species for 14 days led to 80% of the total degradation of crude oil in water⁶. Whole-genome sequencing of Alcaligenes sp. strain MMA showed that this strain was able to degrade amoxicillin and remove a variety of heavy metals⁷. These findings highlight the potential of Acinetobacter strains as powerful tools in the bioremediation arsenal, capable of detoxifying environments compromised by hydrocarbons and other xenobiotic compounds, thereby fostering the restoration of ecological balance and integrity. Nevertheless, since many Acinetobacter species have the potential to cause infectious diseases in humans or animals⁸a thorough safety assessment of these strains is essential before utilizing them for bioremediation.

In this study, we isolated an Acinetobacter strain, designated A1-4-2, from the aquatic environment of a hairy crab farming base, phylogenetically representing a novel Acinetobacter species. The environmental distribution of A1-4-2-like strains was studied using large-scaled metagenomic datasets. The metabolic potentials and xenobiotic degradation capacity were systematically explored through genomic and experimental approaches. Moreover, the antibiotic resistance, biosafety, and feasibility of genetic manipulation of strain A1-4-2 were further evaluated, reinforcing its potential for bioremediation. Our study demonstrated the proficiency of strain A1-4-2 in degrading diverse aquatic pollutants, highlighting its applicability for ecological restoration.

Materials and methods

Sample source and bacterial isolation

Water samples for the isolation of this experimental strain were collected from Jinghu Hairy Crab Farming Base, Huaian City, Jiangsu Province, China (33.012°N, 119.168°E) on August 10, 2022. The samples were filtered through sterile gauze with a pore size of 1 mm ×1 mm to remove impurities and then subjected to bacterial isolation using plates containing the following components: CH₃COONa·3H₂O 3.6 g/L, MgSO₄·7H₂O 0.132 g/L, Na₂HPO₄·2H₂O 0.029 g/L, K₂SO₄ 0.027 g/L, NH₄Cl 0.054 g/L, CaCl₂·2H₂O 0.017 g/L, HEPES buffer 12 mL/L, and agar powder 20 g/L, adjusted to pH 7.0. Once the medium was autoclaved, 1 mL/L of trace element solution was added, which had been sterilized by filtration through a 0.1 μm filter. Subsequently, we employed LB-agar plates for streaking experiments to purify each single colony, and verified their identity as a single species by conducting 16 S rRNA PCR and sequencing. The purified strain A1-4-2 (= MCCC1K09560) was cultivated in LB liquid medium and preserved in a -80℃ freezer after the addition of 30% glycerol.

Extraction, sequencing, and assembly of bacterial genomic DNA

The genomic DNA of A1-4-2 was extracted following the method of Fang et al.⁹and the whole-genome sequencing was conducted by MajorBio (Shanghai Majorbio Bio-pharm Technology Co., Ltd., Shanghai, China) using the PacBio Sequel IIe and Illumina sequencers. The genomic DNA was fragmented to approximately 400 bp fragments for the Illumina library construction, and the NEXTFLEX Rapid DNA-Seq Kit was used for library preparation. For the PacBio library construction, the genomic DNA was fragmented to approximately 10 kb fragments, followed by end-repairing and circular single-stranded adapters ligation at both ends according to the PacBio protocol (Pacific Biosciences, CA).

For Illumina sequencing, the prepared libraries were subjected to paired-end sequencing (2 × 150 bp) on the Illumina sequencer. For PacBio sequencing, the single-stranded circular library was annealed and hybridized to the polymerase attached at the bottom of the fixed ZMW (zero-mode waveguides) on the PacBio Sequel IIe sequencer¹⁰. Sequencing reagents were added, and after each base pairing and synthesis, a corresponding light signal was emitted and detected. With a high-resolution optical detection system, real-time detection was performed, with each base incorporation appearing as a pulse peak¹¹.

The raw data from the Illumina sequencing is stored in the fastq format. To ensure more accurate subsequent assembly, the software fastp v0.23.0 is used to perform quality trimming on the raw data, resulting in high-quality clean data¹². The data from the PacBio Sequel IIe is HiFi reads. The quality-controlled Illumina data and HiFi reads are assembled using Unicycler v0.4¹⁰. Then, Pilon v1.22 is used to map the Illumina short sequences onto the assembled genome for correction¹³. The GenBank number of strain A1-4-2’s genome is GCA_040233925.1.

Gene annotation

The NCBI prokaryotic genome annotation pipeline (PGAP) was employed for ORF prediction and gene annotation¹⁴. Moreover, we also use Prodigal¹⁵ for local ORF prediction to find ORFs that might be missed in the PGAP process. The predicted protein sequences were further aligned with the Clusters of Orthologous Groups of proteins (COG)¹⁶ and the Transporter DB 2.0 database using the BLASTp software with the following parameters: identity of 50%, query coverage of 80%, e-value of 1 × 10^− 5, and a score of 40 ¹². InterProScan 5 was used for the functional annotation of proteins to predict features such as protein domains and important sites using default parameters¹⁷. BlastKOALA was utilized to assign Kyoto Encyclopedia of Genes and Genomes (KEGG) annotations¹⁸.

Genomic Island and prophage prediction

Genomic islands were identified using IslandViewer 4¹⁹. For prophage prediction, VirSorter2 (version 2.2.2)²⁰ was firstly employed to predict viral sequences within strain A1-4-2, utilizing the parameters “—include-groups dsDNAphage, ssDNA—min-length 5000—min-score 0.5”. This configuration was selected to optimize the identification of temperate viruses, with a minimum score threshold set at 0.5 for enhanced sensitivity. Following this, CheckV (version 0.8.1)²¹ was applied to assess and ensure the quality of the predicted viral sequences. Subsequently, the sequences curated by CheckV were resubmitted to VirSorter2 for further refinement and formatting, preparing them for input into DRAM-v (version 1.2.4) for comprehensive viral annotation²². The sequences that passed through these filters were then evaluated based on established empirical criteria, which consider the counts of viral and host genes, the VirSorter2 score, the number of hallmark genes, and the length of the sequence. The classification of proviruses adhered to the guidelines outlined in the online protocol available at https://www.protocols.io/view/viral-sequence-identification-sop-with-virsorter2-5qpvoyqebg4o/v2, accessed on 16 November 2022.

To investigate whether these prophages are replicating, we aligned the Illumina reads of strain A1-4-2 to its complete genome sequences using Bowtie 2²³, and the resulting SAM file was converted to BAM files using Samtools²⁴. Subsequently, we imported the BAM file into the Integrative Genomics Viewer (IGV, version 2.16) to ascertain coverage details for each genomic region²⁵. If the coverage of a prophage is more than two times the average coverage depth of the chromosome, we proposed that the prophage could be in an active state of replication.

Phylogenetic analysis

The Genome Taxonomy Database (GTDB) taxonomy was used to study the phylogeny of strain A1-4-2, “A. haemolyticus” w12 from GenBank database, and all Acinetobacter genomes from NCBI RefSeq database (before October 2024). The sequences of 120 concentrated proteins in the genomes were predicted and aligned using GTDB-Tk (database version: R220)²⁶. The phylogenetic tree was constructed using FastTree2 with the neighbor-joining method²⁷and a bootstrap analysis with 1,000 replicates was performed to assess the robustness of the tree. Finally, the subtree encompassing strain A1-4-2 was extracted using MAGA, and further decorated using iTOL²⁸.

Identification and acquirement of strain A1-4-2-like genomes from Large-scaled metagenomic datasets

To study the environmental distribution of A1-4-2-like strains, we firstly identified the “core genome” of Clade_I Acinetobacter using OrthoMCL pipeline²⁹with following parameters: identity 50%; coverage, 80%; e-value, 1e-5; and mcl inflation, 1.4. Next, we aligned the core-genome sequences against NCBI nr database. This analysis led to a unique autotransporter beta-barrel protein (encoded by ABJ384_RS12135 in strain A1-4-2), which showed > 97% sequence identity and 100% coverage among the Clade_I strains, but was absent in other nr-indexed microbes (BLASTp: identity ≥ 50%, coverage ≥ 50%, e-value ≤ 1e-5), only except for Acinetobacter sp. CAAS 2–6. Then, we aligned the signature proteins from all Clade_I genomes against MGnify and nr databases³⁰as well as our in-house database which encompasses the protein sequences predicted from ~ 12,000 metagenomic assemblies, with BLASTp parameters of identity, 90%; coverage, 95%; and e-value, 1e-30.

Samples containing the signature protein underwent read trimming using Sickle v1.33 and subsequent assembly using MEGAHIT³¹with the following settings: a minimum kmer length of 31, a maximum kmer length of 149, and a kmer increment of 6. Bowtie 2 was then applied to align the reads to the assembled contigs²³. Samtools was utilized to convert SAM files into BAM format, sort the BAM files, and establish their indexes²⁴. The BamM³² was employed to filter through the aligned reads, retaining only those with a coverage of at least 90% and an identity of at least 95%. The average coverage per base pair of the contigs was determined by “parse” subcommand of BamM, which facilitated the exclusion of the highest and lowest 10% coverage areas using the “tpmean” parameter³². Contigs exceeding 2.5 kb in length were subjected to further analysis for coverage calculation and binning.

Prodigal¹⁵ was used to predict protein sequences from the contigs with the parameter “-p meta”, which were then aligned against the signature protein. Dimensionality reduction based on the tetranucleotide frequency matrix was performed using the t-SNE algorithm, with visualization facilitated using the R package mmgenome2³³. During the visualization, contigs containing the signature protein were highlighted, and manual binning process was used to ensure precise demarcation of boundaries within metagenome-assembled genomes (MAGs). Rebinning involved plotting all contigs from each MAG based on their abundance and GC content, with manual curation to remove contigs showing inconsistent coverage or GC content.

Notably, the raw reads for the NCBI assemblies with accession numbers GCA_016277765.1, GCA_016278185.1, GCA_016294955.1, and GCA_008770255.1 are not publicly available. Therefore, we determined the coverages of their contigs by aligning the Illumina reads of strain A1-4-2 and SRR23126129 to these assemblies. Using the reads of SRR23126129 is because that the strain A1-4-2-like MAG obtained from it (SRR23126129_bin1) showed superior completeness and N50 value. The subsequent rebinning process was specifically based on such two data of coverage, with manual curation to eliminate contigs that exhibited inconsistent coverage patterns.

The MAGs were phylogenetically checked using GTDB-tk²⁶and those identified as “Acinetobacter sp002135245” were retained for further analysis. The integrity and contamination levels of the MAGs were evaluated using CheckM³²with preference given to MAGs exhibiting completeness above 30% and contamination below 10%.

Chemotaxonomy analysis

Strain A1-4-2 was cultivated in LB liquid medium for a period of 48 h at a temperature of 30℃. Cellular fatty acid analysis was conducted following the MIDI (Sherlock Microbial Identification System, version 6.0) protocol, which involved saponification, methylation, and extraction. The fatty acid methyl esters were subsequently examined using gas chromatography with an Agilent Technologies 6850 system and identified against the RTSBA6.0 database of the Microbial Identification System, as Athalye et al.³⁴.

For the polar lipid analysis of strain A1-4-2, extraction was performed, and separation was achieved using silica gel 60 F254 aluminum-backed thin-layer chromatography plates (10 × 10 cm; Merck 5554). The analysis was conducted in accordance with the methodology of Dobson et al.³⁵. The solvent system for the first dimension consisted of a chloroform/methanol/water mixture (65:24:4, by volume), while the second dimension utilized a chloroform/glacial acetic acid/methanol/water mixture (80:15:12:4, by volume). The plates were then sprayed with a 5% phosphomolybdic acid solution (w/v in alcohol) and heated to 160℃ for a duration of 10–15 min to visualize the total lipids.

The extraction of respiratory quinones was carried out following the procedure outlined by Dobson et al.³⁵and the analysis was performed using high-performance liquid chromatography (HPLC), as described by Costa et al.³⁶.

Utilization of plant oils, peptone and glycerol

We first used the Biolog GenIII kit (Biolog, Inc., the United States) to preliminarily study the substrate utilization ability of strain A1-4-2. Moreover, we also used various substrates as the sole carbon and/or sulfur sources to assess the substrate metabolic capabilities of strain A1-4-2. In detail, strain A1-4-2 was cultivated in an LB liquid medium for 24 h. Then, the bacterial cells were centrifuged and washed three times with an equal volume of sterile PBS and then transferred to a liquid culture medium of 1% basal salt medium (MSM). Each medium was supplemented with one of the following carbon sources: peanut oil at 1% v/v, sunflower oil at 1% v/v, linseed oil at 1% v/v, glycerol at 5 g/L, or peptones at 5 g/L. All groups were incubated in a shaker at room temperature with a shaking speed of 150 revolutions per minute. The condition of the medium was monitored and documented each day. Each experimental setup comprised three replicate groups and one blank control.

Lignin degradation and aromatics utilization

Microbial degradation of lignin was tested in triplicate in 200 mL of MSM liquid medium with lignin as the sole carbon source, and lignin was added at a rate of 5 g/L. Cultures of strain A1-4-2 were washed three times with sterile PBS, inoculated at a 1% inoculum size into medium with lignin, and incubated at 30 ℃.

Since certain lignin peroxidases such as laccase, DyP, LiP, and MnP are known to decolorize methylene blue, we verify its lignin-degrading capability via dropping 0.1 mL of the culture broth onto LB-agar plates added with 3% aniline blue. Then, the plates were incubated at 30℃, and the decolorization of aniline blue was clearly observed around the colonies after 24 h. Furthermore, we leveraged the property of laccase to cause color change in guaiacol to verify its production by strain A1-4-2. In detail, 0.2 mL of the bacterial culture broth was dropped on the LB-agar plates at 30℃. After incubation for 30 h, a 9 mm bacterial colony was taken using a sterile punch, and inoculated at the center of another LB-agar plates containing 0.5% guaiacol. After incubation at 30℃ for 48 h, the appearance of a reddish-brown colored circle around the colony indicated positive laccase activity.

To confirm the utilization of aromatic compounds, the cells of strain A1-4-2 were centrifuged and washed three times with an equal volume of sterile PBS and then transferred to a liquid culture medium of 1% MSM supplemented with 3-hydroxybenzoic acid and sodium benzoate (both with 0.5 g/L). Each of the aforementioned experiments was performed three times.

Antibiotic sensitivity test

The antibiotic susceptibility testing for strain A1-4-2 was conducted via spreading 200 µL of the culture broth onto the surface of LB agar plates. Subsequently, the plates were positioned with a disc containing antibiotic-sensitive paper, and incubated at 30℃ for a duration of 24 h. The measurement of the inhibition zone diameter was carried out in accordance with the guidelines set by the European committee for the testing of pharmacological sensitivities (EUCAST), employing the paper disk diffusion method³⁷.

Biosafety test

To assess the safety of strain A1-4-2 for potential applications, a safety experiment was conducted using zebrafish. We selected zebrafish that were 2 to 2.5 cm in length and 2 to 3 months old. The experiment included an immersion group with three varying concentrations of strain A1-4-2: 1.1 × 10⁷ CFU/mL, 1.1 × 10⁶ CFU/mL, and 1.1 × 10⁵ CFU/mL, along with a blank control group (LB medium). In each group, eight zebrafish were placed randomly into separate beakers. During the experimental period of 5 days, the fish were maintained under normal breeding conditions, and mortalities were recorded daily. Each experimental group was replicated three times to ensure the reliability of the results.

Results

Description of strain A1-4-2

Growth on LB-agar plates, the colonies of strain A1-4-2 exhibited clear, round colonies with regular, slightly raised edges. Gram staining results showed that it was Gram-negative. Cultured in the LB broth, its optimum growth temperature was 30 ℃. Transmission electron micrograph (TEM) showed that the cells were measured 1.5 μm by 0.8 μm, lacking flagella (Fig. 1).

Fig. 1.

Characteristics of Strain A1-4-2, including A the colony appearance, B Gram stain, C, D scanning electron microscope images, and E growth curves at various temperatures

The major cellular fatty acids of strain A1-4-2 (> 5.0%) included C15:0 (39.78%), C17:0 (24.88%), Summed Feature 3 (C16:1ω7c and/or C16:1ω6c; 6.53%; Supplementary Table S1). The polar lipids of strain A1-4-2 were composed of phosphatidylglycerol (PG) and phosphatidylethanolamine (PE) (Supplementary Figure S1, Supplementary Table S1). The respiratory quinones of strain A1-4-2 include ubiquinone Q-8 and Q9 (Supplementary Figure S2).

The genomic features of strain A1-4-2

The complete genome of strain A1-4-2 consisted of one chromosome and three plasmids with a total length of 3,272,818 base pairs (bp), which is smaller than that of closely related strains, Acinetobacter sp. CS-2 (Supplementary Figure S3). The G + C content of strain A1-4-2 genome was 41.99%. The genome of strain A1-4-2 contained 3,005 genes, including 2,296 protein-coding genes, 80 tRNAs, and 7 rRNA operons (Table 1). All of the seven 16 S rDNA sequences showed 100% identity with each other.

Table 1.

Genome features of strain A1-4-2

Items	Description
Size (bp)	3,272,818
G + C content (%)	41.99%
Total genes	2,985
Protein-coding genes	2,296
Genes assigned to COG	2,334
rRNA operons	7
tRNA genes	80
Gene island	9

Table 2.

A1-4-2 antibiotic susceptibility test

Class	Diameter (mm)	Sensitivity
Kanamycin	0	R
Ampicillin	21	S
Clindamycin	24	S
Cefuroxime	28	S
Levofloxacin	30	S
Ceftazidime	18	I
Polymyxin B	17	I
Vancomycin	21	S
Erythromycin	0	R
Doxycycline	22	S
Streptomycin	23	S
Ciprofloxacin	32	S
Lincomycin	0	R
Florfenicol	33	S
Piperacillin	18	I
Cephalexin	24	S
Oxacillin	9	R
Cefazolin	34	S
Ceftriaxone	21	S
Tetracycline	12	R
Norfloxacin	24	S
Amikacin	20	S
Azithromycin	0	R
Fosfomycin	34	S
Cefoperazone	28	S
Penicillin	12	R
Chloramphenicol	36	S

Based on the results of the antibiotic susceptibility test, bacterial sensitivity was categorized into three levels: For “susceptible” (S), the inhibition zone diameter must be 20 mm or greater; for “intermediate” (I), the inhibition zone diameter ranges between 15 mm and 20 mm; and for “resistant” (R), the inhibition zone diameter is no more than 14 mm

Moreover, nine genomic islands were predicted from strain A1-4-2’s chromosome. The genome of strain A1-4-2 also harbored two predicted proviruses, located at genomic coordinates 984,197 to 1,012,801 and 1,168,692 to 1,180,345, which were of medium and low quality, respectively. The regions containing these proviruses did not show a notable increase in sequencing coverage when compared to other chromosomal regions (as shown in Supplementary Figure S3), suggesting that these proviruses may have compromised their ability to replicate.

Upon COG classification, 2,296 protein-coding genes were assigned to 22 categories (Supplementary Table S2). The major COG categories were translation, ribosomal structure and biogenesis (COG-J, 9.54%), general function prediction only (COG-R, 8.14%), amino acid transport and metabolism (COG-E, 7.93%), lipid transport and metabolism (COG-I, 6.84%), and cell wall/membrane/envelope biogenesis (COG-M, 6.79%). The graphical representation of the A1-4-2 genome is shown in Fig. 2.

Fig. 2.

Diagram of Acinetobacter A1-4-2 genome. Genes on the forward (shown in the outer circle) and reverse (shown in the inner circle) strand are colored according to their cognate gene cluster (COG) categories: RNA genes are highlighted in different colors (blue for tRNA and red for rRNA), GC content is shown in yellow/blue, and GC shift is shown in orange/red

The phylogeny of strain A1-4-2: a novel Acinetobacter species

The alignments of 16 S rRNA gene sequences showed that strain A1-4-2 was closely related to Acinetobacter johnsonii IC001, A. johnsonii BIGb0494, A. johnsonii HN020 and Acinetobacter sp. MYb177 with identities of 99.153%, followed by many A. haemolyticus strains with identities of 99.152% (Supplementary Table S3). However, the average nucleotide identity (ANI) values between strain A1-4-2 and these strains were below 85%, lower than the threshold of 95% for the same species.

To validate the phylogeny of strain A1-4-2, we constructed a phylogenetic tree based on 120 conserved protein sequences (known as GTDB taxonomy), which encompassed all Acinetobacter genomes from NCBI RefSeq database, as well as “A. haemolyticus” w12 from Genbank database. As shown in Fig. 3, strain A1-4-2 formed a coherent phylogenetic cluster (as “Clade_I”) associated with 10 Acinetobacter strains, including unclassified ANC_4218, YH01022, YH12201, SA01, STC_101, CS-2, YH16039, YH01016, YH01018 and misclassified “A. haemolyticus w12”. Consistent with its phylogeny, strain A1-4-2 showed ANI values within the following ranges: 96.37–96.76% with Clade_I, at most 91.30% with Clade_II, at most 88.06% with Clade_III, at most 89.22% with Clase_IV, 85.40% with Clade_V, and less than 85% with other Acinetobacter strains (Supplementary Table S3). These pieces of evidence indicated that strain A1-4-2 represents a novel species of Acinetobacter.

Fig. 3.

Phylogeny of Acinetobacter strain and their related species based on 120 protein concentrates. The phylogenetic tree was constructed using all Acinetobacter strains from the NCBI RefSeq database, and this figure only show the subtree which is closely related to strain A1-4-2. The Acinetobacter clades are distinguished by different colors, with red representing Clade_I, blue representing Clade_II, orange representing Clade_III, green representing Clade_IV, and grey representing Clade_V. The nodes with black points mean the bootstrap values ≥  80

The environment distribution of A1-4-2-like strains

To study the environmental distribution of A1-4-2-like strains, we aligned the “core genome” of the Clade_I against NCBI nr database. This analysis identified a unique autotransporter beta-barrel protein (encoded by ABJ384_RS12135 in A1-4-2), which showed > 97% sequence identity and 100% coverage among the genomes of Clade_I, but was absent in other nr-indexed microbes (BLASTp: identity ≥ 50%, coverage ≥ 50%, e-value ≤ 1e-5), only except for Acinetobacter sp. CAAS 2–6 with 75.51% identity. Such specification enables it serve as a molecular marker for quickly searching A1-4-2-like genomes in metagenomics.

Then, we aligned such signature protein against MGnify and nr databases, as well as our in-house database which encompasses the protein sequences predicted from ~ 12,000 metagenomic assemblies (BLASTp parameters: identity 90%; coverage, 95%; e-value, 1e-30). A total of 15 metagenome assembled genomes (MAGs) were finally retrieved, all of which were confirmed as members of Clade_I in phylogeny (Supplementary Table S4 and Supplementary Figure S4).

The environmental distribution of Clade_I species is shown in Fig. 4. They are widely distributed in various environments, including sewage or bioreactor sludge (9), animal feces or the environment surrounding animal farms (4), the water of lake or river (5), reclaimed water (1), petroleum-contaminated wastewater (1), hospital wastewater (1), clayey mud of a drained pond (1), and soil (1). The diversity of environments inhabited by Clade_I members underscores their metabolic versatility, particularly in the degradation of organic matter and waste.

Fig. 4.

Environmental distribution of Acinetobacter Clade I. Pink represents strain A1-4-2, green represents isolated, and blue represents MAG

The metabolic characteristics of strain A1-4-2

We reconstructed the metabolic pathways to study the metabolic characteristics and ecological potentials of strain A1-4-2. We compared them with those of the other Acinetobacter strains in Clade_I (Fig. 5), which are detailed below.

Fig. 5.

Metabolic potentials of strain A-4-2. The black arrows represent metabolic pathways predicted in strain A1-4-2 and the red arrows highlight the pathways in strain A1-4-2 that are absent in the other Acinetobacter strain

The metabolism of amino acids

The COG category COG-E, which pertains to amino acid transport and metabolism, is the third most prevalent class in the genome of strain A1-4-2. Although strain A1-4-2 is predictively capable of synthesizing all 20 types of amino acids, it harbors 21 genes that encode transport systems for methionine, branched-chain amino acids, aromatic amino acids, lysine, proline, glutamate/aspartate, glutamine, serine/threonine, and S-methylmethionine (Supplementary Table S5). Furthermore, at least 14 genes of extracellular peptidases (signalP-fused) were identified in strain A1-4-2, which belong to families M16, M48, M23, M20, S8, S9, S11, and S41 (Supplementary Table S6). Our additional experiments confirm that strain A1-4-2 can utilize L-alanine (Supplementary Table S7) and grow with peptone as the sole carbon source (Fig. 6A). Collectively, these findings suggested that proteolysis might be a significant aspect of strain A1-4-2’s lifestyle.

Fig. 6.

Substrate utilization profile of Strain A1-4-2. A peptone; B D-serine; C D-alanine; D peanut oil; E flax oil; F sunflower oil; G glycerol; H 3-hydroxybenzoic acid; I sodium benzoate. In each subplot, the bottle on the left represents the control group without uninoculation. At the time of initial inoculation, there was no visible difference in the transparency of the medium compared to the control group

In addition, strain A1-4-2 predictively contains two copies of genes encoding D-serine/D-alanine/glycine permease, and can grow on D-serine or D-alanine as the sole carbon and nitrogen source (Fig. 6B, C and Supplementary Table S7). This suggests its potential to metabolize typically refractory organic compounds, such as components of bacterial cell walls.

The utilization of organic acids and plant oils

COG-I (lipid transport and metabolism) is the fourth most abundant COG category of strain A1-4-2. A total of 12 genes were predicted to have putative roles in the transport of a diverse array of organic acids, including acetate, malonate, C4-dicarboxylates, L-lactate, malate, shikimate, oxalate, and gamma-aminobutyric acid (Supplementary Table S5). Most of these transporters are conserved among members of Clade I. We further confirmed that it could metabolize acetate, propionic acid, butyrate, lactic acid, lactate, L-malic acid, methyl pyruvate and acetoacetic acid (Supplementary Table S7).

Besides, strain A1-4-2 contained four genes dedicated to the import of long-chain fatty acids, complemented by 24 genes related to fatty acid beta-oxidation (Supplementary Table S8), including two strain-specific genes (ABJ384_RS07050 and ABJ384_RS07050) encoding acyl-CoA dehydrogenase in strain A1-4-2 which are absent in other Clade I members. Furthermore, a gene cluster was identified on its plasmid_A, ranging from ABJ384_RS15535 to ABJ384_RS15560, which encodes 3-hydroxyisobutyryl-CoA hydrolase (EC:3.1.2.4), enoyl-CoA hydratase (EC:4.2.1.17), butyryl-CoA dehydrogenase (EC:1.3.8.1), acetyl-CoA synthetase (EC:6.2.1.1), 3-hydroxyisobutyrate dehydrogenase (EC:1.1.1.31), and CoA-acylating methylmalonate-semialdehyde dehydrogenase (EC:1.2.1.18 & 1.2.1.27). Additionally, an oleate hydratase gene (ABJ384_RS05410) was detected in strain A1-4-2, indicating its potential capability in utilizing unsaturated fatty acids (Supplementary Table S8)³⁸.

Interestingly, albeit strain A1-4-2 harbors a minimal number of carbohydrate kinases, it does possess a gene for glycerol kinase (ABJ384_RS03635), as well as an operon encoding two extracellular (signalP-fused) triacylglycerol lipases and one lipase secretion chaperone (from ABJ384_RS01635 to ABJ384_RS01645). These findings suggested the strain’s ability to hydrolyze extracellular triacylglycerols into fatty acids and glycerol. We further confirmed that strain A1-4-2 could grow with glycerol and many plant oils (including oils of peanut, sunflower, and flaxseed) as the sole carbon sources, and observed significant emulsification in culture media (Fig. 6D–G).

The utilization of n-alkanes

Strain A1-4-2 harbored the altL gene (ABJ384_RS03085), which encodes the transporter of a broad range of n-alkanes and fatty acids³⁹. This protein is highly conserved among Clade_I and Clade_II strains (with over 85% sequence identities), and exhibits a 78.0% identity with that of Acinetobacter venetianus RAG-1 (F959_RS14525), suggesting a fundamental role of n-alkanes utilization for Acinetobacter species.

Moreover, strain A1-4-2 contained the genes of non-heme diiron integral membrane n-alkane monooxygenases (alkB or alkM), alkB-dependent rubredoxin (alkG), and rubredoxin reductase (alkT), suggesting that it could metabolize middle-chain n-alkanes (Supplementary Table S8). It also contained three ladA encoding long-chain alkane-degrading monooxygenase, a flavin-binding monooxygenase responsible for transforming long-chain alkanes (C₁₀-C₃₆) into their corresponding primary alcohols⁴⁰. Additionally, it possessed two almA genes (ABJ384_RS08330 and ABJ384_RS02065), which are crucial for the initial degradation step of long-chain (C₂₈-C₃₂) and branched-chain alkanes^41,42.

Additionally, a group III alcohol dehydrogenase (encoded ABJ384_RS06095) was found in strain A1-4-2, which exhibited 92.31% sequence identity with that of A. venetianus RAG-1 (WP_004879683.1). This enzyme, characterized in strain RAG-1, has a broad substrate oxidation spectrum, including alkyl alcohols from C₁ to C₃₂ and isomeric alcohols (e.g., isopropanol, isobutanol, isoamyl alcohol, and propanetriol), and is also necessary for RAG-1’s growth with C₂₈ n-alkane^43,44. These pieces of evidence suggested that strain A1-4-2 could have a broad spectrum of n-alkane utilization, including medium-, long-, and branched-chain alkanes.

Utilization of lignin and aromatic monomers

Strain A1-4-2 harbored as many as 7 genes involved in transport of benzoate, including 3 MFS family benzoate transport protein (benK), 3 benzoate: H + symporter (benE) and 1 benzoate transport porin (benP). BenK and BenE are typically considered to be inner membrane proteins, and usually work in concert with outer membrane-specific porins, such as BenF, which facilitate the passage of benzoate through the outer membrane of the bacterium.

Moreover, strain A1-4-2 contained three gene clusters comprising a total of 43 genes related to the degradation of aromatic compounds (Supplementary Table S8). The first cluster, ranging from ABJ384_RS05470 to ABJ384_RS05380, encodes enzymes such as styrene monooxygenase, phenylpropionate dioxygenase and 4-hydroxyphenylacetate 3-hydroxylase, along with others involved in the oxidation of catechol and cis, cis-muconic acid (catABC) and phenylacetic acid (pcaDFJI). The second cluster, spanning from ABJ384_RS07480 to ABJ384_RS07545, contained the genes for degradation of benzoate, catechol and toluene, including benK, benE, catABC, pcaDFJI and benABCD. The third cluster included genes involved in 3-phenylpropionate degradation (hcaDBCFE), as well as genes of 2-hydroxymuconate semialdehyde hydrolase (dmpD), 4-hydroxy-2-oxovalerate aldolase (bphI), and acetaldehyde/propanal dehydrogenase (bphJ). The revelation of these genes suggested that the strain A1-4-2 could metabolize a wide range of aromatic compounds. Experimentally, we confirmed that it could grow with either 3-hydroxybenzoic acid or benzoate as the sole carbon source (Fig. 6H, I), positioning it as a strong candidate for bioremediation applications of aromatic compounds.

Given that aromatic monomers are intermediates in lignin breakdown, and considering that ABJ384_RS13945 in strain A1-4-2, which is predicted to encode dye-decolorizing peroxidase (DyP, EC: 1.11.1.19), a crucial enzyme in lignin depolymerization, we propose that strain A1-4-2 may possess the capability to degrade lignin. In order to confirm this hypothesis, we started the incubation using soluble lignin as the sole carbon source at 30℃, 120r/min for 36 h. We found that the OD (OD₆₀₀) increased from the initial 0.038 to 0.112, indicating that strain A1-4-2 was able to utilize lignin for growth. Moreover, since certain lignin peroxidases such as laccases, DyP, LiP, and MnP are known to decolorize methylene blue⁴⁵we incorporated it into LB agar and observed a distinct decolorized halo surrounding the colonies. Additionally, we noticed the emergence of a reddish-brown circle around the colonies of strain A1-4-2 on LB-agar plates supplemented with 0.5% guaiacol, which confirmed its production of laccases (Supplementary Figure S5). Collectively, these observations illustrated the ligninolytic potential of strain A1-4-2 and highlighted its prospective application in bioremediation of lignocellulosic wastes.

The sulfur metabolism

Strain A1-4-2 harbored genes for sulfate/thiosulfate ABC transporter (ssuABC), sulfite reductase (cysHIJ), and thiosulfate/3-mercaptopyruvate sulfurtransferase (EC:2.8.1.1, 2.8.1.2), suggesting that sulfite and thiosulfate could be important inorganic sulfur sources for A1-4-2. Furthermore, the genes encoding aliphatic sulfonate transporters and FMNH2-dependent alkanesulfonate monooxygenase suggests that strain A1-4-2 may have the capacity to metabolize alkanesulfonates (Supplementary Table S8). To verify this hypothesis, we cultivated strain A1-4-2 using either sodium octanesulfonate or sodium dodecylbenzenesulfonate as the sole sources of carbon and sulfur. Despite no visible medium turbidity or increase in optical density was observed, a white, spherical precipitate was formed at the bottom of each culture flask, with gradually increasing in size. Our subsequent analysis employing the SYTO 9/PI live/dead bacterial double staining method confirmed that the observed particulates were viable cells (Supplementary Figure S6). These findings illustrated that strain A1-4-2 could utilize alkanesulfonates, and form dense cell aggregation as a strategy to resist the effects of surfactants.

Discussion

Bioremediation of aromatic compounds

Aromatic compounds, known for their persistence and lipophilicity, are prone to bioaccumulation within the food chain, leading to significant environmental challenges. They have the potential to induce chronic toxicity, disrupt both aquatic and terrestrial ecosystems, and interfere with the normal functioning of various organisms. These impacts underscore the critical need for the degradation of aromatic compounds to alleviate their long-term ecological effects^46–48.

Strain A1-4-2 exhibits the metabolic capability to degrade a range of aromatic compounds. Interestingly, the benzoate transport genes (ABJ384_RS14745 and ABJ384_RS14750) are in close genomic proximity to the origin of replication (OriC), with a distance of mere 27 kb. This genomic arrangement may reflect an adaptive strategy that facilitates the strain’s ability to regulate the expression of catabolic pathways in concert with its DNA replication cycle, and thereby enhances its efficiency in degrading benzoate. Indeed, the initial optical density (OD₆₀₀) of the medium containing 1 mg/mL sodium benzoate as the sole carbon source was 0.041, exhibiting notable turbidity within just 6 h at room temperature, with the OD₆₀₀ increasing to 0.249 in 24 h. These pieces of evidence collectively highlighted its rapid response to and robust capability in benzoate degradation.

Notably, strain A1-4-2, unlike its phylogenetic relative, strain SA01, known for phenol degradation up to 1 g/L within 60 h at optimal conditions⁴⁹cannot use phenol for growth regardless of concentration. Our further genomic investigation indicated that the high-efficiency phenol degradation observed in strain SA01 is strain-specific, as it possesses the genes encoding phenol/toluene 2-monooxygenase subunits (dmpB/P/M/N), and are absent in other Clade_A strains. Conversely, strain A1-4-2 contains the genes involved in the degradation of styrene and phenylpropionate (ABJ384_RS05225 to ABJ384_RS05380), which are absent in strains SA01 and CS-2. These findings suggested that horizontal gene transfer could significantly influence the substrate specificity for Acinetobacter strains involved in the degradation of aromatic compounds, illustrating the dynamic nature of microbial adaptability in degrading environmental pollutants.

Bioremediation of n-alkanes and oils

In various natural ecosystems, n-alkanes and oils originate from a diverse array of sources, including plant waxes, fauna-derived excretions, metabolic byproducts from sedimentary microbial communities, petroleum seep effluents, marine phytoplankton lipids, submerged aquatic plants, and kitchen wastes^50,51. Despite their low aqueous solubility, these compounds are distinguished by their substantial energy content, making them an attractive substrate for numerous microorganisms skilled in their metabolism, including various Acinetobacter strains. For example, A. pittii SW-1 is capable of effectively metabolizing long-chain alkanes (C₁₈–C₃₆), achieving a degradation rate of up to 91.25% for C₂₀ ⁵². A. venetianus RAG-1 is known for its efficient degradation of three types of crude oils and its remarkable emulsification capabilities³⁹. Additionally, A. junii WCO-9 has been shown to degrade olive oil efficiently, with considerable enzymatic activity in the decomposition of p-nitrophenyl decanoate, reaching levels as high as 3000 U/L⁵³.

In this study, we identified that all members of Clade_I, II, III, IV and V possess comprehensive metabolic pathways for the breakdown of long-chain fatty acids and n-alkanes, suggesting that the ability to utilize these compounds may be a common characteristic of the Acinetobacter genus. Experimentally, we found that strain A1-4-2 can degrade crude oil, which is rich in long-chain hydrocarbons and has high viscosity, at room temperature without the addition of any external carbon sources. This cultivation process resulted in the near-complete degradation of crude oil, which accounted for approximately 5% of the total culture volume in 30 days (unpublished data). Such ability highlighted its robust metabolic capabilities, positioning it as a candidate for bioremediation in oil-contaminated environments.

Besides n-alkane, strain A1-4-2 is also capable of utilizing a variety of plant oils. A strain-specific gene cluster related to fatty acid oxidation, extending from ABJ384_RS15535 to ABJ384_RS15560, was identified on its plasmid_A (Supplementary Table S8). Given that strain A1-4-2 was isolated from a crab farming base, which is likely abundant in plant oils used as food additives, it is reasonable to suggest that these oils have played a role in maintaining the horizontally transferred gene cluster, and potentially enhanced the strain’s capacity for fatty acid utilization.

Additionally, we have observed that strain A1-4-2 possesses emulsifying capabilities against various plant oils (Fig. 6). Genomically, we identified an outer membrane protein A (OmpA, encoded by ABJ384_RS11760) with 84.96% sequence identity to that of strain SA01 (WP_166170865.1), which exhibits emulsification indices (E24) of 60% for toluene, 58% for soybean oil, 56.5% for sunflower oil, 42% for crude oil, 12% for n-hexane, and 8.33% for gasoline as reported⁵⁴. Actually, strain SA01’s OmpA proteins during biogenesis of outer membrane vesicles (OMVs) is posited to enhance its emulsifying activity in response to phenols⁵⁵. However, it remains to be explored whether the emulsifying effect of OmpA in strain A1-4-2 on plant oils also relies on the formation of OMVs and whether the bacterium secretes other emulsifying agents.

The antibiotic sensitivity and biosafety of strain A1-4-2

The genus Acinetobacter encompasses a broad array of species, each with distinct traits. On one hand, the diverse metabolic capabilities of Acinetobacter species suggest a wide range of potential applications in industry and environmental management. On the other hand, the clinical strains of this genus, such as A. baumannii, have been linked to numerous infections^56–58. Significantly, Acinetobacter sp. CS-2 was isolated from hospital wastewater and identified as a super antibiotic-resistant bacterium⁵⁹which shares a very close phylogenetic relationship with strain A1-4-2. Therefore, conducting a thorough safety assessment of strain A1-4-2 is crucial before its implementation in bioremediation efforts.

Our genomic comparison disclosed that strain CS-2 is endowed with multiple antibiotic resistance genes, encompassing the following: (a) a subclass B1 metallo-beta-lactamase NDM-1, designated as JFY49_RS16200, which is implicated in the hydrolysis of β-lactam antibiotics, including penicillins, cephalosporins, and carbapenems⁶⁰; (b) a bleomycin-binding protein Ble-MBL, annotated as JFY49_RS16205, which is associated with bleomycin resistance⁶¹; (c) a carbapenem-hydrolyzing class-D beta-lactamase, identified by the locus tag JFY49_RS17245, contributing to resistance against carbapenem antibiotics⁶²; (d) a chloramphenicol/florfenicol efflux protein, with the gene tag JFY49_RS17090, which is implicated in resistance to phenicol drugs such as chloramphenicol, florfenicol, and thiamphenicol⁶³; (e) a sulfonamide-resistant dihydropteroate synthase, encoded by the gene JFY49_RS17185, involved in resistance to sulfonamide drugs⁶⁴; (f) an Mph(E) family macrolide 2’-phosphotransferase, as indicated by JFY49_RS17105, which is involved in resistance to macrolide antibiotics by catalyzing the phosphorylation of the 2’-hydroxyl group on the macrolide sugar moiety⁶⁵; (g) an ABC-F type ribosomal protection protein Msr(E), tagged as JFY49_RS17110, which plays a role in resistance to macrolide, lincosamide, and streptogramin B group antibiotics by shielding the antibiotic binding site on the ribosome⁶⁶; and (h) a fosfomycin resistance glutathione transferase, with the gene identifier JFY49_RS16125, which is involved in fosfomycin resistance through the catalyzation of the opening of the oxirane ring⁶⁷. Conversely, these ARGs are completely absent in strain A1-4-2.

Moreover, unlike strain A1-4-2, which shows sequence uniformity in its 16 S rRNA genes, strain CS-2 exhibits four unique variants of the 16 S rRNA gene sequences. This genetic heterogeneity in strain CS-2 could conceivably enhance its defensive capabilities against antibiotics that interfere with ribosomal activity, such as aminoglycosides (e.g., gentamicin, streptomycin, neomycin) which bind to specific sites on the 16 S rRNA, disrupting the process of bacterial protein synthesis⁶⁸and tetracyclines (e.g., tetracycline, doxycycline) which bind to the 16 S rRNA on the 30 S ribosomal subunit, preventing the entry of aminoacyl-tRNA into the A site of the ribosome and thereby inhibiting protein synthesis⁶⁹. Conversely, strain A1-4-2 displayed sequence uniformity in its 16 S rRNA genes, suggesting a potentially reduced capacity for resistance to certain antibiotics.

Experimentally, strain A1-4-2 showed resistance (R) to 7 out of the 27 tested antibiotics, including kanamycin, erythromycin, lincomycin, and azithromycin. Conversely, it exhibited an intermediate sensitivity (I) to ceftazidime, polymyxin B, and piperacillin, and demonstrated sensitivity (S) to ampicillin, clindamycin, cefuroxime, levofloxacin, vancomycin, doxycycline, streptomycin, ciprofloxacin, florfenicol, cephalexin, cefazolin, ceftriaxone, norfloxacin, amikacin, fosfomycin, cefoperazone, and chloramphenicol (Table 2, Supplementary Figure S7). This indicates that strain A1-4-2 exhibits sensitivity to a broad spectrum of antibiotics, underscoring its potential for controlled use in environmental applications.

Regarding the biosafety of strain A1-4-2, we utilized the zebrafish model to assess its toxicity, and it did not result in obvious mortality of zebrafish after its introduction to the fish tank (Supplementary Table S9). This evidence suggests that, although Acinetobacter species are considered opportunistic pathogens, strain A1-4-2 could be safe for applications in bioremediation and aquaculture.

Genetic engineering potential of strain A1-4-2 for enhanced bioremediation and biosafety

We used Escherichia coli WM3064’s conjugative system and the pRE112 plasmid to create molecular vectors for genetic exchange with strain A1-4-2. This method proved effective for precise genome editing in strain A1-4-2, including the integration of foreign DNA and removal of native genes (data not shown). It opens the door for enhancing A1-4-2’s bioremediation capabilities by introducing enzymes that target organic pollutants and removing genes linked to pathogenicity or antibiotic resistance, thus improving its safety for ecological restoration in future.

Conclusion

In conclusion, our research presents Acinetobacter strain A1-4-2, derived from a freshwater crab habitat, potentially identifying it as a novel Acinetobacter species. Through the analysis of around 12,000 public datasets, we’ve discovered that this species is ubiquitous across various environments, particularly those with high organic content like sludge, feces, and wastewater. Strain A1-4-2 showcased impressive metabolic versatility, degrading an array of substrates such as amino acids, organic acids, oils, n-alkanes, lignin, and aromatic compounds with high efficiency. Our genomic and experimental data revealed that this strain is resistant to only a narrow range of antibiotics. The strain’s biosafety, validated through zebrafish toxicity tests, made it a suitable candidate for environmental application. Moreover, the genetic tractability of A1-4-2 made it a promising chassis cell for the enhanced biodegradation of organic pollutants through genetic engineering. This study elucidates the genomic and metabolic profile of A1-4-2, shedding light on the biodegradation potential of Acinetobacter and aiding in the formulation of strategies for environmental restoration amidst pollution challenges in the future.

Electronic supplementary material

Below is the link to the electronic supplementary material.

Author contributions

Conceptualization, J.W. and Z.S.; biological experiment, R.W., L.W. and Y.C.; data analysis, J.W.; writing—original draft preparation, J.W. and R.W.; writing—review and editing, J.W., R.W., Z.S., J.F., Y.W., H.L., Y.C., B.C.; project administration, Z.S.; funding acquisition, J.C., Z.S., and J.F. All authors have read and agreed to the published version of the manuscript.

Data availability

Data is provided within the manuscript or supplementary information files. The GenBank accession number for the complete genome sequence of strain A1-4-2 is GCA_040233925.1. The A1-4-2-like metagenome-assembled genomes (MAGs) in the current study have been deposited in eLMSG (an eLibrary of Microbial Systematics and Genomics, https://www.biosino.org/elmsg/index) under accession numbers LMSG_G000044973.1-LMSG_G000044989.1.

Declarations

Competing interests

The authors declare no competing interests.