Introduction

The development of antibiotic resistance in bacteria has increased greatly over time and poses a significant risk to public health. The abuse of antibiotics in human and veterinary medicine, agriculture, and poultry production contributes to the emergence of microorganisms resistant to antibiotics. In addition to being found frequently in healthcare facilities, multidrug-resistant bacteria are increasingly being identified in the community and surroundings environmental sources1,2,3.

Among antibiotic-resistant bacteria, Acinetobacter spp. have emerged as opportunistic pathogens often related to healthcare-associated infections4,5. However, diverse species of Acinetobacter have been isolated from different sources. Even though A. baumannii is unequivocally clinically and epidemiologically the most important Acinetobacter species, other Acinetobacter species have also been linked to human infections and found to be antibiotic-resistant and able to spread among hospitalized patients6,7,8. A study by Touchon et al.9 revealed that the Acinetobacter genus consists of isolates whose core DNA sequences are surprisingly variable and they identified a clade containing members with proteolytic or hemolytic activity. Seven of these members were named as species and six others as genomic species, including one named 13BJ/14TU9. Nemec et al.10 investigated the taxonomic status of 40 isolates of Acinetobacter and named an additional five species. In 2017, Nemec et al.11, investigated the genomic species 13BJ/14TU and found 24 strains with characteristic rpoB/gyrB sequences. These sequences had all been isolated from patients and had high levels of colistin (polymyxin E) resistance that is not observed in any other species within the hemolytic/proteolytic clade6,12. Due to the intrinsic resistance to colistin, Nemec et al.11 renamed the 13BJ/14TU genomic sequence isolate as Acinetobacter colistiniresistens. The genome assembly of the 13BJ/14TU isolate is available as GCF_003227755.1.

The genus Acinetobacter is a strictly aerobic, gram-negative coccobacilli with oxidase-negative and catalase-positive characteristics11. So far, the species A. colistiniresistens has been isolated only from clinical specimens, including sputum5, skin, blood13, vagina, eye, wound swab, catheter, conjunctiva and cerebrospinal fluid11. Its presence is typically linked to severe illnesses such as septicemia14,15. A. colistiniresistens type strain NIPH 2036T (genome assembly GCF_000413935.1) was isolated prior to 1990 from a catheter in Belgium13. There is no evidence of characterization of this species from environmental niches, animals or healthy individuals.

The emergence and prevalence of multidrug-resistant (MDR) Acinetobacter spp has led to the reintroduction of the polymyxin antibiotic colistin as a first-line therapy for such infections12,16. Consequently, colistin resistance in Acinetobacter, and other bacteria, has emerged worldwide, including in Malaysia, reducing the options for treating MDR pathogens. Two mechanisms for colistin-resistance have been described in A. baumannii: (1) The addition of phosphoethanolamine to the lipid A part of lipopolysaccharide (LPS), which is caused by mutations in the genes encoding signalling proteins PmrA and PmrB, and (2) the loss of LPS production, which is caused by mutations in the lpxA, lpxC and lpxD genes17. However, there are no detailed studies on the colistin resistance in A. colistiniresistens strains. Additionally, no virulence-related study has been reported yet on A. colistiniresistens.

In light of the growing clinical significance of A. colistiniresistens and its heightened antibiotic resistance, understanding its potential reservoirs and routes of exposure has become a pressing concern. However, the transmission and genotypic characteristics of A. colistiniresistens within the community remain poorly understood. Here we report the characterization of an A. colistiniresistens isolated from the feces of a healthy individual. To the best of our knowledge, this is the first study of A. colistiniresistens isolated from the community. We phenotypically and genotypically characterized the isolated A. colistiniresistens strain and conducted a comparative whole genome sequence analysis with other A. colistiniresistens strains curated from NCBI. The larvae of the greater wax moth, Galleria mellonella, a relatively simple, non-mammalian model, was used to explore the pathogenicity of the A. colistiniresistens strain18. The data generated from this study provides insights into the genetic diversity within A. colistiniresistens strains and highlights its potential threat to the community.

Methods

Ethics approval

The study was approved by the ethics committee/IRB of Monash University Human Research Ethics Committee (MUHREC, project number: 1516), which is in accordance with the WMA Declaration of Helsinki (WMA and World Medical Association 2013). Written informed consent was obtained from each participant involved in the study. In addition, this study was performed in collaboration with the Southeast Asia Community Observatory (SEACO) based in the Segamat District of the southern Johor state in Peninsular Malaysia.

Sample collection and Isolation of Acinetobacter spp

A single colistin-resistant Acinetobacter isolate was obtained after screening 233 fecal samples from Segamat for bacterial and fungal isolates in 2018. The isolate formed part of a larger cohort of Acinetobacter spp. isolates identified during a community research project that involved the isolation and investigation of ESKAPE pathogens from individuals living in the Segamat District19.

The sample collection and processing steps have been described earlier20. The samples were plated on Leeds Acinetobacter Agar (HiMedia, India) and MacConkey Agar (Oxoid, UK) and subsequently incubated at 37 °C for 24 h. Colony morphology and nature of the strains were observed and recorded. Three colonies with Acinetobacter morphology were selected from each sample and identified by standard biochemical methods (Gram stain, Catalase test and Oxidase reactions).

PCR amplification of a 16S rRNA gene fragment and subsequent sequencing was performed to confirm the Acinetobacter spp. The 16S rRNA gene was targeted using the universal primers described in previous studies21. Bacterial DNA extraction for PCR was carried out by the boiling extraction method described by Dashti et al.22.

Colistin and other antimicrobial susceptibility testing

Antimicrobial susceptibility testing and interpretation were performed using the standard disk diffusion method for 12 different antibiotics on Mueller Hinton agar (Oxoid, UK) according to the Clinical Laboratory Standards Institute (CLSI) guidelines23. The antibiotic disks used in this study were piperacillin, piperacillin-tazobactam, ceftazidime, cefepime, aztreonam, imipenem, meropenem, gentamicin, amikacin, ciprofloxacin and tetracycline.

However, broth microdilution was used for colistin and polymyxin B, the only method CLSI recommends. The minimum inhibitory concentration (MIC) for colistin and polymyxin B was performed following the CLSI, 2015 guidelines and observed breakpoints (≤ 2 mg/L, susceptible; > 4 mg/L, resistant). Acinetobacter baumannii ATCC BAA 1605 and E. coli ATCC 2325 were used as controls with known antibiotic resistance patterns.

Biofilm production and quantification assay

Biofilm production and quantification assays were performed according to Huet et al.20 with slight modifications. In brief, a total of 100 μl Tryptone Soya Broth (TSB) medium (Oxoid, UK) supplemented with 0.2% glucose was added to each well. Using an overnight bacterial culture, the cell suspension was adjusted to 0.5 McFarland standard in TSB supplemented with 0.2% glucose and 100 μl of each suspension was inoculated into each well. Two wells were left uninoculated and used as negative controls. The plates were incubated at 37 °C for 24 h for biofilm production. Following the biofilm production assay, biofilm quantification was carried out using Crystal violet (CV) and XTT assays.

DNA extraction and whole genome sequencing

To achieve the complete genome sequence, hybrid short and long-read based whole genome sequencing (WGS) was performed. Total genomic DNA was extracted using the phenol–chloroform phase-separation method, according to Sambrook & Russell24. Extracted DNA quality and concentration was assessed using a Nanodrop bioanalyzer spectrophotometer (Thermo Scientific, Delaware, USA).

The short-read sequencing data was generated with a Nextera XT library preparation kit (Illumina, San Diego, CA, USA) and sequencing was performed using an Illumina MiSeq sequencer with the MiSeq Reagent Kit v3 as per the manufacturer's protocol (2 × 250 bp paired-end read setting). Additionally, for long-read sequencing, DNA libraries were prepared according to the Ligation Sequencing Kit protocol (SQK-LSK109). The long-read sequencing data was then generated using a MinION FLO-MIN106 flow cell and a MinION MK1B sequencing device (Oxford Nanopore Technology).

De novo genome assembly and annotation

Quality trimming and filtering of the raw MiSeq short-reads were performed using Trimmomatic—0.39 version with the parameters PE, ILLUMINACLIP: adapters/NexteraPE.fa:2:30:10:8, LEADING:3, TRAILING:3, SLIDINGWINDOW:5:20, MINLEN:35 (Bolger et al.25). The short read draft genome was de novo assembled using SPAdes 3.13.026. For the long read, base-calling was conducted using Guppy v3.2.10 through MinKnow v3.6.17, using the fast base calling configuration. The long-read genome assembly was performed using Flye v2.727 and the sequence was later corrected and polished with the trimmed short-reads using Pilon v1.2328. The quality of the corrected assembly was evaluated using BUSCO v4.0.629. Functional annotation was performed using Prokka 1.1330 and the genome map was plotted using BLAST Ring Image Generator (BRIG) v0.9531. Plasmids were detected using Plasmid Seeker32.

Precise species identification

Species identification was carried out through the average nucleotide identity (ANI) based on BLAST and in silico DNA-DNA hybridization (isDDH) using the online server tool JSpeciesWS33 and genome-to-genome distance calculator34, respectively, with default parameters. An ANI value of more than 95% and isDDH values ≥ 70.0% were used as a cut-off to define bacterial species precisely. A phylogenomic analysis of closely related Acinetobacter spp. whole genome sequences was carried out using GToTree program v.1.7.0535. These sequences were retrieved from the National Center for Biotechnology Information (NCBI) based on the presence of single copy genes in each genome, including our isolate C-214.

Genome analysis

A comparative genomic analysis was carried out between our strain C-214 and the genome sequences of 20 other A. colistiniresistens strains obtained from NCBI (GCF_000248195.1, GCF_000369645.1, GCF_000369765.1, GCF_000413935.1, GCF_000876115.1, GCF_003227755.1, GCF_003569565.2, GCF_007713425.1, GCF_008982465.1, GCF_008984005.1, GCF_008987005.1, GCF_008988385.1, GCF_008990765.1, GCF_008992365.1, GCF_008993755.1, GCF_009013055.1, GCF_009013115.1, GCF_009013295.1 and GCF_900406805.1). Two A. baumannii strains (H-10112 and C-98) collected from the same location during the study were used to compare the sequence variation with A. colistiniresistens strains. The strain H-10112 was an MDR hospital strain, and C-98 was a non-MDR community strain36.

The comprehensive antibiotic resistance database (CARD)37 was used to identify acquired antibiotic-resistance genes using Abricate version 1.0.1(https://github.com/tseemann/abricate). Virulence-associated genes were identified using the virulence factor database (VFDB4)38. Mobile genetic elements were detected using ISFinder39. Gene content matrices were obtained using anvi’o40.

An ampC gene was detected in the genome of the C214 isolate. It resembles the UniRef90_N9PW73 cluster (UniRef50_A0A0N1I997 cluster at 50% cutoff), whose protein sequences belonged exclusively to A. colistiniresistens. The AmpC protein sequence from C214 was compiled together with protein sequences of the UniRef50_A0A0N1I997 cluster and Ambler class C beta-lactamases from the BLDB database41 to build a phylogenetic tree using FastTree42. The ampC gene tree was visualized using iTol v643.

In vivo Galleria mellonella killing assay

To determine the virulent nature of A. colistiniresistens, an in vivo killing assay was performed on the greater wax moth, Galleria mellonella. The G. mellonella larvae were purchased from Carolina Biological, US. Larvae showing symptoms of melanization or deformation were omitted from the assay to eliminate the potential for bias. Each larva was weighed and those meeting the criteria of 250 ± 50 mg were used in the study. Killing assay experiments were performed by injecting 10 µl of two different bacterial solutions with 107 and 106 colony-forming units per larva (CFU/larva), respectively, into the last left proleg using a Hamilton syringe. To check for death caused by physical damage, one group of larvae was injected with 10 µl of PBS as a negative control. Another control group did not receive an injection. The larvae were incubated for seven days at 37 °C and checked for symptoms of death every 24 h. Larva that did not respond to tactile stimulation or had a blackish discolouration were reported dead. A. baumannii C-98 and E. coli OP50 were selected as high and low-pathogenicity reference strains, respectively. The experiments were repeated three times, with the average reading taken into account.

Statistical data analysis

All analyses were performed using three separate experiments using GraphPad Prism software 6.01. The significance of differences was determined at p ≤ 0.05. The killing of G. mellonella by A. colistiniresistens was analyzed using the Kaplan–Meier method. Log-rank test was performed.

Results

Characteristics of the isolate

A study of A. baumannii from fecal samples from the community in Segamat district, Johor, Malaysia led to the isolation of a single colistin-resistant Acinetobacter spp. designated as C-214, on selective agar plates. The carrier was a 34-year-old female housewife from the indigenous Orang Asli Jakun community.

For preliminary species identification, PCR was done with the universal primers 27F and 1492R followed by Sanger sequencing to get the nearly complete 16S rRNA gene sequence of the strain36. The use of BLAST for the 16S rRNA sequence against the NCBI database revealed that the isolate belonged to the genus Acinetobacter and is a member of the species colistiniresistens.

FE-SEM imaging was carried out, and the colony morphology was compared to explore any differences in bacterial cell morphology between A. baumannii and A. colistiniresistens. No significant difference was observed in their cell membrane and colony formation. Both were found to have coccobacillus phenotypes. On selective Leeds Acinetobacter agar media, they produced identical colonies and colours (data not shown).

Antibiotic resistance profile and biofilm-forming ability

The isolate, C-214, had a colistin and polymyxin B MIC of 32 and 8 µg/ml, respectively (Table 1). In addition to colistin resistance, this isolate was phenotypically resistant to cefotaxime, amikacin and tetracycline but susceptible to cefepime, ceftazidime, ciprofloxacin, gentamicin, piperacillin/tazobactam and carbapenems based on the disk diffusion antibiotic susceptibility testing (Table 1).

Table 1 Antibiotic resistance profile and biofilm formation of strain C-214.
Full size table

The biofilm-forming ability of C-214 was also assessed. Based on XTT and CV assays, isolate C-214 showed moderate biofilm-forming capability (Table 1).

Genomic features of the strain C-214

C-214 was sequenced using both short-reads (Illumina MiSeq) and long-read (Oxford Nanopore) sequencing technologies. Hybrid genome assembly revealed that the Acinetobacter strain C-214 contained one circular chromosome of 3,865,171 bp (GC content 41.33%) (Fig. 1). The GC content is almost identical to that reported for the A. colistiniresistens sequence represented by GCF_003227755.1 and the genome size is typical for this strain. Besides, three circular plasmids were also detected with sizes of 10,411 bp (p214-1), 5509 bp (p214-2), 8305 bp (p214-3) and GC content of 35.4, 30.79 and 33.65%, respectively (Supplementary B). The average genome size is similar to A. baumannii, whose genomes range between 3.6 and 4 Mbp in size with a GC content of around 39%44,45.

Figure 1
figure 1

Genome map of Acinetobacter colistiniresistens strain C-214 chromosome (CP102099) plotted using BLAST ring image generator (BRIG). The outer coloured circle denotes the GC skew of genomic sequences (purple: negative; green: positive), followed by distributions of antimicrobial resistance genes (red), virulence genes (blue) and CRISPR-Cas system loci (grey).

Full size image

The genome annotation analysis using Prokka detected 3844 genes with 3,705 coding sequences, 75 tRNA sequences, 18 rRNA sequences, 1 tmRNA sequence and 45 misc RNA sequences.

Precise species detection

ANI and in silico DNA-DNA hybridization analyses of strain C-214 were conducted against 79 different Acinetobacter spp along with 19 A. colistiniresistens strains (Supplementary B). The highest ANI (98.08%) and DDH (71.04) values were found against A. colistiniresistens strain NR1165 (Supplementary B). A phylogenetic tree was constructed on the basis of 20 A. colistiniresistens, three A. baumannii, one A. gyllenbergii and one A. proteolyticus genomes, and it is clear that C-214 is a genomovar within the A. colistiniresistens group which is itself composed of two sub-groups (Fig. 2).

Figure 2
figure 2

The phylogenomic tree of Acinetobacter colistiniresistens, annotated with ANI values and percentage identity matrix against antimicrobial resistance genes from CARD. Note that the type strain NIPH2036 is represented by the genome assembly GCF_000413935.1, the 13BJ/14TU isolate by GCF_003227755.1 and three TUM isolates by GCF_009013115.1, GCF_9013295.1 and GCF_9013055.1.

Full size image

Presence of antibiotic resistance genes in C-214 compared with other A. colistiniresistens strains

The distinct patterns of antibiotic resistance phenotypes observed in strain C-214 prompted us to investigate the known genes linked to resistance in the sequenced genome and compare them with other available A. colistiniresistens genomes. In addition, we also compared the presence of AMR genes in three A. baumannii, one A. gyllenbergii and one A. proteolyticus genomes. Detailed results are summarized in Fig. 2. The AMR gene analysis using the CARD database detected five antibiotic resistance genes in the genome of strain C-214, where one tetracycline resistance gene (tet39) was found within a plasmid (p214-1). Besides, one beta-lactam resistance gene blaOXA302, two aminoglycoside resistance genes (ANT(3’’)-IIc, AAC(6’)-Ij) and multidrug efflux pump gene adeB were also present in the isolate. Detection of these genes also supported our AST phenotypic data (Table 1), where the strain C-214 showed resistance against beta-lactams (CTX-30), tetracycline (TE30) and aminoglycoside (AK30). However, even though the isolate is highly resistant against colistin and polymyxin B, the gene responsible for resistance could not be identified. Studies have found that the plasmid-mediated gene mcr is responsible for colistin resistance in Acinetobacter baumannii46. We could not find any mcr gene in any A. colistiniresistens isolates.

After comparison of the putative resistance genes in C-214 with 19 other A. colistiniresistens strains, it was discovered that most of the isolates carried a similar collection of resistance genes (n ≤ 5). However, the tetracycline resistance A gene (tet39) was only detected in the strain, C-214, described in this paper. An insertion sequence ISaba26 was detected in both the chromosome and plasmid (p214-1). In addition, when Acinetobacter derived cephalosporinases (ADCs) were compared between A. colistiniresistens isolates and A. baumannii, considerable differences were observed. Both A. baumannii and the A. colistiniresistens isolates carried the class C intrinsic beta-lactamase gene. However, whereas the A. baumannii isolates analyzed in this study carried an ADC-1 type of gene, the A. colistiniresistens carried an ADC-8 type gene (Fig. 3). The amino acid sequence similarity of these two types is about 50%. Efflux pump related genes were commonly seen in both A. baumannii strains, but only two genes were found in the A. colistiniresistens isolates.

Figure 3
figure 3

Phylogenetic tree based on curated Ambler class C beta lactamases from BLDB and UniRef90_N9PW73 and UniRef50_A0A0M1I997 sequences. AmpC sequence from C-214 clusters together with UniRef90_N9PW73 sequences. In green are the Acinetobacter derived cephalosporinases (ADCs) except ADC-8. The brown group is a clade consisting of UniRef50_A0A0M1I997 sequences. UniRef50 sequences are sequences that form a 50% similarity cluster (50% is supposedly a lenient cutoff here). We can also see that the UniRef50_A0A0M1I997 cluster form two subclades. The C-214 ampC gene (UniRef90_N9PW73 "sub-"subclade, red) does not fall under the ADC-8 subclade. Here, the three-letter names refer to different beta-lactamase’s which are all class C.

Full size image

Virulence factor related genes observed in A. colistiniresistens

Different virulence factor related genes were analyzed in the C-214 strain, along with 19 other A. colistiniresistens, three A. baumannii, one A. gyllenbergii and one A. proteolyticus genomes derived from NCBI, where two A. baumannii genomes from the same project. The results are summarized in Fig. 4. The outer membrane protein gene ompA, which promotes bacterial biofilm formation, eukaryotic cell infection, antibiotic resistance, and immunomodulation, was found in all A. colistiniresistens isolates, including C-21447. Genes related to lipopolysaccharide (LPS) production, such as lpxA, B, C, D and lpxL, were present in all A. colistiniresistens genomes. The lpxA, lpxC, and lpxD genes are primarily involved in the initial stages of lipid A production and the hydrophobic anchor of LPS48. It has been found that mutations in lpxA, lpxC and lpxD may play a role in the development of colistin resistance48. We compared these lpx genes acquired from WGS of all A. colistiniresistens and three A. baumannii colistin-sensitive strains (A. baumannii ATCC19606 type strain, A. baumannii H-10112, A. baumannii C-98) (Supplementary A, Fig. S1). We found similar polymorphisms in lpxA/C/D and lpxL genes in all A. colistiniresistens strains suggesting that alterations in LPS metabolism could be the reason for colistin resistance observed in these strains. We also found other virulence factor related genes including; type VI secretion system, stress adaptation, antiphagocytosis, two-component regulatory system (bfmR, bfmS), serum resistance, iron uptake and adherence genes. While most isolates shared similar virulence-related genes (Fig. 4), certain differences were noted between A. colistiniresistens and A. baumannii isolates. Although both hospital and community A. baumannii isolates possessed a complete set of type VI secretion system genes, only one to three genes encoding this system were found in six A. colistiniresistens isolates, including C-214. Type VI secretion system (T6SS) genes are well recognized as a crucial virulence factor in A. baumannii and toxins produced by T6SS genes could kill other bacteria as well as eukaryotic cells49.

Figure 4
figure 4

Presence of genes involved in virulence in the strain C-214 and 19 other A. colistiniresistens, three A. baumannii, one A. gyllenbergii and one A. proteolyticus genomes. The presence of genes in an isolate is specified by a colored rectangle, colored according to sequence similarity to curated VFDB sequences. The absence of genes is shown as blank spaces with no color.

Full size image

Pangenome analysis

Extensive pan-genome analysis investigations can aid in understanding a bacterial species' functional adaptability50. To get insight into the pan-genome information of A. colistiniresistens, we created different plots to visualize the number of total genes, core accessory genes, and unique genes as a function of the sequenced genomes.

Acinetobacter colistiniresistens is classified into two genomovars. To explore the genomic diversity, we performed a pan-genome analysis using the Anvi’o pan-genome workflow40 (Fig. 5). The pan-genome consisted of a total of 6825 genes, with 2601 core genes shared among all strains and 2179 core genes occurring only once. Furthermore, we investigated the enrichment between the two genomovars51. The enrichment was defined as an enrichment score greater than 15 and an adjusted q-value below 0.01. Functional enrichment analysis was conducted based on COG20 annotation (Supplementary B). Our strain, C-214, was found in clade 2, which exhibited enrichment in 17 genes, while clade 1 showed enrichment in eight genes. The genomes were arranged according to the phylogenomic tree (Fig. 5).

Figure 5
figure 5

Anvi’o pangenome display of 20 A. colistiniresistens genomes. Layers were coloured according to the two proposed A. colistiniresistens genomovars. Genomes were sorted based on the phylogenomic tree from Fig. 2. Gene clusters were sorted based on the presence and absence of genes. Gene clusters that fall under the single copy genes and the core genome were indicated. Barplots representing singleton gene clusters per each genome were also shown.

Full size image

Isolate pathogenicity

The pathogenicity of A. colistiniresistens C-214 was tested in the G. mellonella model and compared using a virulent A. baumannii strain C-98 (unpublished) and a non-virulent E. coli OP-50. Figure 6 highlights the variability of pathogenicity in two different bacterial inoculums.

Figure 6
figure 6

Kaplan–Meier survival distributions for dose-dependent challenges of A. colistiniresistens C-214. A. baumannii C-98 and E. coli OP50 were used as highly virulent and non-virulent strain controls, respectively, at all dosage levels. Three biological repeats of each experiment were pooled and results are shown as a percentage probability of survival. Infection results for all three isolates tested were significantly different (p = 0.001; Mantel–Cox log-rank test, demonstrating that larval survival is dependent on the quantity of bacteria injected.

Full size image

We administered parenteral injections of two different concentrations (107 and 106 CFU) of A. colistiniresistens strain C-214 to examine the impact on larval pathogenicity. The infected larvae exhibited distinct symptoms, including nodulation, blackening of the cuticle, and eventual mortality. Notably, the degree of melanization increased significantly with higher inoculum doses, indicating that the initial infectious inoculum size plays a crucial role in the progression of the infection. To analyze the survival outcomes, we employed Kaplan–Meier survival distributions for each bacterial inoculum and conducted a log-rank (Mantel–Cox) test, which revealed significant differences (p < 0.001). The survival probability of the larvae depended on the number of CFU injected. For larvae injected with an inoculum size of 107 CFU/larvae, the survival rate after 24 h was 40% for A. colistiniresistens C-214, 0% for A. baumannii C-98, and 100% for the non-virulent E. coli strain OP50. However, the survival rate for C-214-treated larvae decreased to 0% after 120 h. In comparison, larvae injected with 106 CFU/larva exhibited a 90% survival rate for C-214 after 24 h, 20% for C-98, and 100% for OP-50. Within the observation period of 168 h post-inoculation, 60% of the C-214-treated larvae survived. Further dilution (105 CFU/larva) resulted in 100% survival but showed melanization in 30% of the population after 168 h of observation.

Discussion

MDR Acinetobacter baumannii is a significant nosocomial pathogen that has been the primary focus of most research on Acinetobacter spp. Aside from A. baumannii, little is known about other Acinetobacter species. However, non-baumannii Acinetobacter species are increasingly being identified as causative agents of nosocomial infections. One such organism, A. colistiniresistens has been isolated from different sources, including sputum, blood, wound swab, catheter and conjunctiva in hospitalized patients5,6,52 (Supplementary B). However, the genomic features of this organism have been rarely discussed11,52. In this study, we focused on characterizing the A. colistiniresistens strain isolated from a healthy human fecal sample. This report represents a significant finding as it is the first known instance of this species being isolated from a healthy individual. The study offers new information on the genomic and virulence traits of A. colistiniresistens, which could be useful in treating this particular pathogen.

Treating infections caused by Acinetobacter spp. has become increasingly challenging due to their multidrug-resistance (MDR) and pan-drug-resistance (PDR) patterns. Colistin is commonly regarded as a last-resort antibiotic against MDR Acinetobacter infections. Hence, the presence of A. colistiniresistens within a healthy community, naturally resistant to colistin11,12, raises significant public health concerns. The C-214 isolate displayed high resistance to both colistin and polymyxin B, with minimum inhibitory concentrations (MIC) of 32 and 16, respectively. This resistance profile is comparable to that of the NR1165 isolate reported in a study conducted in Japan5, further exacerbating concerns regarding the spread of such resistant strains.

Baraka et al.53, identified resistance genes against sulfonamides, macrolides, ABC-F, and beta-lactamases antibiotics in previously isolated A. colistiniresistens. Similarly, WGS of A. colistiniresistens strain C-214 revealed several resistance genes, including beta-lactam resistance gene blaOXA302, tetracycline resistance tet39, aminoglycoside resistance ANT(3'')-IIc, AAC(6')-Ij which also supported our AST data. The presence of plasmids carrying genes such as tet39 makes this strain more threatening to the community as this might enable the strain to confer resistance genes to other species through horizontal gene transfer54.

A comparative WGS study with nineteen other A. colistiniresistens retrieved from NCBI revealed that most of the A. colistiniresistens isolates harboured similar types of AMR genes except two strains (NR1165, DL) carrying more AMR genes, including carbapenemase genes coding for OXA-58, IMP-34 and ESBL gene coding for TEM-181.

Though the A. colistiniresistens isolates shared some virulence properties of A. baumannii, there were some noticeable differences. For instance, the type VI secretion system gene number was lower in the A. colistiniresistens isolates. Though A. baumannii and A. colistiniresistens carry different subclasses of type VI secretion related genes, the number of genes lost in A. colistiniresistens could be of vital importance. Additionally, it has recently been shown that silencing the chromosomally encoded type VI secretion system is crucial for horizontal gene transfer by conjugation, which is essential for disseminating antibiotic resistance55. As such, the type VI secretion system in A. colistiniresistens warrants further investigation for its virulence and resistance properties. It is to be noted that although the individual who carried the A. colistiniresistens isolated in this study did not carry A. baumannii, such an occurrence is a possibility in the future. In such a situation, the transfer of additional resistances into A. baumannii or vice versa could lead to the organism being resistant to all currently used drugs. Consequently, infections by organisms like this would become difficult to treat.

Recently, non-animal in vivo models like G. mellonella have been used to determine the virulence of pathogens such A. baumannii, P. aeruginosa, Burkholderia cepacia, Bacillus cereus, and disease-causing fungi18,56. G. mellonella can tolerate incubation temperatures up to 37 °C, making it preferable for researching human illnesses56. It also breeds quickly and does not require animal ethics clearance. In our study, G. mellonella exhibited dose-dependently sensitivity to A. colistiniresistens (C-214) infection and might be utilized to research its pathogenicity.

In conclusion this study describes the first complete genome sequence of A. colistiniresistens strain isolated from the fecal sample of a healthy adult female individual from Malaysia. Salient genomic features of this strain included the presence of genes relevant to AMR and virulence. MDR A. colistiniresistens is an opportunistic pathogen and is naturally resistant to colistin, which is of great concern as it is an antibiotic of last resort. Further, In vivo G. mellonella killing assay indicated the pathogenic potential of the strain C-214. Carriage of A. colistiniresistens in the asymptomatic community poses a risk to public health, and more attention should be paid to epidemiological surveillance and transmission of this bacteria.