Efflux-dominated resistome architecture and phylogeny-resistance decoupling in pharmaceutical effluent-associated Acinetobacter isolates

1 Introduction

The excessive and inappropriate use of antibiotics in clinical settings, veterinary medicine, and agricultural practices has significantly accelerated the development and spread of Antimicrobial resistance (AMR) worldwide in the 21st Century (La Rosa et al., 2025). AMR develops when microorganisms, including bacteria, fungi, and parasites, gradually acquire the ability to survive exposure to drugs that were once effective against them. In essence, these microbes adapt in ways that render standard treatments less effective or even ineffective (Prestinaci et al., 2015). It has been reported that 2050. AMR is projected to cause up to 10 million deaths annually by 2050 under worst-case scenarios, although recent estimates suggest a range of outcomes depending on intervention strategies (Mirzai et al., 2020). This process is fundamentally evolutionary. Under normal environmental conditions, microorganisms are continuously exposed to various stressors. When antimicrobial agents are present, they create selective pressure, allowing only those microbes with resistant traits to survive and multiply (Ye et al., 2025). Over time, these resistant populations become more dominant, making infections increasingly difficult to treat (Karkman et al., 2019). Wastewater systems have an important part in this dynamic. Effluents discharged from hospitals, pharmaceutical industries, agricultural fields, and livestock operations often contain antibiotic residues and heavy metals, accelerating the spread of multidrug-resistant (MDR) microorganisms (Iyer and Murugan, 2025; Alem et al., 2025; Ye et al., 2025). Resistance genes now considered clinically important were not originally generated in hospital settings; instead, they were mobilised from environmental bacterial communities through HGT, involving mobile genetic elements like plasmids, integrons, and transposons (Chu et al., 2026; Westphal-Settele et al., 2018). Comparative genomic studies indicate that pathogenic bacteria have integrated these resistance determinants from the environmental matrix, especially from soil-dwelling, antibiotic-producing organisms that have maintained such defence mechanisms as part of their natural evolutionary repertoire for millennia (Alem et al., 2025). The most infamous bacterium, Acinetobacter baumanii, tops the list on the World Health Organization’s (WHO) critical priority pathogen list as a superbug for its remarkable capacity to thrive in hospital environments, while rapidly acquiring and accumulating diverse AMR mechanisms (Singh et al., 2013).

In this context, the present study undertakes a comprehensive whole-genome–based investigation of Acinetobacter isolates recovered from pharmaceutical effluent–associated environments to characterise their resistome architecture, distinguish core and accessory resistance determinants, and examine their phylogenetic relationships. By integrating genomic assembly, AMR gene profiling, efflux system analysis, and 16S rRNA–based phylogeny, this work aims to elucidate how environmental selective pressures shape multidrug resistance patterns and to clarify the evolutionary and (HGT) dynamics driving resistance dissemination in non-clinical ecosystems. While environmental reservoirs and HGT) are widely recognised as major drivers of AMR dissemination, the present study is observational and does not directly establish the origin or directionality of resistance gene transfer.

2 Materials and methods

2.1 Data retrieval and genome selection

The Acinetobacter isolates whole-genome sequencing dataset from the pharmaceutical waste-associated environment were retrieved from the publicly available NCBI Sequence Read Archive (SRA) with the accession numbers SRR32616544, SRR32616545, SRR32616546, SRR32616548, SRR36476153, SRR36486889, SRR36487308, and SRR36487335 based on their association with pharmaceutical effluent environments and the availability of high-quality sequencing data. The datasets were selected to enable high-resolution comparative resistome architecture analysis while maintaining stringent assembly quality and consistent environmental origin, environmental origin and relevance to MDR; for this, the raw paired-end FASTQ files were downloaded using the SRA toolkit (v3.3.0) (Din et al., 2025). Although the sample size is modest, it enables detailed comparative resistome analysis under controlled selection criteria.

2.2 Quality control and read pre-processing

The raw data were analysed with FastQC v0.11.9 for nits per-base quality, GC content, adapter contamination, and sequence duplication following the Chen et al. (2018) method. The low-quality reads and residual adapters were trimmed using Trimmomatic v0.39 with parameters SLIDINGWINDOW:4:20 and MINLEN:50 to ensure reliable assembly, and only reads meeting quality thresholds were retained for genome assembly.

2.3 Genome assembly and assembly assessment

The genome assembly was carried out for the high-quality reads using the SPAdes v3.15.5 with k-mer sizes 21, 33, 55, and 77 in careful mode, a short-read genome assembler. Following Xiang et al. (2023) method, the contigs were filtered to remove sequences below predefined length and coverage thresholds. The genome assembly quality was assessed using the total assembly size, contig count, N50, and GC content. Assembly quality was further evaluated based on contiguity metrics; however, completeness and contamination estimates using tools such as BUSCO or CheckM were not performed, representing a limitation.

2.4 Identification of antimicrobial resistance genes

The Comprehensive Antibiotic Resistance Database (CARD) was used as a reference for the identification of the AMR genes using ABRicate v1.0.1. The assembled genome hits were filtered using CARD-recommended percentage identity and coverage thresholds of ≥80% identity and ≥60% coverage. The AMR genes for each isolate were compiled and merged into a single presence-absence matrix across isolates, which was then converted into a binary resistome profile for the comparative analyses. For multiple hits, the highest-scoring match per locus was retained. While other databases, such as ResFinder and ARG-ANNOT, provide complementary insights, integration of multiple databases was beyond the scope of the present study.

2.5 Core and accessory resistome analysis and 16S rRNA gene extraction and alignment

The AMR genes were grouped into a core and accessory resistomes, which determined the conserved resistance determinants from isolate-specific adaptations in the environmental Acinetobacter population. Core resistome genes were defined as those present in 100% of isolates, while accessory genes were present in a subset of isolates. Alongside the 16s rRNA gene sequences were extracted from assembled genomes using Barrnap v0.9 and aligned using MUSCLE v3.8, and the high-quality sequences were pooled and aligned with a multiple sequence alignment method optimised for bacterial rRNA genes.

2.6 Phylogenetic tree construction

FastTree (v2.1.11) was used under the GTR+CAT model with a nucleotide substitution model and SH-like local support values for the preparation of phylogenetic relationships. The resulting Newick-formatted tree displayed an evolutionary relationship based on 16s rRNA sequences. The integration of the phylogenetic data and resistome profiles was done via internal node identifiers, via systematic replacements with the corresponding SRA isolate IDs. Additionally, the phylogenetic trees and AMR gene matrices were visualised in iTOL, and the binary heatmap annotation was generated to map AMR gene distribution directly onto the tree for the direct comparison of the phylogenetic clustering with the resistome composition across isolates.

2.7 Statistical analysis

To quantitatively assess the relationship between phylogenetic relatedness and resistome composition, a Mantel test was performed between pairwise phylogenetic distances and resistome dissimilarity matrices. Phylogenetic distances were calculated from the 16S rRNA-based maximum-likelihood tree using patristic distances. Resistome dissimilarity between isolates was computed from the binary AMR presence–absence matrix using Jaccard distance. The Mantel test was implemented using the scikit-bio framework with 999 permutations to evaluate statistical significance. The correlation coefficient (r) and associated p-value were used to determine the strength and significance of the relationship between phylogeny and resistome composition.

3 Results

3.1 Genome assembly and quality assessment

The whole genome sequencing data of Acinetobacter isolates from the pharmaceutical effluent-associated environment were assembled, and antimicrobial resistance gene counts for all isolates are summarised in Table 1. A high GC content was prevalent across all assemblies, with a consistent genome size ranging from 3.72 to 4.35 Mb, which supports the taxonomic consistency among the isolates. With the contig numbers ranging from 51 to 1, 255, the assembly contiguity varied across the datasets, reflecting the differences in the sequencing depth and read quality. Even with variability, all assemblies showed sufficient coverage and structural integrity for downstream comparative genomic analysis. The large-scale contamination or excessive fragmentation was not detected, which indicated that the genome assemblies were suitable for comprehensive resistome profiling.

Table 1. Genome assembly statistics of Acinetobacter isolates.

3.2 Identification and diversity of antimicrobial resistance genes

ABRicate and CARD databases were used to identify the AMR genes, which displayed highly diverse resistance across all Acinetobacter isolates (given inTable 2). AMR gene counts ranged from 13 to 35 per isolate (mean: 22.9), indicating substantial inter-strain variability despite shared environmental origin. Multiple antibiotic classes, such as β-lactams, aminoglycosides, macrolides, tetracyclines, and sulfonamides, displayed resistance determinants, alongside OXA-type, which are clinically significant β-lactamases, were widely detected. Similarly, the OXA-23, a globally disseminated carbapenemase, and OXA-66, an intrinsic oxacillinase, were co-detected in multiple isolates along with other oxacillinase variants, such as OXA-144, OXA-1010, and OXA-1166, which were detected sporadically, contributing to strain-specific resistance profiles. Also, it was observed that aminoglycoside resistance was common, mediated by multiple modifying enzymes such as ANT(3ʺ)-IIa, APH(3ʺ)-Ib, APH(3ʹ)-Ia, and APH(6)-Id, while high-level resistance was observed in one isolate carrying the armA 16S rRNA methyltransferase. Macrolide resistance genes (mphE and msrE) and tetracycline resistance via tet(B) were consistently detected among several isolates. The presence-absence heatmap (Figures 1, 2) illustrates the heterogeneous yet overlapping distribution of these resistance determinants across genomes (Figures 2, 3).

Table 2. Distribution of antimicrobial resistance (AMR) genes across Acinetobacter isolates recovered from pharmaceutical effluent-associated environments.

3.3 Efflux pump-dominated resistance architecture

In the dominant resistance mechanism across all Acinetobacter genomes, the efflux-mediated resistance was analysed, which resulted in multiple members of the Resistance-Nodulation-Division (RND) efflux family (as shown inTable 3), forming the backbone of the resistome. It was observed that the AdeIJK efflux system (adeI–adeJ–adeK) was universally present and constituted a core resistance feature for the broad-spectrum resistance to β-lactams, fluoroquinolones, macrolides, tetracyclines, and phenicols. The tight transcriptional regulation of efflux activity was also suggested via adeN. The AdeABC system, which is primarily associated with aminoglycoside and fluoroquinolone resistance, was present in most isolates, along with its two-component regulatory system (adeR-adeS), indicating the potential for inducible efflux under antibiotic stress. Also, the AdeFGH system showed variable distribution, functioning as a core-accessory hybrid; the multidrug transporters such as AbeM (MATE family) and AbeS (SMR family) further expanded the efflux repertoire. Collectively, the widespread presence of both structural and regulatory efflux components indicates a resistance that optimised the multidrug extrusion in chemically polluted environments.

Table 3. Core efflux pump–mediated resistance determinants identified in Acinetobacter isolates.

3.4 Core and accessory resistome analysis

Core resistome genes were consistently present across all isolates, whereas accessory genes exhibited variable distribution (as shown inTable 4), which was dominated by the efflux pump genes and their regulators, including AdeABC, AdeIJK, AdeFGH, AbeM and multiple MFS transporters across most isolates, indicating intrinsic resistance traits under sustained environmental selective pressure. Consequently, the accessory resistome comprised strain-specific resistance determinants with variable distribution across isolates, including carbapenemases (OXA-23, OXA-144, OXA-1010, OXA-1166), extended-spectrum β-lactamases (PER-1, TEM-12), aminoglycoside-modifying enzymes, sulfonamide resistance genes (sul1, sul2), and disinfectant resistance (qacEΔ1). The uneven distribution of these genes accounted for the observed variability in total AMR gene counts and suggests ongoing horizontal gene acquisition and loss within the effluent-associated microbial community.

Table 4. Core and accessory antimicrobial resistance genes.

3.5 16S rRNA-based phylogenetic analysis and phylogeny-resistome relationship

The 16s rRNA sequences extracted from assembled genomes were used to infer phylogenetic relationships using FastTree (as shown in Figure 4). The resulting maximum-likelihood tree demonstrated clear clustering of isolates with close relatedness among several strains despite differences in resistome composition. Phylogenetic reconstruction confirmed that isolates with similar resistance profiles were not always the most closely related, indicating that AMR gene content does not strictly follow phylogenetic lineage. Integration of phylogenetic data with AMR profiles revealed a decoupling between evolutionary relatedness and resistance gene content. While core efflux systems were conserved across the phylogeny, accessory resistance genes showed no strict phylogenetic clustering. This pattern highlights that HGT and local selective pressure, rather than vertical inheritance alone, play dominant roles in shaping MDR profiles in pharmaceutical effluent–associated Acinetobacter populations. To statistically validate the observed phylogeny–resistome decoupling, a Mantel test was performed comparing phylogenetic distance and resistome dissimilarity matrices. The analysis revealed a weak negative correlation (r = −0.1228) with no statistical significance (p = 0.678), indicating that resistome composition is not associated with phylogenetic relatedness among the isolates. These results quantitatively support the observed lack of congruence between evolutionary relationships and AMR gene profiles (as shown in Figure 5).

4 Discussion

The resistome architecture of Acinetobacter isolates recovered from a pharmaceutical effluent-associated environment highlights the combined impact of intrinsic resistance mechanisms and horizontally acquired determinants in shaping MDR phenotypes. The whole-genomic data of the Acinetobacter genus displayed GC content and genome size ranging from 3.72- 4.35 Mb, which is similar to a study done by Zhu et al. (2020). The variability in the contig was likely due to the biological divergence, as reported by Jenke et al. (2025), caused by differences in sequencing depth and read quality. It is also important to note that variability in assembly contiguity may influence downstream AMR gene detection. Fragmented assemblies can result in partial or split gene sequences, particularly for plasmid-associated or repetitive regions, potentially leading to underestimation of resistance gene content in some isolates. Significantly, the genome sizes across the isolates were comparable, suggesting that taxonomic consistency while permitting sufficient accessory genome flexibility (Mishra et al., 2025). The continuous exposure to sub-inhibitory antibiotic concentrations, heavy metals, disinfectants, and other xenobiotics is the major characteristic of pharmaceutical effluents (Kotwani et al., 2021), and such a chemical ecosystem develops a sustained pressure that favours both intrinsic resistance optimisation and acquisition of genetic elements as reported by Bradshaw (2025). However, physicochemical parameters of the effluent environments, including antibiotic concentrations, heavy metal content, and pH, were not available for the analysed datasets, limiting direct quantification of selective pressures. Hence, even though all the isolates had a shared environmental origin, the isolates exhibited substantial variability in AMR gene counts (13–35 genes), indicating dynamic resistome plasticity, which was also suggested by Bobate et al. (2023). The board spectrum antimicrobial exposure to β-lactamases, aminoglycoside-modifying enzymes, macrolide, and tetracycline resistance genes was observed. Similarly, OXA-type carbapenemases, including OXA-23 and intrinsic OXA-66, were also detected; these findings were in accordance with Sartori et al. (2025). Additionally, the intrinsic and acquired oxacillinases suggested β-lactam resistance. OXA-144, OXA-1010, and OXA-1166 variants’ sporadic detection further indicated ongoing microevolution and gene flux within the population. Malik et al. (2024) reported that extended-spectrum β-lactamases (PER-1, TEM-12) and aminoglycoside resistance determinants, including the high-level methyltransferase armA, underscore the clinical relevance of these environmental isolates, which supported the amplification of clinically significant MDR determinants. Notably, in the present study, it was observed that the efflux-mediated resistance across the entire genome was dominant. Hence, the RND family, AdeABC, AdeIJK, and AdeFGH members constituted the structural core of the resistome. Also, the presence of AdeIJK suggested that this system forms a conserved intrinsic defence mechanism, which is regulated by adeN. This observation was in accordance with the review by Kornelsen and Kumar (2021). Additionally, the widespread detection of AdeABC and its regulatory components highlights the potential for overexpression under antibiotic pressure. In clinical A. baumannii, AdeABC overexpression is strongly associated with resistance to aminoglycosides, fluoroquinolones, and tigecycline (Lari et al., 2018). Several resistance determinants identified in this study, including OXA-type carbapenemases, aminoglycoside-modifying enzymes, and efflux systems, are widely reported in clinical A. baumannii strains. This overlap suggests that environmental isolates may serve as reservoirs of clinically relevant AMR genes, highlighting the potential for transmission from environmental to clinical settings. Similarly, it was observed in environmental isolates that this regulatory circuitry is present in the effluent-associated strains, which poses the rapid phenotypic escalation of resistance. Furthermore, the identification of MATE (AbeM), SMR (AbeS), and MFS transporters displays an efflux repertoire, supporting the concept that multidrug extrusion represents a primary adaptive strategy in chemically polluted ecosystems (Abdi et al., 2020). Notably, the efflux system and its regulators were highly conserved, as determined via core resistome analysis, which suggested that intrinsic resistance was a shared evolutionary trait. In contrast, the carbapenemases, ESBLs, sulfonamide resistance genes (sul1, sul2), and disinfectant resistance gene qacEΔ1, were shown to be unevenly distributed in accessory resistance determinants. This difference supported the study by Michaelis and Grohmann (2023), suggesting that HGT plays a central role in shaping the accessory resistome. However, the present analysis does not distinguish between plasmid-borne and chromosomal AMR determinants, which limits the resolution of gene mobility and transfer dynamics. Future integration of plasmid reconstruction or long-read sequencing approaches would help to better resolve the genomic context of these resistance genes. Also, the sul and qacEΔ1 genes were often associated with class 1 integrons (Domínguez et al., 2019), supporting mobile genetic platforms further. Although these associations strongly suggest HGT, the absence of explicit mobile element annotation limits definitive attribution of gene mobility mechanisms. However, direct identification of plasmids or transposons was not performed, and therefore, HGT inference is based on gene distribution patterns and known associations rather than direct genomic confirmation. Subsequently, this suggests that genomic fluidity is a characteristic of environments under sustained anthropogenic pressure.

The pharmaceutical effluents, which facilitated gene exchange between the environmental and opportunistic pathogenic strains, hence the co-selection of antibiotic and disinfectant resistance genes, may further enhance persistence and dissemination potential. Phylogenetic reconstruction based on 16S rRNA sequences demonstrated clustering of closely related isolates; however, resistome composition did not follow phylogenetic lineage, which is in accordance with Tao et al. (2022), suggesting that similar resistance profiles do not always relate evolutionarily which might be due to decoupling between the phylogeny and AMR gene supporting the dominant influence of HGT and environmental selective pressures over vertical inheritance. However, 16S rRNA-based phylogeny has limited discriminatory power for closely related Acinetobacter strains and may be influenced by recombination or conserved sequence similarity. While it provides a computationally efficient framework for broad evolutionary inference, higher-resolution approaches such as whole-genome SNP-based phylogeny or core-genome alignment would offer more precise insights into strain-level relationships. Moreover, the core efflux systems were phylogenetically conserved, and the accessory AMR genes showed a mosaic distribution pattern, which suggested that evolutionary relatedness was identified due to intrinsic resistance, whereas the accessory genes were primarily shaped due to ecological exposure and mobile element dynamics, hence, the convergence of intrinsic efflux dominants and acquisition of high-risk carbapenemases highlights the dual adaptive strategy of Acinetobacter isolates. The clinically relevant AMR determinants harboured by environmental strains act as reservoirs for dissemination into healthcare (Ahmad et al., 2021). This overlap between environmental and clinical resistomes underscores the potential for environmental Acinetobacter populations to contribute to the emergence and dissemination of clinically relevant multidrug-resistant strains. Overall, this study highlights the intrinsic efflux system forming a conserved resistance backbone, accessory AMR genes dynamically acquired and lost via HGT, environmental selective pressure driving multidrug optimisation, and that phylogenetic relatedness alone cannot predict resistance burden. This observation was further supported by Mantel test analysis, which demonstrated a weak and non-significant correlation between phylogenetic distance and resistome composition (r = −0.1228, p = 0.678). The lack of statistical association reinforces the hypothesis that HGT and environmental selective pressures play a more dominant role than vertical inheritance in shaping antimicrobial resistance profiles in these isolates. Such decoupling has been increasingly reported in environmental bacterial populations exposed to anthropogenic stressors, where gene mobility overrides lineage-specific constraints. While the present study provides detailed insights into resistome architecture, the relatively small sample size (n = 8) may limit the generalisability of the findings. The isolates analysed were selected based on availability and environmental relevance, which may introduce sampling bias, including potential geographic and ecological underrepresentation. Future studies should incorporate larger and geographically diverse datasets, integrate whole-genome SNP-based phylogeny, and include plasmid reconstruction or long-read sequencing approaches to resolve gene mobility. Additionally, coupling genomic data with environmental physicochemical measurements will be essential to better understand the selective pressures driving resistome evolution.

5 Conclusion

The pharmaceutical effluent-associated Acinetobacter isolates’ resistome is characterised by a conserved efflux-centred intrinsic core and a highly variable accessory gene pool shaped by HGT and chemical selective pressure. The observed phylogeny-resistome decoupling underscores the ecological complexity of MDR evolution in anthropogenically impacted environments. These findings emphasise the importance of environmental surveillance in understanding and mitigating the emergence and dissemination of clinically significant antimicrobial resistance.

Generative AI statement

The author(s) declared that generative AI was not used in the creation of this manuscript.

Any alternative text (alt text) provided alongside figures in this article has been generated by Frontiers with the support of artificial intelligence and reasonable efforts have been made to ensure accuracy, including review by the authors wherever possible. If you identify any issues, please contact us.

Publisher’s note

All claims expressed in this article are solely those of the authors and do not necessarily represent those of their affiliated organizations, or those of the publisher, the editors and the reviewers. Any product that may be evaluated in this article, or claim that may be made by its manufacturer, is not guaranteed or endorsed by the publisher.

Footnotes

Edited by: Selvankumar Thangaswamy, Saveetha Institute of Medical and Technical Sciences (SIMATS), India

Reviewed by: P. Srinivasan, PGP College of Arts and Science, India

Suresh Dhanaraj, Vels Institute of Science Technology & Advanced Studies, India

Footnotes

AMR, Antimicrobial Resistance; MDR, Multidrug Resistance; HGT, Horizontal Gene Transfer; WHO, World Health Organization; NCBI, National Centre for Biotechnology Information; SRA, Sequence Read Archive; FASTQ, Fast Quality (sequence format); CARD, Comprehensive Antibiotic Resistance Database; RND, Resistance-Nodulation-Division; MATE: Multidrug And Toxic compound Extrusion; SMR, Small Multidrug Resistance; MFS, Major Facilitator Superfamily; ESBL, Extended-Spectrum β-Lactamase; rRNA, Ribosomal Ribonucleic Acid; 16S rRNA, 16S Ribosomal Ribonucleic Acid; GC, Guanine-Cytosine.