Skip to main content
NCBI home page
BMC Microbiology logoLink to BMC Microbiology
. 2026 May 25;26:651. doi: 10.1186/s12866-026-05217-y

Genomic and phenotypic characterization of carbapenem-resistant Acinetobacter baumannii isolates from patients and trolleys and their potential connections in an intensive care unit

Pengpeng Min 1,#, Haotian Xu 2,#, Hongjia Pan 1, Siyuan Chen 3, Jiayao Yao 1, Shuyan Li 4, Xuewen Ding 5, Xiaoting Hua 6,7, Xi Li 1, Yunsong Yu 1,6, Hua Zhou 8,✉, Haiyang Liu 1,✉
PMCID: PMC13383449  PMID: 42185782

Abstract

Background

Carbapenem-resistant Acinetobacter baumannii (CRAB) poses significant challenges in intensive care units (ICU) due to its multidrug resistance and environmental persistence. To date, most studies have focused on clinical isolates or patient-environment transmission. However, high-touch mobile surfaces have been less extensively characterized.

Results

We collected 49 CRAB isolates in 2019 and 2021 from ICU patients (n = 32) and trolley surfaces (n = 17). 91.84% (45/49) isolates belonged to ST2Pas and carried blaOXA−23. The capsular polysaccharide (K locus) and lipopolysaccharide (OC locus) types were predominantly KL7 and OCL1, respectively. Comparative genomic analysis revealed a high degree of genomic relatedness between patient and trolley associated isolates, including three clonal clusters (SNP differences < 19). To further characterize environmental persistence, we assessed biofilm formation using Live/Dead staining and scanning electron microscopy (SEM), and evaluated disinfectant susceptibility across all isolates.

Conclusions

These findings demonstrate the coexistence of closely related CRAB on patients and trolley surfaces within the same ICU and highlight the importance of considering high-touch mobile surfaces in infection prevention and control (IPC) strategies.

Trial registration

ClinicalTrials.gov, NCT04310722. Registered on 18 March 2020. Retrospectively registered.

Supplementary Information

The online version contains supplementary material available at 10.1186/s12866-026-05217-y.

Keywords: ICU trolleys, CRAB, Whole-genome sequencing, Genomic analysis, Biofilm

Background

The spread of multidrug-resistant organisms (MDRO) is a serious challenge in global healthcare settings [1]. Among the numerous drug-resistant bacteria, the Acinetobacter calcoaceticus-baumannii (ACB) complex has emerged as a major cause of nosocomial infections, particularly in critically ill patients [2]. The ACB complex includes several closely related species, such as Acinetobacter baumannii, Acinetobacter pittii, and Acinetobacter nosocomialis, which are often difficult to distinguish from each other based on phenotypic properties [3]. Among these, A. baumannii is the most clinically relevant species [4]. The World Health Organization (WHO) listed Carbapenem-resistant Acinetobacter baumannii (CRAB) as a key priority pathogen in the 2024 list of key bacterial pathogen (https://www.who.int/publications/i/item/9789240093461). CRAB is frequently implicated in hospital outbreaks, especially in intensive care units (ICU), due to its persistence and multidrug resistance [5]. It can cause serious diseases in patients, including respiratory infections, bloodstream infections, meningitis, and urinary tract infections [6], especially in immunocompromised patients with high pathogenicity and mortality rates [7]. Therefore, it is crucial to implement effective preventive measures and management methods to control the spread of CRAB [8]. Carbapenem antibiotics have long been regarded as the last line of defense for treating A. baumannii infections [9]. However, owing to the improper use of carbapenem antibiotics, the emergence of CRAB has greatly limited treatment options [10]. Although polymyxin and tigecycline have high bactericidal activity against CRAB, their poor pharmacokinetics, significant nephrotoxicity, and neurotoxicity to humans limit their use [11]. CRAB resistance to carbapenems is mainly caused by the production of carbapenemases [12], including class B metallo-β-lactamases (such as NDM, VIM, and IMP types) and class D oxacillinases (such as OXA-23, OXA-24/40, OXA-58, OXA-143, and OXA-235 families) [13]. Among these, OXA-23 remains the most prevalent class D carbapenemase worldwide and is strongly associated with hospital-acquired CRAB infections [14].

Apart from resistance, biofilm formation is also a critical issue that needs to be addressed in CRAB. CRAB can increase its survival ability in the environment by forming biofilms, making it easier to exist on the surface of medical facilities and leading to further spread among the various equipment in ICU [3, 15]. Of note, after the formation of the CRAB biofilm, it is more difficult to completely remove it in traditional cleaning processes [16]. Additionally, previous studies have shown that CRAB can form biofilms on the surfaces of medical devices such as ventilators and electrocardiogram monitors, which are typically fixed equipment [17, 18]. Compared with fixed medical devices, mobile trolleys (such as dispensing trolleys and crash trolleys) represent high-touch mobile surfaces that are frequently used in the ICU, yet their microbiological characteristics have been less extensively characterized. Colonization and biofilm formation on trolley surfaces may contribute to the environmental persistence of CRAB and complicate routine infection prevention and control (IPC) efforts.

In this study, we performed genomic and phenotypic characterization of CRAB isolates collected from ICU patients and trolley surfaces across different time periods, aimed to provide microbiological insights relevant to IPC.

Methods

Setting and research design

This study was conducted in the ICU of a tertiary hospital in Hangzhou, China. The flow chart of this study was constructed using ProcessOn (https://www.processon.com/) (Fig. 1). The ICU is a comprehensive unit consisting of 13 rooms, including two single rooms (Rooms 1 and 13), ten double rooms (Rooms 2–6 and 8–12), and one six-person room (Room 7). Each room had a dispensing trolley and two crash trolleys in the ICU that moved between different rooms, resulting in a total of 15 trolleys included in the study (Fig. 2A). Photographs of the trolleys were provided in Figure S1. This study was based on a previously published longitudinal genomic surveillance study conducted in the same ICU [19]. Briefly, CRAB sampling from patients and trolley surfaces was performed weekly during two sampling periods: August-October 2019 and May-July 2021. The 2019 sampling period lasted 12 weeks, with 12 sampling rounds, whereas the 2021 sampling period lasted 13 weeks, with 13 sampling rounds. Surveillance sampling was interrupted in 2020 due to COVID-19-related restrictions and resumed in May 2021. In each round, all 15 trolleys were sampled. Trolley sampling was performed at predefined surface sites, including handles, top surfaces, and drawers, and the same sites were consistently sampled throughout the study period. Patient isolates were obtained from both systematic surveillance sampling and clinically indicated diagnostic cultures. Surveillance sampling was conducted at admission (or at the beginning of the study) and weekly thereafter, whenever clinically feasible. In parallel, clinical specimens such as sputum, ascitic fluid, pelvic drainage fluid, hydrothorax, bile, and abdominal incision exudates were collected as part of routine medical practice. Weekly ICU bed occupancy rates are provided in Table S1. Trolley disinfection frequency was increased during the 2021 sampling period.

Fig. 1.

Fig. 1

Open in a new tab

Flowchart of all experiments used in this study. This flow chart was built using the ProcessOn software

Fig. 2.

Fig. 2

Open in a new tab

(A) Schematic of the ICU. The ICU has 28 beds and 13 rooms (two independent rooms, one six-person room, and ten double rooms). The numbers next to the beds represent the bed numbers. B Schematic representation of the antimicrobial susceptibility profiles of the 49 isolates. C Schematic representation of the susceptibility of the 49 isolates to three disinfectants, with ATCC 17978 used as the reference strain. The numbers indicate the corresponding MIC values

Strain isolation and identification

Patient samples were collected using rectal swabs containing Cary-Blair Transport Medium (Gongdong, Taizhou, China), while other specimen types were obtained with swabs moistened in 0.9% saline. All patient-derived swabs were directly inoculated onto Acinetobacter CHROMagar plates (CHROMagar, Paris, France) and incubated at 37 °C for 24 h. Environmental samples were obtained by swabbing trolley surfaces with sterile Copan swabs (Italy) pre-moistened with 2 mL of tryptic soy broth (TSB) supplemented with 0.1% sodium thiosulfate (Sangon Biotech, China). After sampling, each swab was transferred into a 14 mL sterile tube containing 2 mL of the same TSB-thiosulfate solution and incubated at 37 °C for 24 h for enrichment. Subsequently, 20 µL of the enriched culture was streaked onto Acinetobacter CHROMagar and incubated overnight at 37 °C [19]. Presumptive A. baumannii colonies were identified based on their typical red coloration and colony morphology and were subsequently inoculated onto Mueller-Hinton agar (MHA) plates (Oxoid, Hampshire, UK) for purification. A single purified colony was subjected to species identification using matrix-assisted laser desorption/ionization time-of-flight mass spectrometry (MALDI-TOF MS; bioMérieux, Marcy-l’Étoile, France) [20].

Antibacterial agents and disinfectants susceptibility testing

The minimum inhibitory concentrations (MIC) of imipenem, cefoperazone-sulbactam, polymyxin, tigecycline, ciprofloxacin, meropenem, amikacin, and sulbactam were determined using the broth microdilution (BMD) method. MIC testing was performed using commercially manufactured broth microdilution plates designed for susceptibility testing of non-fastidious gram-negative bacteria. All procedures, including preparation of bacterial inocula, plate inoculation, incubation, and MIC determination, were conducted strictly according to the manufacturer’s instructions. The results for imipenem, ciprofloxacin, meropenem, amikacin and sulbactam were interpreted according to the breakpoints of the Clinical and Laboratory Standards Institute (CLSI, 2025) M100 standards. The MICs of cefoperazone-sulbactam were determined at cefoperazone: sulbactam ratios of 1:1 and 2:1. In the absence of CLSI interpretive criteria for cefoperazone-sulbactam, susceptibility interpretation for cefoperazone-sulbactam (2:1) was performed using ampicillin-sulbactam MIC breakpoints as surrogate criteria (S ≤ 8/4 mg/L, I = 16/8 mg/L, and R ≥ 32/16 mg/L), in accordance with previous studies [21]. Cefoperazone-sulbactam (1:1) MIC values were reported descriptively without formal S/I/R categorization. The breakpoint of tigecycline was determined by the US Food and Drug Administration (FDA). The breakpoint of polymyxin was interpreted according to European Committee on Antimicrobial Susceptibility Testing (EUCAST, 2025) clinical breakpoints (https://www.eucast.org/), following the recommendations for colistin testing. Escherichia coli ATCC 25922 served as the quality control strain. Each isolate was tested using the standardized BMD procedure, and MIC results were accepted only when the quality control strain yielded values within the expected range. Moreover, the MICs of chlorhexidine (CHG), p-chlorophenol (p-CP), and calcium hypochlorite [Ca(OCl)2] were determined by the BMD method [22]. The tested concentration ranges for three disinfectants were as follows: CHG 0.125-128 mg/L, p-CP 0.5-512 mg/L, and Ca(OCl)2 2-2048 mg/L. Because no standardized breakpoints exist for these disinfectants, MICs were compared across isolates to assess relative tolerance, with A. baumannii ATCC 17978 included as the reference strain [23]. All antimicrobial and disinfectant MIC assays were performed in triplicate. MIC data were summarized using MIC50 and MIC90 values to describe antimicrobial susceptibility profiles and relative disinfectant tolerance.

Whole‑Genome Sequencing (WGS) and bioinformatic analysis

Genomic DNA was extracted from 49 CRAB using a Qiagen Mini Kit (Qiagen, Hilden, Germany) according to the manufacturer’s recommendations, and WGS was performed using the Illumina HiSeq platform (Illumina, San Diego, USA) by Zhejiang Tianke, China, with an average sequencing depth of approximately 150×. De novo assembly of Illumina reads was performed using Shovill v0.9.0 (https://github.com/tseemann/shovill). Genome quality was assessed using CheckM [24], and all 49 sequenced isolates passed quality assessment and were retained for downstream genomic analysis. Species identity and virulence genes were further confirmed using the fIDBAC platform (http://fbac.dmicrobe.cn/submitTask/pipline) based on genome sequencing data [25]. Multilocus sequence types (MLST) were analyzed using PubMLST (https://pubmlst.org/organisms/acinetobacter-baumannii). Capsular polysaccharides (K locus) and lipooligosaccharides (OC locus) were identified using Bautype v1.0 [26]. Resistance genes were identified using ABRicate v0.8.13, with the ResFinder database (https://github.com/tseemann/abricate) [27]. Plasmid replicons were typed using the Hamidian lab-typing scheme [28]. A phylogenetic tree was constructed based on single nucleotide polymorphisms (SNP) using Snippy v4.6.0 [29] and FastTree v2.1.11 [30]. The tree file was further visualized using the online tool iTOL v7. (https://iTOL.ebbl.de) [31]. An alluvial diagram was generated using RAWGraphs (https://rawgraphs.github.io/). SNP were calculated using SNP-dists v0.6.3 (https://github.com/tseemann/snp-dists), and the matrix was further visualized with RStudio v4.4.3. CRAB isolates with less than 19 SNP differences were classed as the same clone, and clonal transmission was defined as < 19 SNP differences [5].

Live and dead cell staining

Two trolley-derived CRAB isolates from different sampling periods, DETAB-E51 and DETAB-E2174, were selected for biofilm assays. After incubation with 1 mL of 1 × 108 CFU/mL bacterial solution at 37 °C for 1 and 3 days, the suspension was removed, and the biofilms were gently rinsed with phosphate buffered saline (PBS) to eliminate planktonic cells and stained using the LIVE/DEAD® BacLightTM Bacterial Viability Kit (Invitrogen, USA) according to the manufacturer’s instructions in an environment protected from light, where live bacteria were stained green in interaction with fluorescent SYTO-9, and dead bacteria were stained red in interaction with fluorescent propidium iodide (PI). After staining, bacterial growth and biofilm thickness were observed in the 3D mode using confocal laser scanning microscopy (CLSM, Leica, TCS SP8) [32]. CLSM imaging was performed in three independent experiments for each isolate and time point. Biofilm thickness was quantified from CLSM z-stack images and reported as mean values.

Observation of microbial morphology

Biofilm morphology was observed after 1 and 3 days of incubation using scanning electron microscopy (SEM, Sigma 500, Zeiss, Germany). The biofilm preparation procedure was performed as described above. All specimens were first fixed in 2.5% glutaraldehyde solution overnight and then dehydrated using a graded ethanol series (30%, 50%, 70%, 80%, 90%, 95%, and 100%). Subsequently, the samples were freeze-dried and sprayed with gold. Finally, the bacteria were observed by SEM [33].

Results

Overview of isolates collection and drug resistance

From August to October 2019, 31 CRAB isolates were obtained from the ICU (19 patients and 12 trolleys). From May to July 2021, 18 isolates of CRAB were isolated in the ICU (13 patients and 5 trolleys). ICU bed occupancy was comparable between the two sampling periods, with rates of 94.64% in 2019 and 95.33% in 2021 (Table S1). Among the 32 CRAB isolates recovered from patients, 27 (84.38%) were collected from sputum specimens, while the remaining 5 isolates were from pelvic drainage fluid (3.13%), hydrothorax (3.13%), ascitic fluid (3.13%), bile (3.13%), and abdominal incision exudate (3.13%) (Table S2). Among the 17 CRAB isolates from the trolleys, 13 (76.47%) were collected from the dispensing trolley and 4 (23.53%) were isolated from the crash trolley (Table S3). Notably, repeated CRAB recovery from specific trolleys was observed during both sampling periods. During the 2019 sampling period, dispensing trolleys in Rooms 2, 4, and 7 were repeatedly positive, whereas during the 2021 sampling period, the dispensing trolley in Room 3 and crash trolley 2 were repeatedly positive. All MICs data were shown in Table S2 and Table S3. 49 isolates were confirmed to be CRAB, with the MICs of meropenem for all isolates exceeding 16 mg/L. In contrast, polymyxin and tigecycline MICs remained relatively low across the isolate collection, ranging from 1 to 2 mg/L and 0.5 to 2 mg/L, respectively (Fig. 2B).

Sensitivity of CRAB to disinfectants

All 49 CRAB isolates were subjected to disinfectant susceptibility testing. For both Ca(OCl)2 and p-CP, all isolates exhibited identical MIC values. For CHG, seven isolates (14.29%) showed an MIC of 8 mg/L, while 37 isolates (75.51%) exhibited an MIC of 32 mg/L. The remaining isolates displayed MIC consistent with the reference strain A. baumannii ATCC 17978 (Fig. 2C and Table S4).

Genomic characteristics and clonal clusters of CRAB in the ICU

According to the MLST Pasteur scheme, the isolates can be divided into two main STs: ST2Pas (91.84%) and ST164Pas (6.12%) (Figure S2A). According to the Oxford scheme, the isolates could be classified into five main STs: ST208Oxf (51.02%), ST2499Oxf (22.45%), ST191Oxf (6.12%), ST938Oxf (6.12%), and ST1418Oxf (6.12%) (Figure S2B). Bautype revealed that the K locus types were predominantly KL7 (34.69%), KL104 (22.45%), and KL2 (14.29%). For the OC locus, OCL1 and OCL5 accounted for 93.88% and 6.12%, respectively. Interestingly, all ST2Pas isolates belonged to the OCL1 (Fig. 3).

Fig. 3.

Fig. 3

Open in a new tab

Alluvial diagram of the 49 CRAB isolates. MLST, KL, and OCL types were shown. The width of each flow represents the number of isolates sharing specific type combinations

ST2Pas CRAB was independently analyzed based on SNP, and data showed that three groups of ST2Pas isolates were identified, each of which was associated with clonal clusters between patients and trolleys in 2019 and 2021, respectively (Fig. 4A and B). Group 1 consisted of 12 isolates collected in 2019, of which 3 isolates were recovered from dispensing trolley in Rooms 3, 4, and 10, respectively, and 9 isolates were obtained from patients. The median number of SNP differences between the isolates was 9 (SNP range 4–18) (Fig. 4A). Group 2 contained 11 isolates collected in 2021, one was collected from the dispensing trolley in Room 3, and the others were from patients. The median number of SNP differences among these isolates was 4 (SNP range 1–12) (Fig. 4A). Apart from Groups 1 and 2, a small clonal cluster (Group 3) in 2019 (SNP range 3–9) was observed among crash trolleys (DETAB-E226), dispensing trolleys (DETAB-E93 and DETAB-E188), and patients (DETAB-C15 and DETAB-C16) (Fig. 4A).

Fig. 4.

Fig. 4

Open in a new tab

Phylogenetic analysis of CRAB ST2Pas based on SNP. A SNP matrix values among the 45 CRAB ST2Pas isolates. SNP differences are represented by numbers in a box. B 45 CRAB ST2Pas isolates of ST_Pasteur, ST_Oxford, drug resistance genes, plasmid types, drug susceptibility results, and phylogenetic relationships. The blue, yellow, and green rectangles represent three distinct groups based on the SNP analysis. The reference strain used was the earliest isolated strain, DETAB-C1

All ST2Pas isolates carried blaOXA−23, blaOXA−66 and ant(3’’)-IIa (Fig. 4B). Ten isolates (90.91%) in Group 2 carried blaTEM−12, whereas none of the isolates in Group 1 did. The sulphonamide resistance gene sul1 was detected in Group 2 isolates, and sul2 was present in Group 1. The isolates in Group 1 carried only one plasmid, R3-T4, while two plasmid types, R2-T1 and GR25 (except DETAB-D13), were detected in Group 2 isolates (Fig. 4B). All ST2Pas isolates were screened for biofilm-associated determinants, and multiple biofilm genes were identified across the isolates. Among them, 73.33% (33/45) of the isolates simultaneously carried the ompA, pgaABCD, bfmRS, abaIR, and bap genes. Source-specific analysis showed that 66.67% (10/15) of trolley-derived isolates and 76.67% (23/30) of patient-derived isolates carried these biofilm-associated genes. In addition, 44.44% (20/45) of the isolates carried the Csu pili gene cluster (Fig. 4B). Among the three identified clonal clusters, only 50.00% (6/12) of the isolates in Group 1 carried these biofilm-associated genes, whereas all isolates in Groups 2 and 3 harbored them. Notably, isolates in Group 3 additionally carried the Csu pili gene cluster compared with those in Group 2.

Biofilm characteristics of selected trolley-derived CRAB isolates

Biofilm formation of the two isolates (DETAB-E51 and DETAB-E2174) was evaluated by CLSM and SEM. After 1 day of incubation, CLSM revealed that both isolates were able to establish early biofilm structures, with comparable mean thicknesses of 20.8 μm for DETAB-E51 and 20.9 μm for DETAB-E2174 (Fig. 5A). After 3 days, the biofilm structures appeared more developed, with the mean biofilm thickness increasing to 29.9 μm for DETAB-E51 and 22.1 μm for DETAB-E2174 (Fig. 5A). SEM imaging further confirmed these observations. After 1 day of incubation, both isolates produced thin and relatively sparse biofilms characterized by scattered bacterial cells and limited extracellular matrix. By day 3, however, the biofilms had become markedly denser and thicker, displaying extensive bacterial aggregation, multilayered structures, and abundant extracellular material (Fig. 5B).

Fig. 5.

Fig. 5

Open in a new tab

Biofilm structure of CRAB. A Images of CRAB biofilms obtained using Live/Dead cells staining. Green fluorescence represents live bacteria, while red fluorescence represents dead bacteria. Representative CLSM images are shown. Biofilm thickness values represent the mean thickness. B Biofilm structure observed under SEM

Discussion

A. baumannii remains a major therapeutic challenge in clinical practice, posing a serious threat to public health [34]. Previous studies have demonstrated that A. baumannii can persist in diverse hospital environments, particularly on high-touch surfaces frequently contacted by healthcare workers or patients [35]. One study suggested that drug-resistant bacteria persist on the surface of medical devices for a long time and may spread through the device in the ICU without thorough cleaning [36]. Sana Azaiez et al. found that 14% of ICU equipment was contaminated with CRAB [18]. Trolleys are frequently used in ICUs as tools to carry medical supplies. Owing to the presence of various medical equipment and drugs, contact with multiple patients and medical staff, and inadequate cleaning and disinfection measures, trolleys are likely to provide favorable conditions for the environmental persistence and contamination of CRAB.

In this study, we investigated the genomic and phenotypic characteristics of CRAB isolates recovered from ICU patients and trolley surfaces. We found that among the 49 isolates, 45 belonged to ST2Pas, suggesting that ST2Pas was the predominant clone type in this ICU during the study period. All OCL types of ST2Pas belong to OCL1, and OCL1 may be a target for future CRAB vaccine designs. During the study period, there were three clonal clusters, all of which involved ST2Pas isolates. In the first clonal cluster, isolates detected in the three trolleys belonged to the same clone as those from the six patients. In the second clonal cluster, isolates detected in trolleys belonged to the same clone as those from the seven patients. In the third clonal cluster, isolates detected in the three trolleys belonged to the same clone as those from the two patients. These findings demonstrate that trolleys and patients shared highly similar CRAB clones. The strong genomic relatedness between patients and trolleys isolates indicates that both serve as components of the same clonal transmission landscape within the ICU. Trolleys, which are frequently moved between rooms and come into contact with multiple surfaces and personnel, may act as important environmental nodes within this network. The presence of highly similar CRAB clones on trolleys underscores the need to consider mobile equipment as part of environmental hygiene strategies aimed at limiting CRAB dissemination in the ICU.

Resistance of A. baumannii to carbapenems is mainly due to the presence of class D carbapenemases. In our study, all ST2Pas isolates contained blaOXA−23, which is consistent with previous research results [37]. This finding is consistent with the observed carbapenem-resistant phenotypes and may contribute to the persistence of CRAB in the ICU environment. According to the drug sensitivity results, all ST2Pas isolates were resistant to imipenem and meropenem, which was consistent with the results for the resistance genes. In addition, all ST2Pas isolates carried the ant(3’’)-IIa gene, although their phenotypic resistance profiles were primarily associated with carbapenem antibiotics. The persistent presence of the ant(3’’)-IIa gene may contribute to resistance or reduced susceptibility to aminoglycosides, and could be part of a broader multidrug-resistant genetic background [38]. blaTEM−12 was also detected in 90.91% of the isolates in Group 2. We compared the two clonal clusters and found that isolates in Group 2 carried a more complex resistance gene profile. This suggests that these isolates may possess broader antimicrobial resistance and potentially enhanced survival capabilities during the second clonal cluster. This pattern was further supported by antimicrobial susceptibility profiling, which showed higher MIC50/MIC90 values for cefoperazone-sulbactam (1:1), cefoperazone-sulbactam (2:1), and sulbactam in Group 2 than in Group 1 (128/128 vs. 16/32, > 128/>128 vs. 32/64, and 128/128 vs. 16/32 mg/L, respectively), although this pattern was not consistent across all isolates. However, susceptibility interpretation for cefoperazone-sulbactam in this study was performed using ampicillin-sulbactam breakpoints as surrogate criteria due to the absence of CLSI interpretive standards for cefoperazone-sulbactam. Although this approach has been adopted in previous studies, it may not fully reflect the intrinsic activity or clinical relevance of cefoperazone-sulbactam. The detection of genetically similar isolates on trolleys during this period further indicates that trolleys may have contributed to the dissemination of isolates carrying diverse resistance genes.

Notably, isolates from Group 2 carried the qacEdelta1 gene, which has been associated with reduced susceptibility to commonly used hospital disinfectants such as benzalkonium chloride and chlorhexidine [39]. The level of tolerance has been reported to increase with higher gene expression [40], which often coexist with other antibiotic resistance genes (such as blaTEM−12 and sul1) and spread together on the same plasmid or integron, especially in hospital-acquired infections [41]. However, in our susceptibility testing involving all 49 isolates, no significant difference in MIC was observed between Groups 1 and 2. This suggests that the presence of qacEdelta1 does not necessarily result in a measurable increase in MICs under the tested conditions. Nevertheless, the detection of qacEdelta1 may still represent a latent risk factor. Notably, the MICs of Ca(OCl)₂ for all tested isolates exceeded 2048 mg/L, in agreement with previous findings [42], suggesting possible relative tolerance to Ca(OCl)₂ under the experimental conditions used in this study. However, MIC-based testing does not fully represent clinical disinfection practices, where factors such as disinfectant concentration, contact time, and the presence of organic matter may substantially influence disinfectant efficacy.

In addition to determinants associated with higher MICs to disinfectants, the strong biofilm-forming potential of the isolates may also contribute to their persistence on trolley surfaces. Genes known to be closely associated with biofilm formation, including ompA, pgaABCD, bfmRS, abaIR, bap, and the Csu pili gene cluster [43, 44], were prevalent among the isolates in our collection, with 73.33% (33/45) carrying ompA, pgaABCD, bfmRS, abaIR, and bap, and 44.44% (20/45) harboring the Csu pili gene cluster. Notably, these biofilm-associated genes were widely distributed among both trolley- and patient-derived isolates, with comparable prevalence observed between the two sources. In addition, isolates within all identified clonal clusters carried key biofilm-related determinants, although the prevalence varied, with only 50.00% of isolates in Group 1 harboring these genes compared with complete carriage in Groups 2 and 3. Isolates in Group 3 additionally carried the Csu pili gene cluster. This variation among clonal clusters may suggest that differences in biofilm-forming capacity exist even among closely related isolates. Furthermore, the presence of the Csu pili gene cluster in Group 3 may confer an additional advantage in surface colonization, potentially contributing to its persistence in the ICU environment. These genomic findings were consistent with biofilm imaging observations in selected trolley-derived isolates. Live/Dead staining assays demonstrated that DETAB-E51 and DETAB-E2174 were capable of forming dense biofilms, which provide protective microenvironments that shield bacteria from chemical disinfectants. Moreover, SEM revealed dense multilayered biofilm structures and abundant extracellular polymeric substances. These observations provide morphological support for the biofilm-forming potential of environmental CRAB isolates recovered from ICU trolleys.

Together, these results suggest that the persistence of CRAB on ICU trolleys is not solely attributable to disinfectant resistance genes but is also strongly reinforced by robust biofilm-forming capabilities. Given that biofilms can enhance bacterial tolerance to environmental stress, including exposure to disinfectants, routine surface cleaning alone may be insufficient to fully eliminate CRAB under certain conditions. These findings have important implications for infection prevention and control in ICU settings. ICU trolleys contain multiple high-touch surfaces, such as handrails, drawer handles, and worktop edges, which are frequently contacted by healthcare workers during patient care. Repeated recovery of CRAB from specific dispensing and crash trolleys across sampling rounds suggests persistent or recurrent contamination of these high-touch surfaces. Together with frequent handling, intermittent cleaning, and the biofilm-forming potential of CRAB, this pattern may contribute to bacterial persistence on trolley surfaces and increase the potential risk of cross-transmission within the ICU. Therefore, it is necessary to explore more effective cleaning methods and establish more IPC measures. In addition to routine IPC practices, such as regular hand hygiene, proper use of personal protective equipment, and adherence to standard operating procedures for disinfection, specific enhanced measures can also be implemented. For instance, the cleaning frequency of trolleys can be increased, and different disinfectants can be alternately used [45]. Moreover, novel antimicrobial materials have shown promising potential in preventing bacterial colonization and inhibiting biofilm formation on frequently touched surfaces [46]. These materials could serve as an effective supplement to traditional disinfection, especially in high-risk areas like the ICU. Further research and development in this area may contribute significantly to reducing the persistence and spread of MDROs such as CRAB in healthcare settings.

Our study had several limitations. Firstly, only 49 CRAB isolates were included, which may limit epidemiological inference regarding transmission patterns. Surveillance sampling was also interrupted during the COVID-19 pandemic, resulting in a temporal gap between the 2019 and 2021 sampling periods and preventing assessment of intermediate clonal dynamics. Secondly, although SNP-based genomic analysis revealed close relatedness among isolates, it could not determine transmission direction, source attribution, or movement routes between patients and environmental surfaces, particularly given the limited sampling frequency and lack of detailed trolley movement tracking. Thirdly, environmental sampling was restricted to trolley surfaces and did not include other high-touch ICU sites or healthcare workers’ hands, limiting our ability to assess the broader ICU environmental reservoir and potential hand-mediated transmission. Finally, only two selected trolley-derived isolates were used for biofilm imaging, limiting comprehensive assessment of biofilm-forming capacity across all isolates or clonal clusters. Future studies incorporating broader environmental sampling, healthcare worker surveillance, representative biofilm assays, and integrated epidemiological data are needed to better define the role of trolleys in ICU CRAB persistence and dissemination.

Conclusion

In summary, our study provides genomic and phenotypic characterization of CRAB isolates recovered from ICU patients and trolley surfaces. SNP-based analysis revealed a high degree of genomic relatedness between isolates from patients and high-touch mobile surfaces, indicating the coexistence of closely related CRAB populations within the ICU environment. The widespread presence of biofilm-associated genes and strong biofilm-forming capacity may contribute to the persistence of CRAB on high-touch surfaces. These findings highlight the importance of considering mobile medical equipment as part of IPC strategies in healthcare settings.

Supplementary Information

Supplementary Material 1. (865.9KB, docx)

Acknowledgements

We thank the doctors and nurses in the ICU for performing the patient sampling.

Reporting guidelines

This study was reported in accordance with the applicable items of the CONSORT guidelines. A completed CONSORT checklist has been provided as an additional file, with items not applicable to this prospective observational study indicated accordingly.

Authors' contributions

P.M.: writing-original draft, visualization, and data curation. H.X. and H.P.: validation and visualization. S.C. and J.Y.: methodology and software. S.L. and X.D.: data curation and visualization. X.H. and X.L.: conceptualization, resources and software. Y.Y.: project administration, resources and funding acquisition. H.Z.: funding acquisition and writing – review & editing. H.L.: writing-review & editing, funding acquisition and project administration. All authors reviewed and approved the final version.

Funding

This work was funded by the National Natural Science Foundation of China (U22A20338, 82402677, 32270183, 82272338, 32011530116, and 82472323), the Zhejiang Provincial Natural Science Foundation of China (LQ24H200005), and the Medical and Health Science Program of Zhejiang Province (2025HY0129).

Data availability

All sequence data and genome assemblies generated in this study have been submitted to GenBank under the BioProject accession numbers PRJNA738868 and PRJNA1034534.

Declarations

Ethics approval and consent to participate

Ethical approval was obtained from the local ethics committee of Sir Run Run Shaw Hospital (approval number: 20190802-1). Written informed consent was obtained from all participants and/or their legal guardians, authorizing the use of their clinical data for research purposes. All procedures involving human-derived samples and clinical data were performed in accordance with the principles of the Declaration of Helsinki.

Consent for publication

Not applicable.

Competing interests

The authors declare no competing interests.

Footnotes

Publisher’s note

Springer Nature remains neutral with regard to jurisdictional claims in published maps and institutional affiliations.

Pengpeng Min and Haotian Xu contributed equally to this work.

Contributor Information

Hua Zhou, Email: zhouhua1@zju.edu.cn.

Haiyang Liu, Email: liuhaiyanggao@163.com.

References

  • 1.Sy CL, Chen PY, Cheng CW, Huang LJ, Wang CH, Chang TH, Chang YC, Chang CJ, Hii IM, Hsu YL, et al. Recommendations and guidelines for the treatment of infections due to multidrug resistant organisms. J Microbiol Immunol Infect. 2022;55(3):359–86. [DOI] [PubMed] [Google Scholar]
  • 2.Benoit T, Sajjad D, Cloutier M, Lapen DR, Craiovan E, Sykes EME, Kumar A, Khan IUH. Acinetobacter calcoaceticus-baumannii complex prevalence, spatial-temporal distribution, and contamination sources in Canadian aquatic environments. Microbiol Spectr. 2024;12(10):e0150924. [DOI] [PMC free article] [PubMed] [Google Scholar]
  • 3.Peleg AY, Seifert H, Paterson DL. Acinetobacter baumannii: emergence of a successful pathogen. Clin Microbiol Rev. 2008;21(3):538–82. [DOI] [PMC free article] [PubMed] [Google Scholar]
  • 4.Tacconelli E, Carrara E, Savoldi A, Harbarth S, Mendelson M, Monnet DL, Pulcini C, Kahlmeter G, Kluytmans J, Carmeli Y, et al. Discovery, research, and development of new antibiotics: the WHO priority list of antibiotic-resistant bacteria and tuberculosis. Lancet Infect Dis. 2018;18(3):318–27. [DOI] [PubMed] [Google Scholar]
  • 5.Baleivanualala SC, Isaia L, Devi SV, Howden B, Gorrie CL, Matanitobua S, Sharma S, Wilson D, Kumar S, Maharaj K, et al. Molecular and clinical epidemiology of carbapenem resistant Acinetobacter baumannii ST2 in Oceania: a multicountry cohort study. Lancet Reg Health West Pac. 2023;40:100896. [DOI] [PMC free article] [PubMed] [Google Scholar]
  • 6.Bergogne-Berezin E, Towner KJ. Acinetobacter spp. as nosocomial pathogens: microbiological, clinical, and epidemiological features. Clin Microbiol Rev. 1996;9(2):148–65. [DOI] [PMC free article] [PubMed] [Google Scholar]
  • 7.Wang M, Ge L, Chen L, Komarow L, Hanson B, Reyes J, Cober E, Alenazi T, Zong Z, Xie Q, et al. Clinical Outcomes and Bacterial Characteristics of Carbapenem-resistant Acinetobacter baumannii Among Patients From Different Global Regions. Clin Infect Dis. 2024;78(2):248–58. [DOI] [PMC free article] [PubMed] [Google Scholar]
  • 8.Jung J, Choe PG, Choi S, Kim E, Lee HY, Kang CK, Lee J, Park WB, Lee S, Kim NJ, et al. Reduction in the acquisition rate of carbapenem-resistant Acinetobacter baumannii (CRAB) after room privatization in an intensive care unit. J Hosp Infect. 2022;121:14–21. [DOI] [PubMed] [Google Scholar]
  • 9.Nguyen M, Joshi SG. Carbapenem resistance in Acinetobacter baumannii, and their importance in hospital-acquired infections: a scientific review. J Appl Microbiol. 2021;131(6):2715–38. [DOI] [PubMed] [Google Scholar]
  • 10.Hamidian M, Nigro SJ. Emergence, molecular mechanisms and global spread of carbapenem-resistant Acinetobacter baumannii. Microb Genom. 2019;5(10):e000306. [DOI] [PMC free article] [PubMed]
  • 11.Piperaki ET, Tzouvelekis LS, Miriagou V, Daikos GL. Carbapenem-resistant Acinetobacter baumannii: in pursuit of an effective treatment. Clin Microbiol Infect. 2019;25(8):951–7. [DOI] [PubMed] [Google Scholar]
  • 12.Ramirez MS, Bonomo RA, Tolmasky ME. Carbapenemases: Transforming Acinetobacter baumannii into a Yet More Dangerous Menace. Biomolecules. 2020;10(5):720. [DOI] [PMC free article] [PubMed]
  • 13.Ayoub Moubareck C, Hammoudi Halat D. Insights into Acinetobacter baumannii: A Review of Microbiological, Virulence, and Resistance Traits in a Threatening Nosocomial Pathogen. Antibiot (Basel). 2020;9(3):119. [DOI] [PMC free article] [PubMed]
  • 14.Mack AR, Hujer AM, Mojica MF, Taracila MA, Feldgarden M, Haft DH, Klimke W, Prasad AB. Bonomo RA: beta-Lactamase diversity in Acinetobacter baumannii. Antimicrob Agents Chemother. 2025;69(3):e0078424. [DOI] [PMC free article] [PubMed]
  • 15.Yang CH, Su PW, Moi SH, Chuang LY. Biofilm Formation in Acinetobacter Baumannii: Genotype-Phenotype Correlation. Molecules. 2019;24(10):1849. [DOI] [PMC free article] [PubMed]
  • 16.Upmanyu K, Haq QMR, Singh R. Factors mediating Acinetobacter baumannii biofilm formation: Opportunities for developing therapeutics. Curr Res Microb Sci. 2022;3:100131. [DOI] [PMC free article] [PubMed] [Google Scholar]
  • 17.Ng DHL, Marimuthu K, Lee JJ, Khong WX, Ng OT, Zhang W, Poh BF, Rao P, Raj MDR, Ang B, et al. Environmental colonization and onward clonal transmission of carbapenem-resistant Acinetobacter baumannii (CRAB) in a medical intensive care unit: the case for environmental hygiene. Antimicrob Resist Infect Control. 2018;7:51. [DOI] [PMC free article] [PubMed] [Google Scholar]
  • 18.Azaiez S, Haenni M, Cheikh AB, Chalbi MS, Messaoudi A, Tilouch L, Bahri S, Drapeau A, Saras E, Mtibaa M et al. Healthcare Equipment and Personnel Reservoirs of Carbapenem-Resistant Acinetobacter baumannii Epidemic Clones in Intensive Care Units in a Tunisian Hospital. Microorganisms. 2023;11(11):2637. [DOI] [PMC free article] [PubMed]
  • 19.Liu H, Moran RA, Doughty EL, Hua X, Snaith AE, Zhang L, Chen X, Guo F, van Schaik W, McNally A, et al. Longitudinal genomics reveals carbapenem-resistant Acinetobacter baumannii population changes with emergence of highly resistant ST164 clone. Nat Commun. 2024;15(1):9483. [DOI] [PMC free article] [PubMed] [Google Scholar]
  • 20.Chen L, Gao W, Tan X, Han Y, Jiao F, Feng B, Xie J, Li B, Zhao H, Tu H, et al. MALDI-TOF MS Is an Effective Technique To Classify Specific Microbiota. Microbiol Spectr. 2023;11(3):e0030723. [DOI] [PMC free article] [PubMed] [Google Scholar]
  • 21.Niu T, Luo Q, Li Y, Zhou Y, Yu W, Xiao Y. Comparison of Tigecycline or Cefoperazone/Sulbactam therapy for bloodstream infection due to Carbapenem-resistant Acinetobacter baumannii. Antimicrob Resist Infect Control. 2019;8:52. [DOI] [PMC free article] [PubMed] [Google Scholar]
  • 22.Han X, Song M, Yu Q, Zhou J, Hu H, Jiang Y, Shi Q, Chen Y, Yang Q, Du X, et al. Unveiling the clonal dynamics and transmission mechanism of carbapenem-resistant klebsiella pneumoniae in the ICU environment. Int J Antimicrob Agents. 2025;66(3):107532. [DOI] [PubMed] [Google Scholar]
  • 23.Chen YT, Chen LK, Jiang RS, Chang KC, Tseng CC. Phage-based decontamination: Exploring resistance, disinfectant tolerance, and virulence trade-offs in Acinetobacter baumannii. Ecotoxicol Environ Saf. 2025;303:119052. [DOI] [PubMed] [Google Scholar]
  • 24.Parks DH, Imelfort M, Skennerton CT, Hugenholtz P, Tyson GW. CheckM: assessing the quality of microbial genomes recovered from isolates, single cells, and metagenomes. Genome Res. 2015;25(7):1043–55. [DOI] [PMC free article] [PubMed] [Google Scholar]
  • 25.Liang Q, Liu C, Xu R, Song M, Zhou Z, Li H, Dai W, Yang M, Yu Y, Chen H. fIDBAC: A Platform for Fast Bacterial Genome Identification and Typing. Front Microbiol. 2021;12:723577. [DOI] [PMC free article] [PubMed] [Google Scholar]
  • 26.Hua X, Liang Q, Fang L, He J, Wang M, Hong W, Leptihn S, Wang H, Yu Y, Chen H. Bautype: Capsule and Lipopolysaccharide Serotype Prediction for Acinetobacter baumannii Genome. Infect Microbes Dis. 2020;2(1):18–25. [Google Scholar]
  • 27.Siguier P, Perochon J, Lestrade L, Mahillon J, Chandler M. ISfinder: the reference centre for bacterial insertion sequences. Nucleic Acids Res. 2006;34(Database issue):D32–36. [DOI] [PMC free article] [PubMed] [Google Scholar]
  • 28.Lam MMC, Koong J, Holt KE, Hall RM, Hamidian M. Detection and Typing of Plasmids in Acinetobacter baumannii Using rep Genes Encoding Replication Initiation Proteins. Microbiol Spectr. 2023;11(1):e0247822. [DOI] [PMC free article] [PubMed] [Google Scholar]
  • 29.Sun L, Guo H, Chen Y, He F, Xu J. Genomic and phylogenetic analysis of a multidrug-resistant Pseudomonas aeruginosa ST463 strain co-carrying blaKPC-2, blaOXA-246, and blaOXA-486 in China. J Glob Antimicrob Resist. 2023;33:301–3. [DOI] [PubMed] [Google Scholar]
  • 30.Price MN, Dehal PS, Arkin AP. FastTree: computing large minimum evolution trees with profiles instead of a distance matrix. Mol Biol Evol. 2009;26(7):1641–50. [DOI] [PMC free article] [PubMed] [Google Scholar]
  • 31.Letunic I, Bork P. Interactive Tree Of Life (iTOL) v5: an online tool for phylogenetic tree display and annotation. Nucleic Acids Res. 2021;49(W1):W293–6. [DOI] [PMC free article] [PubMed] [Google Scholar]
  • 32.Zhou E, Qiao D, Yang Y, Xu D, Lu Y, Wang J, Smith JA, Li H, Zhao H, Liaw PK, et al. A novel Cu-bearing high-entropy alloy with significant antibacterial behavior against corrosive marine biofilms. J Mater Sci Technol. 2020;46:201–10. [Google Scholar]
  • 33.Li WY, Wang ML, Wang XD, Wang TM, Li TJ, Lu YP. A novel Co-free high-entropy alloy with excellent antimicrobial and mechanical properties. Rare Met. 2024;44(1):1–10. [Google Scholar]
  • 34.Castillo-Ramirez S. Genomic epidemiology of Acinetobacter baumannii goes global. mBio. 2023;14(6):e0252023. [DOI] [PMC free article] [PubMed] [Google Scholar]
  • 35.Li Y, Ge H, Zhou H, Zhou W, Zheng J, Chen W, Cao X. Impact of environmental cleaning on the colonization and infection rates of multidrug-resistant Acinetobacter baumannii in patients within the intensive care unit in a tertiary hospital. Antimicrob Resist Infect Control. 2021;10(1):4. [DOI] [PMC free article] [PubMed] [Google Scholar]
  • 36.Perez S, Innes GK, Walters MS, Mehr J, Arias J, Greeley R, Chew D. Increase in Hospital-Acquired Carbapenem-Resistant Acinetobacter baumannii Infection and Colonization in an Acute Care Hospital During a Surge in COVID-19 Admissions - New Jersey, February-July 2020. MMWR Morb Mortal Wkly Rep. 2020;69(48):1827–31. [DOI] [PMC free article] [PubMed] [Google Scholar]
  • 37.Zhang X, Li F, Awan F, Jiang H, Zeng Z, Lv W. Molecular Epidemiology and Clone Transmission of Carbapenem-Resistant Acinetobacter baumannii in ICU Rooms. Front Cell Infect Microbiol. 2021;11:633817. [DOI] [PMC free article] [PubMed] [Google Scholar]
  • 38.Wareth G, Linde J, Nguyen NH, Nguyen TNM, Sprague LD, Pletz MW, Neubauer H. WGS-Based Analysis of Carbapenem-Resistant Acinetobacter baumannii in Vietnam and Molecular Characterization of Antimicrobial Determinants and MLST in Southeast Asia. Antibiot (Basel). 2021;10(5):563. [DOI] [PMC free article] [PubMed]
  • 39.Kelemen J, Sztermen M, Dakos EK, Budai J, Katona J, Szekeressy Z, Sipos L, Papp Z, Stercz B, Dunai ZA et al. Complex Infection-Control Measures with Disinfectant Switch Help the Successful Early Control of Carbapenem-Resistant Acinetobacter baumannii Outbreak in Intensive Care Unit. Antibiotics (Basel). 2024;13(9):869. [DOI] [PMC free article] [PubMed]
  • 40.Namaki Kheljan M, Hassanzadeh M, Srdari Jabedar M, Mohammadi Gollou A, Ashouri P, Teimourpour R, Arzanlou M. Characterization of disinfectant susceptibility profiles among clinical isolates of Acinetobacter baumannii in Ardabil, Iran. Acta Microbiol Immunol Hung. 2023;70(4):311–7. [DOI] [PubMed] [Google Scholar]
  • 41.Chen S, Fu J, Zhao K, Yang S, Li C, Penttinen P, Ao X, Liu A, Hu K, Li J, et al. Class 1 integron carrying qacEDelta1 gene confers resistance to disinfectant and antibiotics in Salmonella. Int J Food Microbiol. 2023;404:110319. [DOI] [PubMed] [Google Scholar]
  • 42.Zhao F, Liu H, Yao Y, Zhang L, Zhou Z, Leptihn S, Yu Y, Hua X, Fu Y. Description of a Rare Pyomelanin-Producing Carbapenem-Resistant Acinetobacter baumannii Strain Coharboring Chromosomal OXA-23 and NDM-1. Microbiol Spectr. 2022;10(4):e0214422. [DOI] [PMC free article] [PubMed] [Google Scholar]
  • 43.Zeighami H, Valadkhani F, Shapouri R, Samadi E, Haghi F. Virulence characteristics of multidrug resistant biofilm forming Acinetobacter baumannii isolated from intensive care unit patients. BMC Infect Dis. 2019;19(1):629. [DOI] [PMC free article] [PubMed] [Google Scholar]
  • 44.Mendes SG, Combo SI, Allain T, Domingues S, Buret AG, Da Silva GJ. Co-regulation of biofilm formation and antimicrobial resistance in Acinetobacter baumannii: from mechanisms to therapeutic strategies. Eur J Clin Microbiol Infect Dis. 2023;42(12):1405–23. [DOI] [PMC free article] [PubMed] [Google Scholar]
  • 45.Pallos P, Gajdacs M, Urban E, Szabados Y, Szalai K, Hevesi L, Horvath A, Kuklis A, Morjaria D, Iffat W, et al. Characterization of antibiotic and disinfectant susceptibility in biofilm-forming Acinetobacter baumannii: A focus on environmental isolates. Eur J Microbiol Immunol (Bp). 2024;14(2):126–33. [DOI] [PMC free article] [PubMed] [Google Scholar]
  • 46.Ren G, Huang L, Hu K, Li T, Lu Y, Qiao D, Zhang H, Xu D, Wang T, Li T, et al. Enhanced antibacterial behavior of a novel Cu-bearing high-entropy alloy. J Mater Sci Technol. 2022;117:158–66. [DOI] [PMC free article] [PubMed] [Google Scholar]

Associated Data

This section collects any data citations, data availability statements, or supplementary materials included in this article.

Supplementary Materials

Supplementary Material 1. (865.9KB, docx)

Data Availability Statement

All sequence data and genome assemblies generated in this study have been submitted to GenBank under the BioProject accession numbers PRJNA738868 and PRJNA1034534.

Articles from BMC Microbiology are provided here courtesy of BMC

RESOURCES