Genomic Characterization of Carbapenemases in Providencia Species from Croatia: The Results of the Multicenter Study
Jasmina Vraneš, Branka Bedenić, Gernot Zarfel, Josefa Luxner, Andrea Grisold, Rocio Arazo del Pino, Tessa Burgwinkel, Haris Car, Maja Anušić, Vladimira Tičić, Marina Bubonja-Šonje, Sanda Sardelić, Paul G. Higgins
TL;DR
This study analyzed carbapenem-resistant Providencia species in Croatia to understand their resistance mechanisms and spread.
Contribution
The study provides the first genomic characterization of carbapenemases in Providencia species from Croatia.
Findings
OXA-48 and NDM carbapenemases were identified in 15 isolates each.
WGS revealed multiple resistance genes to various antibiotics, including aminoglycosides and tetracyclines.
Isolates were categorized as extensively drug-resistant (XDR) with limited treatment options.
Abstract
Background/objectives: A rise in infections associated with carbapenem-resistant Providencia species (CRPS) has been observed worldwide. This study presents a genomic analysis of CRPS isolates from four hospitals in Croatia and the outpatient setting, in order to determine the extent of the spread of CRPS in Croatia. In the present study, we applied a combination of phenotypic characterization and molecular analysis of resistance traits to determine the mechanisms and the routes of spread of CRPS. Material and methods: The antibiotic susceptibility testing was performed using disk-diffusion and broth dilution methods. The nature of extended-spectrum β-lactamases (ESBLs), carbapenemases, and fluoroquinolone resistance determinants was investigated by polymerase chain reaction (PCR). In order to obtain an insight into the whole resistome, selected isolates were subjected to the Interarray…
Genes, proteins, chemicals, diseases, species, mutations and cell lines named across the full text — each resolved to its canonical identifier and authoritative record.
Click any figure to enlarge with its caption.
Figure 1| PROTOCOL NUMBER | AGE AND GENDER | SPECIMEN | CLINICAL | OUTCOME | CENTER AND | TZP | CAZ | CTX | CRO | FEP | IMI | MEM | AMI | CIP | β-LACTAMASE CONTENT | |
|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|
| 1 | 45767/2-24 | 93 | Urine- | UTI | D | OUTPATIENT | ≥128 (R) | ≥128 (R) | ≥128 (R) | 32 (R) | 32 (R) | 16 (R) | 16 (R) | 16 (S) | 32 (R) | NDM, |
| 2 | 39520/2-24 | 80 M | Urine- | UTI | D | OUTPATIENT | 128 (R) | ≥128 (R) | ≥128 (R) | 64 (R) | ≥128 (R) | 8 (R) | 8 (R) | 16 (S) | 64 (R) | NDM |
| 3 | 48059/2-24 | 86 | Urine- | Dementia | NA | SV. IVAN | 128 (R) | 16 (R) | ≥128 (R) | 64 (R) | ≥128 (R) | 8 (R) | 16 (R) | ≥128 (R) | 128 (R) | OXA-48, |
| 4 | 49097/2-24 | 81 | Urine- | Dementia | NA | SV. IVAN | 128 (R) | 4 (R) | ≥128 (R) | ≥128 (R) | 32 (R) | 16 (R) | 32 (R) | ≥128 (R) | ≥128 (R) | OXA-48, |
| 5 | 46920/2-24 | 68 | Urine- | Dementia | NA | SV. IVAN | 128 (R) | 4 (R) | ≥128 (R) | 8 (R) | 64 (R) | 8 (R) | 32 (R) | ≥128 (R) | ≥128 (R) | OXA-48, |
| 6 | 30022/2-24 | 75 | Urine- | Dementia | NA | SV. IVAN | ≥128 (R) | 4 (R) | ≥128 (R) | 32 (R) | 128 (R) | 16 (R) | 64 (R) | ≥128 (R) | ≥128 (R) | OXA-48, |
| 7 | 19729/2-24 | 82 | Urine- | Dementia | NA | SV. IVAN | ≥128 (R) | 4 (R) | ≥128 (R) | ≥128 (R) | 32 (R) | 32 (R) | 32 (R) | ≥128 (R) | ≥128 (R) | OXA-48, |
| 8 | 47950/2-24 | 87 | Urine- | Dementia | NA | SV. IVAN | ≥128 (R) | 8 (R) | ≥128 (R) | ≥128 (R) | 64 (R) | 16 (R) | 32 (R) | ≥128 (R) | 64 (R) | OXA-48, |
| 9 | 28170/2-24 | 83 | Urine- | Dementia | NA | SV. IVAN | ≥128 (R) | 4 (R) | ≥128 (R) | ≥128 (R) | 16 (R) | 16 (R) | 32 (R) | ≥128 (R) | 64 (R) | OXA-48, |
| 10 | 50091/2-24 | 89 | Urine- | Dementia | NA | SV. IVAN | ≥128 (R) | 8 (R) | ≥128 (R) | 32 (R) | 32 (R) | 32 (R) | 32 (R) | ≥128 (R) | 64 (R) | OXA-48, |
| 11 | 50257/2-24 | 95 | Urine- | Dementia | NA | SV. IVAN | ≥128 (R) | 8 (R) | ≥128 (R) | 64 (R) | 128 (R) | 64 (R) | 128 (R) | ≥128 (R) | ≥128 (R) | OXA-48, |
| 12 | 48272/-24 | 95 | Urine- | Dementia | NA | SV. IVAN | ≥128 (R) | 128 (R) | ≥128 (R) | 16 (R) | 16 (R) | 16 (R) | 32 (R) | ≥128 (R) | 128 (R) | OXA-48, |
| 13 | 30472/2-24 | 95 | Urine- | Dementia | NA | SV. IVAN | ≥128 (R) | 4 (R) | ≥128 (R) | 32 (R) | 64 (R) | 16 (R) | 32 (R) | ≥128 (R) | ≥128 (R) | OXA-48, |
| 14 | UG3084 | 55 M | Urine- | Chronic myeloid leukemia | S | UHS | ≥128 (R) | 4 (R) | ≥128 (R) | 64 (R) | 64 (R) | 16 (R) | 32 (R) | 16 (S) | 128 (R) | OXA-48, |
| 15 | UG48005 | 53 M | Urine | Non-Hodkin lymphoma | S | UHS | ≥128 (R) | 32 (R) | ≥128 (R) | ≥128 (R) | 64 (R) | 8 (R) | 32 (R) | 8 | 64 (R) | OXA-48, |
| 16 | 68511 PS | 85 F | Urine | UTI | D | CHCR | ≥128 (R) | ≥128 (R) | ≥128 (R) | ≥128 (R) | 32 (R) | 32 (R) | 32 (R) | 16 (S) | 4 (R) | NDM |
| 17 | 69188 PS | 70 M | Urine | Bowel cancer, UTI | S | CHCR | ≥128 (R) | ≥128 (R) | ≥128 (R) | ≥128 (R) | 128 (R) | 32 (R) | 32 (R) | 32 (R) | 4 (R) | NDM |
| 18 | 60953 | 83 F | Urine | Cerebral stroke, UTI | D | CHCR | ≥128 (R) | ≥128 (R) | ≥128 (R) | ≥128 (R) | 128 (R) | 32 (R) | 32 (R) | 16 (S) | 4 (R) | NDM |
| 19 | 62849PR | 74 | Urine | Diabetes mellitus, UTI | D | CHCR | ≥128 (R) | ≥128 (R) | ≥128 (R) | ≥128 (R) | 128 (R) | 32 (R) | 32 (R) | 16 (S) | 8 (R) | NDM |
| 20 | 63703 PR | 60 M | Urine | Convulsions | S | CHCR | 128 (R) | ≥128 (R) | ≥128 (R) | ≥128 (R) | 128 (R) | 32 (R) | 32 (R) | 16 (S) | 4 (R) | NDM |
| 21 | 61738 PR | 77 | BC | UTI | S | CHCR | 128 (R) | ≥128 (R) | ≥128 (R) | ≥128 (R) | 128 (R) | 32 (R) | 32 (R) | 16 (S) | 4 (R) | NDM |
| 22 | 54062 PR | 78 | Rectum swab | Cranial trauma | S | CHCR | ≥128 (R) | ≥128 (R) | ≥128 (R) | ≥128 (R) | 64 (R) | 32 (R) | 32 (R) | 16 (S) | 4 (R) | NDM |
| 23 | 35649 PR | 68 | Urine | Acute respiratory failure | S | CHCR | ≥128 (R) | ≥128 (R) | ≥128 (R) | ≥128 (R) | 128 (R) | 16 (R) | 32 (R) | 16 (S) | 128 (R) | NDM |
| 24 | 42237 PR | 77 | Urine | Cerebral stroke, UTI | D | CHCR | 64 (R) | ≥128 (R) | ≥128 (R) | 64 (R) | 128 (R) | 32 (R) | 128 (R) | 16 (S) | 4 (R) | NDM |
| 25 | 32488 PS | 52 M | Rectum | UTI | S | CHCR | ≥128 (R) | ≥128 (R) | ≥128 (R) | ≥128 (R) | ≥128 (R) | 32 (R) | 32 (R) | 16 (S) | 4 (R) | NDM |
| 26 | 29664 PS | 88 M | Urine | Acute respiratory failure, UTI | S | CHCR | ≥128 (R) | 64 (R) | ≥128 (R) | ≥128 (R) | ≥128 (R) | 16 (R) | 32 (R) | 16 (S) | 64 (R) | NDM |
| 27 | 22994 PR | 81 | Urine | UTI, | D | CHCR | ≥128 (R) | 128 (R) | ≥128 (R) | ≥128 (R) | ≥128 (R) | 16 (R) | 32 (R) | 16 (S) | 64 (R) | NDM |
| 28 | 53151 PR | 52 | Tracheal aspirate | Brain damage, pleural effusion | H | CHCR | 64 (R) | 128 (R) | ≥128 (R) | 64 (R) | 32 (R) | 16 (R) | 32 (R) | 8 | 4 (R) | NDM |
| 29 | 14811 PS | 44 M | Wound swab | Wound infection | S | UHSM | ≥128 (R) | ≥128 (R) | 32 (R) | 16 (R) | 128 (R) | 32 (R) | 32 (R) | ≥128 (R) | ≥128 (R) | OXA-48, CTX-M-15 |
| 30 | 20585 PS | 32 M | Tracheal aspirate | S | UHSM | ≥128 (R) | ≥128 (R) | ≥128 (R) | ≥128 (R) | 64 (R) | 8 (R) | 64 (R) | 4 (S) | 32 (R) | OXA-48, CTX-M-15 |
- —Extended-spectrum β-lactamases, plasmid-mediated β-lactamases, and carbapenemases in Proteus mirabilis
- —University of Zagreb
- —Croatian Science Foundation
Peer Reviews
No public reviews on file for this paper yet. If you reviewed it on a platform where reviews are public (OpenReview, ICLR, NeurIPS, ICML), you can paste yours below so the community can read it here.
Videos
No videos yet. Explain this paper in a talk, walkthrough, or lecture? Add one.
Taxonomy
TopicsAntibiotic Resistance in Bacteria · Bacterial Identification and Susceptibility Testing · Nosocomial Infections in ICU
1. Introduction
Providencia spp. belong to the genus Morganellaceae within the family Enterobacterales and are increasingly identified as causative agents of urinary tract infections (UTI), wound infections, and skin and soft-tissue infections. Recently, a rise in infections caused by carbapenem-resistant Providencia species (CRPS) has been observed worldwide [1].
The species most frequently involved in human infections are Providencia stuartii and Providencia rettgeri [1]. They have intrinsic resistance to natural and semisynthetic penicilins, β-lactam-inhibitor combinations, and first-generation cephalosporins, due to the production of AmpC Ambler class C β-lactamase. Overproduction of AmpC can be due to induction or derepression and is associated with resistance to second- and third-generation or extended-spectrum cephalosporins (ESCs) such as ceftazidime, cefotaxime, and ceftriaxone. Resistance to ESC can also be mediated by production of extended-spectrum β-lactamases (ESBL) or overproduction of chromosomal or acquisition of plasmid-mediated AmpC β-lactamases (p-AmpC) [2].
The three main families of ESBLs are: TEM (Temoniera), SHV (suphydryle-variable), and CTX-M (cefotaximase-Munich). TEM and SHV variants are derived by mutations of parental broad-spectrum TEM-1, TEM-2, and SHV-1 β-lactamases [2]. CTX-M β-lactamases are class A enzymes, which are not closely related to TEM or SHV β-lactamases [3,4]. In contrast to TEM and SHV β-lactamases, which rely on amino acid substitutions to extend their substrate profile, CTX-M enzymes are native ESBLs. CTX-M β-lactamases are divided into five phylogenetic clusters based on their protein sequences: CTX-M-1, CTX-M-2, CTX-M-8, CTX-M-9, and CTX-M-25. AmpC β-lactamases hydrolyze ESC, monobactams, and cephamycins but spare fourth-generation cephalosporins and carbapenems. Unlike most ESBLs, they are not susceptible to inhibition with clavulanic acid, sulbactam, or tazobactam [5]. Carbapenems are the antibiotics of choice for treating infections caused by derepressed Amp-C or ESBL producers. However, increased use of carbapenems has led to the emergence of carbapenemases in Enterobacterales, including Providencia spp. The widespread occurrence of ESBL-producing Providencia spp. was noticed earlier, but the dissemination of CRPS has been observed much later [1].
Resistance to carbapenems in Providencia spp. is mediated primarily by two main mechanisms. The first involves overexpression of β-lactamase, such as AmpC or an ESBL, combined with decreased permeability due to porin loss or alteration. The second mechanism is related to the production of carbapenem-hydrolyzing β-lactamases belonging to Ambler class A serine β-lactamases (KPC, GES), class B metallo-β-lactamases (MBL) of the IMP, VIM, or NDM family, and class D OXA-48-like β-lactamases, also known as carbapenem-hydrolyzing oxacillinases (CHDL) [6,7]. Production of carbapenemases in Providencia species poses a serious therapeutic problem because of their intrinsic resistance to polymyxins and tigecycline [1].
Croatia belongs to the Southeastern European countries with a high rate of carbapenem-resistant Enterobacterales (CRE). While carbapenemases in other Enterobacterales species have been extensively investigated before, there are no bibliographical data referring to CRPS. This study presents a genomic analysis of CRPS isolates from four hospitals in Croatia and outpatients setting, in order to determine the extent of spread of CRPS in Croatia. In the present study, we applied a combination of phenotypic characterization and molecular analysis of resistance traits in order to determine the mechanisms and routes of spread of CRPS. All isolates were first screened by phenotypic methods and conventional PCR to determine resistance phenotype and presence of β-lactam and fluoroquinolone resistance genes. Selected isolates were then analyzed using new molecular techniques to determine the whole resistome. The second goal was to compare resistance patterns among isolates with different carbapenemase types.
2. Materials and Methods
2.1. Bacterial Isolates
The isolates exhibiting reduced susceptibility to at least one carbapenem (imipenem, meropenem, or ertapenem) licensed in Croatia were collected from Dr Andrija Štampar Teaching Public Health Institute (ASPH) covering the outpatient setting in Zagreb and Sv. Ivan Zelina psychiatric hospital in Zagreb, University Hospital Center Sestre Milosrdnice (UHCSM) in Zagreb, University Hospital of Split (UHS), and Clinical Hospital Center Rijeka (CHCR) located in the coastal region of Croatia, from 19 March 2023 to 28 June 2025. The isolates were sent to the Clinical Department for Clinical Microbiology, Infection Prevention and Control of the University Hospital Centre Zagreb for phenotypic and molecular analysis. The taxonomy was investigated using MALDI-TOF MS (matrix-assisted laser desorption ionization–time of flight mass spectrometry) (Vitek MS, Bruker, Bremen, Germany). The selected isolates were subjected to the whole resistome analysis by an Interarray CarbaResist Kit, Bad Langensalza, Germany at the Diagnostic and Research Institute for Hygiene, Microbiology, and Environmental Medicine of the University in Graz, Austria, and to the Institute for Medical Microbiology, Immunology, and Hygiene of the University in Cologne, Germany, for whole genome sequencing (WGS).
2.2. Antimicrobial Susceptibility Testing and Phenotypic Tests for Detection of ESBLs, AmpC β-Lactamases, and Carbapenemases
The antimicrobial susceptibility profiles were verified by a disk-diffusion test (DDT) according to EUCAST guidelines [8] and the broth dilution method, following the latest version of the Clinical Laboratory Standard Institution (CLSI) guidelines [9]. The panel for routine DDT contained: amoxicillin (AMX 25 µg), amoxicillin-clavulanate (AMC-20/10 µg), piperacillin-tazobactam (TZP-110 µg), cephalexin (CN-30 µg), cefuroxime (CXM-30 µg), ceftazidime (CAZ-10 µg), cefotaxime (CTX-10 µg), ceftriaxone (CRO-10 µg), cefepime (FEP-30 µg) imipenem (IMI-10 µg), meropenem (MEM-10 µg), ertapenem (ERT-10 µg), gentamicin (GM-10 µg), amikacin (AMI-30 µg), norfloxacin (NOR), ciprofloxacin (CIP-5 µg), levofloxacin (LVX-5 µg), chloramphenicol (CHL-30 µg), cefoxitin (FOX-30 µg), ceftazidime-avibactam (CZA-14 µg), and cefiderocol (FDC-30 µg) and, for urine samples, additionally, sulphamethoxazole-trimethoprim (SMX-25 µg) (Oxoid, Basingstoke, UK). The minimum-inhibitory concentrations (MICs) were determined by the broth microdilution method in Mueller–Hinton broth (MHB) (Biolife Italiana, Milan, Italy) to assess susceptibility to 13 antibiotics: amoxicillin alone and in combination with clavulanic acid, piperacillin-tazobactam, cefuroxime, ceftazidime, cefotaxime, ceftriaxone, cefepime, imipenem, meropenem, gentamicin, amikacin, and ciprofloxacin according to CLSI [9] except for amoxicillin alone, for which EUCAST criteria were applied (resistance breakpoint 8 mg/L) due to the lack of CLSI recommendation. Serial dilutions were prepared ranging from 0.06 to 128 µg/mL. MICs were read by naked eye, with the plates at the mirror, as the lowest concentration of an antibiotic that inhibited visible growth after 18 h at 37 °C. Escherichia coli ATCC 25922 and Klebsiella pneumoniae ATCC 700603 were used as quality control strains. The isolates were classified as multidrug-resistant (MDR), extensively drug-resistant (XDR), or pandrug-resistant (PDR), as described previously by Magiorakos et al. [10].
The double disk synergy test (DDST) [8] was carried out in the frames of routine laboratory analysis of the isolates in line with EUCAST guidelines. In brief, overnight culture of the test strain was adjusted to MacFarland 0.5 and inoculated on Mueller–Hinton (MH) agar (Biolife Italiana, Milan, Italy). A disk containing AMC (20/10 µg) was placed in the middle of the plate, and the surrounding ESC disks were placed 2 cm apart from the central disk. Distortion of the inhibition zone around the cephalosporin disk toward the central disk with clavulanic acid was considered suspicious of an ESBL. ESBL production was confirmed by CLSI combined disk test using disks with ESC alone or with the addition of clavulanic acid [9]. A difference of ≥5 mm between the zone diameters of either of the cephalosporin disks and their respective cephalosporin clavulanate disk is taken to be phenotypic confirmation of ESBL production. E. coli ATCC 25922 and K. pneumoniae ATCC 700603 were used as negative and positive control strains, respectively. AmpC β-lactamase overproduction or plasmid-mediated AmpC (p-AmpC) positivity was screened based on reduced susceptibility to FOX and confirmed by the combined disk test using ESC disks supplemented with cloxacillin [11]. A modified Hodge test (MHT) [12] and the carbapenem-inactivation method (CIM) were used to screen for the presence of carbapenemases [13]. MHT was carried out according to the previously published protocol. In brief, E. coli ATCC 25922, susceptible to carbapenems, was suspended in saline and inoculated on MacConkey agar (Biolife, Milan, Italy) to produce confluent growth. An MEM disk (10 µg) was placed in the center of the plate. The test isolates were streaked from the edge of the disk to the edge of the plate. Carbapenemase was suspected if the clover-leaf indentation of the indicator organism was observed toward the meropenem disc. Since MHT was found to have limited sensitivity, we applied an alternative method. CIM was performed according to the CLSI recommendations. In brief, a suspension of the test strain was adjusted to McFarland 0.5 (10^8^ CFU/mL), and an MEM disk was placed in the suspension. The suspension was incubated for 2 h at 37 °C. E. coli ATCC 25922 was inoculated on MH agar, and the disk was placed in the middle of the plate. The plates were read after overnight incubation. The lack of an inhibition zone, decreased inhibition zone (<15 mm), or colonies within the inhibition zone indicated the presence of carbapenemase.
As a part of routine diagnostic procedures, the isolates with reduced susceptibility to carbapenems in the disk-diffusion test were subjected to the immunochromatographic OKNVI (OXA-48, KPC, NDM, VIM, IMP) (O.K.N.V.I. Resist-5, Coris Bioconcept, Gembloux, Belgium) [14] in the participating centers to determine the type of carbapenemase. Testing was performed as per the manufacturer’s instructions. Briefly, one bacterial colony was taken from the 5% Sheep Blood agar (Biolife Italiana, Milan, Italy) and homogenized in the LY-A buffer. Three drops were then added to the sample well of each of the two cassettes (KPC/OXA-48 and NDM/VIM/IMP). Results were recorded after 15 min.
2.3. Conjugation
Conjugal transfer of ertapenem and cefotaxime resistance was attempted in mixed broth cultures at 35 °C. Donor and recipient strains were grown to the exponential stage, mixed in a ratio of 1:1, and incubated overnight at 37 °C. We used E. coli J65 resistant to sodium azide as the recipient [15]. The ESBL and carbapenemase-producing transconjugants were selected on MacConkey agar containing either ERT (0.5 mg/L) or CTX (2.0 mg/L) to inhibit the growth of the donors and sodium azide (100 mg/L) to suppress the recipient strain. Serial dilutions of donor and recipient strains and the mating mixture were prepared in order to determine the frequency of conjugation. It was determined relative to the number of donor cells.
2.4. Molecular Detection of Resistance Genes
The test was completed on all 30 isolates. An in-house DNA extraction was performed. A fresh bacterial colony was suspended in 100 µL of sterile, distilled water and boiled at 100 °C for 15 min. After centrifugation, the supernatant was used as a DNA template. Genotypic confirmation of resistance genes was completed by PCR targeting genes, conferring resistance to β-lactams, including broad and extended-spectrum β-lactamases (blaSHV, blaTEM, blaCTX-M, and blaPER-1) [16,17,18,19] and fluoroquinolone resistance genes (qnrA, qnrB, and qnrS) [20]. Genes encoding five phylogenetic clusters of CTX-M β-lactamases (CTX-M-1, CTX-M-2, CTX-M-8, CTX-M-9, and CTX-M-25) [21], plasmid-mediated AmpC β-lactamases [22], and carbapenemases belonging to the class A (blaKPC), class B carbapenemases (blaVIM, blaIMP, and blaNDM), and CHDL (blaOXA-48-like) [23] were detected by multiplex PCR according to the previously published protocols. The primer sequences, annealing temperature, and amplicon sizes are provided in Table 1. The primers and the master mix were provided by Takara Bio Inc. San Jose, CA, USA. PCR products were separated by electrophoresis in a 1.5% agarose gel for 1 h at 100 V in 1× Tris-acetate-EDTA (TAE) buffer and stained with ethidium bromide. A 100 bp DNA ladder (Medical Intertrade, Zagreb, Croatia) was used as a molecular mass marker. DNA bands were visualized using a UV Transilluminator. Genetic context of blaCTX-M genes was determined by PCR mapping with forward primer for ISEcp1 and IS26 combined with primer MA-3 (universal reverse primer for blaCTX-M genes) [24]. Flanking regions of blaOXA-48 genes were analyzed by PCR using a forward primer for IS1999 and a reverse primer for blaOXA-48 [25]. Amplicons were sequenced in both forward and reverse directions in Eurofins Genomic Service (https://eurofinsgenomics.eu/, accessed on 1 October 2025).
2.5. Interarray Genotyping Kit CarbaResist
The genotyping of four randomly selected P. stuartii isolates was carried out with the microarray-based CarbaResist Genotyping Kit, in line with the manufacturer’s instructions, version 1012012100004 (INTERARRAY, fzmb GmbH, Bad Langensalza, Germany). Genomic DNA was extracted from monoclonal overnight cultures with the Qiagen DNeasy Blood and Tissue Kit (Qiagen, Hilden, Germany). The unfragmented DNA was then amplified by one primer for each target sequence (antisense) and was internally labeled with biotin dUTP. The obtained ssDNA products were then transferred into the ArrayWells to perform the hybridization. The wells contained 230 probes for the distinct genes encoding the most relevant carbapenemases, ESBL and AmpC, aminoglycoside (aac, aad, aph, arm, and rmt), sulfonamide (sul), trimethoprim (dfr), and fluoroquinolone resistance (qnrA, B, C, and S). The wells were washed to remove any unbound DNA, and horseradish peroxidase (HRP)-conjugated streptavidin was bound to all of the hybridized sections, resulting in dark spots on the chip due to an enzymatic reaction. The detection of these spots and data acquisition was carried out automatically by the INTER-VISION Reader.
2.6. Characterization of Plasmids
Plasmids were extracted with the Qiagen Plasmid Mini Kit (Qiagen, Hilden, Germany) according to the manufacturer’s instructions. After staining with ethidium bromide, the DNA was visualized by ultraviolet light. PCR-based replicon typing (PBRT) [26] was applied to determine the plasmid content of the tested isolates. PBRT can be inefficient in identifying IncL/M plasmid types, and thus, we used an updated method designed to identify and distinguish between IncL and IncM plasmids [27]. Positive control strains for PBRT were kindly provided by Dr A. Carattoli (Instituto Superiore di Sanita, Rome, Italy).
2.7. Whole Genome Sequencing (WGS)
Nine representative isolates (six with OXA-48 and three with NDM) had their genomes sequenced as previously described [28]. Briefly, isolates were grown overnight on blood agar plates and had their DNA isolated using the DNeasy UltraClean Microbial Kit (Qiagen, Hilden, Germany). Sequencing libraries were prepared with the NEB Next^®^ Ultra™ II FS DNA Library Prep Kit for Illumina^®^ (New England Biolabs, Frankfurt, Germany). The libraries were sequenced, 2 × 250 bp paired end, in an Illumina MiSeq (Illumina GmbH, Munich, Germany). The resulting Fasta files were assembled using the SKESA assembler (Version 2.4.0) as part of the Ridom SeqSphere+ software (Ridom GmbH, Münster, Germany, version 10.5.5) SKESA (version 2.4.0) and AMRFinderPlus (version 2024-12-18.1 (4.0)) are integrated into MBioSEQ Ridom version 10.5.5 (formerly known as SeqSphere) [29].
A core-genome MLST scheme (cgMLST) developed for P. stuartii, based on a core genome of 3079 target alleles (core) and 665 target alleles (accessory), was used to determine clonality. The 7-loci MLST sequence types were extracted from the assembled genome using PubMLST. Antimicrobial resistance genes were identified using NCBI AMRFinderPlus [30].
3. Results
3.1. Isolates and Patients
In total, 30 non-duplicate (one per patient) isolates were collected: 13 from ASPH (11 from Sv. Ivan Zelina Hospital and 2 from outpatients) and CHCR, respectively, and 2 from UHS and UHCSM, respectively. There were 22 P. stuartii and 8 P. rettgeri isolates. The patient’s age ranged from 32 to 95 (mean value 74.8), and the study included 21 females and 9 males. Most of the isolates originated from the urinary tract (14 from urinary catheter and 10 from midstream urine). All patients with urinary tract infection (UTI) had >10^5^ CFU/mL of Providencia spp. in a pure culture and pyuria (>5 white blood cells/high power field) or a positive test for leukocyte esterase. There was one isolate from blood culture and wound swab, respectively, and four from surveillance cultures (rectum swab and endotracheal aspirate). Twenty-six patients had an infection (urinary tract, wound, and bloodstream infection), whereas four were only colonized. The clinical and epidemiological data are summarized in Table 2.
3.2. Antimicrobial Susceptibility Testing and Phenotypic Tests for Detection of ESBLs, AmpC β-Lactamases, and Carbapenemases
The isolates were fully resistant to CXM, CAZ, CRO, CTX, FEP, TZP, IMI, MEM, ERT, SMX, and CHL, in addition to the antibiotics to which they are intrinsically resistant (AMX, TET, and GM), with MICs of most cephalosporins exceeding 128 mg/L, as shown in Table 2. Moderate resistance rates were observed for AMI (40%) (Table 2). One isolate tested resistant to FOX. There was a marked difference in the AMI-resistant rates between OXA-48 and NDM-producing organisms (73% vs. 0%). Interestingly, 47% of the isolates were susceptible to GM in spite of the intrinsic resistance of Providencia spp. According to initial DDT, all isolates were resistant to AMX, AMC, TZP, CN, CXM, CAZ, CTX, FEP, IMI, MEM, ERT, and CIP. The rates of AMI, CZA, and FDC resistance were 40% (12/30), 50% (15/30), and 37% (11/30), respectively. The immunochromatographic test demonstrated OXA-48 and NDM carbapenemases in 15 isolates each. KPC carbapenemases were not detected. Phenotypic tests for ESBLs were positive in all OXA-48 and one NDM-positive organism (16 isolates). Screening for p-AmpC was positive in one isolate, which was confirmed as p-AmpC positive by an inhibitor-based test with cloxacillin. The isolates were categorized as XDR. OXA-48-producing isolates were susceptible only to CZA, whereas NDM producers tested susceptible to FDC and the majority to AMI (93%) n = 14) as well. MHT successfully identified both OXA-48 and NDM carbapenemase. In contrast, the CIM test demonstrated high sensitivity only for OXA-48 (93%, n = 14) but very poor performance with NDM producers (20%, n = 3).
3.3. Conjugation
The isolates failed to transfer either imipenem or cefotaxime resistance to the E. coli recipient strain.
3.4. Molecular Detection of Resistance Genes
PCR confirmed the results of the preliminary immunochromatographic testing. blaOXA-48-like and blaNDM genes were found in 15 isolates, respectively. blaKPC genes were not detected. All 16 isolates phenotypically positive for an ESBL yielded a PCR product with primers specific for the blaCTX-M gene, belonging to the phylogenetic cluster 1. blaTEM genes were identified in 11 isolates harboring OXA-48, all from Zagreb (Table 2). The cefoxitin-resistant strain turned out to be positive for the blaCMY gene. Sequencing of blaCTX-M genes revealed CTX-M-15 allelic variant. The ISEcp insertion element was associated with blaCTX-M genes. The IS1999 was not found.
3.5. Interarray CarbaResist Kit
An Interarray CarbaResist Kit identified identical resistance gene content in four selected OXA-48-producing organisms: blaCTX-M-15, bla_TEM_, and blaOXA-10 for β-lactam resistance (Table 3). There was a plethora of aminoglycoside resistance genes: aadA1 and aadA2 encoding adenyltransferases and arm gene for 16S rRNA methylase present in all isolates, with one isolate harboring aac(6″)Ib for acetyltransferase as well. Sulfonamide resistance was mediated by sul1 and sul2 genes, generating dihydropteroate synthetase, whereas dfrA12 and dfrA14 were linked to dihydrofolate reductase, an enzyme responsible for trimethoprim resistance (Table 3). All four isolates were positive for class 1 integron.
3.6. WGS
There were two allelic variants of blaNDM genes: blaNDM-1 and blaNDM-5 and only one blaOXA-48 variant. One NDM-positive organism harbored the plasmid-mediated AmpC, β-lactamase bla_CMY-16,_ in addition to a carbapenemase. WGS identified additional phosphorylases (aph(3″)-Ib and aph((6)-Id) in six isolates, not detected with the Interarray CarbaResist Kit, which are in line with elevated MICs of AMI (Table 4). Genes for aadA36 adenylase were present in all isolates, while aadA1 and aadA2 were found only in five OXA-48-producing organisms (Table 4). The sul1 gene was present in all but one isolate, while sul2 was carried by six isolates. Isolates harbored four allelic variants of dfr genes: dfrA1, dfrA10, and dfrA14. dfrA1 and dfrA10 were found only in NDM-positive organisms, while dfrA12 was identified in five OXA-48 producers. Tet(A) and tet(B) were responsible for TET resistance, with tet(B) being present in all tested isolates, while catA3 and catA5 encoded chloramphenicol acetyltransferase, causing CHL resistance (Table 4). One OXA-48 positive isolate tested positive for aac(6′)-Ib-cr5, which is associated with combined aminoglycoside and fluoroquinolone resistance (Table 4).
Molecular typing reveals that there were six ST22 isolates, two ST3 isolates, and one ST5 isolate (Table 4). cgMLST analysis reveals that there are two transmission clusters and three singletons (Figure 1). One cluster of four isolates was ST22, all from Sv. Ivan Zelina hospital in Zagreb, also closely related to the other two ST22 isolates, but with 17–29 alleles outside the threshold of 15 alleles. The two ST3 isolates differed in three alleles.
3.7. Plasmid Analysis
PBRT tested negative for plasmid replicons, but WGS identified IncC plasmid replicon in five of the sequenced isolates: the OXA-48-positive isolates PS7, PS9, PS8, PS4, and PS10 and the NDM-1-positive isolate PS2. The two NDM-5 isolates have a Col3M plasmid. Short-read sequencing data are unreliable for plasmid analysis; however, MOB-suite predicts that only the isolate PS14 has a carbapenemase (OXA-48) encoded on a plasmid.
4. Discussion
The overuse of carbapenems has driven the rise of CRPS in Europe, including Croatia. The rise of NDM-producing CRPS in Europe was noticed following the onset of war in Ukraine and the subsequent arrival of refugees in the European Union [1]. NDM was previously found among CRPS in Romania [31,32], Bulgaria [33], and Portugal [34], with some studies pre-dating the current conflict. The isolates from Bulgaria harbored additional CMY-4 p-AmpC. In our study, a CMY-16 allelic variant was found as an additional β-lactamase in one NDM-1-producing organism. It seems that the spread of CRPS in Europe occurred first in the Iberian Peninsula and later in Eastern Europe. Outside Europe, NDM-positive CRPS were described in Afghanistan much earlier, back in 2012 [35], and later in Israel [36]. In Greece, carbapenem resistance was due to another family of MBLs—VIM [37]. In contrast, in South America, KPC-2 was found responsible for carbapenem resistance in CRPS [38], with additional CTX-M-2 and CTX-M-9 contributing to ESC resistance. IMP variants are the rarest among CRPS and are reported only in Mexico, alongside OXA-24 and OXA-58, CHDLs, more associated with Acinetobacter baumannii [39], demonstrating the ability of P. stuartii to acquire resistance determinants from other Gram-negative bacteria. Coproduction of OXA-48 and NDM-1 was described recently in PDR P. rettgeri isolated in Turkey [40], showing the amazing capacity of the species to accumulate resistance traits. Thus, resistance determinants conferring carbapenem resistance depend on the local epidemiology. It seems that the spread of CRPS occurred in countries with a high rate of carbapenemase-producing Enterobacterales (CPE), indicating the probable transfer of blaCARB genes from K. pneumoniae and E. coli. Croatia has a high rate of CPE of approximately 6% [41], and excessive colistin usage in the hospitals provides a survival advantage to intrinsically colistin-resistant organisms like Proteus spp. and Providencia spp. [42]. With the growing burden of CPE and carbapenem-resistant A. baumannii (CRAB), colistin therapy as the salvage treatment is on the rise. The urinary tract, particularly urinary catheters, was shown to be the dominant source of CRPS. This is in line with the capacity of this genus to produce biofilm [1]. Interestingly, P. rettgeri was identified only in CHCR, whereas in the other three centers, only P. stuartii was found to harbor carbapenemase-encoding genes.
The main finding presented in this study is the equal presence of NDM and OXA-48 carbapenemases among CRPS in Croatia. All our isolates were shown to be XDR. OXA-48-positive isolates were susceptible to CZA, whereas NDM producers demonstrated sensitivity to FDC and the majority to AMI. Interestingly, FDC resistance was noted only among OXA-48-producing organisms from Zagreb. There was no FDC resistance among MBL-producing organisms from Rijeka. Thus, FDC could be considered a good therapeutic option for infections with NDM producers. However, there is also an issue with the standard used. According to EUCAST, which is used in Europe, the majority of OXA-48-positive isolates were resistant, but if CLSI criteria were applied with a lower breakpoint, all isolates would be assessed as susceptible. The isolates exhibited very variable MICs of carbapenems and ESC, but within the resistance range. This variability may be due to differential expression of ESBL and carbapenemase-encoding genes, probably due to mutations in the promotor region of the genes or variable gene copy numbers. In general, the level of ESC resistance was much higher among OXA-48 producers, likely due to the co-production of an additional ESBL. Among NDM-positive organisms, there was disconcordance between high MICs of MEM, IMI, and FEP—indicating resistance—and large inhibition zones, falling within the susceptible or intermediate susceptible range.
MHT proved to be a reliable phenotypic testing option for both OXA-48 and NDM, in contrast to CIM, which yielded a lot of false negative results. NDM is a membrane-bound lipoprotein connected to the outer membrane. Therefore, it inhibits the release of NDM carbapenemase into the extracellular medium, unlike soluble MBLs [43].
Phenotypic tests proved the production of an ESBL in all OXA-48- and one NDM-positive isolate. PCR identified the gene encoding CTX-M belonging to clonal lineage 1 with CTX-M-15 as the only allelic variant. A small outbreak of urinary tract infections with CTX-M-15-producing P. stuartii was reported in the University Hospital Centre Split in 2012 [44]. In the present study, blaCTX-M genes were preceded by the ISEcp insertion element, important for the mobilization of the ESBL coding gene. One isolate positive for NDM coharbored CMY-16. This type of p-AmpC was previously identified among Proteus mirabilis isolates from University Hospital Split [45]. Additional ESBLs and p-AmpC are often reported in CRPS [46], contributing to ESC resistance. Since OXA-48 does not hydrolyze cephalosporins, ESC resistance was attributed to CTX-M-15, the variant that efficiently hydrolyzes all cephalosporins. The fact that IS1999 was not identified by PCR mapping indicates that blaOXA-48 genes in CRPS were located in a different genetic environment compared to K. pneumoniae and E. coli [25].
The majority of isolates originated from within the hospital setting, but two NDM-1 isolates were isolated from outpatients, indicating the spread of the CRPS to the community. Community-acquired NDM-positive isolates harbored additional β-lactam resistance determinants such as ESBL or p-AmpC, while MBL genes were the sole β-lactam resistance genes among hospital isolates. Despite the presence of aminoglycoside resistance genes, all NDM-producing organisms were susceptible to amikacin, in contrast to OXA-48 producers, which, in the majority of cases, displayed high-level resistance with MICs above 128 mg/L. The phenomenon could be linked to the lack of aph genes among the majority of NDM-positive isolates. Even the MICs of gentamicin, which is intrinsically ineffective against Providenciae, were within the susceptible range among NDM-positive organisms. The majority of isolates did not carry acquired fluoroquinolone resistance genes, suggesting that chromosomal mutations of gyrA and parC genes may explain elevated MICs of ciprofloxacin and lack of inhibition zones around norfloxacin and levofloxacin disks. Only one strain tested positive for the acquired aminoglycoside and fluoroquinolone resistance determinant aac(6′)-Ib-cr5. High level of ciprofloxacin resistance with MIC values exceeding 128 mg/L was linked to OXA-48 carbapenemase producers, whereas the majority of NDM producers exhibited low-level resistance with MIC values around 4 mg/L. The presence of sul and dfr genes was in line with phenotypic resistance to SMX. Discrepancies between Interarray CarbaResist Kit and WGS were observed for isolate 9, which lacked the armA gene in WGS, and isolate 10, for which bla_OXA-1_, aac(6″), and armA genes were found by the Interarray CarbaResist Kit, but not by WGS. This could be attributable to the loss of resistance genes when the isolates were subcultured without selection pressure.
The genetic relatedness was evaluated by WGS, and two transmission clusters were found. This suggests clonal cross-transmission of isolates and the possibility of transmission between patients and medical staff by hand colonization and by environmental contamination. It seems that the isolates belonging to ST3 and ST5 carried NDM coding genes, while blaOXA-48 genes were carried by ST22 isolates. In Romania, NDM-positive organisms belonged to ST28, ST46, and ST233 [31].
Although the IncC plasmid replicon was found in five OXA-48-positive isolates subjected to WGS, there is no proof that it harbors carbapenemase-encoding genes for several reasons: the failure to transfer resistance via conjugation experiments and the MOB-suite prediction. However, these data were generated by short-read sequencing, which is not always a reliable predictor of plasmid-located resistance genes. The association of IncC plasmid and blaNDM genes was described in Romanian and Italian isolates [31,47]. Due to the problems with resistance transfer, we could not determine if the spread of carbapenemase-producing CRPS was due to vertical expansion of related isolates or horizontal transmission of IncC plasmids harboring blaCARB genes. NDM carbapenemase-producing isolates are often associated with hospital outbreaks, which are difficult to control. However, in our study, the endemic presence of CRPS in the participating centers, over a prolonged time, was noticed. Regarding the therapeutic options, CZA appears effective against OXA-48-positive isolates, whereas FDC and AMI demonstrated activity against NDM-producing isolates.
The strength of the present study is the large-scale susceptibility data obtained according to the standardized protocols. In addition, we tested antimicrobial susceptibility to many last resort antibiotics. However, this study has several limitations. The participating centers do not cover all regions in Croatia, for example, Slavonia. Isolates were collected retrospectively. Systematic screening for carriage of CRPS was not performed. Moreover, the study used a relatively small number of isolates, and the genetic comparison between isolates was not carried out for all isolates. Furthermore, aminoglycoside, sulfonamide, trimethoprim, tetracycline, and chloramphenicol resistance genes were not determined for the whole collection of isolates.
5. Conclusions
This study found equal presence of OXA-48 and NDM carbapenemases in Providencia spp. In other Enterobacterales in Croatia, OXA-48 is the dominant carbapenemase, with only sporadic occurrence of NDM. This study revealed regional differences in the prevalence of carbapenemase types: OXA-48 was dominant in Zagreb and Split, while NDM prevailed in Rijeka. Identification of NDM-1 in the outpatients’ samples indicates diffusion of this resistance determinant to the community. Resistance phenotypes did not differ between the two outpatients and the hospital isolates, indicating diffusion of the nosocomial isolates into the community. Spread of CRPS poses a serious therapeutic challenge for clinicians due to the limited therapeutic options and lack of new antibiotics. Moreover, identification of carbapenemases in CRPS may be a challenge because of the poor performance of phenotypic tests. Future studies on a larger collection of isolates should elucidate the modes and routes of spread of these increasingly resistant pathogens.
The effective use of molecular epidemiology methods such as WGS provides useful data for elucidating the transmission pathways of this rare, but increasingly important hospital pathogen. Providencia spp., in spite of being a rare pathogen, should be included in the surveillance studies across the medical centers in Croatia. Since there was a clonal expansion detected in Sv. Ivan Zelina hospital in Zagreb, more attention should be paid to hospital hygiene measures to avoid cross-infections.
The reference list from the paper itself. Each links out to its DOI / PubMed record.
- 1Witteven S. Hendrickx A.P.A. Providencia stuartii Trends. Microbiol.20253381081110.1016/j.tim.2025.02.00940121130 · doi ↗ · pubmed ↗
- 2Bradford P.A. Extended-spectrum beta-lactamases in the 21st Century: Characterization, Epidemiology, and Detection of this Important Resistance Threat Clin. Microbiol. Rev.20011493395110.1128/CMR.14.4.933-951.200111585791 PMC 89009 · doi ↗ · pubmed ↗
- 3Bonnet R. Growing Group of Extended-spectrum beta-lactamases: The CTX-M Enzymes Antimicrob. Agents. Chemother.20044811410.1128/AAC.48.1.1-14.200414693512 PMC 310187 · doi ↗ · pubmed ↗
- 4Rossolini G.M. D’Andrea M.M. Mugnaioli C. The Spread of CTX-M-type Extended-spectrum β-lactamases Clin. Microbiol. Infect.200814334110.1111/j.1469-0691.2007.01867.x 18154526 · doi ↗ · pubmed ↗
- 5Jacoby G.A. Amp C β-lactamases Clin. Microbiol. Rev.20092216118210.1128/CMR.00036-0819136439 PMC 2620637 · doi ↗ · pubmed ↗
- 6Poirel L. Pitout J.D. Nordmann P. Carbapenemases: Molecular Diversity and Clinical Consequences Future Microbiol.2007250151210.2217/17460913.2.5.50117927473 · doi ↗ · pubmed ↗
- 7Cantón R. Akóva M. Carmeli Y. Giske C.G. Glupczynski Y. Gniadkowski M. Livermore D.M. Miriagou V. Naas T. Rossolini G.M. European Network on Carbapenemases. Rapid Evolution and Spread of Carbapenemases among Enterobacteriaceae in Europe Clin. Microbiol. Infect.20121841343110.1111/j.1469-0691.2012.03821.x 22507109 · doi ↗ · pubmed ↗
- 8European Committee for Antimicrobial Susceptibility Testing Breakpoint Tables for Interpretation of MI Cs and Zone Diameters. Version 15.00. Valid from 2025-01-01(accessed on 1 October 2025)Available online: http://www.eucast.org