Exploring the Antimicrobial Efficacy of Mesenchymal Stem Cell Secretome Against Klebsiella pneumoniae in Urinary Tract Infections

ABSTRACT

Antimicrobial resistance (AMR) has become a global health crisis, with Klebsiella pneumoniae ( K. pneumoniae ) emerging as a major driver of AMR and a leading cause of urinary tract infections (UTIs). Mesenchymal stem cell (MSC)‐derived secretomes have gained attention for their antimicrobial properties. This study aimed to investigate the antibacterial activity of MSC‐derived secretomes against urinary K. pneumoniae isolates. The secretomes derived from adipose tissue (AD‐MSCs), bone marrow (BM‐MSCs), and umbilical cord (UC‐MSCs) were tested against 50  K. pneumoniae clinical strains isolated from urine, including extended‐spectrum β‐lactamase (ESBL)‐, AmpC‐, and carbapenem‐resistant (CRE) strains. This secretome inhibited 68% of isolates overall, including 81.5% of K. pneumoniae (non‐ESBL, non‐AmpC, non‐CRE), 46.7% of ESBL‐, and 100% of AmpC‐producing isolates, but showed no effect on CRE isolates. Genotypic analysis revealed that all ESBL‐producing isolates (100%) harbored the blaSHV gene and expressed the fimH, wabG, uge, ureA, and entB virulence factors. These findings demonstrate, for the first time, the antibacterial potential of three distinct MSC secretome sources against K. pneumoniae , highlighting their promise as a novel adjunctive therapy for drug‐resistant UTIs.

Study Highlights

What is the current knowledge on the topic?

• Klebsiella pneumoniae is a leading cause of urinary tract infections and is increasingly associated with multidrug resistance, creating an urgent need for alternative therapeutic strategies.

• Its virulence factors are well‐known contributors to urinary tract colonization and persistence.

• In parallel, mesenchymal stem cell‐derived secretomes have emerged as a promising cell‐free alternative with reported antimicrobial properties.

• However, despite this significant interest in various diseases, the potential application of these treatments in K. pneumoniae urinary tract infections remains limited.

What question did this study address?

• This study investigates the antimicrobial potential of secretomes derived from adipose (AD‐MSCs), bone marrow (BM‐MSCs), and umbilical cord (UC‐MSCs) mesenchymal stem cells against K. pneumoniae strains isolated from urine samples, including ESBL, non‐ESBL, AmpC β‐lactamase‐producing, and carbapenem‐resistant (CRE) strains.

• Moreover, it explores the potential associations between the presence of ESBL genes, virulence factors, the hypermucoviscosity phenotype, and corresponding antibiogram profiles.

What does this study add to our knowledge?

• The findings demonstrate that the antibacterial effect of MSC‐derived secretomes against K. pneumoniae is not uniform across isolates but instead varies between strains.

• Importantly, similar antibacterial effects were observed across secretomes derived from different MSC sources, indicating that the observed activity is largely independent of MSC origin.

How might this change clinical pharmacology or translational science?

• These findings underscore the therapeutic potential of MSC‐derived secretomes as a safe, cell‐free alternative that can target resistant pathogens through mechanisms beyond conventional antibiotics, providing a new direction for combating antibiotic‐resistant infections and thus advancing the fields of regenerative and translational medicine.

1. Introduction

Antimicrobial resistance (AMR) is projected to surpass cancer as a leading cause of death by 2050, making it one of the most critical global health challenges of the 21st century [1]. Currently, AMR‐related infections result in approximately 700,000 deaths annually, a number projected to rise dramatically to nearly 20 million deaths per year by 2050, with economic costs expected to exceed $2.9 trillion [2]. Among the pathogens driving this crisis, Gram‐negative bacteria stand out as particularly troublesome due to their intrinsic and acquired resistance mechanisms, including limited membrane permeability, β‐lactamase production, efflux systems, and biofilm formation [3]. Within this group, Klebsiella pneumoniae ( K. pneumoniae ) has emerged as a major cause of healthcare‐associated infections worldwide, intensifying the AMR crisis.

This encapsulated, non‐motile, Gram‐negative rod bacterium belongs to the Enterobacteriaceae family [4]. It is a prominent cause of hospital‐acquired infections, such as pneumonia, urinary tract infections (UTIs), bloodstream infections, and liver abscesses. These infections are particularly dangerous for ICU patients, immunocompromised individuals, and those who need mechanical ventilation or catheters [5]. Traditionally, K. pneumoniae was regarded as a hospital‐acquired pathogen primarily affecting immunocompromised individuals; however, in recent decades, significant evolutionary changes have occurred [6]. The hypervirulent (hvKp) and multidrug‐resistant (MDR) strains have emerged, posing an important public health challenge [7]. While MDR strains exhibit extensive resistance to one or more classes of antimicrobial agents [8], hvKp demonstrates increased invasiveness and virulence. The emergence of these strains is driven by an extensive accessory genome, including both plasmid‐encoded and chromosomal genes that confer antimicrobial resistance and enhanced virulence [9].

Moreover, infections caused by antibiotic‐resistant K. pneumoniae are particularly challenging due to widespread resistance to multiple antibiotic classes, particularly β‐lactams [9]. Some strains are capable of expressing extended‐spectrum β‐lactamases (ESBLs), which render the bacteria resistant to cephalosporins and monobactams. While others can express carbapenemases that render bacteria resistant to almost all available β‐lactams, including the carbapenems [6]. To add to that, K. pneumoniae has evolved several virulence factors that can cause severe and invasive infections, mainly relying on specific virulence‐associated genes encoding multiple factors that boost bacterial survival, colonization, and pathogenicity [10]. Among these critical factors are capsular polysaccharide regulators, iron acquisition systems, fimbrial genes, and colibactin‐producing genes [11, 12]. These factors help K. pneumoniae to increase capsule formation, compete for essential nutrients within the host, contribute to bacterial adhesion and biofilm formation, and enhance invasiveness [13].

Annually, UTIs lead to high healthcare costs surpassing $6 billion, impacting around 150 million people worldwide [14]. In both community‐ and hospital‐acquired infections, members of the Enterobacteriaceae family account for the majority of cases, with Escherichia coli ( E. coli ) remaining the primary causative agent and K. pneumoniae rising as the second most commonly isolated Gram‐negative pathogen in UTIs, particularly with the increasing prevalence of MDR and carbapenem‐producing strains [15, 16]. Recognizing the severity of this escalating threat, the World Health Organization (WHO) has classified K. pneumoniae as a critical priority pathogen, emphasizing the urgent need for the development of new antimicrobial agents and alternative therapeutic strategies [17].

In response to this public health challenge, mesenchymal stem cells (MSCs) have offered a promising alternative therapeutic approach. These cells are multipotent adult stem cells that have been isolated from various sites, most notably the bone marrow, adipose tissue, and umbilical cord. Several studies have demonstrated that MSCs exhibit antimicrobial properties through the secretion of antimicrobial peptides (AMPs), in turn enhancing and upregulating antimicrobial activity [18, 19]. Remarkably, these AMPs have previously been reported to display pharmacodynamic curves that are similar to those of conventional antibiotics and are less prone to resistance development [20]. Moreover, it has been demonstrated that AMPs have a high selectivity index, meaning that they display strong affinity for bacterial targets while bypassing host cells and reducing toxicity [21]. This property, together with their unique mechanisms of action, renders MSCs promising alternatives to conventional antibiotics.

Another distinguished feature of MSCs is their ability to engage with the host's innate immune system, thus enhancing the antibacterial activity through increasing phagocytosis and killing properties of monocytes and neutrophils [22, 23, 24]. Despite their therapeutic potential, MSC‐based therapies still face substantial challenges, including variability in stemness and differentiation, heterogeneity, homing, and migration variability, thus hindering their widespread clinical adoption [25]. This has led to a growing interest in cell‐free therapeutic approaches, particularly the MSC‐derived secretome. The secretome, composed of bioactive molecules such as cytokines, growth factors, extracellular vesicles, and antimicrobial peptides, offers a safer and more practical alternative to traditional cell‐based therapies.

It is evident from the significant rise in preclinical and clinical studies of stem cell‐based and cell‐free therapies over the past ten years that their therapeutic application for several disorders has been thoroughly investigated. Previous studies have demonstrated the antimicrobial activity of MSCs and their secretomes against a variety of Gram‐negative pathogens; they were mainly focusing on overall bacterial growth inhibition [26, 27, 28, 29]. Despite this significant interest in various diseases, their potential application in treating K. pneumoniae urinary tract infections remains limited. Our previous work has explored the antimicrobial activity of MSC‐derived secretome against urinary pathogens, particularly E. coli [28]. However, K. pneumoniae represents a clinically more challenging pathogen because of the rising frequency of ESBL‐, AmpC‐, carbapenem‐resistant phenotypes, contributing to its higher propensity for multidrug resistance and more complex virulence traits [30].

In this context, the present study goes beyond simple pathogen substitution by evaluating the antimicrobial activity of secretomes derived from adipose (AD‐MSCs), bone marrow (BM‐MSCs), and umbilical cord (UC‐MSCs) mesenchymal stem cells against K. pneumoniae strains isolated from urine samples, including ESBL‐, non‐ESBL, AmpC β‐lactamase‐producing, and carbapenem‐resistant (CRE) strains. Moreover, the current study extends previous work by investigating possible correlations with resistance genes, virulence factors, hypermucoviscosity phenotype, and corresponding antibiogram profiles.

2. Methodology

2.1. Ethics Statement

This study complied with the relevant ethical regulations. Urine samples, bacterial strains, and information used in this study were anonymized and labeled with unique codes to ensure patient confidentiality. This study was approved by the Institutional Review Board (IRB) of the Lebanese Hospital Geitaoui–University Medical Center (LHG‐UMC) under IRB approval number 2024‐IRB‐021.

2.2. Bacterial Culture and Identification

Fifty K. pneumoniae isolates were collected from urine samples at LHG‐UMC between February and April 2025. All inpatients and outpatients of any age or gender who presented with urinary tract infection symptoms and had positive urine cultures (> 1000 CFU/mL of K. pneumoniae ) were included. To avoid duplication, only one isolate per patient was analyzed. Isolates were first streaked on UriSelect4 Agar, then subcultured on MacConkey agar plates and incubated at 37°C for 24 h. Identification of K. pneumoniae was confirmed using the API 20E system in accordance with the manufacturer's instructions. The string test was then employed to evaluate the hypermucoviscosity of the isolates, which were grown overnight on MacConkey agar at 37°C. The test was interpreted as positive when a sterile loop could form a viscous string longer than 5 mm by stretching bacterial colonies on the agar plate [31].

2.3. Antimicrobial Susceptibility Testing

Using the Kirby–Bauer disk diffusion method, the antimicrobial susceptibility profiles of the K. pneumoniae isolates were examined in accordance with the Clinical and Laboratory Standards Institute (CLSI) guidelines [32]. Overnight cultures were adjusted to a 0.5 McFarland turbidity standard and evenly spread over sterile Mueller–Hinton agar plates. Twenty‐three antibiotic discs obtained from Bio‐Rad Laboratories (USA) were then tested, including Ampicillin (10 μg), Piperacillin (100 μg), Amoxicillin‐clavulanic acid (30 μg), Piperacillin/tazobactam (110 μg), Cefuroxime (30 μg), Cefoxitin (30 μg), Cefotaxime (30 μg), Ceftazidime (30 μg), Cefixime (5 μg), Cefepime (30 μg), Aztreonam (30 μg), Imipenem (10 μg), Ertapenem (10 μg), Gentamicin (10 μg), Amikacin (30 μg), Tigecycline (15 μg), Tetracycline (30 μg), Prulifloxacin (5 μg), Ciprofloxacin (5 μg), Norfloxacin (10 μg), Fosfomycin (200 μg), Nitrofurantoin (300 μg), and Trimethoprim‐sulfamethoxazole (25 μg). Inhibition zones were measured and interpreted according to CLSI guidelines after a 24 h incubation at 37°C.

2.4. Phenotypic Detection of ESBL‐Producing Isolates

Extended‐spectrum β‐lactamase (ESBL) production in K. pneumoniae isolates was determined using double‐disk synergy testing on Mueller–Hinton agar. The amoxicillin–clavulanic acid (AMC) disc (30 μg) was inserted at the center of each plate, with disks of ceftazidime (30 μg), cefotaxime (30 μg), cefoxitin (30 μg), cefepime (30 μg), and aztreonam (30 μg) placed 20 mm away. The isolates showing an increased zone of inhibition around any of the antibiotic disks toward the AMC disc, shaping a keyhole zone, were interpreted as ESBL producers [33].

2.5. Molecular Detection of ESBL Genes

Fifteen of the 50  K. pneumoniae isolates were screened for ESBL‐associated genes (bla _OXA , bla _CTX‐M , bla _TEM , bla _SHV ) by conventional polymerase chain reaction (PCR). Genomic DNA was extracted by heat lysis: two colonies from each isolate were suspended in 200 μL of sterile distilled water, heated at 95°C for 10 min, and centrifuged at maximum speed for 3 min to collect the supernatant. PCR reactions (20 μL) contained 4 μL of reaction master mix, 5 μL of DNA template, 4 μL of primer mix (0.5 μL each of forward and reverse primer per gene; sequences in Table 1), and 7 μL of nuclease‐free water. Amplification was performed in a thermal cycler under the following conditions: initial denaturation at 95°C for 15 min; 30 cycles of denaturation at 94°C for 30 s, annealing at 62°C for 90 s, and extension at 72°C for 60 s; with a final extension at 72°C for 10 min [28].

TABLE 1.

Sequences of primers and expected PCR amplicon sizes for bla _OXA , bla _CTX‐M , bla _TEM , bla _SHV .

Target	Sequence (5′ – 3′)	Size (bp)	Reference
OXA	F: ACACAATACATATCAACTTCGC	813	[28]
	R: AGT GTG TTT AGA ATG GTG ATC
CTX‐M	F: ATGTGCAGYACCAGTAARGTKATGGC	593
	R: TGGGTRAARTARGTSACCAGAAYCAGCGG
TEM	F: CGCCGCATACACTATTCTCAGAATGA	445
	R: ACGCTCACCGGCTCCAGATTTAT
SHV	F: CTTTATCGGCCCTCACTCAA	237
	R: AGGTGCTCATCATGGGAAAG

2.6. Detection of Virulence Genes

Six virulence genes associated with capsule regulation, adhesion/biofilm formation, and siderophore production (wabG, uge, fimH, kfuBC, entB, and ureA) were detected by conventional PCR. Simplex PCR was carried out for all genes except ureA and kfuBC, which were amplified together by multiplex PCR. Genomic DNA was extracted using the same heat‐lysis method described previously. PCR reactions were set up in 20 μL volumes using the same master mix composition as that used for ESBL gene amplification, with the volume of nuclease‐free water adjusted to 10 μL for simplex reactions and 9 μL for the multiplex reaction. Primer sequences, expected amplicon sizes, and annealing temperatures are listed in Table 2, while full thermal cycling conditions for each gene, including those for the multiplex reaction, are provided in Table S1.

TABLE 2.

Sequences of primers and expected amplicon sizes for the wabG, uge, fimH, entB, ureA, and kfuBC virulence genes.

Target	Sequence (5′ – 3′)	Size (bp)	Annealing Temp (°C)	References
WabG	F: CGGACTGGCAGATCCATATC	683	53	[34]
	R: ACCATCGGCCATTTGATAGA
Uge	F: GAT CAT CCG GTC TCC CTG TA	534	53	[34]
	R: TCT TCA CGC CTT CCT TCA CT
FimH	F: ATGAACGCCTGGTCCTTTGC	688	55	[35]
	R: GCTGAACGCCTATCCCCTGC
UreA	F: GCT GAC TTA AGA GAA CGT TAT G	337	55	[34]
	R: GAT CAT GGC GCT ACC T(C/T) A
KfuBC	F: GAA GTG ACG CTG TTT CTG GC	797	55	[34]
	R: TTT CGT GTG GCC AGT GAC TC
EntB	F: ATTTCCTCAACTTCTGGGGC	371	57	[35]
	R: AGCATCGGTGGCGGTGGTCA

2.7. Mesenchymal Stem Cell Culture, Characterization, and Secretome Preparation

Human mesenchymal stem cells (MSCs) derived from adipose tissue (AD‐MSCs), bone marrow (BM‐MSCs), and umbilical cord (UC‐MSCs) were donated by the Reviva Regenerative Medicine Center (Bsalim, Lebanon). MSCs were cultured in Dulbecco's Modified Eagle Medium (DMEM/F12; Sigma‐Aldrich, Great Britain; D8437‐500ML) supplemented with 10% fetal bovine serum (FBS; Sigma‐Aldrich, Paraguay; F9665‐500ML) and 1% antibiotic–antimycotic solution (BIOWEST, France; L0022‐100) at 37°C in a humidified atmosphere containing 5% CO₂. Cells were seeded separately in 90‐mm Petri dishes with 5 mL of supplemented medium and allowed to reach 70%–90% confluence. Once this confluence was achieved, cultures were washed and incubated for an additional 48 h in serum‐free DMEM/F12 medium. The secretome was collected by centrifugation at 300 × g for 5 min to remove cells, followed by additional centrifugation to eliminate residual debris. Secretomes were then lyophilized, reconstituted in deionized distilled water, filtered through a 0.45 μm sterile syringe filter, and stored at −20°C until use [28, 36]. MSCs were characterized according to the minimal criteria established by the International Society for Cell and Gene Therapy (ISCT), including plastic adherence under standard culture conditions, tri‐lineage differentiation (adipogenic, osteogenic, and chondrogenic), and a surface marker profile positive for CD73, CD90, and CD105 and negative for hematopoietic and endothelial markers such as CD34 and CD45 [26].

2.8. Antibacterial Effects of the Secretome

The antibacterial activity of the MSC secretomes against K. pneumoniae was evaluated by the Kirby–Bauer disk diffusion method. Bacterial suspensions were prepared in sterile saline and adjusted to a 0.5 McFarland turbidity standard, then spread onto Mueller–Hinton agar plates. CITOTEST sterile 6‐mm filter paper disks were placed on the agar surface, one for each MSC secretome, one for the positive control, and one for the negative control. Twenty microliters of each test or control solution were applied to the respective discs. Secretome samples were applied at a fixed volume (20 μL), and normalization based on total protein concentration was not performed. Plates were then incubated at 37°C for 24 h, after which inhibition zones were measured in millimeters. Discs without measurable inhibition zones were recorded as 6 mm [37, 38], and these isolates were considered resistant to the MSC‐derived secretome.

2.9. Statistical Analysis

Data were entered into IBM SPSS Statistics version 30.0.0.0 and analyzed. Normality of the inhibition zone data for each bacterial group and each MSC secretome source (AD‐MSCs, BM‐MSCs, UC‐MSCs) was checked using the Shapiro–Wilk test. Because most datasets showed significant departures from normality, non‐parametric tests were used. Differences between secretome sources within each bacterial group were first explored with the Kruskal–Wallis test. To look at main and interaction effects across all groups, an Aligned Rank Transform (ART) ANOVA, to analyze non‐parametric factorial data while allowing for interaction effects, was run with bacterial group (Klebsiella, ESBL, AmpC, CRE) and secretome source (AD, BM, UC) as fixed factors and inhibition zone size as the dependent variable. Post hoc pairwise comparisons were carried out using estimated marginal means with Bonferroni correction for multiple testing. Heat maps summarizing the antibiotic susceptibility profiles and inhibition zone data were generated using Python and executed in Google Colaboratory.

3. Results

3.1. Phenotypic and Biochemical Identification of K. pneumoniae Urinary Isolates

After bacterial incubation, all isolates on the UriSelect agar appeared glossy, blue‐purple in color, with similar colony sizes and shapes (Figure 1A). On MacConkey agar, they exhibited the typical pink, mucoid appearance of K. pneumoniae (Figure 1B). For biochemical identification, the API 20E system produced profiles consistent with K. pneumoniae across all isolates (Figure 1C). Furthermore, assessment of the hypermucoviscosity phenotype by the string test showed that 8% (4/50) of the isolates were hypermucoid, producing a viscous string greater than 5 mm (Figure 1D).

FIGURE 1.

(A): Representative colonies of K. pneumoniae on UriSelect4 agar showing characteristic blue coloration. (B): Typical pink, mucoid colonies of K. pneumoniae grown on MacConkey agar. (C): API 20E biochemical profile confirming K. pneumoniae identification. (D): Positive string test (arrow) demonstrating the hypermucoviscosity phenotype in K. pneumoniae isolates.

3.2. Demographic and Clinical Data of Patients With K. pneumoniae Urinary Isolates

Between February and April 2025, 50  K. pneumoniae strains isolated from urine samples were collected from patients aged 1–96 years (mean age 67.2 years). From these isolates, 30% (15/50) were ESBL‐producing, 10% (5/50) were AmpC‐producing, and 6% (3/50) were CRE, while the remaining 54% (27/50) were non‐ESBL, non‐AmpC, and non‐CRE (non‐resistant isolates). The highest K. pneumoniae UTI incidence was observed in elderly patients (≥ 65 years) across all groups, representing 70.4% (19/27) of non‐resistant isolates, 66% (9/15) of ESBL‐producing isolates, 60% (4/5) of AmpC‐producing isolates, and 66.7% (2/3) of CRE‐producing isolates. In contrast, UTIs in children (< 18 years) were rare, with only one case (~3%) among non‐resistant isolates and none in the resistant groups.

Female patients outnumbered male patients in the non‐resistant, ESBL‐, and AmpC‐producing groups, comprising 74.1% (20/27), 66.7% (10/15), and 60% (3/5) of cases, respectively, whereas all CRE isolates were recovered from male patients. Regarding patient status, most non‐resistant and AmpC‐producing isolates were obtained from outpatients (59.3% [16/27] and 60% [3/5], respectively). However, this pattern reversed in ESBL‐ and CRE‐producing isolates, where a higher proportion of cases was observed among inpatients (66.7% [10/15] and 66.7% [2/3], respectively) (Table 3). There was no statistical significance association between the number of ESBL‐producing isolates and patient settings (inpatient vs. outpatients) (Fisher's exact test, p = 0.217).

TABLE 3.

Demographic and clinical characteristics of patients with K. pneumoniae urinary isolates.

Demographic factors	Categories	K. pneumoniae*		ESBL‐producing K. pneumoniae		AmpC‐producing K. pneumoniae		CRE‐producing K. pneumoniae
		n	(%)	n	(%)	n	(%)	n	(%)
Age group	Children (< 18)	1	3.7%	0	0%	0	0%	0	0%
	Adult (19 – 64)	7	25.9%	6	40%	1	40%	1	33.3%
	Elderly (≥ 65)	19	70.4%	9	60%	4	60%	2	66.7%
	Total	27	100%	15	100%	5	100%	3	100%
Gender	Male	7	25.9%	5	33.3%	2	40%	3	100.0%
	Female	20	74.1%	10	66.7%	3	60%	0	0.0%
	Total	27	100%	15	100%	5	100%	3	100%
In‐ or out‐patient	In‐patient	11	40.7%	10	66.7%	2	40%	2	66.7%
	Out‐patient	16	59.3%	5	33.3%	3	60%	1	33.3%
	Total	27	100%	15	100%	5	100%	3	100%

ESBL: extended‐spectrum β‐lactamase; AmpC: AmpC β‐lactamase‐producing isolates; CRE: carbapenem‐resistant Enterobacterales. K. pneumoniae* represents non‐ESBL‐, non‐AmpC‐, and non‐CRE isolates. Values are presented as n (%).

3.3. Antibiotic Profiles of K. pneumoniae Strains

For the 50 isolates, antimicrobial susceptibility testing was performed (Figure 2). All isolates, regardless of classification, were 100% resistant to ampicillin (AMP). In the K. pneumoniae group (non‐ESBL, non‐AmpC, non‐CRE), 22% (6/27) of isolates were classified as multidrug‐resistant (MDR), whereas all ESBL‐producing (100%, 15/15) and AmpC‐producing (100%, 5/5) isolates were MDR. In contrast, all CRE isolates were extensively drug‐resistant (XDR, 100%, 3/3) . Resistance in the K. pneumoniae group was most frequent to nitrofurantoin (F) at 59.3% (16/27), followed by piperacillin (PIP) at 40.7% (11/27), amoxicillin–clavulanic acid (AMC) at 25.9% (7/27), tetracycline (TE) at 18.5% (5/27), prulifloxacin (PFX) at 18.5% (5/27), ciprofloxacin (CIP) at 18.5% (5/27), norfloxacin (NOR) at 18.5% (5/27), trimethoprim–sulfamethoxazole (SXT) at 18.5% (5/27), and fosfomycin (FOS) at 7.4% (2/27). Besides, all isolates in this group remained completely susceptible to the other 13 antibiotics tested.

FIGURE 2.

Antibiotic susceptibility profiles of K. pneumoniae isolates. Heatmap illustrating the antimicrobial susceptibility of 50  K. pneumoniae urinary isolates against 23 antibiotics across four groups: K. pneumoniae * (non‐ESBL/non‐AmpC/non‐CRE), ESBL‐producing, AmpC‐producing, and carbapenem‐resistant (CRE) isolates. Each row represents one isolate, and each column corresponds to an antibiotic tested. Colors indicate susceptibility categories according to CLSI guidelines. AK, amikacin; AMC, amoxicillin‐ clavulanic acid; AMP, ampicillin; ATM, aztreonam; CAZ, ceftazidime; CIP, ciprofloxacin; CTX, cefotaxime; CXM, cefuroxime; ETP, ertapenem; FEP, cefepime; FOS, Fosfomycin; FIX, cefixime; FOX, cefoxitin; GEN, gentamicin; IMP, imipenem; F, nitrofurantoin; NOR, norfloxacin; PFX, prulifloxacin; PIP, piperacillin; SXT, trimethoprim‐ sulfamethoxazole; TE, tetracycline; TGC, tigecycline; TZP, piperacillin‐ tazobactam.

The ESBL‐producing K. pneumoniae isolates showed complete resistance to AMP, PIP, cefuroxime (CXM), cefotaxime (CTX), ceftazidime (CAZ), cefixime (FIX), cefepime (FEP), and aztreonam (ATM). Variable resistance was observed for nitrofurantoin (F) at 80% (12/15), SXT at 80% (12/15), TE at 66.7% (10/15), AMC at 46.7% (7/15), piperacillin–tazobactam (TZP) at 20% (3/15), PFX at 20% (3/15), CIP at 20% (3/15), NOR at 20% (3/15), and gentamicin (GEN) at 13.3% (2/15). All ESBL‐producing isolates were fully susceptible to the remaining six antibiotics: cefoxitin (FOX), imipenem (IMP), ertapenem (ETP), amikacin (AK), tigecycline (TGC), and fosfomycin (FOS).

AmpC‐producing isolates showed 100% resistance to AMP, PIP, AMC, CXM, FOX, CTX, CAZ, FIX, and F, with variable resistance to TZP (60%, 3/5), FEP (40%, 2/5), FOS (40%, 2/5), and ATM (20%, 1/5). These isolates remained fully susceptible to IMP, ETP, GEN, AK, TGC, TE, PFX, CIP, NOR, and SXT.

On the other hand, the CRE isolates demonstrated complete resistance to almost all tested antibiotics, retaining susceptibility only to tigecycline (TGC). Given this extensive resistance, especially among ESBL‐, AmpC‐, and CRE isolates, the next step was to evaluate the antibacterial activity of MSC‐derived secretomes against the same strains to explore their potential as an alternative therapeutic option.

3.4. Antibacterial Profiles of AD‐MSC, BM‐MSC, and UC‐MSC Secretomes Across Klebsiella Subgroups

Using the Kirby–Bauer disk diffusion method, the antibacterial activity of AD‐MSC, BM‐MSC, and UC‐MSC secretomes against different Klebsiella groups was assessed. Results revealed that the secretomes were effective against 68% of isolates overall, including 81.5% of K. pneumoniae (non‐ESBL, non‐AmpC, non‐CRE), 46.7% of ESBL‐producing strains, and 100% of AmpC‐producing K. pneumoniae . However, it was totally resistant against CRE‐producing strains. The heat map in Figure 3 summarizes the average inhibition zones (mm) for each secretome source across K. pneumoniae , ESBL‐producing, AmpC‐producing, and CRE K. pneumoniae groups.

FIGURE 3.

Average inhibition zones of MSC secretomes against K. pneumoniae groups. Heatmap showing the mean diameters (mm) of inhibition zones produced by secretomes from adipose (AD‐MSCs), bone marrow (BM‐MSCs), and umbilical cord (UC‐MSCs) mesenchymal stem cells against four groups of K. pneumoniae : K. pneumoniae* (non‐ESBL/non‐AmpC/non‐CRE), ESBL‐producing, AmpC‐producing, and carbapenem‐resistant (CRE) isolates. Each cell represents the average inhibition zone (mm) for a given secretome source and bacterial group. Warmer colors indicate larger zones of inhibition; cooler colors indicate smaller or absent inhibition. No significant differences among secretome sources within each bacterial group (p > 0.05) were detected.

For the non‐resistant K. pneumoniae group, the average zones of inhibition were 15.4 mm (±4.8 mm), 15.3 mm (±4.8 mm), and 14.4 mm (±4.5 mm) for AD‐MSC, BM‐MSC, and UC‐MSC secretomes, respectively. While for ESBL‐producing K. pneumoniae , the average inhibition zones were lower at 10.3 mm (±5.3 mm), 10.2 mm (±5.6 mm), and 9.6 mm (±5 mm), respectively. AmpC‐producing K. pneumoniae showed inhibition zones of 15.0 mm (±3 mm), 14.6 mm (±1.7 mm), and 13.2 mm (±2.4 mm) for AD‐MSC, BM‐MSC, and UC‐MSC secretomes, respectively. While no inhibition zones were detected for CRE K. pneumoniae across all secretome sources.

Statistical analysis using Kruskal–Wallis tests indicated no significant differences among the secretome sources within each bacterial group (p > 0.05). Taken together, the data suggest that variation in antibacterial activity was attributable primarily to differences among bacterial groups, with no significant effect detected from the MSC secretome source.

3.5. Distribution of ESBL Resistance Genes and Virulence Factors Among K. pneumoniae Isolates

The 15 ESBL‐producing K. pneumoniae isolates convey the presence of resistance genes, including blaSHV, blaCTX‐M, and blaTEM. All isolates carried blaSHV as the most predominant gene (100%). Of these, 80% (12/15) co‐carried blaCTX‐M, while 13.3% (2/15) co‐carried blaTEM. However, only 6.7% of the isolates harbored the blaSHV alone. Notably, none of the isolates expressed the blaOXA gene (Figure 4).

FIGURE 4.

Distribution of ESBL resistance genes among K. pneumoniae isolates. Stacked bar chart showing the percentage of ESBL‐producing K. pneumoniae isolates carrying different resistance gene profiles. All isolates (100.0%) harbored the SHV gene either alone or in combination with other genes. Percentages represent the proportion of isolates positive for each gene profile.

Furthermore, virulence factors were screened among ESBL‐producing K. pneumoniae isolates, and those non‐ESBL, non‐AmpC, and non‐CRE isolates that were resistant to the MSC secretome were identified, aiming to explore potential differences and interpret mechanisms underlying secretome resistance. All isolates were positive for the adhesin gene fimH as well as wabG and uge, which are linked to host colonization and virulence. Moreover, all isolates also harbored ureA, which is associated with the urease enzyme. Regarding iron acquisition systems, entB was detected in 100% of isolates, while kfuBC was largely absent and only present in 4 isolates (20%). These distributions showed no apparent association with the isolates' susceptibility to MSC secretomes under the conditions tested (Table S2).

4. Discussion

The pipeline for effective, safe antibiotics remains alarmingly limited, despite repeated global appeals for action, exposing healthcare systems to escalating risk. AMR has emerged as a worldwide crisis, often described as a ticking time bomb highlighting the critical necessity for innovative strategies that go beyond traditional antibiotics to combat multidrug‐resistant K. pneumoniae [39]. Against this backdrop, MSC secretomes have gained attention as potential candidates because of their rich content of extracellular vesicles, cytokines, and antimicrobial peptides [40]. Earlier research has shown that factors derived from MSCs possess both immunomodulatory and antimicrobial properties; however, evidence for their activity against urine‐isolated Klebsiella species and their relationship to virulence and antibiotic resistance determinants remains scarce.

In this study, the antibacterial activity of AD‐MSC, BM‐MSC, and UC‐MSC secretomes against diverse K. pneumoniae groups was thoroughly examined; however, the study was conducted on a relatively small number of isolates from a single center, which may limit the generalizability of the findings. Overall, the data revealed that all the secretome sources exhibited inhibitory activity against 68% of the UTI‐causing K. pneumoniae strains. This activity was represented by the inhibition zone sizes, which differed significantly among bacterial groups but not between the three secretome sources. Although the differences were not statistically significant, the AD‐MSC secretomes produced the largest inhibitory zones, followed by BM‐MSC and UC‐MSC secretomes. Interestingly, the only consistent pattern observed between antibiotic susceptibility and secretome response was that all K. pneumoniae (non‐ESBL/non‐AmpC/non‐CRE) isolates resistant to the three types of secretome were also resistant to the antibiotic tetracycline, whereas no other consistent correlations emerged. The complete lack of response observed in CRE‐producing isolates is a notable finding that warrants further consideration. One possible explanation may be related to the structural and behavioral traits of each CRE isolate, such as increased capsule formation and hypermucoviscosity, which may restrict the entry of secretome components [41]. Furthermore, the increased ability of biofilm formation and alterations in the outer membrane may further reduce susceptibility [42]. These factors, although not studied in this work, may contribute to the observed resistance of CRE‐producing isolates. It should also be noted that the limited number of CRE isolates included in this study restricts the ability to draw definitive conclusions. The screening for ESBL resistance genes and important virulence factors showed high prevalence rates, with no obvious correlation with susceptibility to MSC secretomes. These results imply that although MSC secretomes demonstrate significant antibacterial activity against multidrug‐resistant K. pneumoniae isolates, their effects seem to be unrelated to conventional antibiotic resistance or virulence factors, indicating their unique, varied mechanism of action. It should also be noted that the secretome dosing was standardized by volume (20 μL) rather than protein concentration. All secretome samples were lyophilized and reconstituted to a defined concentration (62.5 mg/mL) prior to use to ensure consistency across preparations.

It has been proven that the MSC secretome encompasses a complex mix of extracellular vesicles, antimicrobial peptides, enzymes, iron‐binding proteins, and immunomodulatory cytokines that may act through multiple pathways [43, 44, 45]. This complex mode of activity has the potential to evade established resistance mechanisms, explaining why even isolates harboring ESBL genes or multiple virulence factors showed inhibition zones similar to those of less‐resistant strains. However, this is not always the case, since secretome responsiveness could be influenced by the bacteria themselves. For example, K. pneumoniae is known for its ability to form biofilms, its quorum‐sensing (QS) mechanisms, and its cell envelope characteristics [13], which may influence responsiveness to the secretome. Therefore, further mechanistic studies are required to better understand the factors governing susceptibility to the secretome.

Several preclinical studies have reported robust antibacterial effects of MSCs and their secretomes across multiple bacterial species without correlating with resistance genes or virulence profiles [29, 33]. In line with this, a previous study reported no statistically significant differences in antibacterial activity among secretomes derived from AD‐, BM‐, and UC‐MSCs [28]. This observation may be further supported by findings from Hodgson‐Garms et al., who showed that proteomic analyses of secretomes from these three MSC sources revealed a conserved core signature, meaning that they have the same recurring group of proteins or bioactive factors, despite some variation under inflammatory licensing [46]. This could support the reason behind the lack of significant differences among the three secretome sources.

Conversely, prior studies suggested that BM‐MSC secretomes mainly carry out their antimicrobial activities via inducing the secretion of hepcidin and human cathelicidin LL‐37, especially after bacterial or inflammatory activation [18, 19, 47]. Whereas UC‐MSC secretomes express less LL‐37 and lipocalin 2 (Lcn2) and rely more on β‐defensin‐2 overexpression through TLR4 signaling [45, 47]. Besides, AD‐MSC proteomic profiles are less well established; their secretomes have been shown to improve the effectiveness of antibiotics, indicating a potential synergistic function in combination treatments [19].

Despite its strengths, this study has several limitations that warrant consideration. The relatively small number of isolates collected from a single hospital may limit the generalizability of the findings. Moreover, all experiments were performed in vitro using disk diffusion assays, which, while useful for screening antibacterial effects, fail to capture the full complexity of in vivo infections. Moreover, the diffusion properties of the secretome components may differ from those of conventional antibiotics, potentially influencing inhibition zone sizes. Furthermore, this method does not allow differentiation between bacteriostatic and bactericidal effects. It is also important to note that MIC and MBC analyses were not performed in this study, which limits the ability to accurately assess antimicrobial potency. Besides, no clear established association between the antibiotic resistance profiles of K. pneumoniae, their resistance genes, virulence factors tested, and their susceptibility to MSC‐derived secretomes was concluded. In addition, the MSC secretomes were not profiled at the molecular level to identify specific active components responsible for the antibacterial activity. Future investigations should integrate the study of MSCs' proteomics, transcriptomics, and metabolomics to help identify the core antimicrobial factors and clarify their mechanisms of action. Lastly, testing secretomes against biofilms and in vivo UTI infection models will provide benchmark data before proceeding to clinical studies.

Since dose–response and MIC/MBC analyses were not included in this investigation, it is impossible to directly compare the effectiveness of MSC‐derived secretomes with traditional antibacterial drugs. Furthermore, this study did not assess safety profiles or toxicity. Future research should focus on these factors since they are crucial for evaluating the translational potential of secretome‐based treatments. To better understand the connection between exposure and antimicrobial activity, more research using a range of secretome concentrations and dose–response analyses is needed.

Taken together, these findings demonstrate that secretomes derived from adipose, bone marrow, and umbilical cord MSCs exert measurable antibacterial effects against multidrug‐resistant K. pneumoniae strains isolated from urine, including ESBL‐ and AmpC‐, but not CRE‐producing strains. This MSC secretome effect suggests a complex mechanism different from that of conventional antibiotics, since no significant differences were observed between MSC sources under the tested conditions. The secretome also appears to be independent of virulence determinants and classical resistance. These findings emphasize the promise of MSC secretomes as a new, broad‐spectrum antibacterial approach. However, they also point out that more mechanistic and translational research is needed to fully realize their therapeutic potential.

Author Contributions

L.A. wrote the manuscript. F.A.S. and E.S.S. designed the research. C.K. and L.A. performed the research. F.A.S., E.S.S., and L.A. analyze the data. All authors reviewed and approved the final manuscript.

Funding

The authors have nothing to report.

Conflicts of Interest

The authors declare no conflicts of interest.

Supporting information

Table S1: PCR cycling conditions for the wabG, uge, fimH, entB, ureA, and kfuBC virulence genes.

Table S2: Distribution of Virulence‐Associated Genes (wabG, uge, ureA, KfuBC, fimH, and entaB) Among K. pneumoniae isolates.