Antimicrobial resistance (AMR) represents a growing threat to human health all around the world. A recent systematic analysis reported an estimate of almost 5 million deaths associated with bacterial AMR in 2019, of which 1.27 million were directly attributable to bacterial AMR (Murray et al., 2022). The current antibiotic development pipeline does not adequately address the emergence of multi-, extensively-, and pan-resistant bacterial pathogens. As a result, the development of innovative antimicrobial strategies is widely recognized as a research priority (Ajose et al., 2024; Piddock et al., 2024; Ralhan et al., 2024).

Pseudomonas aeruginosa is an opportunistic Gram-negative bacterium currently representing one of the major threats in clinical settings due to its ability to express a variety of virulence factors, form biofilms, and infect disparate body districts. In addition to its multifaceted pathogenicity, the bacterium expresses multiple resistant mechanisms even toward last-resort antibiotics, rendering problematic its eradication with the current arsenal of antimicrobial drugs (Horcajada et al., 2019; Kothari et al., 2023). In hospital settings, P. aeruginosa is often involved in highly deadly acute pneumonia (Barp et al., 2023), but it can also cause chronic lung infections, especially in patients with underlying diseases such as cystic fibrosis (CF), non-CF bronchiectasis (NCFB), and chronic obstructive pulmonary disease (COPD) (Martínez-Solano et al., 2008; Wilson et al., 2016; Alqasmi, 2024).

The challenges in treating multidrug P. aeruginosa infections have stimulated the search for antibiotic alternatives to target this bacterium, including quorum sensing inhibitors (Rodríguez-Urretavizcaya et al., 2024), phages (Martinho et al., 2024), antimicrobial peptides (Maisetta et al., 2017; Grassi et al., 2019), and others. Among these alternatives, lung delivery of probiotics has recently gained an intense interest as it could correct the dysbiotic microbiota at the pulmonary level while controlling the growth and virulence of respiratory pathogens (Batoni et al., 2022; Byun et al., 2023; Glieca et al., 2024b). As proof of concept, Fangous and coworkers demonstrated that intra-tracheal administration of lactobacilli protects mice from P. aeruginosa pulmonary infection (Fangous et al., 2019). Likewise, intranasal delivery of probiotics was shown to synergistically interact with an anti-infective antibody and improve mice survival after primary and secondary infection with P. aeruginosa while exerting an immune-modulating effect (Sécher et al., 2024).

Inhalation powders containing lactic acid probiotics (i.e., Lactiplantibacillus plantarum, Lacticaseibacillus rhamnosus, or Lactobacillus acidophilus) have been recently developed in order to assess the feasibility of their administration to the lungs of patients with P. aeruginosa respiratory infections (Glieca et al., 2024a). The feasibility of nebulization as a modality to deliver L. rhamnosus to the lung has also been recently investigated, demonstrating that the efficiency of the process depends on the nebulization type and the liquid medium of the probiotic suspension (Byun et al., 2023). Interestingly, we have previously reported that in an in vitro system, L. plantarum and L. rhamnosus exert strong antibacterial and antibiofilm properties against lung isolates of P. aeruginosa (Batoni et al., 2023a), while L. acidophilus effectively prevents adhesion of the same isolates to lung epithelial cells (Batoni et al., 2023b), suggesting possible mechanisms of probiotic-dependent P. aeruginosa control at the lung level.

Despite these promising findings, current research supports the view that the beneficial health effects of probiotics are only partially ascribed to their vitality, as probiotic-derived inanimate preparations (e.g., non-viable microbes, their cell components, or their metabolites) can also exert biological activities with therapeutic potential (Aggarwal et al., 2022; da Silva Vale et al., 2023). These findings have led to the evolving concept of postbiotics, originally defined as metabolic products of probiotics, although this definition has recently been revised (Aguilar-Toalá et al., 2021; Vinderola et al., 2022). As postbiotics may overcome some of the safety concerns related to the use of live probiotics (e.g., transmission of antibiotic resistance, possibility of systemic infections in vulnerable individuals, challenges in maintaining vitality during processing and storage), their preclinical characterization as anti-infective agents is worth investigating.

We previously reported that postbiotics from L. rhamnosus in the form of cell-free supernatants (CFS) exert strong antibacterial, antibiofilm, and antivirulence properties against clinical isolates of P. aeruginosa (Pompilio et al., 2024). Importantly, we have also demonstrated that they act even against antibiotic-refractory persister cells (Bianchi et al., 2024) and do not induce resistance following multiple passages at sub-inhibitory concentrations, unlike what was observed with a conventional antibiotic (Kaya et al., 2024).

In this study, we sought to investigate further, at a preclinical level, the translational potential of CFS for the local therapy of lung infections by P. aeruginosa as an alternative to live bacteria. To this aim, we tested the antipseudomonal activity of CFS in an Air-Liquid interface (ALI) lung infection model in order to mimic in vitro the conditions found in the infected airways. In the same model, the cytotoxic potential toward a human lung epithelial cell line was also explored. Furthermore, the use of L. rhamnosus postbiotics as a solution for nebulization was investigated, evaluating its aerodynamic distribution under normal adult and CF patients’ breathing profiles. Finally, physicochemical stability studies were performed under stressed temperature conditions, and the antimicrobial activity maintenance of postbiotics under different storage conditions was investigated.

Overall, our findings demonstrate that CFS exhibit effective anti-bacterial activity at non-toxic concentrations in the ALI lung infection model against clinical isolates of P. aeruginosa and the potential to reach the deeper regions of the lung. In addition, they exhibit high stability in a number of storage conditions. These results form the basis for innovative strategies for the local treatment of P. aeruginosa lung infections.

There is growing interest in postbiotics as an alternative to live probiotics in clinical applications. This is largely due to their ease of production, safety, extended shelf life, patentability, and lack of risk for antibiotic resistance transfer. Moreover, postbiotics offer additional advantages over live bacteria, such as the possibility of encapsulation for targeted delivery and safe administration to immunocompromised individuals (Kumar et al., 2024).

Despite ongoing challenges in postbiotics development and applications, numerous clinical trials have yielded promising results (for a recent and comprehensive review on the topic, see Kumar et al., 2024). Most of these trials have focused on the digestive tract; however, there is increasing interest in exploring postbiotics for other body sites, including the respiratory system (Mosca et al., 2022; Kumar et al., 2024).

In this study, we investigated the potential of postbiotics from L. rhamnosus, in the form of CFS, as an innovative strategy to treat P. aeruginosa lung infections. The first step of the work was to test the ability of CFS to reduce P. aeruginosa load in an in vitro model of lung infection strictly recapitulating the conditions found in human airways. The use of in vitro cell culture models is increasingly favored to adhere to the 3R principle (reduce, refine, replace) and limit experimental animal testing for drug screening (Baldassi et al., 2021). Herein, we employed an ALI model, which is considered more physiologically relevant than traditional submerged models (Lacroix et al., 2018). Our group has previously used this model to successfully assess the therapeutic potential of non-conventional antimicrobials (Grassiri et al., 2024). The key feature of ALI culture is that the apical surface of cells is exposed to air while the basal side is in contact with a liquid cell culture medium. This configuration allows cell polarization and differentiation toward a mucociliary phenotype, simulating a scenario close to in vivo conditions (Baldassi et al., 2021). The model effectively supported the growth of lung isolates of P. aeruginosa (both non-mucoid and mucoid), which caused significant toxicity to the human NCl-H441 distal lung epithelial cell line within 25 h. This observation aligns with the previously reported ability of P. aeruginosa to cause lung injury through the secretion of its virulence factors that may be even injected directly into the target cells via a specialized TTSS (type III secretion system) widely expressed in environmental as well as clinical isolates (Ader et al., 2005). Nevertheless, CFS markedly reduced (2–3 Log) the P. aeruginosa viable count in approximately 7 h, a time point at which the epithelial layer remained intact. In contrast, untreated samples showed clear signs of bacterial aggregation within the same time frame, resembling early biofilm formation as confirmed by SEM imaging. This finding is consistent with the rapid biofilm-forming ability observed in certain clinical P. aeruginosa isolates, which can initiate biofilm development in as little as 2 h in vitro (Olivares et al., 2016). This suggests that similar dynamics may occur in vivo, under conditions comparable to those replicated in this study. SEM imaging of untreated samples also showed the formation of a sort of “bridges” among bacterial cells that may indicate the early production of extracellular polysaccharides (e.g., Psl) previously demonstrated to hold bacteria together at the early stage of biofilm development on a surface (Ma et al., 2009). The rapid mechanisms of action demonstrated in the ALI infection model could therefore represent an important property of CFS, capable of preventing the initial stages of biofilm formation before the acquisition of a highly tolerant phenotype by the bacterium. The rapid mechanism of action of CFS corroborates findings from a previous study conducted in a cell-free system, which demonstrated the absence of bacterial regrowth even at extended time points (up to 24 h) (Pompilio et al., 2024).

In a recent study, we sought to identify the components of the CFS responsible for their antipseudomonal activity by fractionating the CFS and analyzing their constituents using a proteomic approach (Bianchi et al., 2025). Our findings indicate that lactic acid—present at approximately 125 mM in the undiluted CFS—plays a significant role in their antibacterial effect. However, additional components (proteins) are likely to contribute to the overall antimicrobial activity (Bianchi et al., 2025).

Importantly, the anti-pseudomonal activity of CFS in the ALI infection model occurred at non-toxic concentrations (dilution 1:8) for lung epithelial cells as assessed in cytotoxicity assays. However, the therapeutic window was quite narrow, as at one dilution lower (1:4), a detectable toxicity was observed. It is worth noting that in vitro toxicity assays may not accurately replicate the complex interactions established in a living organism where cells lie on the basal membrane, are protected within a tissue and are surrounded by the host extracellular matrix. Supporting this, our recent in vivo study using the Galleria mellonella model showed no cytotoxic effects from L. rhamnosus GG CFS, which also conferred protection against subsequent P. aeruginosa infection (Pompilio et al., 2024). Additionally, previous research has reported higher cytotoxic potential of CFS from lactobacilli toward immortalized cell lines than primary cells (Motevaseli et al., 2013). Despite these considerations, the possible toxic effects of CFS should be further investigated and, eventually, strategies aimed at minimizing them should be implemented. For instance, we previously demonstrated that encapsulation of a relatively toxic antimicrobial peptide in chitosan nanoparticles led to a reduction of its toxic potential against murine cells while preserving its antibacterial potential that was rather reinforced by the presence of chitosan (Piras et al., 2015). Similarly, Sharaf et al. (2019) applied this approach to produce seven lactic acid culture-free supernatants in nano-chitosan-loaded forms. In another recent study, CFS of Bifidobacterium bifidum and L. acidophilus were successfully encapsulated in chitosan nanoparticles by the ionic gelation technique (Derakhshan-Sefidi et al., 2024) for oral administration. Encapsulation enhanced their antibacterial activity against Vibrio cholerae, improved their stability during simulated gastric transit and enabled their controlled release at the site of interaction between the host and bacteria (Derakhshan-Sefidi et al., 2024). Thus, although still relatively under-investigated, CFS encapsulation strategies may contribute to accelerating the clinical application of CFS by reducing their toxicity and enhancing their activity.

The next phase of our study focused on evaluating the delivery of CFS to the lungs as part of their preclinical assessment. The International Scientific Association of Probiotics and Prebiotics (ISAPP) defines postbiotic, “a preparation of inanimate microorganisms and/or their components that confers a health benefit on the host” (Salminen et al., 2021). Historically, however, the term postbiotics has often referred to secreted metabolites. Despite postbiotics that act purely through physical mechanisms could be classified as medical devices (European Parliament, 2017), many postbiotics, especially bacterial lysates, have been used for medical purposes, in vaginal, nasal and oral routes of administration (Vinderola et al., 2023). Among postbiotics, bacterial lysate-based medicines are widely used for the treatment or prevention of upper and lower respiratory tract infections, as well as chronic respiratory disease. These medicines are mostly administered orally (capsules, tablets, granules/powder to form a mixture, or oral drops), or inhaled through the nose (as a liquid) (European Medicines Agency, 2018). Moreover, biological medicinal products, as defined in Annex 1 of Directive 2001/83/CE, are those whose active substances are biological in origin. Since postbiotics are derived from living organisms, they can be considered in the biological medicinal products category and, as such, require physical–chemical-biological testing as part of stability and quality programs. Although specific guidelines for postbiotics are lacking, developers can follow existing protocols for biological medicinal products (Vinderola et al., 2023).

In this contest, the translational potential of CFS as a solution for nebulization, was evaluated in accordance with Eu. Phar. (European Pharmacopoeia, 2023) indications and EMA guidelines (European Medicines Agency, 2024). Firstly, sterility was guaranteed by the last step of the preparation method, when CFS are filtered through 0.22 μm filters. The solution’s pH (3.8) fell within the Eu. Phar. range of pH 3–8.5, and its osmolarity was indicative of a slightly hypertonic solution (438 mOsm/Kg), within the tolerability range for inhalable formulations (150–550 mOsmol/kg) (Law, 2001).

Mesh nebulization technology was chosen because it generally does not compromise the stability of labile products and has recently demonstrated superior performance when tested for the delivery of Lacticaseibacillus rhamnosus GG suspension (Byun et al., 2023). A preliminary evaluation confirmed the CFS as suitable for aerosol administration, leading to almost 80% of the solution recovered in the second stage of a TSI. Given that the epithelial lung lining fluid (ELF) physiologically dilutes inhaled droplets, the use of undiluted CFS for nebulization in this study was chosen to reflect clinically relevant exposure conditions. A 2 mL dosage was applied as a representative volume, allowing for both realistic delivery simulation and accurate analytical quantification during the aerodynamic particle size distribution (APSD) studies. This strategy will inform the design of future studies aimed at determining the appropriate clinical dosage, based on the effective and cytocompatible 1:8 dilution. The setting of the therapeutic protocol will take into account that the ELF volume is estimated to range from 10 to 70 mL, with distribution varying by lung region and disease state; specifically, in the respiratory region—where CFS is intended to act—it is estimated at 7–20 mL (Hastedt et al., 2016; Hastedt et al., 2022). In view of future applications, additional investigations were performed to ensure CFS’s effective delivery to the deep lungs. Particularly, APSD is considered one of the Critical Quality Attributes (CQA) of inhalation products, crucial to ensure consistency for further in vivo studies. Instead of using the percentage of emitted dose for the TSI analysis, a multistage impactor or impinger should be used to measure the distribution on each stage and cumulative undersize mass, as these provide more accurate dose variations. A cumulative percentage versus cut-off diameter plot should also be included, from which MMAD and GSD can be determined (European Medicines Agency, 2024). For the APSD measurements performed, mass balance reconciliation was always in the 98–85%, confirming the reproducibility of the analysis. Furthermore, it is recognized that depending on the disease and disease severity, patient-breathing capability and in particular, the flow rate can make a difference in the inhaled and fine particle doses. Lastly, it affects the impact of the drug on lung regions. To assess the complete delivery profile of the medicinal product for in vivo studies (pivotal clinical and/or comparative), in vitro characterization for both the delivery rate and total amount delivered should be provided. A recognized validated technology to ensure accurate in vitro/in vivo correlation is a breath simulator coupled to NGI (European Medicines Agency, 2024). In this study, the aerosol was generated using the same nebulizer and settings as those supposed for future use in adult CF patients. The performed analysis evidenced the good aerodynamic distribution of the CFS both under simulated healthy adult and adult CF patients’ breathing profiles. The most affected parameter is the delivered dose, which is drastically reduced for the simulated CF patient’s profile and suggests a dose volume adjustment to provide a comparable effective dose impact.

In the case of inhalation therapy development, animal models are necessary and widely used for preclinical safety and pharmacological testing of drugs. Nevertheless, fundamental anatomical differences exist between the airways of rodents and humans, emphasizing a substantial lack of functional homology regarding drug deposition rates and localization (Baldassi et al., 2021). For instance, mice only have 6–8 levels of branching airways while humans have up to 20 or more. Furthermore, mice do not have respiratory bronchioles comparable to humans, which are characterized by interruptions on their walls that project into the alveoli. They only have short terminal bronchioles opening straight into the alveolar ductules (Bal and Ghoshal, 1988). Therefore, while data derived from rodent models cannot be directly translated to human contexts, the use of a breathing simulator coupled with NGI is essential to confirm the feasibility of human administration and assess dose impact. In this study, the combination of in vitro cellular models and in vitro ASPD under simulated breathing conditions has advanced our understanding, demonstrating that the product is effective and suitable for inhalation delivery. The necessary animal testing will follow to confirm the pharmacological and safety aspects.

In view of the future development of CFS as a solution for inhalation, stability issues were also investigated. A key factor driving interest in postbiotics is their natural stability, both during production and storage. Live microorganisms present a technological challenge because they are often sensitive to oxygen and heat. However, postbiotics, being non-living, can be easily developed into products with a long shelf life. Postbiotics may also be better suited for regions that lack reliable cold chains or face challenges with ambient temperatures that hinder the storage of live microorganisms (Salminen et al., 2021). For example, the marketed product Lantigen B (Bruschettini S.r.l.), an oral suspension of bacterial lysate, has an expiration date of 3 years in unopened packaging, under storage at temperatures not exceeding 25°C. Several literature studies related to postbiotics are actually focused on isolated metabolites, antimicrobial peptides, and bacteriocins, which actually do not fall within the ISAPP definition of postbiotics. These molecules have a fair stability at diverse pH, UV exposure, and temperatures (Hasan et al., 2025), e.g., confirming a 30-day stability at room temperature (Janakiraman, 2012). CFS may offer even greater stability, with several studies reporting 1 to 5 months of stability at temperatures ranging from 4 to 35°C (Koohestani et al., 2018; Mani-López et al., 2022). Thanks to the interest in CFS application in food processing, it was assessed that when the antimicrobial activity is mainly due to organic acids, phenyllactic acid, fatty acids and heat-resistant proteinaceous compounds, the CFS supports baking (up to 14 min at 210°C) and heat sterilization (Cortés-Zavaleta et al., 2014; Schmidt et al., 2018).

In this study, physicochemical stability studies of CFS were undertaken under stress temperature conditions. Specifically, the Arrhenius equation was applied after exposing CFS sealed into glass vials to extreme temperatures for short periods. The goal was to identify the primary degradation kinetics and extrapolate its behavior under the conditions outlined in the International Council for Harmonization of Technical Requirements for Pharmaceuticals for Human Use (ICH) guidelines (European Medicines Agency, 1995, 2003). These studies are crucial not only for assessing the stability of CFS but also for understanding how it may respond to prolonged storage and heat stress. In addition to functionality tests, the evaluation included acceptance criteria related to appearance and physical attributes—such as color changes and phase separation—which must be monitored and reported for the final drug product (European Medicines Agency, 1995, 2003).

A main macroscopical change observed in heat-stressed CFS was a temperature and time-dependent darkening. This phenomenon is attributed to Maillard reactions between glucose and protein residues, which can also lead to the formation of a visible particulate when the highest darkening is reached (Jiang et al., 2019). This effect was monitored by instrumental acquisition of UV–Vis absorbance. The results evidenced that 10% color alteration would be expected at 9 years of storage at 25°C, thus predicting a good stability performance of CFS.

Biological activity is a fundamental criterion in stability assessment. Firstly, it was observed that no statistically significant differences were observed between heat-treated and untreated CFS in their antibacterial activities, despite the heat-induced degradation. These findings agree with previous reports in which the antibacterial activity of CFS was retained after heating at 95°C for 10 min (Pompilio et al., 2024). It is noticeable that the forced degradation induced the formation of Maillard reaction products (MRPs), which are widely investigated as antimicrobial agents and could then participate in the antimicrobial effect for the heat-stressed CFS (Habinshuti et al., 2019; Nooshkam et al., 2020). To complete the stability studies, the antipseudomonal activity of CFS at one-year storage at room temperature, at 4°C and at −20°C was assessed. The result obtained demonstrated that the killing ability of CFS against P. aeruginosa remained unaffected under all three long-term storage conditions, thereby providing additional evidence in support of their pre-clinical characterization.

In conclusion, in an ALI lung infection model, CFS from L. rhamnosus demonstrated effective anti-bacterial activity against both non-mucoid and mucoid strains of P. aeruginosa isolated from CF lung. The liquid formulation developed demonstrated favorable physicochemical characteristics for pulmonary administration, with optimal aerodynamic particle size distribution for deep lung penetration, essential for addressing chronic infections. Furthermore, CSF exhibited a marked long-term stability, along with the lack of cytotoxicity at microbicidal concentrations. These findings support the development of postbiotic-based inhalation therapies in the context of rising antimicrobial resistance.