Staphylococcus epidermidis MB614 and Staphylococcus hominis MB621 from human milk as up regulators of TNF-α induced IL-8 production by Caco-2 cell line

Abstract

Breast milk harbors many bacteria that shape the newborn microbiota. The present study aimed to investigate the immunomodulatory properties of bacteria from selected Pakistani mother’s breast milk samples. The probiotic strains studied are associated with immune modulation in intestinal epithelial cells, showing potential probiotic properties to be used in a simulated human milk formula. Identification was carried out through 16 S rRNA gene sequencing, which revealed the presence of 50% diversity of Staphylococcus hominis and 35% of Staphylococcus epidermidis among all the isolates. Probiotic bacteria are considered to protect the gastrointestinal tract (GIT) by adhering to the gut epithelial cells and reducing the population of pathogens. The cytotoxicity of Staphylococcus hominis MB614 and Staphylococcus epidermidis MB621 against epithelial cells was tested by co-culturing the strains with Caco-2 cells. Tested probiotic bacteria at concentrations, 1 × 10⁶, to 1 × 10² CFU/ml for 24 h exhibited no noticeable cytotoxic effect on Caco-2 cells. Staphylococcus hominis MB614 and Staphylococcus epidermidis MB621 were examined for modulation of the immune system in human colon carcinoma cell lines (Caco-2) after the effect of the proinflammatory cytokine tumour necrosis factor α (TNF-α) by regulating the proinflammatory interleukin 8 (IL-8). Staphylococcus hominis MB614 and Staphylococcus epidermidis MB621 regulate TNF-α induced IL-8 production in a time-dependent manner. In conclusion, Staphylococcus hominis MB614 and Staphylococcus epidermidis MB621, isolated from human milk, when administered in sufficient amounts and for a certain duration, may be used to modulate inflammation inside the colon by up-regulating the production of IL-8, potentially enhancing innate immunity. This study with these probiotics suggested the modulation of the immune system and disease prevention in the human body.

Supplementary Information

The online version contains supplementary material available at 10.1186/s12896-025-01079-w.

Introduction

The worldwide market for probiotic products is enormous, expected to range over $46 billion by 2020 [18, 32] but now according to allied market research, the 2021 market for products containing probiotic species is estimated to gather $73.9 billion by 2030, registering a CAGR (compound annual growth rate) of 8.6% (https://www.alliedmarketresearch.com/probiotics-market). Formulated poorly, supplements containing probiotics proposed for oral intake frequently fail to shield probiotic bacteria from the harsh environment of the human digestive tract, resulting in bacteria not reaching a viable state to the small intestine and any health benefits time and again shrinks to a placebo effect. Additionally, the capability of probiotic species to impact conventional gut microbiota has not been established. Professionals providing healthcare and the general public consider probiotics to improve clinical symptoms with disbelief [31].

Human breast milk is a thriving supply of nutrients and bioactive ingredients [43]. It does a remarkable job of providing the necessary ingredients for the baby’s healthy development and growth. Infants’ developing immune systems have several increasing deficits that put them at higher risk of illness. Although that infant’s intestinal microflora is linearly designed by the microbiome of breast milk, the development of those gut bacteria is aided by the oligosaccharides found in human milk. Novel “omic” methods such as glycomics, proteomics, metabolomics, and genomes are being applied to identify this mutualistic relationship. Research outlines the expanded function of antimicrobial peptides found in breast milk in innate immune defense. Breast milk plays a crucial role in delivering antioxidative and anti-inflammatory dynamics to the gastrointestinal tract, which contributes to enhanced immunity and decreased illness. Data from creative investigations support the notion that mucosal-associated lymphoid tissue influences the cellular component of breast milk. A new finding concerns human milk stem cells (hMSCs) and their direct contribution to the renewal and restoration of homeostasis in the developing organism. These hMSCs could demonstrate the use of multi-lineage stem cells for tissue regeneration and cancer research. When an infant’s immune system and GIT develop, breast milk also evolves similarly, delivering more calories and fewer immunological components with time [6]. About 90% of the infant’s gut microbiome is colonized by oligosaccharides, which also play a significant role in the development of a stable and diverse microflora, which is essential for appropriate innate and adaptive immunity [53].

Though little is known about the commensal or probiotic bacteria found in human milk, common isolates from this biological fluid in human breast milk include Staphylococci, Streptococci, Lactobacilli, Micrococci, and Enterococci [43]. Microbes from these genera can be easily isolated from the fresh milk of a variety of healthy women. Hence, these bacterial communities ought to be regarded as the typical core microbiota of human milk rather than as common contaminants [3, 30].

According to previous studies (Mártin et al., 2003; 2005; 2006; Arboleya et al., [1, 45]; Mártin et al., 2019), human milk is now known to include commensal and potentially probiotic bacteria. We do not yet know the exact mechanisms that lead to the development of the human milk microbiota. Nonetheless, other theories exist to explain how the milk-bacterium association came to be. Undoubtedly, certain microorganisms are exclusive to the mother’s skin or the baby’s oral cavity, and they may become an important part of the microbiota in the mother’s milk through the reverse flow of milk into the mammary ducts during breastfeeding [39]. For instance, this technique could confirm the frequency of cutaneous and oral microbes found in the milk’s microbiota like Staphylococcus spp. & Streptococcus spp [19]. Interestingly, breast milk also contains an excessive number of intestinal bacteria that may have entered through the mother’s intestines through dendritic cells and CD18 + cells (DC). The ability of CD18 + and DC to retain bacteria from the gut and intestinal tract and transport them to the mammary ducts through translocation led to an increase in microorganisms towards the end of pregnancy and during breastfeeding [39, 42].

Investigations represented that in the human milk samples of mothers who are overweight, the quantity of Staphylococcal strains was higher than Bifidobacterium counts as compared to normal-weight mothers. The body mass index (BMI) and its impact on the composition of the breast milk microbiota during lactation were investigated using the mixed-models approach. Researchers find that during the first six months of breastfeeding, moms who are overweight have higher overall bacterial counts and higher amounts of Lactobacillus and Staphylococcus bacteria, while mothers who are average weight have fewer Bifidobacterium germs. The importance of weight growth during pregnancy had an impact on the composition of the breast milk microbiota in the first six months of breastfeeding. Extreme weight gain during pregnancy was found to be associated with higher amounts of bacteria from the Staphylococcus group in colostrum and lower amounts of bacteria from the Bifidobacterium group in samples taken one month apart [10]. Higher amounts of the Bifidobacterium and Staphylococcus groups in human milk were associated with higher sCD14 concentrations, and higher IL6 concentrations were likewise associated with Staphylococcus bacteria [9]. When comparing the breast milk samples of overweight moms to those of normal weight mothers, researchers found lower levels of CD14 and TGF-β. Samples of milk from women who are overweight generally had lower Bifidobacterium group sums and higher Staphylococcus group counts than mothers who are normal weight. This demonstrated that there are differences in the microbiota of human breast milk compared to the gut microflora in inflammatory, obese, and/or allergy conditions, as was previously known [12, 27].

It is known that TNF- 𝛼 stimulates neutrophils and other inflammatory cells, as well as IECs to produce the proinflammatory mediator IL-8 (Ma et al., 2004). Certain gastric bowel disorders and cell damage are the outcomes of these activities. To determine the ideal dose and duration for observing the inflammatory response, the proinflammatory conditions of TNF-𝛼-induced IL-8 production were replicated in Caco-2 cells. The Caco-2 model demonstrated a gradual increase in IL-8 production upon exposure to TNF-𝛼, with TNF-𝛼 significantly stimulating IL-8 synthesis at a dose of 10 ng/mL. A comparable suppression that is linked to that of live entire cells was also sustained when the Caco-2 cells were co-incubated with Lactobacilli cell extract or cell debris. TNF-𝛼-induced IL-8 production was reduced by MRS medium and cell-free habituated medium of Lactobacilli [7, 9, 42]. According to a publication by Ren et al., in 2013 IL-8 production in Caco-2 cells was effectively suppressed by L. plantarum and L. salivarius when TNF-𝛼 was present [49]. As a result, the two distinct Lactobacilli strains shown probiotic potential for producing ecological agents and functional foods.

Probiotics can be divided according to their capabilities to prevent pathogen adhesion and hinder the production of pro-inflammatory cytokines. Capability of probiotic strains to modulate immunity is considered as a benchmark for assessment of probiotic [17, 40]. Probiotic strains possibly could affect pro-inflammatory responses like against tumor necrosis factor-α (TNF-α) or inter leukine-8 (IL-8) and anti-inflammatory responses like for inter leukine-10 (IL-10) of eukaryotic epithelial cells (Caco-2, HT-29 or other eukaryotic cells) treated with lipoprotein, pathogens or other factors [36, 48]; Sun et al., 2012). Some probiotic strains can down regulate the production of the proinflammatory cytokine IL-8 and stimulate the production of the anti-inflammatory IL-10 by intestinal epithelial cells (Fig. 1) [2, 4, 5]; Yanfeng et al., 2018).

Fig. 1.

Pathway for immune response imparted by probiotics while interacting with epithelial lining

Intestinal epithelial cells upon infection releases proinflammatory cytokines and chemokines along with some monocyte chemoattractant protein (MCP-1) triggering a well-organized immune response. After infection a wide range of probiotics can overpower epithelial cell proinflammatory chemokine production. Lactobacilli species lessen Caco-2 cells CXCL-8 (IL-8) secretion in response to TNF-𝛼 and Salmonella encounter and Bifidobacterium longum subsp. infantis BB-02 decreases ulcerative colitis in a critical mouse experimental model of inflammatory bowel disease by reducing KC/CXCL-1 (IL-1) levels [38]; Boonma et al., 2014; Elian et al., 2014; Yahfoufi et al., [54]. Lactobacillus plantarum L9 inhibit TNF-α induced trans-cellular microbial transport and production of IL-8 in Caco-2 cells. Lactobacillus plantarum L9 displayed the potential to shield enterocytes (cells of the intestinal lining) from a serious inflammatory response and consequently could be a fine potential prophylactic (intended to prevent disuse) agent in responding to bacterial translocation across gut (Wang et al., 2017, Malik et al., [29].

The study described in this article aimed to examine potential immunomodulatory effects caused by potential probiotics candidates isolated from human milk samples. The specific objective of this study was to determine proinflammatory cytokine IL-8 production by intestinal epithelial cells line model Caco-2 in response to TNF-α by certain lactic acid bacteria from breast milk. The strains designated for this study were Staphylococcus hominis MB614 and Staphylococcus epidermidis MB621.

Methodology

Isolation, identification and characterisation

The potential probiotic candidates’ strains employed in the present study were initially isolated during the early phases of the research and subsequently identified using the 16 S rRNA gene sequencing technique. Following their molecular identification, these strains were comprehensively characterized for their probiotic potential through a series of in vitro assays assessing their functional and safety attributes. The findings from this initial isolation, identification, and characterization phase have already been documented and published in our previous works [40, 42, 43].

Immunomodulatory assay

Inoculum Preparation

Caco-2 cells (provided by Microbiology and Cell Science Lab, University of Florida, USA) were grown in an overnight culture with an initial inoculum of absorbance 0.045 at 600 nm in 250 ml to prepare the cells for the stimulation of the innate immune response. The bacterial suspension was spun down at 14,000 rpm for 2 min at -4 °C after 16 h, and the resulting pellet was then re-suspended in 25 ml of PBS (pH 7.4) to create a 10X suspension. The 10X culture suspension was separated into aliquots using phosphate buffered saline (PBS) and kept for later use at -80 °C. To determine the precise inoculum concentration, one aliquot of each strain was removed after 24 h, thawed, and plated at various dilutions [25].

Caco-2 cell culture conditions

Caco-2 cell passage

The previously cultivated cell line flask supernatant was discarded into a beaker with diluted bleach once everything was ready to use. The cells were cleaned using 10 ml of PBS, which was then poured into diluted bleach after being twirled around inside the flask to wash the cells. The trypsin concentration was then brought to 0.05% by thoroughly mixing the solution in the flask after adding 8 ml of PBS and 2 ml of trypsin (0.25%) stock. Once closed, the flask was incubated for five to ten minutes at 37 °C. Following that, the flask was examined under a microscope to check for floating cells, which suggested that the cells were ready to pass. Once all of the cells had been extracted from the flask’s bottom, 10 ml of the full medium was added and thoroughly mixed using a serological pipet. The solution was split into two 15 ml tubes (10 ml in each tube) using a serological pipet. After that, the falcon tubes were centrifuged for five minutes at ambient temperature at 700 rpm (99 x g). Each tube’s supernatant was disposed away in bleach. Pallets from each tube were reconstituted in two milliliters of full medium (one milliliter in each tube), and the contents of one tube were then moved into the other. A 0.4% solution of trypan blue (azo dye) was produced using a syringe filter. Subsequently, 10 µl of trypan blue were extracted, and 10 µl of reconstituted solution was introduced. Trypan blue was used to stain the cells, and a hemocytometer was used to count them. Ten ml of dye-resuspended Caco-2 cells were applied to each side of the hemocytometer slide. Each quadrant of the slide’s surface had its cells counted, and the average number of cells was calculated and multiplied by the dilution factor. In 15 ml of full medium, 1 × 10⁶ Caco-2 cells were planted for passage to the new flask. Up until the experiment date, the cells were allowed to develop for 18–21 days per flask, with media changes occurring every other day.

The Microbiology and Cell Science lab at the University of Florida with courtesy, provided the human colon adenocarcinoma, Caco-2 cell lines. Caco-2 cells were grown in Dulbecco’s modified Eagle’s medium (DMEM, HyClone, Laboratories Inc., Logan, UT, USA). At 37 °C in an incubator with 5% (v/v) CO₂ and 95% (v/v) humidified air, DMEM was supplemented with 15–20% heat-inactivated foetal bovine serum (FBS) (HyClone), penicillin (100,000U/ml), streptomycin (10 mg/ml), and amphotericin B (25 µg/ml) (Sigma). Every three days, Caco-2 cells were seeded and then subculture at a concentration of 1 × 10⁶ cells/ml. Every other day, the medium was changed during the culturing procedure. Between passages 12 and 15, or after 80% confluence, were the cells used for this investigation. To explore whether isolated strains of the proinflammatory mediator TNF-α could prevent or exacerbate inflammatory organ or tissue damage, its effects were measured. Wells were treated with bacteria for one hour before being supplemented with 10 ng/ml of TNF-α (R&D Systems) during the experimental process. The medium was taken out for ELISA after 12 and 24 h had passed. At least three independent experiments were conducted with triplicate results for each test.

Hood Preparation

Before starting work, the hood was disinfected by exposure to UV light for ten to fifteen minutes, and the blower was left to operate for ten minutes. The hood was prepared for use by heating PBS and full Dulbecco’s modified Eagle’s medium (DMEM) with 10% fetal bovine serum (FBS) to 37 °C in a water bath, defrosting trypsin, and inspecting the cell line for contamination, such as clear patches (indicating fungal contamination) and floating cells (indicating cell death). It was ensured that all the necessary materials were available. When the hood was prepared for usage, 70% ethanol was sprayed inside to clean and disinfect everything that was placed inside.

Caco-2 cell viability assay in interaction with isolated strains

Caco-2 cell viability after exposure to isolated strains was evaluated by trypan blue exclusion using a hemocytometer. Subsequent to the third Caco-2 cell passage, 1 × 10⁵ cells were planted onto six well culture plates. Following 80% confluence, Caco-2 cells were co-cultured for 24 h with isolated strains at 1 × 10⁶, 1 × 10⁵, 1 × 10⁴, 1 × 10³, and 1 × 10² CFU/ml. After treatment, cells were trypsinized, resuspended in complete medium, and mixed 1:1 with 0.4% trypan blue; viable (unstained) and non-viable (blue) cells were counted under a light microscope. Viability was calculated as (viable cells / total cells) × 100. All conditions were assessed in three independent biological replicates and 3 technical counts per replicate. For the current study we set an a priori acceptance threshold of ≥ 90% viable cells (1 × 10⁵) to exclude cytotoxicity as a confounder for cytokine readouts [35, 47].

Immunostimulant identification on human Caco-2 cell line

Co-culturing of isolated bacteria with Caco-2 cells

After centrifuging pre-prepared cell stocks at 2600 X g for 7 min at 4 °C, different concentrations of selected probiotics were added to pre-seeded Caco-2 monolayers (15–21 days). After one hour, antibiotics and proinflammatory mediators (TNF-α) were administered to halt the development of bacteria. Then, using an ELISA kit (BD OptEIATM Human IL-8 ELISA Set Cat. No. 555244), medium was collected after 12 and 24 h to quantify IL-8 production. By using ELISA from culture media supernatants, interleukin was determined [25].

IL-8 ELISA

ELISA (enzyme-linked immunosorbent assay) was used to measure the amount of IL-8 that the isolated strains produced in response to colon cells. The method used was the one prescribed by the supplier, BD biosciences’ Human IL-8 ELISA Set (cat # BD 555244). In the experimental setup, controls were carefully defined to ensure reliable outcomes. Negative controls consisted of untreated Caco-2 cells, devoid of bacteria or TNF-α, establishing a baseline for comparison. Positive controls utilized TNF-α alone, inducing a robust IL-8 response of 178 pg/mL at 12 h, validating the inflammatory response model. Bacterial concentrations, ranging from 10² to 10⁶ CFU/mL, were selected based on human milk bacterial loads reported in studies by Mehanna et al. [3, 30], as well as prior work by Kingma et al. [25, 38], ensuring physiological relevance. The choice of 12-hour and 24-hour time points for analysis was supported by pilot data indicating peak IL-8 production at 12 h (Fig. 2), consistent with findings from Ren et al. [25, 38], optimizing the temporal framework for assessing inflammatory responses.

Fig. 2.

Time-dependent production of IL-8 by Caco-2 cells in the presence or absence of TNF-α stimulation. Caco-2 cells were incubated with or without TNF-α (10 ng/mL) and IL-8 secretion was quantified at 0, 12, and 24 h by ELISA. Results are expressed as mean ± standard deviation of triplicate experiments. TNF-α stimulation markedly increased IL-8 secretion, peaking at 12 h (~ 170 pg/mL) before declining at 24 h, whereas unstimulated cells produced only low basal levels of IL-8 throughout the incubation period

Standard Preparation

After bringing it to room temperature and carefully opening the vial to avoid material loss, the standard was reconstitution was carried out. To create a stock standard, the lyophilized standard was reconstituted using 1 ml of deionized and distilled water. Before making any dilutions, the stock had to stabilize for at least fifteen minutes. Mixed by gentle vortex. Storage/ handling of reconstituted standard was done after reconstitution by immediately aliquot standard stock in polypropylene vials at 50 µl per vial and freeze at -80 °C for up to 6 months (the reconstituted standard cannot be left at room temperature). To prepare standards for assay, a 200 pg/ml standard from the stock standard was prepared. Then 300 µl Assay Diluent was added in six tubes labeled as 100, 50, 25, 12.5, 6.3, and 3.1 pg/ml. Serial dilutions were performed by adding 300 µl of each standard to the next tube and vortex between each transfer. Assay Diluent served as the zero standard (0 pg/ml).

The 96 well plates were coated with 100 µl per well of Capture Antibody (provided) diluted in Coating Buffer. The plates were incubated overnight at 4 °C. The wells were aspirated and washed 3 times with ≥ 300 µl/well Wash Buffer. After last wash, the plates were inverted and blotted on absorbent paper to remove any residual buffer. The plates were blocked with ≥ 200 µl/well Assay Diluent and incubated at room temperature (RT) for 1 h. Then the plates were aspirated 3 times with ≥ 300 µl/well Wash Buffer. The standard and sample dilutions were prepared in Assay Diluent. Then 100 µl of each standard, sample, and control was pipetted into appropriate wells. The plates were then sealed and incubated for 2 h at RT. Total 5 washes were given to plates. Then 100 µl of prepared Working Detector (Detection Antibody + SAv-HRP reagent) (Provided) was added to each well. The plates were sealed and incubated for 1 h at RT. The plates were washed as mentioned before, but with 7 total washes (in this final wash step, wells were soaked in wash buffer for 30 s to 1 min for each wash). Then 100 µl of substrate solution (Pierce™ TMB Substrate Kit, Cat. No. 34021) was added to each well. The plate (without plate sealer) was incubated for 30 min RT in the dark. Later, 50 µl of stop solution was added to each well. The absorbance was measured at 450 nm within 30 min of stopping reaction. For wavelength correction, absorbance 570 nm was subtracted from absorbance 450 nm. Concentration of IL-8 was calculated taking values of standard curve (Figure S1) in consideration [46].

Statistical analysis

Bacterial isolates from human milk were evaluated through morphological, biochemical, and phenotypic characterization. Data were statistically analyzed using conventional variation statistics methods implemented in the Microsoft Excel statistical analysis package. To assess the significance of differences in arithmetic means and standard deviations, one-way analysis of variance (ANOVA) was performed using Excel’s built-in tools, followed by two-way ANOVA for comparison between variables i.e. no simulation, simulation via TNF-α, simulation by bacterial isolate, and simulation by bacterial isolate and TNF-α, post-hoc tests, including Tukey’s Honestly Significant Difference (HSD) test for pairwise comparisons, to enhance the robustness of p-value analysis (Table S1). Differences were deemed statistically significant at p < 0.05, consistent with methodologies described by Khaziakhmetov et al. [23].

Results

Isolation, identification and characterisation

In the initial phase of this research, multiple bacterial strains were isolated from human milk and subjected to molecular identification by 16 S rRNA gene sequencing, confirming their taxonomic placement as Staphylococcus hominis, Staphylococcus epidermidis, Bacillus sp., Streptococcus salivarius and Streptococcus lacterius [40–43]. These isolates exhibited key probiotic attributes, including tolerance to acidic pH and bile salts, antimicrobial activity against selected pathogens, and the ability to adhere to intestinal epithelial cells. The combined findings from these publications established the probiotic potential of these strains and provided the basis for their use in the present study.

Human cell lines are typically used to simulate the interaction of bacteria with abiotic surfaces or epithelial cells in order to assess the potential adherence and activity of probiotics in vitro. In the current study, the ability of isolated probiotic bacterial strains to modulate immunity was assessed using the Caco-2 cell line. The goal of the experiment was to measure the amount of proinflammatory cytokines i.e. interleukine-8 (IL-8), produced by human epithelial cells, specifically line model Caco-2, in response to TNF-α, by some probiotic bacteria that have been isolated from human milk.

Antibiotics growth curve

A growth curve comparing the effects of antibiotics and non-antibiotic controls was conducted using a cocktail of penicillin, streptomycin, and antimycotes, which is commonly employed in Caco-2 cell growth conditions (Fig. 3).

Fig. 3.

Growth curve with and without antibiotics where A: is representing Staphylococcus hominis MB614 and B: represent Staphylococcus epidermidis MB621

Susceptibility analysis of test strains and Caco-2 cells co-culturing

A six-well culture plate was seeded with 1 × 10⁵ Caco-2 cells following passage three. Following 80% confluence, Caco-2 were co-cultured for 24 h with separate strains at 1 × 10⁶, 1 × 10⁵, 1 × 10⁴, 1 × 10³, and 1 × 10². A hemocytometer was used to count the viable cells (Fig. 4). Cell count indicates that the chosen isolates have no discernible cytotoxic effect on Caco-2 cells (Fig. 5). There is no discernible cytotoxic effect of certain isolates on Caco-2 cells based on cell count at all the concentrations.

Fig. 4.

Caco-2 cells viability counting through hemocytometer where the transparent cells are the living Caco-2 cell and dyed blue cells are dead Caco-2 cell

Fig. 5.

Viability of Caco-2 cells following exposure to different dilutions of probiotic strains MB614 and MB621. Caco-2 cells were treated with serial dilutions (10⁻⁶ to 10⁻² g/mL) for 24 h, and viable cell counts were determined and expressed as colony forming units per milliliter (CFU/mL) on a logarithmic scale. Bars represent the mean ± standard deviation of triplicate experiments. No significant reduction in Caco-2 cell viability was observed across the tested concentrations compared with the untreated control, indicating that both strains were non-toxic to Caco-2 cells at the concentrations evaluated

Immunosimulation of Caco-2 cells by isolated probiotic strains

Co-culturing was done using pre-made bacterial (probiotics) aliquots and pre-seeded Caco-2 monolayer (15–21 days). Once an hour had passed, proinflammatory mediators (TNF-α) and antibiotics (to prevent bacterial growth) were introduced. After 12 and 24 h, the supernatant medium was gathered to assess the innate immune system response’s production of IL-8, a crucial immune response mediator released by epithelial cells.

IL-8 ELISA

Neutrophils and other inflammatory cells are known to be activated by TNF-𝛼, which also causes IECs to release the proinflammatory mediator IL-8. Certain intestinal tract disorders and cell damage can result from these actions. The best time to look into the inflammatory response was found by tracking the time course of TNF-𝛼-induced IL-8 production in Caco-2 cells (a schematic representation of the IL-8 ELISA is shown in Fig. 6). This allowed us to simulate a proinflammatory backdrop. The Caco-2 cell lines do not create IL-8 at time zero, as Fig. 2 illustrates. When TNF-𝛼 was induced at 10ng/mL, IL-8 production rose and peaked at 12 h (178 pg/mL). Therefore, in the IL-8 induction trials that followed, the stimulation duration of 12 h was determined to be accurate. Previous studies showed that at 10 ng/mL, TNF-𝛼 significantly stimulated IL-8 production. Therefore, this concentration was considered apt for in subsequent studies.

Fig. 6.

Diagram showing experimental design used in the study for the measurement of IL-8 production by probiotic bacteria from human milk in response to TNF-α inside epithelial cell model

To investigate the response of Caco-2 cells to test strains in an inflammatory context, Caco-2 cells were treated with TNF-𝛼 to mimic an inflammatory background after treatment with live test strains at different concentration to get statistically significant results. However, co-incubation of the TNF-𝛼-stimulated Caco-2 cells with both test strains significantly decreased IL-8 production in a time dependent manner. Error bars in the experiment showed standard deviation of the experiment done in triplicate.

By measuring the linear trend of optical density (450–570 nm) vs. known concentration of IL-8 (200pg, 100pg, 50pg, 25pg, 12.5pg, 6.2pg and 3.1pg/ml of IL-8 standard) the liner model was generated by excluding the points far away from the trend line (Figures S3 & S4). The liner model then was used measure the enzyme concentration in pg/ml. Results for the strains used for the study are presented below.

Staphylococcus hominis MB614

Wet lab results showed that the probiotic strain Staphylococcus hominis MB614, NCBI Accession # MG751372.1, interacted with Caco-2 cell line epithelial cells for an hour. The isolate’s lipopolysaccharides (LPS) acted upon toll-like receptors (TLRs) on the epithelial surface. They activated the myeloid differentiation primary response MyD88 pathway, a known connector for the inflammatory signaling pathway, producing nuclear factor kappa B (NF-κB) subunits, namely P65 and P50, as well as an inhibitor of kappa B kinase (IKK). Interleukine 8 (IL-8) is a proinflammatory cytokine produced when IKK phosphorylates an inhibitor of kappa B (IκB), causing NF-κB to translocate into the nucleus. To inhibit the growth of the bacterial cells, a particular antibiotic was applied after the strain was co-cultured with epithelial cells for one hour. The addition of tumor necrosis factor-alpha (TNF-α) causes oxidative stress by binding to tumor necrosis factor receptors (TNFR) on epithelial cells and redox-regulated gene expression (Ref-1 & ROS). IL-8 nuclear expression was subsequently induced by this expression. The Ref-1-regulated transcription factors NF-κB and AP-1 had their DNA binding activity to the IL-8 prompter boosted by TNF-α, which in turn led to an increase in IL-8 production (Fig. 7). The pathway’s findings demonstrated Staphylococcus hominis MB614’s pro-inflammatory activity in the Caco-2 cell line.

Fig. 7.

IL-8 secretion in Caco-2 cells following exposure to Staphylococcus hominis MB614 under different conditions. Bars represent IL-8 concentrations (pg/mL) after 12-h and 24-h incubations with or without TNF-α stimulation. The p-value (< 0.05) reflects the significant correlation between IL-8 levels measured at 12 h and 24 h

Staphylococcus hominis MB614 at 10⁶ CFU/ml induced a robust response in the CXCL8 gene (IL-8) at concentrations by 2.5-fold after 12 h and 5.5-fold after 24 h of incubation. TNF-α stimulation led to a two-fold increase in CXCL8 chemokine concentrations in untreated Caco-2 cells after 12 h and a 2.5-fold increase after 24 h, as shown by the chemokine synthesis. TNF-α increased CXCL8 expression by two times (P < 0.05) when added to Caco-2 monolayers treated with 1 × 10⁶ CFU/ml of Staphylococcus hominis MB614, but the inflammatory response was not exacerbated by incubation with 10⁵ or 10⁴ CFU/ml (Table S1). The results were calculated by measuring the IL-8 concentration. It was observed that the amount of IL-8 that Caco-2 cells produced in the absence of TNF-α stimulation showed a significant proliferation, but that concentrations of TNF-α at 1/10, 1/100, and 1/1000 showed a decrease. According to these findings, Staphylococcus hominis MB614 may not activate the immune system when it comes into contact with intestinal epithelium as indicated in Fig. 7. The results indicated that the measurable proliferation in specific chemokines may be associated with the identification by IECs of RNA and/or DNA from Staphylococcus hominis MB614.

Staphylococcus epidermidis MB621

To control the production of proinflammatory cytokines like IL-8, the results for the probiotic strain Staphylococcus epidermidis MB621, NCBI Accession # MG751379.1 demonstrated the same interaction with the MyD88 pathway as was previously discussed for Staphylococcus hominis MB614. The next step of the experiment was the addition of TNF-α and its interaction with TNFRs on Caco-2 cells to generate oxidative stress through the production of the redox-regulated gene Ref-1. IL-8 is subsequently expressed inside nucleus as a result of this expression. To increase the synthesis of IL-8, TNF-α increased the DNA binding action of transcription factors regulated by Ref-1 to the IL-8 prompter. The pathway’s findings demonstrated Staphylococcus epidermidis MB621’s pro-inflammatory activity in the Caco-2 cell line.

The production of IL-8 was increased to roughly 29 pg/ml in biological replicates incubated for 12 h and to 65 pg/ml in samples incubated for 24 h (Fig. 8), following a 3 × 10⁸ CFU/ml treatment of Staphylococcus epidermidis MB621 to trigger a response in the amounts of CXCL8. TNFα stimulation of untreated Caco-2 cells led to an increase in CXCL8 chemokine levels of roughly 80 pg/ml. When TNFα was added to Caco-2 monolayers treated with 3 × 10⁸ CFU/ml Staphylococcus epidermidis MB621, the result was an increase in CXCL8 expression, resulting in 115 pg/ml of IL-8 after 2 h and 158 pg/ml after 24 h (P < 0.05) (Table S1). In contrast, incubation with 10⁷ to 10⁵ CFU/ml did not worsen the inflammation. The IL-8 concentration was measured to derive these conclusions. Measurements revealed that, in the absence of TNFα activation, Caco-2 cells showed a significant increase in IL-8 concentration, which is correlated with the amount of Staphylococcus epidermidis MB621 that was supplied. The results suggested that the measurable increase in a specific chemokine may be related to the genetic material from Staphylococcus epidermidis MB621 that IECs detected. A 96-well plate with various IL-8 doses for MB621 is shown visually in Figure S2.

Fig. 8.

IL-8 secretion in Caco-2 cells following exposure to Staphylococcus epidermidis MB621 under different conditions. Bars represent IL-8 concentrations (pg/mL) after 12-h and 24-h incubations with or without TNF-α stimulation. The p-value (< 0.1) reflects the correlation between IL-8 levels measured at 12 h and 24 h

Discussion

The impact of probiotics on regulating inflammatory responses is a crucial component of their therapeutic efficacy, as numerous inflammatory disorders are attributed to gut abnormalities [15]. In an in vitro cell culture model exposed to the simulator of the human intestinal microbial ecosystem (SHIME) media after dosing with Symprove (a balanced microbiome), no deterioration in epithelial barrier integrity or reduction in inflammation markers occurred. Instead, inflammatory chemokines IL-8, CXCL10, and MCP-1 levels were reduced [28, 31].

At this stage of the study, Caco-2 cell viability was assessed using the trypan blue dye exclusion method, which provides a direct and reliable measure of membrane integrity to distinguish viable from non-viable cells. As the primary objective was to confirm the non-cytotoxic nature of probiotic strains MB614 and MB621, this method was sufficient to demonstrate that no significant reduction in viable cell counts occurred across the tested concentrations. Similar approaches have been reported in early cytotoxicity screening studies, where trypan blue exclusion was effectively used to evaluate cell integrity prior to performing metabolic assays such as MTT [47].

Kingma et al. [25] observed strong responses in CXCL8 (IL-8) mRNA, which increased 90-fold with 10¹¹ CFU/L of Lactobacillus johnsonii (Ljo). TNF-α stimulation of untreated Caco-2 cells increased CXCL8 mRNA, but incubation with 10¹¹ CFU/L Ljo did not worsen the inflammatory response, while TNF-α with 10¹⁰ CFU/L Ljo increased CXCL8 expression. A similar pattern was seen in our study for Staphylococcus hominis MB614 and Staphylococcus epidermidis MB621. In Caco-2 cells without TNF-α, IL-8 levels rose proportionally with bacterial dosage, indicating that both strains stimulate innate immunity via IL-8 upregulation. Additional TNF-α did not exacerbate this response, implying these strains trigger early immune activation that may enhance host defense (Mazziotta et al., 2023).

These findings from human milk isolates contradict some reports on epithelial models. IL-8, one of the strongest neutrophil activators, was our focus (Noh et al., 2015; Schneider et al., 2016). Following Clostridium difficile infection, colonic epithelial cells secrete less IL-8 in the presence of Lactobacillus rhamnosus L34 and Lactobacillus casei L39 [54]. However, like our results, Lactobacillus plantarum strains have been assessed for immunomodulatory potential by measuring IL-8 release in vitro. Streptococcus cerevisiae similarly suppressed ETEC-induced proinflammatory cytokines and chemokines such as IL-6, IL-8, CCL20, CXCL2, and CXCL10 (Zanello et al., 2011, [52]).

Mueller et al. (2013) and Schneider et al. (2016) examined whether Lactobacillus plantarum adherence influences IL-8 release from NCM460 cells exposed to an inflammatory cytokine cocktail (IL-1β, TNF-α, INF-γ). Numerous studies confirm that intestinal epithelial cells respond to bacteria by secreting cytokines and chemokines such as TNF-α, IL-8, MCP-1, and macrophage inflammatory protein-1, which recruit neutrophils, T cells, monocytes, and eosinophils (Borruel et al., 2003). IL-8, a C-X-C chemokine, plays a central role in mucosal inflammation and intestinal barrier disruption (Daig et al., 1996; Seydel et al., 1997). We thus examined IL-8 secretion pathways in Caco-2 cells after TNF-α treatment. Although Caco-2 cells alone secrete little IL-8, pre-incubation with S. hominis MB614 and S. epidermidis MB621 greatly increased TNF-α-triggered IL-8 production (Wang et al., 2017).

Similar findings have been reported using HT-29 cells to test anti-inflammatory activities of GI-digested bean milk and yogurt fractions. TNF-α induced IL-8 secretion, yet some milk fractions (10–50 kDa) showed no significant inhibition at 0.5 mg/mL (Chen et al., 2019). Caco-2 cells, expressing intestinal enzymes and transporters, are widely used to model these interactions (Jochems et al., 2018). In our study, human milk isolates S. hominis MB614 and S. epidermidis MB621 increased IL-8 secretion in intestinal cells treated with inflammatory cytokines, consistent with previous reports (Reilly et al., 2007; Ren et al., [38]; Schneider et al., 2016).

IL-8 secretion is strain-dependent. Human-derived Lactobacillus plantarum WCFS1 and IMC513 R did not impact IL-8 secretion in another study (Garcia-Gonzalez et al., 2018). Receptor interactions, particularly via TLR9, are key to probiotics’ anti-inflammatory effects. Probiotic ligands activate NF-κB through MyD88 after binding to TLR2 and TLR6, initiating signaling [20, 21, 26, 33]. TLR2 recognizes peptidoglycan, and Lactobacillus strains enhance TLR2 expression in Caco-2 cells. In Salmonella-free mice, L. casei produced IL-10, FN-γ, and TNF-α while eliciting TLR expression [34, 37, 54]. Such mechanisms help normalize the host immune system and strengthen pathogen defenses [11, 24, 50].

Other probiotics demonstrate antiviral effects. Enterococcus faecium promotes antiviral responses by increasing IL-8 and IL-6 mRNA [8] and suppresses coronavirus formation causing epidemic diarrhea in pigs [16, 44]. Probiotics may enhance vaccine efficacy, prevent viral entry, and modulate cytokine production during infection [51].

Beyond gut models, Staphylococcus epidermidis from skin produces butyric acid. In vitro, its glycerol fermentation pattern was studied, and in UVB-exposed mice, S. epidermidis with glycerol or butyric acid altered IL-6 production [22]. In contrast, our human milk isolate S. epidermidis MB621 could not suppress IL-8 generation in colon epithelial cells after 24 h. Dong et al. [14] explained this by strain specificity: Staphylococcus epidermidis sepsis strains induced higher IL-10 and IL-6 than biofilm-forming clinical isolates, which increased IL-10 but not IL-6. Mutant strains for extracellular virulence factors and biofilm development showed no effect on cytokine expression. S. epidermidis also induces a pro-inflammatory response in lung epithelial cells, with strain-to-strain variation. Biofilm-forming strains produce stronger respiratory pro-inflammation than non-biofilm strains [13]. In this study TNF-α primes Caco-2 cells results in increased IL-8 production and S. hominis/epidermidis further boosts IL-8 resulting in synergistic neutrophil recruitment signal from gut epithelium, suggesting these milk-derived staphylococci enhance mucosal immune vigilance against pathogens.

Conclusion

Breast milk contains a distinct microbiota dominated by Staphylococcus hominis and Staphylococcus epidermidis, which may influence early gut colonization. This study used an in vitro Caco-2 epithelial cell model to assess the immunomodulatory effects of S. hominis MB614 and, more prominently, S. epidermidis MB621 on TNF-α-induced IL-8 production. Although neither strain significantly reduced TNF-α-induced IL-8 secretion, S. epidermidis MB621 demonstrated a stronger capacity to modulate epithelial immune responses, suggesting its potential as a candidate for probiotic starter cultures or supplements. Further in vivo studies are warranted to confirm these effects and evaluate its clinical utility.

Supplementary Information

Below is the link to the electronic supplementary material.

Author contributions

A.S was the principle investigator and conducted project designing, experimentations and initial draft writing, whereas all the other authors were A.S.‘s research team.

Funding

This study not financially funded by any organization.

Data availability

The datasets used and/or analysed during the current study available from the corresponding author on reasonable request.

Declarations

Ethical approval

This study was conducted in accordance with the ethical principles outlined in the Declaration of Helsinki and the guidelines of the FJWU Institution’s Ethical Review Committee. The MB614 and MB621 strains of S. hominis and S. epidermidis were isolated in 2017 from human milk collected by manual expression, with informed consent obtained from the donor for research use.

Competing interests

The authors declare no competing interests.