Caco-2 cells (American Type Culture Collection, Middlesex, UK) were used as a model of mature and differentiated enterocytes. Cells were grown in high glucose DMEM with 10% fetal calf serum (FBS), 1% non-essential amino acids, 50 mg/ml streptomycin, 50mU/ml penicillin. The cells were grown for 15-18 days after confluence on polycarbonate Snapwell filters (pore size 0,4 micron) (Costar Italia, Milan, Italy).

The infection of Caco-2 cell monolayers was performed with the simian rotavirus strain SA11 (RV) at a multiplicity of infection (MOI) of 25. RV activation was performed with 20 µg/mL trypsin for 1 hour at 37°C. Then, viral sample was added to the apical side of the Caco-2 cell monolayers for 1 hour at 37°C, then the cells were rinsed 3 times and incubated in serum–free medium for 1 hour. Then, Caco-2 cells were mounted in Ussing chamber system as described below. This model has been currently used to test the effects of antimicrobial molecules (Buccigrossi et al., 2014) and has become a standard tool for studying the direct effects of drugs and toxins in human intestinal epithelial cells. Cells were incubated with LGG in different conditions as described below before and after RV infection.

A preparation of LGG supernatant (6x10⁹ u.f.c.). LGG were grown in Dulbecco’s modified Eagle essential medium (DMEM; Life Technologies Italia, Monza, Italy) with a high glucose concentration (4.5 g/L); 10% fetal bovine serum (FBS, Life Technologies Italia, Monza, Italy) and 1% non-essential amino acids were added. LGG were cultured for 72 h at 42°C. The cells-free culture supernatant (mLGG) was obtained by centrifugation and the subsequent passage through a 0.22 mm filter. The stock used in this study has a protein content of 1.5 mg/mL

The Ussing chamber system is used to evaluate ion transport in polarized cell monolayers, grown on permeable supports. Caco-2 cells were short-circuited by a voltage clamp in Ussing chambers (Physiological Instruments, San Diego, CA). Electrical parameters analyzed are: short circuit current [Isc represents the short circuit required to bring Vt to 0 mV expressed in µA/cm², epithelial membrane resistance expressed as tissue ionic conductance (G measured in mS/cm²) and, finally, transmembrane voltage (Vt, expressed in mV)]. Every 20 sec current pulses were passed across the epithelium and the current was measured and transepithelial resistance (R) was calculated.

Isc is the parameter that is a measure of the secretory/absorptive result. The integrated response was calculated between 10 and 50 mins. In addition, the peak of Isc response obtained in the time period considered (after subtraction of baseline Isc; named Δ Isc) was used as a measure of the secretory/absorptive response. The proabsorptive effect is indicated as negative value of Δ Isc whereas the positive values indicate an active ion secretion. The maximal increase of Isc (Δ Isc) is considered the peak effect (Buccigrossi et al., 2020). Caco-2 cell monolayers were stimulated with LGG or mLGG 1 hour before RV infection and for the duration of the experiment.

DCFH-DA spectrofluorometry was used to measure ROS production in our model. DCFH-DA (20 µM) was added for 30 minutes at 37°C in the dark after stimulation. Intracellular ROS levels was measured in a fluorometer (Kontron Instruments, Japan). For DCF fluorescence imaging, Caco-2 cells were grown for 3 days on the cover glass, then fixed and permeabilized with paraformaldeyde 4% and Triton 0,2% for 30 minutes at 4°C. DCF-HA 20 µM was added for 30 minutes at 37°C in the dark. Fluorescence images from multiple fields were obtained using a Nikon Eclipse e 80i microscopy. The images were analyzed using NiS Elements D imaging software (Nikon Instruments Inc., NY, USA).

GSH/GSSG ratio at intracellular level was measured by a fluorimetric assay kit (Biovision Milpitas, CA). The values were normalized for protein content and expressed as % of control.

Transepithelial electrical resistance (TEER) of cell monolayers grown on filters was measured using a Millicel-ERS resistance monitoring apparatus (Millipore). The net TEER (in Ohms/cm²) was calculated by subtracting the background from the actual value and multiplying the value obtained by the area of the filter (4.9 cm²).

Actin staining: fixation and permeabilization was performed with 4% paraformaldehyde and 0.2% Triton X-100 for 30 min at + 4°C. Cells were treated with 50 µg/ml solution of Alexa Fluor 594-phalloidin (Sigma-Aldrich) for 40 min.

Occludin staining: cells were fixed by adding 100% methanol for 10 min at room temperature and probed with anti-occludin antibody (Abcam ab59720) over night at +4°C. Bound antibody was detected with Alexa fluor 488 conjugated anti-rabbit IgG antibody (Invitrogen, A21206).

Immunofluorescence assay for NF-kB: Caco-2 cells were grown on the chambered cover glass for 3 days, fixed with 4% buffered paraformaldehyde (pH 7.4) for 30 minutes followed by blocking and permeabilization for 1 h in PBS with 1% BSA and 0.1% Triton X-100 at room temperature. Caco-2 cells were then incubated at 4°C with anti-NFkB p65 antibody (Santa Cruz Biotechnology, Santa Cruz, CA, USA) in a humidified chamber and incubated with a FITC-conjugated goat anti-rabbit antibody at room temperature and with Hoechst 33342 (10 µg/mL) for 5 min.

For all immunofluorescence studies, slides were mounted with Vectashield Mounting Medium with DAPI (Vector laboratories, Ltd, UK). Fluorescence images from multiple fields were obtained using a Nikon Eclipse e 80i microscopy. The images were analyzed using NiS Elements D imaging software (Nikon Instruments Inc., NY, USA).

We used caspase-3 as a marker of apoptosis (Furuya-Kanamori et al., 2015). An apoptosis assay kit was used to determine caspase-3 activity, according to the manufacturer’s instructions (Biovision, Mountain View, CA). Caspase-3 activity was investigated in Caco-2 cells by the release of the chromophore pNA after substrate cleavage. Modifications of caspase-3 activity were determined by comparing the sample optical density (OD) with the control.

Total cell lysates were obtained by homogenization of cell pellets in cold lysis buffer (20 mM Tris, pH 7.5 containing 300 mM sucrose, 60 mM KCl, 15 mM NaC1, 5% (v/v) glycerol, 2 mM EDTA, 1% (v/v) Triton X-100, 1 mM PMSF, 2 mg/ml aprotinin, 2 mg/ml leupeptin and 0.2% (w/v) deoxycholate) for 1 min at 4°C and then sonication for 30 sec at 4°C. Equal amounts of protein were separated on 10% (v/v) SDS-PAGE and transferred to a PVDF membrane (Millipore). The membrane was blocked with 5% (w/v) skim milk. Then, incubation with primary antibody followed by an HRP-conjugated secondary antibody were performed. Separated proteins were visualized with the ECL detection system (GE-Healthcare). The following antibodies were used for Western blot analysis: anti-caspase-3, anti-NFkB p65 and anti-Tubulin antibodies (Santa Cruz Biotechnology).

GraphPad Prism Software (San Diego, CA) was used to evaluate the two-tailed unpaired Student t test. In addition, a 2-tailed paired Student t test to evaluate statistical significance. The alpha value of 0.05 was set for statistical significance. p-values for each analysis are indicated in figure legends.

As previously reported in the basic model of Caco-2 cells, RV-induced chloride secretion is NSP4-dependent and involves oxidative stress (De Marco et al., 2009; Buccigrossi et al., 2014). To investigate the effects of LGG on RV-induced chloride secretion, we preincubated Caco-2 cell monolayers with living LGG (LGG) or the conditioned medium from LGG (mLGG) before RV infection. In Ussing chamber experiments, the magnitude of short circuit current (ΔIsc), which reflects the intensity of RV-induced chloride secretion, was significantly reduced by LGG and completely abolished by mLGG ( Figure 1 ). In the latter condition, a pro-absorptive effect was also observed. LGG and mLGG alone did not induce Isc changes (data not shown).

Since RV-induced enterotoxic damage is oxidative stress-dependent (De Marco et al., 2009), the redox state was evaluated in Caco-2 cells infected with RV following preincubation with LGG or mLGG. RV induced ROS production that was completely inhibited by LGG and mLGG at the same levels of N-acetylcysteine (NAC), a potent antioxidant ( Figures 2A, B ). To evaluate the role of antioxidant defenses, we determined the levels of reduced and oxidized glutathione (GSH/GSSG ratio) in RV-infected Caco-2 cells following preincubation with LGG or mLGG. The GSH/GSSG ratio was maintained when RV-infected cells were exposed to LGG or mLGG ( Figure 2C ). In all experiments, NAC was used as antioxidant factor. Our results indicated that NAC preincubation results in an anti-oxidant preventive effect on ROS production and GSH/GSSG ratio ( Figure 2 ).

Since NSP4 induces chloride secretion by Ca²⁺ pathways, we investigated the role of Ca²⁺ in our model. Thapsigargin, an inhibitor of the ER Ca²⁺-ATPase (Michelangeli et al., 1995; Díaz et al., 2012), selectively depletes the endoplasmic reticulum Ca²⁺ stores. Thapsigargin induced an increase in Isc (0.9 ± 0.5 vs 16.1 ± 1.5; p<0.05) but no difference was found when cells were preincubated with LGG or mLGG (control 16.1 ± 1.5; LGG 14.9 ± 0.9; mLGG 13.7 ± 1.2; p=NS) suggesting that the mechanism of action is upstream Ca²⁺ release.

To investigate LGG effects on RV-induced epithelial damage, RV infected Caco-2 cell monolayers were exposed to LGG or mLGG and intestinal epithelial integrity was evaluated.

Occludin staining looked like a network in untreated and well differentiated cells whereas RV induces a major break of the junctions between the cells. The presence of LGG, but not mLGG, preserved cells from RV-induced damage ( Figure 3A ). This effect was also confirmed by transepithelial resistance data. RV induced a strong decrease of TEER that was significantly prevented by LGG but not by mLGG ( Figure 3B ). Finally, actin staining also showed that LGG living microorganisms but not mLGG, protects cells by cytoskeleton destruction induced by RV ( Figure 3C ).

Apoptosis was evaluated by nuclei staining and caspase activation. RV infected cells have smaller nuclei with irregular edges ( Figure 4A ). Because DAPI staining is an indirect marker of apoptosis, caspase-3 activation was evaluated. We observed that RV infection increases caspase-3 activity suggesting that apoptosis is one of the mechanisms of cell damage. Both LGG and mLGG significantly reduced apoptosis induced by RV with a similar efficacy ( Figures 4A, B ).

NF-kB activation has a key role in intestinal epithelial permeability (Dominguez et al., 2013; Yang et al., 2014) as well as in infections (Santoro et al., 2003), including RV (Rossen et al., 2004) and SARS-CoV-2 (Poeta et al., 2021) infections. NF-kB is found in the cytoplasm as an inactive factor, a heterodimer of p50 and p65 subunits bound with a member of the IkB inhibitor protein family. Removal of IkB, induced by several stimuli, results in NF-kB activation through p65 phosphorylation and its translocation to the nucleus (Campbell and Perkins, 2006). To see whether NF-kB was activated in response to RV infection, we evaluated the levels of phospho-p65 and its nuclear localization in RV-infected Caco-2 cells.

As shown in Figure 5A , the levels of phospho-p65 were increased in cell extracts following RV infection. We also observed p65 translocation from cytoplasm to the nucleus with immunofluorescence ( Figure 5B ). Phospho-p65 intracellular levels was inhibited in RV-infected cells exposed to LGG and mLGG with a similar efficacy ( Figure 5A ) and the nuclear translocation did not ensue ( Figure 5B ).