Angelo Sisto, Institute of Sciences of Food Production (ISPA), National Research Council (CNR), Via Amendola 122/O, 70126 Bari, Italy. E-mail: firstname.lastname@example.org
Aims: This study was performed to ascertain the immunomodulatory effect of Lactobacillus paracasei strains. These strains were also genetically characterized.
Methods and Results: The strains were genetically differentiated by using the fluorescent-amplified fragment length polymorphism technique, which led to the identification of several molecular markers unique to each strain. To determine the immunomodulatory properties, we evaluated the effect of strains on dendritic cell maturation, dextran uptake, ability to induce proliferation of allogenic T cells and cytokine secretion. The results indicated that all the strains stimulated phenotypic maturation of dendritic cells (DCs), but they acted differently on DCs in relation to the other tested properties; notably, a different effect on cytokine secretion was detected.
Conclusions: The results of this study revealed different immunomodulatory properties of strains of the species Lact. paracasei. Strain IMPC 4.1 showed an interesting anti-inflammatory ability. Probiotic strains IMPC 2.1 and LMG P-17806 were characterized by a similar and intermediate ability to induce cytokine secretion in contrast to the very low ability of strain LMG 23554.
Significance and Impact of Study: Our results confirm that each single strain of a bacterial species appears to influence the immune system in a peculiar manner. The evaluation of the different types and/or levels of cytokines whose secretion is induced by each strain could be relevant to define its pro- or anti-inflammatory properties and its more appropriate clinical use.
Probiotic bacteria have been defined by FAO and WHO as ‘live micro-organisms which, when administered in adequate amounts, confer a health benefit on the host’ (FAO/WHO 2001). Species of the genera Bifidobacterium and Lactobacillus are contained in many commercial products including fermented foods and food complements and are widely used as probiotics. A general assumption is that the health-promoting capacities of probiotic bacteria are based on one of the following, sometimes overlapping mechanisms such as: (i) restoration of microbial homeostasis through microbe–microbe interactions, (ii) pathogen inhibition, (iii) enhancement of epithelial barrier function and (iv) modulation of immune responses. In this regard, many studies have provided evidence that probiotics can interact with the immune system modulating the immune functions (Borchers et al. 2009). Actually, probiotic bacteria influence both the development and regulation of intestinal immune and nonimmune defences (Tlaskalová-Hogenováet al. 2004). The stimulatory effects of probiotics are not limited to innate immunity but can also influence a pre-established adaptive immunity. In particular, several probiotic strains possess the ability to induce interleukin (IL)-10 production in immune cells (Medina et al. 2007; Ménard et al. 2008), thus regulating the magnitude of immune activation in response to pathogens. Moreover, probiotic bacteria interact with dendritic cells (DCs), which play an essential role in both innate and adaptive immunity (Borchers et al. 2009). In fact, probiotics appear to play a role in lymphocyte differentiation (Forchielli and Walker 2005), a process mediated by DCs, essential for a balanced conversion of naïve lymphocytes into effector T cells or regulatory Treg cells (Baumgart and Carding 2007). Notably, probiotic strains can induce cytokine patterns in DCs different from those created by pathogenic bacteria (O’Mahony et al. 2006). In this regard, IL-10 production and CD83 expression were induced in DCs by Bifidobacterium bifidum, Bifidobacterium longum and Bifidobacterium pseudocatenulatum strains, thus provoking a shift of the immune response towards a T helper type (Th) 2 profile (Young et al. 2004). There are also in vitro studies providing evidence that bacterial strains belonging to different species displayed distinct immunomodulatory effects on human DCs (Hart et al. 2004). However, different strains of the same species may also differentially polarize the phenotype of the immune response. For example, some strains of Bif. longum were shown to induce high levels of IL-4 and IL-10 (Th2-like response) in splenocytes, while others promoted a Th1-like phenotype characterized by interferon-γ and tumour necrosis factor-α (TNF-α) secretions (Medina et al. 2007; Ménard et al. 2008). In contrast, similar studies comparing immunomodulatory capacities of strains that belong to a single species of the genus Lactobacillus are thus far lacking. We recently reported the ability of probiotics to exert modulatory effects both on the innate and on the adaptive immunity in a mouse model of food antigen sensitivity (D’Arienzo et al. 2009). In particular, among the tested strains, Lactobacillus paracasei induced the highest levels of maturation in DCs. Therefore, we focused our further studies on this bacterial species with the aim to reveal the diverse immunomodulatory capacities of different strains. In particular, herein we examine the immune activity on murine DCs of five strains belonging to the species Lact. paracasei. These strains have also been clearly characterized and differentiated from a molecular point of view by using the fluorescent-amplified fragment length polymorphism (F-AFLP) technique, the latest version of the AFLP method based on the use of fluorescent dye-labelled primers (Savelkoul et al. 1999; Mortimer and Arnold 2001).
Materials and Methods
BALB/c and B10.M mice were maintained under pathogen-free conditions at the animal facility of the Institute of Food Sciences. Mice were used at the age of 6–12 weeks and killed by inhalant anaesthesia with isoflurane. These studies were approved by the National Institutional Review Committee.
Bacterial strains and culture conditions
Lactobacillus paracasei IMPC 2.1 (from human intestine; deposited as strain LMG P-22043 in the Belgian Coordinated Collections of Microorganisms, Ghent, Belgium) and IMPC 4.1 (from human intestine) were obtained from the Culture Collection at the Istituto di Microbiologia, Università Cattolica, Piacenza, Italy. Lactobacillus paracasei ATCC 334 (from a dairy product) was obtained from the American Type Culture Collection. Lactobacillus paracasei LMG P-17806 (from human intestine) and LMG 23554 (from human blood) were obtained from the Belgian Coordinated Collections of Microorganisms, Belgium. Strains used in this study were also confirmed to belong to the species Lactobacillus paracasei by using the species-specific PCR assay based on the tuf gene developed by Ventura et al. (2003). Based on their characteristics, strains IMPC 2.1 (Valerio et al. 2006, 2010) and LMG P-17806 (= F19) (Crittenden et al. 2002; Ohlson et al. 2002) are considered probiotic strains, while potential probiotic features of strains ATCC 334 and IMPC 4.1 are unknown. Resembling previous studies (Daniel et al. 2006), strain LMG 23554 (= strain YS8866441) was used as a certain nonprobiotic (Daniel et al. 2006) as it was isolated from a blood culture of a patient with infective endocarditis and was also shown to be capable of exacerbating colitis in TNBS-treated mice (under severe inflammatory conditions) and of translocation to extra-intestinal organs (Daniel et al. 2006). Working cultures were grown in de Man Rogosa Sharpe (MRS) broth (Difco, Detroit, MI, USA) for 24 h at 37°C under aerobic conditions without shaking and were subcultured twice before use in experiments. The cell concentration of individual strains was evaluated by checking the optical density value at 600 nm. For long-term storage, stock cultures were prepared by mixing 8 ml of a fresh culture with 2 ml of Bacto glycerol (Difco) and then freezing 1-ml aliquots of this mixture at −80°C in 2-ml sterile cryovials (Nalgene, Rochester, NY, USA).
Isolation of bone marrow dendritic cells
DCs were generated according to a published method (Lutz et al. 1999) from BALB/c mice. In brief, bone marrow cells from the femurs and tibiae of mice were flushed, and cell aliquots (2 × 106) were diluted in 10 ml of RPMI 1640 medium (Sigma, St Louis, MO, USA) supplemented with antibiotics (penicillin 100 IU ml−1; streptomycin 100 IU ml−1), 10% foetal calf serum and 20 ng ml−1 granulocyte–macrophage colony-stimulating factor (GM-CSF) (culture medium) before to be seeded into 100-mm petri dishes (Falcon, Heidelberg, Germany). On day 3, 10 ml of culture medium was added, and on day 7, 10 ml of the culture medium was replaced by freshly prepared medium.
Dendritic cell treatments
On day 10, nonadherent DCs were harvested by gently pipetting. Cell aliquots (1 × 106 ml−1) were then placed in 24-well plates and incubated in culture medium with 5 ng ml−1 GM-CSF in the presence of LPS and/or different concentrations of bacterial strains (1 × 104, 1 × 105, 1 × 106 and 1 × 107 CFU ml−1). Before incubation bacterial strains were irradiated with 2800 Gy (Gray) γ-irradiation (MDS Nordion γ-cell 1000) to prevent their proliferation during the dendritic cell growth. Following incubation, DCs were analysed by using a vital stain (1% Nigrosin) and found >90% live cells.
Allogeneic T-cell activation
Increasing number of bacteria-treated DCs were cultured (in triplicate) in 96-well plates with 105 allogeneic CD4+ T lymphocytes purified from the spleen of B10.M mice by using Dynal Mouse T-cell-negative isolation kit (Invitrogen, Carlsbad, CA), according to the manufacturer’s instructions. After 48 h bacterial stimulation, DCs were irradiated with 25-Gy (Gray) γ-irradiation to prevent their proliferation during the course of the mixed leucocyte reaction (MLR). DCs were added to the 96-round bottom plate in halving dilution from 5·0 × 104 to 2·5 × 103 cells per well. Responder T cells were added at 1 × 105 cells per well, such that stimulator/responder ratio ranged from 2 : 1 to 40 : 1. Proliferation was determined 5 days later by uptake of [3H] thymidine (1 μCi per well) for the last 18 h of culture. Results were expressed as Δ cpm (cpm T cells elicited by DCs-cpm T cells alone).
Mouse DCs were stained with phycoerythrin (PE) or fluorescein isothiocyanate (FITC)-conjugated Abs (purchased from BD-Pharmingen, San Diego, CA, USA), i.e. FITC-conjugated anti-CD11c, PE-conjugated anti-CD86, anti-CD80 or anti-CD40. Cell staining was analysed by using a CyFlow Space flow cytometer (Partec, Munster, Germany) and FlowJo software (Tree Star Inc., Ashland, OR, USA). For each Ab, an isotype control of appropriate subclass was used.
Analysis of cytokine pattern
Bone marrow DCs were cultivated for 2 days in the presence of different concentrations of bacterial strains (1 × 104, 1 × 105, 1 × 106 and 1 × 107 CFU ml−1) and/or LPS (1 μg ml−1). Supernatants were collected and analysed for IL-12, TNF-α, IL-2 and IL-10 protein levels by ELISA, essentially according to a published ELISA sandwich protocol (Mauriello et al. 2007). In brief, 100 μl of capture antibody solution (BioLegend, San Diego, CA, USA) was plated into ELISA wells (Nunc Maxisorb; Cat. no. 446469) and incubated overnight at 4°C. After the removal of the antibody solution, 100-μl aliquots of PBS supplemented with 2% BSA were added and incubated at room temperature for 2 h. Next, cytokine standard and samples, diluted in blocking buffer supplemented with 0·05% Tween-20, were incubated overnight at 4°C. At the end of the incubation, 100-μl aliquots of biotinylated antibody solution were plated and left for 2 h at room temperature. Streptavidin–horseradish peroxidase conjugate IgGs (1 : 2000 dilution) (BD-Pharmingen) were incubated for 1 h at room temperature and, finally, 100-μl aliquots of 63 mmol l−1 Na2HPO4, 29 mmol l−1 citric acid (pH 6·0) containing 0·66 mg ml−1o-phenylenediamine/HCl and 0·05% hydrogen peroxide were dispensed into each well and allowed to develop. The absorbance was finally read at 450 nm, and the cytokine concentrations were calculated by using appropriate standard curves.
After 48 h incubation of DCs with the various bacterial strains, fluorescein isothiocyanate-dextran (FITC-dextran; Sigma FD40) was added at 1 mg ml−1 and cells were incubated for 30′–60′ at 37°C. Controls were incubated at 4°C. At the end, cells were collected and resuspended in PBS and 2% foetal calf serum and analysed by FACS.
Analyses were performed in triplicate, and each result was expressed as mean ± SD. Statistical significance was determined by anova test using the GraphPad prism 4.0 software (GraphPad Software, Inc, La Jolla, CA, USA). A P-value of 0·05 or less was considered to be significant.
Genomic DNA was extracted using the Wizard® Genomic DNA Purification Kit (Promega Corporation, Madison, WI, USA) according to supplier’s specifications. DNA quality and quantity were controlled using an ND 1000 spectrophotometer (NanoDrop Technologies, Inc., Wilmington, DE, USA) as well as by 0·7% (w/v) agarose gel electrophoresis in TAE buffer stained with 0·5 μg ml−1 of ethidium bromide (Sambrook et al. 1989) and photographed under ultraviolet light.
F-AFLP analysis was performed by using the AFLP microbial fingerprinting kit (Applied Biosystems, Foster City, CA) according to manufacturer’s recommendations. For each strain, a preselective PCR was carried out using EcoRI+0 and MseI+0 primers and PCR products of each reaction were diluted 20 : 1 with TE buffer. For selective PCR, 1·5 μl of the resulting diluted PCR samples was further selectively amplified using the following six combinations of EcoRI and MseI adaptor-specific selective primers with base selection: EcoRI+A/MseI+A, EcoRI+A/MseI+T, EcoRI+A/MseI+G, EcoRI+A/MseI+C, EcoRI+G/MseI+G and EcoRI+G/MseI+T. The EcoRI primers were labelled with fluorescent dye (Applied Biosystems). PCR amplifications were performed in a model PE 9700 Perkin-Elmer thermocycler (Applied Biosystems). After amplification, 1 μl of reaction product was mixed with 20 μl of formamide and 1 μl of GeneScan-500 (ROX) size standard (Applied Biosystems), ranging from 35 to 500 bp in size. The mixture was heated for 3 min at 95°C and cooled on ice. The amplification products were separated by capillary electrophoresis on an ABI PRISM 310 Genetic Analyzer (Applied Biosystems).
Fragments were sized by using ABI Genescan version 2.1 software (Applied Biosystems), and all the electropherograms were visually inspected and compared for peak detection with the same software. The peak height threshold was set at 100 fluorescent arbitrary units, and any peak below this value was not included in the analysis. To ascertain the reproducibility of AFLP fingerprints, F-AFLP analyses were repeated twice. The data of the AFLP fingerprints obtained by utilizing ABI Genescan software were used as input for BioNumerics ver. 5.1 software for fragment matching.
Five strains of Lact. paracasei were genetically characterized and compared by means of F-AFLP using six selective primer pairs. These primers were suitable to obtain clear molecular profiles because they generated a number of fragments (peaks) per strain ranging from 27 to 58 (Table 1). Replicate AFLP experiments generated highly reproducible peak patterns and clearly differentiated the strains. Each strain was characterized by an overall number of fragments (AFLP markers) ranging from 239 to 273 (Table 1) many of which (46–52%) were polymorphic (i.e. fragments not present in all the analysed strains as determined on the basis of their different sizes). Moreover, the comparison of AFLP profiles led to the identification of a number of AFLP markers unique to each strain.
Table 1. AFLP markers of Lactobacillus paracasei strains. The markers were obtained by using selective primers with base selection EcoRI+A/MseI+A, EcoRI+A/MseI+G, EcoRI+A/MseI+C, EcoRI+A/MseI+T, EcoRI+G/MseI+T and EcoRI+G/MseI+G
Total number of AFLP markers
Polymorphic markers n (%)
Markers unique to each strain
Effects of Lactobacillus paracasei strains on the maturation of bone marrow DCs derived from BALB/c mice
The maturation status of bone marrow DCs was preliminary assessed by in vitro incubation with LPS. Collected cells were MAb stained and assessed for the expression of dendritic cell marker CD11c and costimulatory molecules B7-1 (CD80), B7-2 (CD86) and CD40, a member of the TNF receptor family, by FACS analysis. LPS challenge essentially induced the expression of CD86 (Fig. 1). In addition, the costimulatory molecules CD40, CD80 and the dendritic cell-specific marker CD11c were also induced but to a lesser degree.
To determine the effect of Lact. paracasei strains on dendritic cell maturation, bacteria (1 × 107 CFU ml−1) were added to the culture medium for the last 2 days before cell harvesting in the presence/absence of LPS. All strains increased the surface expressions of CD11c and CD80 but not of CD40, in immature dendritic cells (iDCs) at levels comparable with LPS stimulation (Fig. 1). Interestingly, CD86 surface expression following incubation with all strains was higher than with the positive control. On the other hand, co-incubation with bacteria did not further increase the expression of any examined markers (not shown).
Next, FITC-dextran uptake was assessed as a measure of the DCs phagocytic activity. In iDCs we found high levels of uptake, whereas following LPS incubation, this parameter decreased, as expected (Fig. 2; P < 0·001). iDCs’ challenge with all tested strains (1 × 107 CFU ml−1) was also associated to reduction in dextran uptake (P < 0·001). In particular, ATCC 334 was particularly effective in this sense, still operative on mature dendritic cells (mDCs) (9·75 ± 0·1 vs 33·94 ± 0·8, dextran uptake, mean ± SD, ATCC 334 vs DCs alone, P < 0·05). Lower concentration of bacterial strains was found unable to influence the examined activities of DCs (data not shown).
Effects of Lactobacillus paracasei strains on the functionality of bone marrow DCs
To determine whether Lact. paracasei strains influenced the activity of DCs as antigen-presenting cells (APC), we evaluated their capacity to induce proliferation of allogenic T cells in a MLR. We found that all bacterial strains, tested at 1 × 107 CFU ml−1, significantly increased [3H]-thymidine uptake of T cells co-incubated with iDCs (Fig. 3, upper panel); in particular, IMPC 2.1 and IMPC 4.1 were the stronger stimulators (22857 ± 1580 and 22525 ± 5565 vs 4027 ± 257, Δcpm, mean ± SD, IMPC 2.1 and IMPC 4.1 vs DCs alone at 10 : 1 T : DCs ratio; P < 0·01). In LPS-stimulated DCs, APC activity was enhanced and bacterial strains did not further stimulate T-cell proliferation (Fig. 3, bottom panel). Lower bacterial concentrations did not influence the APC activity of DCs (data not shown).
The analysis of the cytokine profile showed that no strain significantly increased TNF-α secretion in iDCs (Fig. 4a), and levels were much lower than those induced by LPS. Moreover, co-administration of LPS and bacterial strains did not further enhance TNF-α production. Similar results were obtained by testing strains in the whole concentration range (1 × 104 to 1 × 107 CFU ml−1; data not shown). On the other hand, only the highest concentration (1 × 107 CFU ml−1) was able to stimulate the expression of the other examined cytokines (data not shown). All strains induced IL-12 secretion both in iDCs and in mDCs (Fig. 4b), with the exception of strain LMG 23554 that showed a reduced stimulatory ability. Interestingly, a marked differential expression of IL-10 was induced by bacteria. In particular, IMPC 4.1 and ATCC 334 stimulated the highest levels (Fig. 4c). In contrast, strain LMG 23554 did not induce IL-10 expression. Interestingly, a synergistic activity with LPS was reported for all examined strains. IMPC 4.1 and ATCC 334 were stronger inducers also in mDCs. Finally, all strains induced IL-2 production both in iDCs and in mDCs. Again, IMPC 4.1 and ATCC 334 were the most inductive strains also for this crucial regulatory cytokine (Fig. 4d).
An increasing number of studies clearly indicate that probiotic bacterial strains interact with components of the immune system exerting a modulatory effect, which may be different depending on the bacterial strains but also on the different experimental models (Delcenserie et al. 2008; Borchers et al. 2009). Although it seems now evident that even strains of the same bacterial species may behave differently, comparison of immunomodulatory features of strains is often difficult because they have been evaluated in different studies using diverse experimental systems. In this study, to compare the immunomodulatory effect of five strains of the species Lact. paracasei, we examined their interaction with DCs. In fact, DCs play a central role in regulating immunity. These cells are important in earliest bacterial recognition and in determining the subsequent T-cell responses. DCs have specialized functions in the intestine, contributing to oral tolerance induction by generating regulatory T cells through production of cytokines such as IL-10 and transforming growth factor-β (Stagg et al. 2003). As intestinal DCs are difficult to isolate in sufficient numbers, we used murine bone marrow DCs that correspond rather well to the results of in vivo studies (Borchers et al. 2009).
We evaluated the effect of Lact. paracasei strains on DCs maturation, and the results indicated that all strains had the property to stimulate phenotypic maturation of iDCs increasing the surface expressions of CD11c and CD80 but not of CD40, at levels comparable with LPS stimulation. In contrast, surface expression of CD86, following incubation with bacterial strains, was even higher than the expression caused by LPS stimulation. Up-regulation of CD86 is consistent with other studies in which probiotic Lactobacillus species caused a similar enhancement (Christensen et al. 2002). The differential effect between members of the B7 family of co-stimulatory molecules may be related to the difference in proportions of DCs expressing CD80 and CD86 (Hart et al. 2004).
The results of this study differentiated the Lact. paracasei strains on the basis of their ability to induce proliferation of allogenic T cells in a MLR. In fact, we found that all strains increased proliferation of T cells co-incubated with iDCs, but strains IMPC 2.1 and IMPC 4.1 were the stronger stimulators. These results seem to be in contrast with inhibition of CD4+ T-cell proliferation by Lact. paracasei NCC2461; on the other hand, this difference could be also due to different experimental conditions (von der Weid et al. 2001), different strains of Lact. paracasei or a diverse T-cell population.
The analysis of the cytokine profile also revealed a different effect of Lact. paracasei strains on cytokine secretion by DCs as they caused a diverse secretion of IL-2, IL-10 and IL-12. In particular, IMPC 4.1 and ATCC 334 stimulated the highest levels of IL-2 and IL-10. The latter has an anti-inflammatory effect and primarily acts to inhibit the Th1 response (Moore et al. 2001). It is a critical cytokine for the maintenance of tolerance to commensal intestinal bacteria, and in its absence mice developed severe intestinal inflammation (Kuhn et al. 1993). On the contrary, IL-12 has functions mutually antagonistic to IL-10 and induction of a high ratio IL-10/IL-12 is considered indicative and predictive of efficient in vivo anti-inflammatory properties (Grangette et al. 2005; Foligne et al. 2007). In fact, the ability of LAB strains to induce a high ratio of IL-10/IL-12 production in human PBMC correlated with their capacity to provide significant protection from TNBS-induced colitis (Grangette et al. 2005; Foligne et al. 2007). In this regard, our results indicated that based on the IL-10/IL-12 ratio and a sustained IL-2 production, strain IMPC 4.1 can be considered particularly efficient as an anti-inflammatory/regulatory strain. Notably, this activity was operative also on mDCs. This feature was not common to the other Lact. paracasei strains tested in this study as well as to Lact. paracasei strain NCC2461, which induced high levels of both IL-12 and IL-10 in murine splenocytes (von der Weid et al. 2001). Interestingly, a strong anti-inflammatory effect on DCs was attributed to Lact. paracasei B21060 when co-incubated with Salmonella (Mileti et al. 2009). This strain, different from Lactobacillus plantarum and LGG which exacerbated a DSS-induced colitis, was protective against the disease. However, strain B21060 did not induce a high secretion of IL-10 when incubated alone with DCs, but when co-incubated with Salmonella, reduced the ability of Salmonella to induce IL-12p70 and TNF-α while not altering its ability to promote IL-10 production (Mileti et al. 2009). Therefore, the anti-inflammatory properties and potential protective ability of strain IMPC 4.1, which was able to stimulate high secretion of IL-10 and a low level of IL-12 in iDCs and more markedly in mDCs, could be even higher than those of strain B21060. A peculiar cytokine pattern also differentiated strain LMG 23554 (= strain YS8866441) from the other tested strains. It was characterized by a very low ability to induce IL-10 and IL-12 secretion in iDCs. This feature could be related to the behaviour of strain LMG 23554 which can be considered atypical in comparison with other strains of the same species. In fact, strain LMG 23554 (isolated from a blood culture of a patient with infective endocarditis) was able – when severe inflammatory conditions occurred – to exacerbate colitis in TNBS-treated mice and to translocate to extra-intestinal organs (Daniel et al. 2006). In this regard, a low ability to stimulate the immune system could explain the ability of that strain to cross the intestinal mucosal barrier and/or persist in the extraintestinal organs or circulation.
Lactobacillus paracasei strains analysed in this study were also genetically characterized by means of F-AFLP, which allowed their differentiation on the basis of a very high number of molecular markers, some of which were also unique to each strain. These markers could be related to the peculiar immunomodulatory properties of each strains, and their isolation and sequencing could allow the identification of the genetic determinants of these relevant phenotypes.
In conclusion, the results of this study revealed the different immunomodulatory properties of strains of the same bacterial species (Lact. paracasei), also showing the interesting anti-inflammatory ability of one of them (strain IMPC 4.1). Moreover, our results suggest that the assessment of the ability to induce cytokine secretion can be helpful in evaluating the safety of new potential probiotic bacteria. Probiotic strains IMPC 2.1 and LMG P-17806 seem to be characterized by both a similar and an intermediate ability to induce cytokine secretion. In fact, they induced a low pro-inflammatory response, not resulting in an inflammatory outcome, but enough to induce a state of host immune system alertness. On the other hand, the unusual behaviour of Lact. paracasei strain LMG 23554 could be related to its low ability to stimulate cytokine secretion. Our results confirm that each single probiotic strain appears to influence the immune system in a particular fashion. Therefore, for each new potential probiotic strain, the profiles of the cytokines whose secretion is induced should be established to allow certification of its pro- or anti-inflammatory properties and to define the more appropriate clinical use; moreover, a strain with a very low ability to induce cytokine secretion should require further investigation to evaluate its possible use as a probiotic.
This research was supported by the Italian Ministry of Education, University and Research (art. 12 D.M. 593/2000 – D.D. 3300 – 22 December 2005 – tema 2) Project ‘Ortobiotici pugliesi’ (D.M. 28830).