Aims: To evaluate the positive influence of the probiotic strain Lactobacillus paracasei LMGP22043 carried by artichokes into the human gut with special reference to faecal bacterial balance, short-chain fatty acid concentrations and enzyme activities in a randomized, double-blind human trial in comparison with probiotic-free artichokes (control).
Methods: Twenty subjects were randomized into two groups, which consumed daily 180 g of the artichoke product (probiotic or control) during two 15-day study periods (periods 1 and 2) separated by a 15-day washout in a crossover manner. Faecal samples were subjected to microbiological and biochemical analyses, and a strain-specific PCR was performed to monitor the probiotic strain.
Results: The probiotic strain, transported by the vegetable matrix, transiently colonized the gut of 17/20 subjects (median 6·87 log CFU g−1 faeces), antagonized Escherichia coli and Clostridium spp. and increased the genetic diversity of lactic population based on REP-PCR profiles, mainly after period 1.
Conclusions: The probiotic L. paracasei LMGP22043 successfully colonized the human gut and positively influenced faecal bacteria and biochemical parameters.
Significance and Impact of the Study: The association of the probiotic L. paracasei with a food carrier rich in fibre can represent a new strategy for favouring a daily supply of probiotics and attracting more consumers to vegetable food fortified with probiotic strains.
Probiotic bacteria are widely consumed for improving human well-being and for their beneficial influence on intestinal microbial balance. Currently, treatments with probiotics are generally managed following protocols of administration with pharmaceutical formulations, while the latest trends are moving towards natural and traditional foods fortified with probiotic strains. Moreover, additional health effects can be obtained in the presence of indigestible molecules that are potentially available for gut microbiota acting as prebiotics. These molecules can be added to the probiotic formulation originating a symbiotic or can be naturally contained in the food matrix used as a carrier for transporting the bacterial cells into the human gut. The symbiotic product combines the efficacy of the probiotic and prebiotic mainly in modulating the intestinal microbiota and its activity, thus resulting in a complex balance of effects. However, the real mechanisms of action of probiotics and prebiotics in the human intestine are not easy to understand and distinguish.
An important factor influencing probiotic efficacy is the ability of the strain to survive food processing and storage and, after human consumption, the gastrointestinal tract (GI). The food carrier should contain functional ingredients that positively interact with the probiotic, sustaining its survival and growth in the food product, buffering bacterial cells through the GI tract and regulating their colonization (Ranadheera et al. 2010). For example, some fruits and plants such as table olives and artichokes represent an especially good source of nutrients which can protect microbial strains and sustain their growth (Lavermicocca et al. 2005; Valerio et al. 2006). In particular, artichoke-based foods provide great advantages upon consumption as they contain the prebiotic inulin and other indigestible organic molecules that reach the colon and can be used by the intestinal microbiota (Goňi et al. 2005; Roberfroid 2007). Besides, a protective action is performed by the micro-architecture of artichoke surface which sustains adequate populations of viable bacteria and their anchorage, improving bacterial survival in gastrointestinal digestion (Valerio et al. 2006). The protective action exerted by these food matrices has been demonstrated by the performance of selected lactobacilli –Lactobacillus plantarum and Lactobacillus paracasei– during a simulated gastrointestinal digestion and in vivo trials (Valerio et al. 2006, 2010; Lavermicocca et al. 2010). Recently, an in vivo post-test trial on eight human subjects suffering from constipation suggested the efficacy of L. paracasei-enriched ready-to-eat artichokes in modulating at individual level potentially harmful bacteria, faecal enzyme activity and short-chain fatty acid (SCFA) production as well as symptom profile (Valerio et al. 2010). Moreover, the ability of this strain to antagonize in vitro the human pathogen Yersinia enterocolitica and to inhibit the related ureolytic activity has also been demonstrated (Lavermicocca et al. 2008). In fact, the mechanisms by which consumption of lactobacilli and bifidobacteria with probiotic characteristics positively affect human health include the increase in the relative numbers of ‘beneficial bacteria’ of gut microflora and pathogen antagonism. Actually, probiotic strains successfully compete with pathogens for space and/or nutrients at the site of action or hamper the outgrowth of the pathogens through negative interactions such as the production and excretion of inhibitory products. In addition, increases in lactobacilli and bifidobacteria may result in the acidification of the gut, improvements to the nutritional status of gut epithelium and the decrease in intestinal permeability to toxic molecules. Dietary components, in particular carbohydrates not digested in the upper gut, are subjected to bacterial metabolism in the large intestine where they are transformed to lactic acid, hydrogen, carbon dioxide and SCFAs and then rapidly adsorbed (Cummings et al. 1987; Brouns et al. 2002; Bekkali et al. 2007). The three major SCFAs are acetate, propionate and butyrate, and the last is supposed to maintain gut integrity and to prevent the bacterial translocation occurring in several gastrointestinal diseases (Perez Chaia and Oliver 2003).
Finally, the regular consumption of probiotics has also been associated with the reduction in faecal enzymatic activity, including that of β-glucuronidase and β-glucosidase involved in the generation of potentially carcinogenic metabolites in the colon (Goldin and Gorbach 1977; Tanaka 1997; Nakamura et al. 2002).
The aim of the current study was to evaluate the positive influence of the probiotic strain L. paracasei LMGP22043 carried by artichokes into the human gut with special reference to faecal bacterial balance, SCFA concentrations and enzyme activity in comparison with control artichokes not containing the bacterial strain, during a double-blind randomized trial on 20 human subjects.
Materials and Methods
Selection of subjects
Twenty subjects (three men and 17 women, age 37·8 ± 13·9 years) completed the study. Information on the healthy status of subjects was obtained at the moment of enrolment during an interview on the current diet, lifestyle and medical history followed by a physical examination.
Exclusion criteria were as follows: BMI outside the range 20–29 kg m−2, alcohol intake above 30 g day−1, smoking habit, hypertension, diabetes mellitus and other pathologies (e.g. systemic, endocrine, collagen-related diseases, familial hypercholesterolaemia, and hypertriglyceridaemia).
Subjects had not taken antibiotics, probiotics, prebiotics, vitamins, minerals, nonsteroidal anti-inflammatory or prokinetic drugs, bismuth, antacids, H2-receptor antagonists, omeprazole, sucralfate or misoprostol in the last 6 months prior to the examination and had no previous history of gastric or duodenal ulcers or gastric surgery. The protocol was approved by the local Scientific and Ethics Committee of I.R.C.C.S. ‘Saverio de Bellis’, (approval number 261/DSC08), and all subjects gave informed consent to take part in the study.
Design of the trial
Participants integrated their usual lifestyles and dietary intakes throughout the study period with probiotic containing artichokes (PR) or control artichokes (CTR) supplied by Copaim Spa, Albinia, Italy. Probiotic artichokes were prepared as reported in the patented procedure by Lavermicocca et al. (2004). This product – authorized for commercialization by the Italian Ministry of Health – contained approximately 1·20 × 108 CFU of live probiotic cells of L. paracasei LMGP22043 (IMPC 2·1) per gram corresponding to a daily dose of about 2 × 1010 CFU (Lavermicocca et al. 2003, 2004; Valerio et al. 2006). Viable counts of the probiotic bacteria in the food product were monitored throughout the study as reported in the study of Valerio et al. (2006). Probiotic and control artichokes were both lightly seasoned with olive oil and packed in modified atmosphere (MAP) to obtain a ready-to-eat artichoke product (about 180 g) that was stored at refrigerated temperature. Final products had identical shape, texture and appearance. All test products were labelled and randomized by the physicians.
The dietary intervention study lasted 45 days and was performed in a double-blind, randomized, crossover manner with probiotic-free artichokes as a control (Fig. 1). The subjects consumed each artichoke type for 15 days, interspersed with a washout period of 15 days. Each subject was identified by a serial number, and they were randomly allocated into two groups, namely A and B. One group (A, n = 10) consumed artichokes in the order PR-CTR and the other group (B, n = 10) followed the order CTR-PR. During the washout period, volunteers maintained their usual food intake and lifestyle.
For microbiological and biochemical investigations, stool samples were collected before the start of the study (no treatment, T0) and after each 15-day treatment period with probiotic or control products (T1 and T2, Fig. 1). A further sampling was performed after the 15-day washout period (Tw) to assess the absence of the probiotic strain in faecal samples of subjects.
Fresh faecal samples (about 1 g each) were collected in sterile plastic containers and immediately mixed (1 : 10, w/v) with Amies transport medium (Oxoid Ltd., Basingstoke, Hampshire, UK) for microbiological analysis. For analysis of SCFAs and faecal enzyme activities, fresh faecal samples were hand-mixed with a glass rod without medium and 1-g samples were frozen at −20°C. Samples were maintained at 4°C and transferred under anaerobic conditions (Anaerojar; Oxoid) to the Laboratory of Microbiology within 4 h.
Analysis of faecal bacteria
The faecal slurry was serially diluted in reduced peptone water (0·1% w/v peptone with 0·05% w/v cysteine-HCl). Decimal dilutions (100 μl for spread plating or 1 ml for pour plating) were plated in duplicate on specific agar media for the detection and enumeration of micro-organisms. For total aerobic and anaerobic counts (Hartemink and Rombouts 1999), dilutions were spread plated on Reinforced Clostridial Agar (RCA; Difco Laboratories, Detroit, MI) and incubated for 3 days at 37°C. For lactic acid bacteria (LAB) enumeration, dilutions were spread plated onto de Man-Rogosa-Sharpe medium (MRS; Oxoid) and incubated anaerobically at 37°C for 48 h; for presumptive lactobacilli and bifidobacteria, dilutions were spread plated on Rogosa SL agar (RG; Difco) enumerating and marking lactobacilli after 2 days and bifidobacteria after 4 days (Tannock et al. 2000); for total Enterobacteriaceae, dilutions were pour-plated in Violet Red Bile Glucose Agar (VRBGA; Difco) and plates incubated at 37°C for 24 h in aerobiosis; Escherichia coliβ-glucuronidase–positive colonies (typically blue-green coloured) were enumerated by pour-plating dilutions in Tryptone Bile X-Glucuronide Agar plates (TBX; Oxoid) after incubation at 44°C for 48 h; for sulfite-reducing Clostridium spp., including cells and spores, dilutions were pour-plated in Sulfite Polymyxin Sulfadiazine Agar (SPS; Difco) and plates incubated at 37°C in anaerobiosis. Clostridium spores were obtained heating faecal suspensions at 80°C for 20 min.
Bacteria were characterized on the basis of colony appearance, Gram’s stain, catalase reaction and cell morphology. Colony counts were obtained and expressed as a log10 of the colony-forming units per gram of fresh faeces (CFU g−1).
Additionally, in order to evaluate the genetic diversity of lactic population, the 20% of total colonies randomly selected from countable MRS agar plates were isolated and checked for purity. Bacterial DNA was extracted from overnight cultures of selected and analysed by repetitive extragenic palindromic PCR (REP-PCR) as described in the study of De Bellis et al. (2010). The genetic diversity within LAB population was calculated by the Shannon diversity index (H) (Shannon 1948) following the equation: H = Σ−(Ni/N*ln Ni/N), in which N represented the total number of isolates and Ni was the number of isolates showing each REP-PCR profile.
Genetic identification of Lactobacillus paracasei LMGP22043 from faecal samples
To identify strain LMGP22043, the 20% of total colonies randomly selected from countable MRS agar plates were isolated and checked for purity. Bacterial DNA was extracted from overnight cultures of selected colonies using a Clonsaver Card Starter Kit (Whatman, Maidstone, UK) and amplified by PCR primers Cp2for (5′-CCCAATCTGTTAGGTCTGAAGG-3′) and Cp2rev (5′-GGAGAAACTAAGCGAAACCAG-3′), which allowed the strain-specific identification of strain LMGP22043, as previously described (Sisto et al. 2009).
This strain-specific PCR was also performed after an enrichment procedure to assess the presence of L. paracasei LMGP22043 in the stool samples in which it was not detected by plating on MRS agar and in those collected after the washout period (Tw). In particular, 5 ml of MRS broth was added to an aliquot (0·3 g) of each stool sample. After 18 h of incubation at 37°C under anaerobic conditions, 300 μl of faecal suspensions were spread in triplicate onto MRS agar plates and incubated under the same conditions. Bacterial cells grown after 48 h on each plate were suspended in 3 ml of sterile saline (0·85% NaCl), and the suspensions from the three replicates were combined. Genomic DNA was extracted from 1 ml of bacterial suspension using the Wizard® Genomic DNA Purification kit (Promega Corporation, Madison, WI, USA), and L. paracasei LMGP22043 was detected by strain-specific PCR (Sisto et al. 2009).
To evaluate the detection limit of L. paracasei LMGP22043 in faecal samples (number of cells g−1 of faeces), 300 μl of six bacterial serial dilutions containing from 105 to 1 CFU ml−1, as determined by plating, was added to 0·3 g of stool samples not containing strain LMGP22043, then 4·7 ml of MRS broth were added, and the samples were processed as above. A negative (water) control was also used.
Short-chain fatty acids analysis
Frozen faecal samples were prepared as described by Chen and Lifschltz (1989). Briefly, 1-g samples were thawed and homogenized for 2 min in 10 ml of 0·15 mmol l−1 H2SO4 in Milli-Q-purified water with a Stomacher (Seward, London, UK). A 5-ml portion of the homogenate was centrifuged at 9000 g, 2°C for 20 min. The supernatant fluid was filtered through a microconcentrator (Ultracel-3k; Amicon, Danvers, MA, USA) with a molecular mass cut-off of 3000 Da (7000 g, 2°C, 1 h) and then through a 0·22-μm filter unit (Millipore, Bedford, MA, USA).
SCFAs were separated as recently described (Valerio et al. 2010) with slight modifications using a HPLC system (AKTABasic10, P-900 series pump; Amersham Biosciences AB, Uppsala, Sweden) using Rezex ROA-organic acid H+ (8%) (7·80 × 300 mm; Phenomenex, Torrance, CA, USA) and a three-channel UV detector (Amersham Biosciences 900) set at 210 nm. The mobile phase was 0·005 mol l−1 H2SO4 (Fluka, Deisenhofen, Germany) pumped at a flow rate of 0·6 ml min−1 through the column heated to 70°C. SCFA amounts were determined by integrating calibration curves obtained from standards and expressed as μmol ml−1 of faecal suspension. Quantification limits (LOQ) of acetic, propionic, butyric, valeric and isovaleric were 1·825, 0·147, 0·380, 0·463 and 0·368 μmol ml−1, respectively.
Frozen faecal samples (1 g each) were homogenized in phosphate buffer (pH 7, 0·1 mol l−1), sonicated 2 × 1 min at 0°C and subsequently centrifuged at 500 g for 15 min at 4°C. The faecal supernatant was used for the assessment of β-glucosidase and β-glucuronidase activities by spectrophotometric methods (spectrophotometer Labsystem Multiskan MS, version 3.0, type 352) using sterile, disposable multiwell microdilution plates (96 wells; IWAKI Scitech Div. Asashi Techno Glass, Tokyo, Japan).
For the β-glucosidase activity, a slightly modified method from Goosens et al. (2003) was used. Ten microlitres of two- or threefold diluted faecal supernatant or 0·1 mol l−1 phosphate buffer saline pH 7 (blank) was added to 40 μl of the reaction mixture (0·1 mol l−1 phosphate-buffered saline pH 7, 1 mmol l−1p-nitrophenyl β-pyranoside) in the multiwell plate and incubated for 1 h at 37°C; reaction was stopped by adding 250 μl of 0·1 mol l−1 NaOH in each well. Readings were taken at 405 nm against blank.
β-Glucuronidase activity was measured using the method reported by McConnell and Tannock (1993) with slight modifications. Briefly, the reaction mixture – 25 μl of faecal supernatant (diluted 1 : 2 or 1 : 3) or 0·1 mol l−1 phosphate buffer pH 7 (blank), 25 μl of distilled water, 100 μl of 0·1 mol l−1 phosphate buffer pH 6·5, and 50 μl of 20 mol l−1p-nitrophenyl-β-glucuronide – was incubated at 37°C for 10 min afterwards the reaction was stopped by the addition of 100 μl of 1 mol l−1 potassium carbonate. An aliquot of 300 μl of mixture was dispensed in multiwell plates, and absorbance values were registered at 420 nm against blank.
Concentrations of p-nitrophenol released during the reactions were calculated from a standard curve and expressed as nmol of p-nitrophenol released per min per gram of faeces.
All chemicals were purchased from Sigma Aldrich (Milan, Italy).
Data were analysed using statistica 6.0 software (StatSoft software package, Tulsa, OK, USA) and Stata Statistical Software (StataCorp. 2005, Release 9; StataCorp, College Station, TX, USA). All statistical analyses were performed only on subjects with proven faecal colonization. SCFAs and enzymatic activities data were analysed by nonparametric Wilcoxon test.
Bacterial counts, expressed as means log10 CFU g−1 faeces ± standard deviations, and Shannon diversity indexes based on REP-PCR profiles were analysed using the analysis of variance with repeated measures (rm-anova) taking into account the crossover study. LSD Fisher post hoc analysis for multiple comparisons of normally distributed data was used. The comparison between microbial data and genetic diversity from the first period of intervention (period 1: T1 vs T0) for each feeding group (probiotic group A and control group B) was performed by the Student’s t-test for dependent variables. A presumptive quantification of the ability of the probiotic strain to influence the faecal microbial balance was obtained calculating the modulatory index (MIPR/CTR) for each bacterial population after the consumption of probiotic artichokes in both treatment groups (A and B) in comparison with the treatment with control artichokes. This index was calculated for each subject as follows:
MIPR/CTR = [log10 CFU g−1 after probiotic intake (T1 A or T2 B)/log10 CFU g−1 after control artichoke intake (T2 A or T1 B)] − 1.
MIPR/CTR equations were extrapolated by the prebiotic index equation calculated by Palframan et al. (2003). The difference between means was considered significant at P <0·05.
Lactobacillus paracasei LMGP22043 was not detected in faecal samples at the beginning of the study (T0) and after the washout period (Tw) as demonstrated by plating, strain-specific PCR and the enrichment procedure. In particular, the last method showed a detection limit of L. paracasei cells corresponding to 2 log10 CFU g−1 of faecal sample. These procedures allowed to establish that 17 of 20 subjects (85%) were transiently colonized by the probiotic strain at a median value of 6·87 log10 CFU g−1 faeces (range: 6·30–8·52 log10 CFU g−1). All the further statistical analyses were performed on subjects (17) with proven faecal colonization.
From the crossover study, consumption of L. paracasei LMGP22043 significantly (P <0·05) increased the genetic diversity of faecal LAB population after the probiotic artichoke intake, as measured by the Shannon diversity index (HPR = 1·48 ± 0·6, HCTR = 1·08 ± 0·85, HT0 = 1·04 ± 0·65) based on the REP-PCR profiles. In particular, the change in diversity after the probiotic intake was significantly higher with respect to the control treatment and to the start of the study (T0). However, no significant differences for each bacterial population within the study period (T0, probiotic and control artichoke interventions) were observed, except for LAB populations which decreased after control artichoke intake respect to the T0 time (Table 1). Additionally, when data from the two treatment groups A and B (Fig. 1) were considered separately, the modulatory index (MIPR/CTR) calculated for each microbial group confirmed that probiotic intake in the group A (period 1) caused the inhibition of potentially harmful bacteria and stimulated the growth of LAB and of presumptive lactobacilli and bifidobacteria, while probiotic intake in treatment group B in period 2 was associated with positive values for the Clostridium spp. group and negatives for lactobacilli (Fig. 2). These data indicated a carryover effect exerted by the control artichokes after the washout period. In fact, the 15-day washout period was sufficient to remove the probiotic strain from the gut of subjects, while it was not enough for eliminating the residual effect of the carrier matrix which contains indigestible substances available for intestinal bacteria. As a result, probiotic artichokes showed a different effect on faecal bacteria during the two treatment periods (period 1 and period 2).
Table 1. Faecal bacterial populations (log10 CFU g−1 faeces) after probiotic and control artichoke intake in a double-blind randomized, crossover human trial
log10 CFU g−1 faeces
T0 (n = 17)
Probiotic (n =17)
Control (n =17)
*P values were obtained analysing data from each microbial group by anova model of repeated measures, taking into account the crossover study. P < 0·05 were considered significant. All values in one row with a common symbol are significantly different from each other (P < 0·05, LSD Fisher post hoc test).
Lactic acid bacteria
Total Clostridium spp.
Clostridium spp. spores
In order to assess the role of the probiotic strain in the efficacy of the probiotic artichoke product and exclude the carryover effect, only the first period of intervention (period 1, Fig. 1) was examined and data were analysed considering separately the two treatment groups as a probiotic group A (n =8) and a control group B (n =9). As a result, the efficacy of the probiotic L. paracasei LMGP22043 was observed (Table 2 and Fig. 3); the consumption of probiotic artichokes (probiotic group A) caused a general reduction in Enterobacteriaceae at T1 sampling time, a significant decrease (P <0·05) in counts of E. coli and Clostridium spp. with respect to the start of the study (T0) and a general increase in presumptive lactobacilli and bifidobacteria (Fig. 3a,b, Table 2). When subjects enrolled in control group B were considered (Table 2, Fig. 3c), a different microbial behaviour was observed with a trend to the increase in potentially harmful bacteria and in particular of E. coli counts (P <0·05). Furthermore, the consumption of control artichokes allowed a slight increase in presumptive lactobacilli and unvaried counts of presumptive bifidobacteria (Fig. 3d), while the total number of LAB resulted to be lower (Table 2). Interestingly, the Shannon diversity index confirmed a higher (P <0·05) genetic diversity of LAB at T1 sampling time only in the probiotic group (1·46 ± 0·60) with respect to the start of the study (T0, 0·82 ± 0·65), while this index remained almost unvaried in the control group (from T0 1·26 ± 0·61 to T1 1·32 ± 0·82).
Table 2. Microbiological counts (log10 CFU g−1 faeces) in faecal samples after period 1 of intervention, relevant to the probiotic group A and control group B
Log10 CFU g−1 faeces
*P values were obtained analysing data by Student’s t-test by comparing data obtained from the probiotic group A and control group B with those obtained by the relevant T0. P < 0·05 were considered significant.
Probiotic group A (n = 8)
Lactic acid bacteria
Total Clostridium spp.
Clostridium spp. spores
Control group B (n = 9)
Lactic acid bacteria
Total Clostridium spp.
Clostridium spp. spores
Short-chain fatty acids
The consumption of artichokes containing the probiotic strain slightly stimulates the production of SCFAs even if without significant differences (Table 3). Each acid and total SCFA concentrations increased slightly in the probiotic group A in comparison with T0 values. In contrast, a general reduction in the total SCFA concentrations was observed in the control group B with a significant (P <0·05) decrease in valeric acid.
Table 3. Concentrations (μmol ml−1 of faeces) of short-chain fatty acids in faecal samples after period 1 of intervention, relevant to the probiotic group A and control group B
μmol ml−1 of faecal suspension (median values and range)
*P values were obtained analysing data from each acid by Wilcoxon test by comparing data obtained from the probiotic group A and control group B with those obtained by the relevant T0. P < 0·05 were considered significant.
Probiotic group A (n = 8)
Control group B (n = 9)
Faecal enzyme activity
The dietary intervention did not cause changes in faecal enzyme activities as assessed by the spectrophotometric method, although a slight trend to the reduction in values was observed after the consumption of the probiotic food (Table 4).
Table 4. Faecal enzymatic activity (nmol min−1 g−1) after period 1 of intervention, relevant to the probiotic group A and control group B
nmol min−1 g−1 ± SD
*P values were obtained analysing data from each enzyme by Wilcoxon test comparing data obtained from the probiotic group A and control group B with those obtained by the relevant T0.
Probiotic group A (n = 8)
Control group B (n = 9)
The present study was designed as a randomized double-blind crossover human study, and a 15-day washout was applied to remove the probiotic strain and the prebiotic residual effects because periods of 1–2 weeks are generally used (Alander et al. 2001; Tannock et al. 2004; Kolida et al. 2007; Valerio et al. 2010). The ready-to-eat artichoke product not containing the probiotic strain, similar in appearance and texture to the test product, was used as a control. Subjects showed a good compliance with both products (probiotic and control) which were integrated in the usual diet during the whole period of trial and appreciated as lightly dressed artichoke salads.
This study demonstrated that the probiotic strain, transported by the vegetable matrix, transiently colonized the gut of 85% of subjects (median 6·87 log10 CFU g−1 faeces). Data from the crossover study indicated that consumption of the probiotic strain caused an increase in the genetic diversity of LAB based on the REP-PCR profiles when compared with the treatment with control artichokes and with the start of the study. However, the modulatory index (MIPR/CTR) suggested a carryover effect of the control product when it was consumed during the first period. In particular, the MIPR/CTR indicated that only the consumption of probiotic artichokes during the first 15-day period (group A) caused a shift of microbial counts towards lower values of potentially harmful bacteria (Enterobacteriaceae, E. coli, total Clostridium spp.) and higher values of LAB and of presumptive lactobacilli and bifidobacteria. These results demonstrated that the 2-week washout period was sufficient for removing the probiotic strain from the gut of subjects but seemed to be not long enough to eliminate the residual effect of the control food, maybe because of the amount of indigestible compounds contained in the plant matrix. The nature of the matrix used as a carrier to transport the probiotic cells into the GI tract can greatly influence the efficacy of a probiotic food by playing a double role in sustaining the survival of the probiotic cells in food and during the GI passage, as well as in stimulating the growth of colonic microflora (Roberfroid 1993, 2007; Lavermicocca et al. 2010; Ranadheera et al. 2010; Valerio et al. 2010). In the current study, the reduction in LAB count as well as SCFA concentrations observed after ready-to-eat control artichoke intake confirmed the modulatory activity of the probiotic strain and suggests that the prebiotic activity of indigestible molecules can be influenced by food processing (i.e. exposure to high temperature, low pH, etc.) as also reported by Huebner et al. (2008). In particular, inulin, which is contained in several plants, can stimulate the growth of beneficial bacteria even if also potentially harmful micro-organisms may benefit from this substrate (Böhm et al. 2006; Loh et al. 2006). Therefore, the real mechanisms by which prebiotic molecules contained in plant-derived food products become available for colonic bacteria need to be investigated. The slight increase in total SCFAs after the probiotic artichoke intake can be related to the evidence that most of SCFAs, and particularly butyrate, is rapidly adsorbed by the intestinal epithelium or utilized by colonic bacteria (Neish 2009).
In conclusion, the consumption of the probiotic strain L. paracasei LMGP22043, resulting in gut colonization, determined the inhibition of E. coli and Clostridium spp., the increase in the genetic diversity of LAB and a slight change in SCFA concentrations and enzyme activity, whereas the consumption of the probiotic-free artichoke product stimulated E. coli growth and caused more than 1 log decrease in LAB population and a concomitant lowering in total SCFAs.
The proficient association of the strain with a food carrier rich in fibre can represent a new strategy for favouring a daily supply of probiotics and attracting more consumers to vegetable foods fortified with probiotic strains.
This work was supported by the Italian Ministry for University and Research (art. 12 D.M. 593/2000 – D.D. 3300 – 22 December 2005 – tema 2) Project ‘Ortobiotici pugliesi’ (D.M. 28830).