Fernández de Palencia, 2008

Page 1 of 31 European Food Research and Technology
1
2
3
4
5
6
7
8
9
10
PROBIOTIC STRAINS: SURVIVAL UNDER SIMULATED
11
12
13 GASTROINTESTINAL CONDITIONS, IN VITRO ADHESION TO CACO-2
14
15 CELLS AND EFFECT ON CYTOKINE SECRETION
16
17
18
19
20
Fo
21
22 Pilar Fernández de Palencia1,2*, Paloma López2, Angel L. Corbí2, Carmen Peláez1,
23
24
Teresa Requena1
rP
25
26
27
28
29
ee
Addresses
30
31 1
32 Instituto del Frío (C.S.I.C.), Jose Antonio Novais 10, and 2Centro de Investigaciones
rR
33
34 Biológicas (C.S.I.C.), Ramiro de Maeztu 9, 28040 Madrid, Spain
35
36
37
ev
38
39
40
41
iew
42
43
44
*Correspondence to: P. Fernández de Palencia. Mailing address: Centro de
45
46 Investigaciones Biológicas (C.S.I.C.), Ramiro de Maeztu 9, 28040 Madrid, Spain; e-
47
48 mail address: psp@cib.csic.es; telephone +34918373112 ext. 4202; fax number:
49
50
51 +34915360432.
52
53
54
55 Running title: Evaluation of probiotic strains
56
57
58
59
60
http://mc.manuscriptcentral.com/efrt
European Food Research and Technology Page 2 of 31
1
2
3 1 Abstract
4
5
6 2 This study evaluated three probiotic strains (Lactobacillus paracasei subsp. paracasei
7
8 3 LC-01, L. acidophilus LA-5, Bifidobacterium lactis Bb-12) and two yoghurt strains (L.
9
10
4 delbrueckii subsp. bulgaricus LBY-27 and Streptococcus thermophilus STY-31) with
11
12
13 5 regard to their resistance to simulated gastrointestinal stress, and their ability to interact
14
15 6 with human intestinal epithelial cells. The viability of strains was analyzed by
16
17
18
7 measurements of fluorescence-stained cells and their growth by plate colony-counts.
19
20 8 The results reveal that for all tested strains, gastric emptying (above pH 3.0) would
Fo
21
22 9 release a large number of viable cells ranging from 91% for L. paracasei to 53% for S.
23
24
10 thermophilus into the intestinal tract, and that between 12%-23% of them subsequently
rP
25
26
27 11 survive intestinal stress. Among them L. paracasei showed the highest resistance to
28
29
ee
12 gastric stress. All the bacteria adhered to the Caco-2 cell line, with the highest adhesions
30
31
32 13 being observed for L. delbrueckii subsp. bulgaricus (9%) and L. acidophilus (7%).
rR
33
34 14 Binding of all strains to Caco-2 cells did not result in a significant increase in the
35
36
15 production of IL6 and IL8 cytokines, suggesting that these bacteria do not trigger an
37
ev
38
39 16 overt inflammatory response in human intestine epithelial cells.
40
41
iew
17
42
43
44 18 Keywords: Probiotic bacteria; lactic acid bacteria, gastrointestinal stress, adhesion;
45
46 19 immunomodulation.
47
48
49
50
51
52
53
54
55
56
57
58
59
60
1
1
2
3 20 Introduction
4
5
6 21 Probiotics are defined as live micro-organisms which, when administered in
7
8 22 adequate amounts, confer a health benefit to the host [1]. There is clinical evidence that
9
10
23 supports the health-promoting characteristics of Lactobacillus and Bifidobacterium spp,
11
12
13 24 which are often included in fermented milk products [2]. However, the minimum
14
15 25 amounts of probiotics needed to obtain a clinical effect have not been established. As
16
17
18
26 more information on probiotics becomes available, it seems likely that these amounts
19
20 27 will vary as a function of the strain and the health effect desired [3], and of the
Fo
21
22 28 probiotic´s capability to display specific responses at various sites along the intestine
23
24
29 [4]. Probiotic product specifications require strains to be designated individually,
rP
25
26
27 30 appropriately classified as to species, and retain an acceptable viable count at the end of
28
29
ee
31 their shelf life in the designated product formulation [1]. The CODEX standard for
30
31
32 32 fermented milks [5] establishes that the minimum counts of these micro-organisms at
rR
33
34 33 the time of consumption should be 106 cfu g-1 [5]. In addition to their ability to survive
35
36
34 in the product, many criteria have been suggested for the selection of probiotics, among
37
ev
38
39 35 them the tolerance of gastrointestinal conditions (acid and bile) and ability to adhere to
40
41
iew
36 intestinal mucosa. Bacterial viability is reduced along the entire tract, but it is apparent
42
43
44 37 that the acidic environment of the stomach and the presence of bile in the duodenum are
45
46 38 the major factors affecting viability [6]. In vivo studies are too complex to be used for
47
48 39 high throughput screening of bacteria viability. Several studies have been reported by
49
50
51 40 other authors about the viability cell after simulated gastrointestinal tract conditions [7,
52
53 41 8]. Therefore, several in vitro multi-compartmental models, which simulate different
54
55 42 parts of the human gastrointestinal tract, have been developed to study the survival rate
56
57
58 43 of potential probiotic strains. Among them Mainville et al. [9] recently developed a
59
60 44 dynamic in vitro model simulating the events of food ingestion and digestion, allowing
45 the addition of a food matrix before or along with the probiotic strain to be tested.
2
1
2
3 46 Bacterial growth viability is typically assessed by plate counting. However there are a
4
5
6 47 number of disadvantages associated with this approach, such as the relatively long time
7
8 48 needed to form visible colonies, and the possible underestimation of viable micro-
9
10
49 organisms, which do not form colonies because they are sublethally damaged, dormant
11
12
13 50 (inactive but ultimately culturable) or active but non-culturable [10, 11]. To resolve
14
15 51 these problems other technologies such as fluorescent detection of cells have been
16
17
18
52 developed [12, 13]. Thus, the combined use of SYTO9 and propidium iodide
19
20 53 fluorescent dyes allows the differential staining of the nucleic acids of intact cells and of
Fo
21
22 54 those with compromised membranes, though there has been some debate as to the
23
24
55 efficacy of these methods in complex matrices. Nevertheless, they have been
rP
25
26
27 56 successfully used for detection of probiotic bacteria [14; 12] and lactic acid bacteria
28
29
ee
57 (LAB) [15] in a food matrix.

30
31
32 58 Adhesion to intestinal mucosa is also regarded as a prerequisite for probiotic
rR
33
34 59 micro-organisms, allowing possible colonization of the intestinal tract [16]. The
35
36
60 difficulties of studying bacterial adhesion in vivo, have led to the development of in
37
ev
38
39 61 vitro model systems for the preliminary studies of adherent strains [17]. These models
40
41
iew
62 are based on adhesion to tissue culture cell lines such as Caco-2 and HT-29, which
42
43
44 63 differentiate and closely resemble the enterocytes of human small intestine, and to
45
46 64 human intestinal mucus [18]. Adhesion is affected by many factors such resident flora
47
48 65 in the gastrointestinal tract, pH, growth phase of the bacteria, density of the bacterial
49
50
51 66 suspension, intensity of washing out the unbound cells, etc. [19]. Furthermore, the
52
53 67 adhesion capacity may be correlated with transient colonization of intestinal cells,
54
55 68 which could be a factor for their immunomodulation by probiotics [20]. It is well known
56
57
58 69 that one of the potential benefits of probiotic therapy is the suppression of the
59
60 70 inflammatory process [21, 22]. The secretion of the pro-inflammatory cytokines such as
71 IL-6 and IL8 is therefore a hallmark of the inflammatory response in the intestine [23].
3
1
2
3 72 In this study, the viability of various probiotic and LAB strains in acidified milk
4
5
6 73 was assessed after exposing the cells in vitro to gastric, or to gastrointestinal, stress
7
8 74 conditions. Moreover, human intestinal epithelial-like Caco-2 cells were used to
9
10
75 examine the adhesion of the bacteria and their ability to stimulate pro-inflammatory
11
12
13 76 cytokine production in this cell line.
14
15 77
16
17
18 78 Materials and methods
19
20 79 Microorganisms, growth conditions and preparation of milk-cell suspension
Fo
21
22 80 The bacterial strains used were Lactobacillus paracasei subsp. paracasei LC-01,
23
24
81 Lactobacillus acidophilus LA-5, Bifidobacterium lactis Bb-12, Lactobacillus
rP
25
26
27 82 delbrueckii subsp. bulgaricus LBY-27 and Streptococcus thermophilus STY-31. The
28
29
ee
83 strains were isolated from a commercial synbiotic product (Synbiotic Drink; Priégola,
30
31
32 84 Madrid, Spain) and identified by molecular typing described previously [24].
rR
33
34 85 Lactobacilli and bifidobacteria strains were propagated on MRS broth (Pronadisa,
35
36
37
86 Madrid, Spain). The medium was supplemented with 0.05 % L-cysteine hydrochloride
ev
38
39 87 (Merck, Darmstad, Germany) for all bacteria (except for L. paracasei) and
40
41
iew
88 supplemented with 0.1% Tween for L. acidophilus and L. delbrueckii subsp. bulgaricus
42
43
44 89 growth. S. thermophilus was grown in ETSY medium (Pronadisa) containing 0.5%
45
46 90 lactose. All incubations were performed at 37 ºC except for L. delbrueckii subsp.
47
48 91 bulgaricus and S. thermophilus, which were grown at 42 ºC. B. lactis. L. acidophilus
49
50
51 92 and L. delbrueckii subsp. bulgaricus were grown anaerobically in jars (AnaeroGenTM,
52
53 93 Oxoid Unipath Ltd. Basingstoke, Hampshire, UK).
54
55
56
94 For preparation of milk-cell suspension, the strains were grown in fresh medium
57
58 95 until they reached the late exponential phase (approximately 109 cfu mL-1 for L.
59
60 96 paracasei, B. lactis, and S. thermophilus; 108 cfu mL-1 for L. delbrueckii subsp.
97 bulgaricus and 107 cfu mL-1 for L. acidophilus). Cells from 25 mL of culture were
4
1
2
3 98 sedimented by centrifugation, 10,000 × g for 10 min, and resuspended in the same
4
5
6 99 volume of 10% reconstituted skim milk powder (autoclaved 110 ºC, 15 min) acidified
7
8 100 with 1 M HCl (Merck) to pH 4.6. (corresponding to the pH of the commercial synbiotic
9
10
101 drink)
11
12
13 102 For adhesion experiments, bacteria were grown until they reached the late
14
15 103 exponential phase and they were sedimented as described above. Then, they were
16
17
18
104 resuspended in the appropriate volume of Dulbecco´s modified Eagle medium (DMEM,
19
20 105 Invitrogen) to give 1.25×106 cfu mL-1.
Fo
21
22 106
23
24
107 Gastric and gastrointestinal transit tolerance assay
rP
25
26
27 108 The gastrointestinal tract model is depicted schematically in Fig. 1. The
28
29
ee
109 experiments were performed in triplicate; three independent cultures of each bacterium
30
31
32 110 were analyzed as follows. Milk-cell suspensions were prepared as described above and
rR
33
34 111 samples of 2.5 mL were withdrawn prior (sample G1, untreated control) or after
35
36
112 (samples G2 through G7) the indicated treatments to determine cell survival by the tests
37
ev
38
39 113 described below. The gastric and gastrointestinal solutions were prepared fresh daily
40
41
iew
114 according to the protocols described by Marteau et al. and Huang and Adams [6, 25]. To
42
43
44 115 simulate the in vivo dilution of saliva, 5 mL of a sterile electrolyte solution (6.2 g L-1
45
46 116 NaCl, 2.2 g L-1 KCl, 0.22 g L-1 CaCl2, 1.2 g L-1 NaHCO3, all purchased from Merck)
47
48 117 was added to 22.5 mL of cell suspension. An aliquot was withdrawn (control G1) and
49
50
51 118 then lysozyme (Sigma-Aldrich, Chemie Gmbh P.O. Steinheim, Germany) was added to
52
53 119 give a final concentration of 0.01%. To simulate the gastric environment, 3 mL of
54
55 120 electrolyte solution containing 0.3% pepsin (final concentration) (Sigma-Aldrich) at pH
56
57
58 121 5.0 was added to the cell suspension and an aliquot was taken without further incubation
59
60 122 (sample G2). Then, the pH curve in the stomach was reproduced by adding 1 M HCl
123 (Merck) to the cell suspension, at an initial pH of 5.0 which was then decreased to 4.1,
5
1
2
3 124 3.0, 2.1 and 1.8. To mimic normal gastric emptying [6], aliquots of the suspension were
4
5
6 125 collected after successive incubations of 20 min at 37 ºC at each pH (samples G3 to
7
8 126 G7). To simulate the intestinal stress, samples 3, 4 and 5, were adjusted to pH 6.5 with 1
9
10
127 M NaHCO3 (Merck), then mixed with 4 mL of a sterile electrolyte solution (5 g L-1
11
12
13 128 NaCl, 0.6 g L-1 KCl, 0.3 g L-1 CaCl2, all purchased from Merck), containing 0.45% bile
14
15 129 salts and 0.1% pancreatin (final concentrations, both from Sigma-Aldrich) at pH 8.0.
16
17
18
130 After 120 min of incubation at 37 ºC, simulating the conditions of the duodenum,
19
20 131 fractions of suspensions were collected (samples GI3, GI4, GI5). To avoid interference
Fo
21
22 132 of milk proteins in the fluorescence determinations of cell viability, all samples (gastric
23
24
133 or gastrointestinal) were treated as follows. First, pH of the sample was neutralized to
rP
25
26
27 134 6.5 with 1 M NaOH (Merck). Then, 1% C6H5Na3O7 2H2O (Merck) was added to
28
29
ee
135 provoke casein micelle dispersion, as previously described [26]. The bacterial cells were
30
31
32 136 then sedimented by centrifugation (10,000 × g for 10 min), washed twice by
rR
33
34 137 resuspension in 2.5 mL of PBS buffer pH 7.5 (10 mM Na2HPO4, 1mM KH2PO4, 140
35
36
138 mM NaCl, 3mM KCl, all purchased from Merck) and sedimented as described above.
37
ev
38
39 139 Finally, cells were resuspended in 2.5 mL of PBS buffer pH 7.5 and analyzed for cell
40
41
iew
140 survival as detailed below.

42
43
44 141
45
46 142 Cell survival analysis
47
48 143 Untreated and treated suspensions were analyzed for growth on solid media by
49
50
51 144 plating and for viability by use of fluorescent dye staining.
52
53 145 For measurement of colony forming units (cfu), samples were plated on the
54
55 146 appropriate culture media as detailed above supplemented with 1.5% bacteriological
56
57
58 147 agar (Scharlau, Barcelona, Spain) and colonies were counted after incubation for 48 h.
59
60 148 Simultaneously, to test bacterial viability, samples were stained with the LIVE/DEAD®
149 BacLightTM bacterial viability kit (Molecular Probes, Inc.AA Leiden, The Netherlands)
6
1
2
3 150 as described by Alakomi et al. [27]. This kit is based on a combination of two probes,
4
5
6 151 SYTO9 and propidium iodide (PI). SYTO9 is a membrane-permeable nucleic acid stain
7
8 152 (emission of green fluorescence), whereas PI enters only the cells with compromised
9
10
153 membranes (emission of red fluorescence). Since PI has a higher affinity for DNA than
11
12
13 154 SYTO9, it is able to displace it. Thus, viable cells are detected by green emission and
14
15 155 damaged or dead cells by red emission.
16
17
18
156 To assses viability, the staining solution was prepared by diluting the commercial stock
19
20 157 in 0.085% NaCl (Merck) to final concentrations of 0.167 mM SYTO9 and 1 mM PI.
Fo
21
22 158 Then, 1 mL of each bacterial suspension in PBS (approximately 1×109 cfu mL-1 of L.
23
24
159 paracasei and B. lactis or 3.5×108 cfu mL-1 of L. acidophilus, L. delbrueckii subsp.
rP
25
26
27 160 bulgaricus and S. thermophilus) were mixed with 33 µL of the staining solution. The
28
29
ee
30
161 samples were incubated at room temperature in the dark for 15 min. Three aliquots of
31
32 162 200 µL of each mix were pipetted into three separate wells of a 96-well microplate. The
rR
33
34 163 green and red fluorescences of the bacterial suspensions were measured in a LS-50B
35
36
37 164 automated fluorometer (Perkin-Elmer, Boston, MA, USA) by detection of emission of
ev
38
39 165 SYTO9 and PI at 530 nm and 620 nm, respectively, upon excitation at 488 nm and with
40
41
iew
42
166 slits of 5.0.
43
44 167 In order to calibrate LIVE/DEAD® BacLightTM for viability assessments, standards
45
46 168 were prepared by mixing non-viable cells with viable cells at 0, 20, 40, 60, 80 and
47
48
49 169 100%. Heat treatment (70ºC, 30 min) was used to prepare non-viable cells, whereas
50
51 170 non-heat-treated cells from fresh culture were used as viable cells. The results showed
52
53 171 that the Green/Red ratio of all the strains analyzed correlated linearly with the number
54
55
56 172 of viable cells in the suspensions (R2 = 0.97-0.99) (as an example the data obtained for
57
58 173 L. paracasei is depicted in Fig. S1).
59
60
174 To automatically correct for possible pipetting errors, the ratio of green and red
175 fluorescences obtained for control cells G1 was considered as 100% and the change in
7
1
2
3 176 this ratio was used to calculate the viability in the G-stress samples G2-G7. Similarly,
4
5
6 177 for the GI-stress experiments, the Green/Red ratios for samples G3, G4 and G5 were
7
8 178 considered as 100% and changes in these ratios was used to calculate the viability in
9
10
179 samples GI3, GI4 and GI5.
11
12
13 180
14
15 181 Confocal laser scanning microscopy
16
17
18
182 To confirm some results, treated and stained bacteria as describe above, were
19
20 183 analyzed with a confocal laser scanning microscope (CLSM), Leica TCS-SP2-AOBS
Fo
21
22 184 model (Leica Microsystems GmbH, Wetzlar, Germany). Confocal illumination was
23
24
185 provided with a X63 magnification objective and numerical aperture of 1.4-0.60 and by
rP
25
26
27 186 Argon laser (488 nm laser excitation) with a long pass 520-565 nm filter (for green
28
29
ee
187 emission) and long pass 630-685 nm filter (for red emission). Image analysis was
30
31
32 188 performed using FRET and FRAP software.
rR
33
34 189
35
36
190 Caco-2 cell culture and adhesion assay
37
ev
38
39 191 Caco-2 cells, originating from human colonic adenocarcinoma, were obtained
40
41
iew
192 from the human cell bank at the Centro de Investigaciones Biológicas (Madrid, Spain).
42
43
44 193 These cells were used in their terminally differentiated state to mimic small intestine
45
46 194 mature enterocytes. Caco-2 cells were grown in Men-Alpha Medium (Invitrogen,
47
48 195 Barcelona, Spain) supplemented with 10% (v v-1) heat-inactivated fetal bovine serum at
49
50
51 196 37 ºC in an atmosphere containing 5% CO2. Intestine cells were seeded in 96-well tissue
52
TM
53 197 culture plates (Falcon Microtest , Becton Dickinson, Franklin Lakes, NJ, USA) at
54
55 198 1.25×104 cells per well and grown during 15 days to obtain a monolayer of
56
57
58 199 differentiated and polarized cells. The culture medium was changed every 2 days. To
59
60 200 study the adhesion of each strain, cells were overlaid with bacteria resuspended in
201 DMEM medium (0.1 mL/well) and at a multiplicity of infection of 10 bacteria per
8
1
2
3 202 epithelial cell. After a 1 h incubation period at 37 ºC under 5% CO2 atmosphere, cells
4
5
6 203 were washed three times with phosphate-buffered saline pH 7.1 (PBS), resuspended in
7
8 204 0.1 mL of PBS, and Caco-2 cells detached after addition of 40% glycerol and freezing
9
10
205 at –70 ºC. To determine the number of cell-associated bacteria, appropriate dilutions
11
12
13 206 were plated onto agar plates containing media specific for each strain. Each adhesion
14
15 207 assay was conducted in triplicate.
16
17
18 208
19
20 209 Cytokine quantification: Enzyme-Linked Immunoabsorbent Assay (ELISA)
Fo
21
22 210 To analyze cytokine secretion, supernatants from bacteria-treated cells were
23
24
211 collected after 8 and 24 h and frozen at –70 ºC until assayed. The concentration of the
rP
25
26
27 212 pro-inflammatory cytokines IL-6 and IL-8 respectively in the supernatants of 8 and 24 h
28
29
ee
213 (optimal conditions of detection) were determined using commercially available ELISA
30
31
32 214 kits (Immuno tools GmbH, Heidelberg, Germany). Each assay was performed in
rR
33
34 215 triplicate.
35
36
37
216
ev
38
39 217
40
41 Results
iew
218
42
43
44 219 The tolerance to digestive tract stress of three probiotic bacteria and the two
45
46 220 yoghurt strains that constitute the microbiota of a commercial synbiotic product
47
48 221 (Synbiotic Drink) was investigated. Bacteria, included in skim milk acidified at pH 4.6,
49
50
51 222 were incubated in conditions that simulated the major factors influencing the survival of
52
53 223 the ingested micro-organisms during their passage through the gastrointestinal tract
54
55 224 (Fig. 1). We have considered three relevant factors, the effect of lysozyme, the influence
56
57
58 225 of acid pH values with pepsin (gastric stress, G-stress) and the further action of bile salts
59
60 226 and pancreatin (intestinal stress, GI-stress), simulating successive gastric delivery of
227 bacteria to the intestine during digestion [6]. These analyses were performed using
9
1
2
3 228 bacteria suspended in acidified milk (pH 4.6) to evaluate the matrix effect on ingestion
4
5
6 229 of bacteria in fermented milks. The methodology is shown schematically in Fig 1. The
7
8 230 survival of each strain was analyzed by fluorescence measurements and plate counts at
9
10
231 different time intervals. The effect of lysozyme alone had no pronounced effect on
11
12
13 232 viability of any strain (data not shown). Fig. 2 illustrates the effect of the G-stresses at
14
15 233 different pH values on the viability of the studied bacterial strains. Presentation of the
16
17
18
234 data as percentages of the values obtained for untreated cultures revealed that the
19
20 235 pattern of cell survival detected by fluorescence (Green/Red ratio) was similar to that
Fo
21
22 236 observed by plate counting for most of the strains analysed. However, there were
23
24
237 greater discrepancies between survival as measured by plate counting as opposed to
rP
25
26
27 238 fluorescence readings for L. acidophilus, L. delbrueckii subsp. bulgaricus and S
28
29
ee
239 thermophilus, when exposured to pH values below 4.1 (G5 through G7). The samples
30
31
32 240 containing lysozyme and pepsin at pH 5.0 did not drastically affect the viability of any
rR
33
34 241 of the strains (sample G2 versus control sample G1). Under these conditions, no
35
36
242 significant reduction of plate counts was observed for any of the bacteria, and only a
37
ev
38
39 243 30% reduction of the Green/Red ratio was detected in the S. thermophilus suspension
40
41
iew
244 (Fig. 2). Exposure of cell suspensions to decreasing pHs (5.0, 4.1, and 3.0) and further
42
43
44 245 incubation for 20 min at 37° C (samples G3, G4, and G5), caused a slight but
45
46 246 progressive reduction of viability in the yoghurt strains (Fig. 2). After incubation at pH
47
48 247 3.0 (sample G5), the percentage of Green/Red ratio remained higher than 80% for L.
49
50
51 248 paracasei and L. acidophilus, whereas it decreased to 55%, 66%, and 72% for S.
52
53 249 thermophilus, L. delbrueckii subsp. bulgaricus, and B. lactis respectively. These
54
55 250 decreases correlated approximately with reductions in plate counts to 17%, 33 % and
56
57
58 251 67% of the values of control untreated cells (Fig. 2). When the simulated G-stress was
59
60 252 set at pH 2.1 (sample G6), all the strains except L. acidophilus, showed a sharp
253 reduction of the Green/Red ratio (86-76% loss of viability) (Fig. 2). Similarly, the
10
1
2
3 254 capability of all bacteria (except for L. acidophilus), to form colonies was reduced
4
5
6 255 between 4 to 5-log-units after incubation of cell suspensions at pH 1.8, which simulated
7
8 256 the last gastric emptying to the intestine, and the Green/Red ratio was reduced by 82%
9
10
257 to 86%. By contrast, after G-stress at pH 1.8, L. acidophilus still showed 53% viability
11
12
13 258 as estimated by fluorescence and only approximately 1-log-unit reduction in plate
14
15 259 counts. The above results indicate that for most of the analyzed bacteria, gastric
16
17
18
260 emptying at pH values below 3.0 should deliver low doses of viable cells into the
19
20 261 intestine.
Fo
21
22 262 Therefore, to evaluate the transit tolerance of strains in conditions simulating the
23
24
263 duodenal, aliquots taken from G3, G4 and G5 where further incubated with 0.45% bile
rP
25
26
27 264 salts and 0.1% pancreatin. Table 1 and Table S1 show comparative results of resistance
28
29
ee
265 to G and GI-stress. In general, the reduction of plate counts caused by duodenal
30
31
32 266 secretions could be correlated with the degree of severity of the previous gastric
rR
33
34 267 incubation of cell suspensions. The highest sensitivity to the simulated intestinal stress
35
36
268 was found for B. lactis, with a drastic reduction of survival (4.4 log units decrease,
37
ev
38
39 269 Table 1) even after G-stress at pH 5.0. This loss of capability to form colonies was also
40
41
iew
270 associated with reduction of the Green/Red ratio to 12% of those of the delivered cells
42
43
44 271 (Table S1). By contrast, L. acidophilus, with a cell survival of 64% after gastric
45
46 272 treatment at pH 3.0, as measured by plate colony counts, was not further affected by
47
48 273 intestinal stress (Tables 1 and S1). These results correlated with its high survival rate
49
50
51 274 observed at low pH values (Fig. 2). However, this was not in strict agreement with the
52
53 275 Green/Red ratio, which decreased from 88% (G5) to 18% (GI5) under these
54
55 276 gastrointestinal stress conditions (Table S1). Surprisingly, S. thermophilus showed a
56
57
58 277 good rate of survival to intestinal stress even after previous gastric exposure to pH 3.0
59
60 278 with a reduction of 2.4 log-units (Table 1), corresponding to a 3% growth ability and a
279 23% viability of the cells surviving G-stress (Tables 1 and S1).
11
1
2
3 280 As shown in Figure 3, the adhesion of strains ranged from 1% to 9%, with L.
4
5
6 281 delbrueckii subsp. bulgaricus and L. acidophilus (7%) showing the highest levels of
7
8 282 adherence, whereas B. lactis and L paracasei showed the lowest capability (2%). The
9
10
283 ability of the five bacterial strains to interact with the enterocytes prompted us to
11
12
13 284 determine their influence on the secretion of pro-inflammatory cytokines by Caco-2
14
15 285 cells. Caco-2 cells constitutively secreted detectable levels of both IL-6 and IL-8 (Table
16
17
18
286 2), but these were not modulated by any of the strains tested (Table 2). Therefore the
19
20 287 adherence of the bacterial cells does not appear to modify the basal state of activation of
Fo
21
22 288 the epithelial cell line. Nevertheless, it should be borne in mind that this lack of
23
24
289 stimulation of cytokine expression in vitro may not be indicative of what occurs in the
rP
25
26
27 290 in vivo situation.
28
29
ee
291
30
31
32 292
rR
33
34 293 Discussion
35
36
294 Fermented milks are a widely used vehicle for delivering probiotic bacteria in food.
37
ev
38
39 295 Strains of L. acidophilus, L. casei group and B. lactis predominate in commercial
40
41
iew
296 probiotic products [28, 29]. In this work, we have analyzed three strains of the
42
43
44 297 aforementioned species and two yogurt strains for their sensitivity to digestive tract
45
46 298 conditions, and their ability to interact with human intestinal cells.
47
48 299 To assess the resistance of these strains to GI-stress, we have studied their in vitro
49
50
51 300 survival under conditions that simulated the normal physiological conditions of the GI
52
53 301 tract, such as the presence of lysozyme and pepsin, sequential gastric emptyings at
54
55 302 increasingly lower pH values, and the presence of bile salts and pancreatin. In addition,
56
57
58 303 we have analyzed the applicability of a microplate fluorochrome assay as a rapid
59
60 304 assessment of bacterial viability. This test has been previously used for analysis of
305 bacterial viability in probiotic preparations and dairy products [14, 30, 27, 15] as well as
12
1
2
3 306 for testing gastrointestinal tolerance of Bifidobacterium [31, 8]. However, to our
4
5
6 307 knowledge, this is the first report on the general application of this fluorescence system
7
8 308 for the assessment of viability of probiotic strains belonging to different genera after
9
10
309 gastrointestinal stress.
11
12
13 310 Strain survival estimated by fluorescence or by plate counts behaved generally in a
14
15 311 similar manner after G-stress (Fig. 2). However, after GI-stress two different behaviours
16
17
18
312 were observed. In the case of L. acidophilus, a higher cell recovery was detected by
19
20 313 plate counts than by fluorescent measurements whereas, for the other four bacterial
Fo
21
22 314 strains analyzed, the cell survival detected by fluorescent measurements was higher than
23
24
315 that observed by plate count (Tables 1 and S1). A survival rate that is higher when
rP
25
26
27 316 measured by Green/Red fluorescence ratio, than when measured by plate colony
28
29
ee
317 counting is normally indicative that a proportion of the cells are still viable but not
30
31
32 318 readily cultivable (e.g. L. rhamnosus strains subjected to acid and bile salt stress only
rR
33
34 319 yielded countable colonies after 168 h of incubation [32]) An addition explication
35
36
320 would be that the GI-stress conditions provoke the formation of chains of cells or cell
37
ev
38
39 321 clumping; each chain or clump would only give rise to single colony on plate counts but
40
41
iew
322 would still be correctly enumerated by fluorescence. The possibility that high
42
43
44 323 fluorescence measurements could be due to non-specific staining of residual caseins
45
46 324 present in the samples was investigated by confocal microscopy which showed that for
47
48 325 all strains tested only the bacteria were stained by SYTO9 or propidum iodide (Fig. 4).
49
50
51 326 As expected, green (viable) cells were predominant in untreated cultures, whereas red
52
53 327 (non-viable) cells were in the majority in suspensions subjected to gastrointestinal
54
55 328 stress. The situation of L. acidophilus, where the colony counts were substantially
56
57
58 329 higher than the fluorescence measurements, is harder to explain, though it is possible
59
60 330 that GI-stress causes cell surface / cell wall changes in this organism that radically alter
331 the penetrability of the fluorescent stains.
13
1
2
3 332 Recent reports give evidence of the survival of yogurt bacteria in the upper
4
5
6 333 compartments of human digestive tract [33; 34]. Our study revealed the significant
7
8 334 impact of the pH on the survival of the strains analyzed. The results showed that most of
9
10
335 the strains did not tolerate pH values below 3 (Fig. 2); but that above this value
11
12
13 336 significant numbers of viable bacteria reach the intestine from the gastric content.
14
15 337 It has been proposed that damage of bacterial cell envelope by low pH could make
16
17
18
338 the cells more susceptible to bile action on cell membranes [35]. The analysis of GI-
19
20 339 stress performed here showed that indeed bacterial suspensions sublethally damaged by
Fo
21
22 340 acid stress at pH 3, were more sensitive to intestinal secretions than those exposed to pH
23
24
341 5 (Table 1) Also, cell suspensions treated at pH values below 3 were highly susceptible
rP
25
26
27 342 to intestinal secretions, yielding counts close to zero (results not shown) and
28
29
ee
343 fluorescence measurements below the detection limits. This contrasts with the findings
30
31
32 344 of Noriega et al. [36] who suggested that in Bifidobacterium a previous exposure to low
rR
33
34 345 pH in the stomach, might cause a transitory rise of bile resistance, thus increasing its
35
36
346 survival in the duodenum. This kind of adaptation has recently been demonstrated to be
37
ev
38
39 347 due to an increase of the intracellular ATP reserve [37]. Mättö et al. [38] have described
40
41
iew
348 that B. lactis Bb-12 also survived through the human gastrointestinal tract, since the
42
43
44 349 bacterium was detected in the faeces of 79% of subjects consuming probiotic yoghurt.
45
46 350 In this study B. lactis Bb-12 showed a drastic reduction of its ability to form colonies in
47
48 351 samples at pH 5.0 (Table 1). However, its viability (around 10% of the untreated control
49
50
51 352 as estimated by fluorescent measurements) remained constant even in samples treated at
52
53 353 lower pH values, supporting resistance of the strain to bile stress. It has been reported
54
55 354 that in some cases the possession of bile or acid resistance alters the ability of the strains
56
57
58 355 to adhere to human mucus [39; 40]. Our analysis of adherence to Caco-2 cells showed
59
60 356 that B. lactis Bb-12, along with the other strains analyzed, is able to interact with
357 epithelial cells. Their detected adherence (2%-10%) was within the lower range of the
14
1
2
3 358 values observed for mucus adhesion of other Bifidobacterium strains from faecal human
4
5
6 359 origin [40].
7
8 360 The potential immuno-stimulating properties of bifidobacteria and LAB have
9
10
361 attracted attention in recent years [41, 42]. IL-6 and Il-8 are multifunctional cytokines
11
12
13 362 that play a major role in the acute-phase response to inflammatory stimulus [23]. Morita
14
15 363 et al. [21] investigated 28 Bifidobacterium and lactobacilli strains for their ability to
16
17
18
364 stimulate cytokine IL-6 and IL-8 production. Some of the Bifidobacterium strains
19
20 365 induced secretion of IL-8, but this was not associated with the strains’ binding affinity
Fo
21
22 366 to Caco-2 cells. In this study, we have detected that the five strains studied have
23
24
367 different adherence to Caco-2 cells (Fig. 3). However, none of the strains, including B.
rP
25
26
27 368 lactis Bb-12, exhibited the undesired property of inducing either IL-6 or IL-8 (Table 2).
28
29
ee
369 The combination of S. thermophilus and L. delbrueckii subsp. bulgaricus is widely

30
31
32 370 used as a starter for the production of yoghurt. L. delbrueckii subsp. bulgaricus LBY-27
rR
33
34 371 and .S. thermophilus STY-31 strains showed relatively high levels of adhesion to Caco-
35
36
372 2 cells, whereas L. paracasei susbp. paracasei LC-01 displayed a lower adhesion. (Fig.
37
ev
38
39 373 3). Values approximately ten fold higher have been observed for L. paracasei susbp.
40
41
iew
374 paracasei strains ACA-DC221,333 and 3335 [43].

42
43
44 375 Of all the strains analyzed, the highest tolerance to GI-stress was observed in L.
45
46 376 acidophilus. A high tolerance to acid pH has been frequently observed in different L.
47
48 377 acidophilus strains [44], but this does not seem to induce significant bile salt, heat, or
49
50
51 378 ethanol tolerance in the strains [45]. Recently, L. acidophilus bile tolerance has been
52
53 379 linked to the expression of a bile-inducible operon encoding a two component
54
55 380 regulatory system [46], and the results obtained in this work show that L. acidophilus
56
57
58 381 LA-05 also possesses a high resistance to intestinal stress (Table 1). A recent study
59
60 382 demonstrated that cell surface proteins of L. acidophilus NCFM can contribute to this
383 organism’s ability to attach to intestinal cells in vitro [47], and our results show that L.
15
1
2
3 384 acidophilus LA-5 has a significant level of adhesion capability. We have also observed
4
5
6 385 by microscopy (results not shown) that L. acidophilus LA-5 can form aggregates, which
7
8 386 may be related to its adhesion to Caco-2 cells since, for the L. acidophilus M92 strain, a
9
10
387 relationship has been shown between autoaggregation and adhesiveness, both mediated
11
12
13 388 by proteinaceous components of the cell surface [48].
14
15 389
16
17
18 390
19
20 391 Conclusions. The overall results indicate that most of gastric emptying would
Fo
21
22 392 release a large number of viable probiotics and yogurt strain cells into the intestinal
23
24
393 tract, and demonstrate differences between bacterial species with respect to their
rP
25
26
27 394 sensitivity to gastric and intestinal secretions. It also indicates that the in vitro model
28
29
ee
395 and the fluorescent detection used to evaluate the gastrointestinal stress responses of
30
31
32 396 probiotic and yogurt bacteria is a reliable tool for high throughput screening of bacterial
rR
33
34 397 survival. Moreover, the high levels of adhesion to epithelial cells observed for L.
35
36
398 delbrueckii subsp. bulgaricus and L. acidophilus highlight the interest to elucidate the
37
ev
38
39 399 molecular mechanisms of these interactions. Finally, the absence of induction of the
40
41
iew
400 cytokines IL-6 and Il-8 in human intestinal epithelial cells, indicates that all the bacterial
42
43
44 401 strains tested in this study can attach to the epithelial cells without triggering local
45
46 402 inflammatory response.
47
48 403
49
50
51 404
52
53 405 Acknowledgements
54
55 406 We thank to Dr. Stephen Elson for the critical reading of the manuscript. This work
56
57
58 407 was supported by the following Spanish entities: Ministerio de Educación y Ciencia
59
60 408 (MEC) (grants: AGL2004-07285-C02-01 and AGL2006-11932-C05-01), by MEC-
409 Consejo Superior de Investigaciones Científicas (grant 2006 7 0I026) and by
16
1
2
3 410 Comunidad de Madrid (grant S-0505/AGR-0153). The research leading to these results
4
5
6 411 has also received funding from the European Community’s Seventh Framework
7
8 412 Programme (FP7/2007-2011) under grant agreement no 211441.
9
10
413
11
12
13 414
14
15 415 References
16
17
18
416 1. FAO/WHO (2002) ftp://ftp.fao.org/es/esn/food/wgreport2.pdf
19
20 417 2. Párvez S, Malik KA, Ah Kang S, Kim H-Y (2006) J Appl Microbiol 100:1171–1185
Fo
21
22 418 3. Roy D (2005) Lait 85:39–56
23
24
419 4. Marco ML, Bongers RS, de Vos WM, Kleerebezem M (2007) Appl Environ
rP
25
26
27 420 Microbiol 73:124–132
28
29
ee
421 5. CODEX Alimentarius Commission (2003) Codex Stan 243–2003

30
31
32 422 http://www.codexalimentarius.net/download/standards/ 400/CXS_243e.pdf
rR
33
34 423 6. Marteau P, Minekus M, Havenaar R, Huis In´t Veld JHJ (1997) J Dairy Sc 80:1031–
35
36
424 1037
37
ev
38
39 425 7. Saarela M, Virkajärvi I, Alakomi H-L, Mattila-Sandholm T, Vaari A, Suomalainen T,
40
41
iew
426 Mätto J (2005) J Appl Microbiol 99:1330-1339

42
43
44 427 8. Masco L, Crockaert C, Van Hoorde K, Swings J, Huys G (2007) J Dairy Sci 90:
45
46 428 3572-3578
47
48 429 9. Mainville I, Arcand Y, Farnworth, ER (2005). Int J Food Microbiol 99:287–296
49
50
51 430 10. Kell DB, Kaprelyants AS, Weichart DH, Harwood CR, Barer MR (1998) Antonie
52
53 431 van Leeuwenhoek 73:169–187
54
55 432 11. Breeuwer P, Abee, T (2000) Int J Food Microbiol 55:193–200
56
57
58 433 12. Lahtinen SJ, Gueimonde M, Ouwehand AC, Reinikainen, JP, Salminen SJ (2006)
59
60 434 Food Microbiol 23:571–577
435 13. Brehm-Stecher BF, Johnson EA (2004) Microbiol Mol Biol Rev 68:538–559
17
1
2
3 436 14. Auty MAE, Gardiner GE, McBrearty SJ, O´Sullivan EO, Mulvihill DM, Collins JK,
4
5
6 437 Fitzgerald GF, Stanton C, Ross RP (2001) Appl Environ Microbiol 67:420–425
7
8 438 15. Moreno Y, Collado MC, Ferrús MA, Cobo JM, Hernández E, Hernández M (2006)
9
10
439 Int J Food Sci Tech 41, 275-280
11
12
13 440 16. Alander M, Satokari R, Korpela R, Saxelin, M, Vilpponen-Salmela T, Mattila-
14
15 441 Sandholm T, Von Wright A (1999) Appl Environ Microbiol 65:351-354
16
17
18
442 17. Vesterlund S, Paltta J, Karp M, Ouwehand AC( 2005) J Microbiol Meth 60:225-233
19
20 443 18. Ouwehand AC, Salminen S (2003) Microb Ecol Health Dis 15:175-184
Fo
21
22 444 19. Riedel CU, Foata F, Goldstein DR, Blum S, Elkmanns J (2006) Int J Food
23
24
445 Microbiol 110:62-68
rP
25
26
27 446 20. Salminen S, Bouley C, Boutron-Ruault M-C, Cummings JH, Franck A, Gibson GR,
28
29
ee
447 Isolauri E, Moreau M-C, Roberfroid M, Rowland I (1998) Br J Nutr 80:S147-S171

30
31
32 448 21. Morita H, He F, Fuse T, Ouwehand AC, Hashimoto H, Hosoda M, Mizumachi K,
rR
33
34 449 Kurisaki J-I (2002) Microbiol Immunol 46:293-297
35
36
450 22. Isolauri E, Kirjavainen PV, Salminen S (2002) Gut 50:III54-III59
37
ev
38
39 451 23. Isolauri E (1999) Curr Opinion Gastroenterol 15:534-537
40
41
iew
452 24. Tabasco R, Paarup T, Janer C, Peláez C, Requena T (2007) Int Dairy J 17:1107–
42
43
44 453 1114
45
46 454 25. HuangY, Adams, MC (2004) Int J Food Microbiol 91:253–260
47
48 455 26. Fernández de Palencia P, Peláez C, Requena T, Martín-Hernández C (1995)
49
50
51 456 Zeitschrift für Lebensmittel-Untersuchung und-Forschung 201:87–90
52
53 457 27. Alakomi H-L, Mättö J, Virkajärvi I, Saarela M (2005) J Microbiol Meth 62:25–35
54
55 458 28. Fasoli S, Marzotto M, Rizzotti L, Rossi F, Dellaglio F, Torriani S (2003). Int J Food
56
57
58 459 Microbiol 82:59–70
59
60 460 29. Masco L, Huys G, De Brandt E, Temmerman R, Swings J (2005) Int J Food
461 Microbiol 102:221–230
18
1
2
3 462 30. Bunthof CJ, van Schalkwijk S, Meijer W, Abee T, Hugenholtz J (2001) Appl
4
5
6 463 Environ Microbiol 67:4264–4271
7
8 464 31. Collado MC, Moreno Y, Hernández E, Cobo JM, Hernández M (2005) Food Sci
9
10
465 Tech Int 11:307-314
11
12
13 466 32. Succi M, Tremonte P, Reale A, Sorrentino E, Grazia L, Pacifico S, Coppola R
14
15 467 (2005) FEMS Microbiol Lett 244:129–137
16
17
18
468 33. Mater DDG, Bretigny L, Firmesse O, Flores MJ, Mogenet A, Bresson JL, Corthier
19
20 469 G (2005) FEMS Microbiol Lett 250:185–187
Fo
21
22 470 34. Elli M, Callegari ML, Ferrari S, Veis E, Cattivelli D, Soldi S, Morelli L, Feuillerat
23
24
471 NG Antoine J-M (2006) Appl Environ Microbiol 72:5113–5117
rP
25
26
27 472 35. Van de Guchte M, Serror P, Chervaux C, Smokvina T, Ehrlich SD, Maguin E
28
29
ee
473 (2002) Antonie van Leeuwenhoek 82:187–216

30
31
32 474 36. Noriega L, Gueimonde M, Sánchez B, Margolles A, De los Reyes-Gavilán CG
rR
33
34 475 (2004) Int J Food Microbiol 94:79–86
35
36
476 37. Sánchez B, De los Reyes-Gavilán CG, Margolles A (2006) Appl Environ Microbiol
37
ev
38
39 477 8:1825–1833
40
41
iew
478 38. Mättö K, Fondén R, Tolvanen T, von Wright A, Vilpponen-Salmela T, Satokari R,

42
43
44 479 Saarela M (2006) Int Dairy J 16:1174–1180
45
46 480 39. Gueimonde M, Noriega L, Margolles A, De los Reyes-Gavilán CG, Salminen S
47
48 481 (2005) Int J Food Microbiol 101:341-346
49
50
51 482 40. Collado MC, Gueimonde M, Sanz Y, Salminen S (2006) J Food Prot 69:1675-79
52
53 483 41. Vinderola CG, Medici M, Perdigón G (2004) J Appl Microbiol 96:230-243
54
55 484 42. Von der Weid T, Bulliard C, Shiffrin EJ (2001) Clin Diagn Lab Immunol 8: 695-
56
57
58 485 701
59
60 486 43. Maragkoudakis PA, Zoumpopoulou G, Miaris C, Kalantzpoulos G, Pot B,
487 Tsakalidou E (2006) Int Dairy J 16:189-199
19
1
2
3 488 44. Chou LS, Weimer B (1999) J Dairy Sci 82:23–31
4
5
6 489 45. Azcárate-Peril MA, Altermann E, Hoover-Fitzula RL, Cano RJ, Klaenhammer TR
7
8 490 (2004) Appl Environ Microbiol 70:5315–5322
9
10
491 46. Pfeiler EA, Azcárate-Peril MA, Klaenhammer TR (2007) J Bacteriol 189:4624-4634
11
12
13 492 47. Buck BL, Altermann E, Svingerud T, Klaenhammer TR (2005) Appl Environ
14
15 493 Microbiol 71:8344-8351
16
17
18
494 48. Kos B, Suskovic J, Vukovic S, Simpraga M, Frece J, Matosic S (2003) J Appl
19
20 495 Microbiol 94:981-987
Fo
21
22
23
24
rP
25
26
27
28
29
ee
30
31
32
rR
33
34
35
36
37
ev
38
39
40
41
iew
42
43
44
45
46
47
48
49
50
51
52
53
54
55
56
57
58
59
60
20
1
2
3 496 Legends to the figures
4
5
6 497 Figure 1. Schematic representation of the in vitro digestive tract model. Aliquots of
7
8 498 reconstituted milk at pH 4.6 were individually inoculated with cellular pellets of each of
9
10
499 the indicated strains after growth in their specific media and conditions as described in
11
12
13 500 Materials and Methods. Subsequently, the cells were subjected to G- and/or GI-stress at
14
15 501 37º C. To simulate G-stress, cells were exposed to pH 5.0 at initiation (sample G2) and
16
17
18
502 to sequential incubations at 37 ºC for 20 min at each pH 5.0, 4.1, 3.0, 2.1 and 1.8
19
20 503 (samples G3, G4, G5, G6 and G7, respectively). To simulate GI-stress, cells were
Fo
21
22 504 subjected to G-stress as described above (conditions 2, 3, 4 and 5) and then, samples
23
24
505 were adjusted to pH 6.5 and treated with 0.45% porcine bile salts and 0.1% pancreatin
rP
25
26
27 506 (samples GI3, GI4 and GI5).
28
29
ee
507
30
31
32 508 Figure 2. Analysis of cell survival after gastric stress. The indicated bacterial strains
rR
33
34 509 were subjected to various G-stresses (G2, G3, G4, G5, G6 or G7) as described in Fig. 1
35
36
510 and in Materials and Methods. Cell viability was analysed by fluorescence (black bars)
37
ev
38
39 511 and by plate counting (white bars).The values are the mean of three independent
40
41
iew
512 experiments, and are expressed as a percentage of the values for untreated control
42
43
44 513 samples. 100% control values for Green/Red fluorescence ratio for L. paracasei subsp.
45
46 514 paracasei, L. acidophilus, B. lactis, L. delbrueckii subsp. bulgaricus and S.
47
48 515 thermophilus were respectively 8.65, 9.87, 10.80, 9.88 and 9.04. 100% control values
49
50
51 516 for plate counting for L. paracasei subsp. paracasei, L. acidophilus, B. lactis, L.
52
53 517 delbrueckii subsp. bulgaricus and S. thermophilus were respectively 1.1x109, 6.0x107,
54
55 518 1.4x109, 3.0x108 and 4.0x108 cfu mL-1
56
57
58 519
59
60 520 Figure 3. Adhesion of bacterial strains to Caco-2 cells after infection with 10 bacteria
521 per epithelial cell, followed by 1 h incubation at 37 ºC in an atmosphere containing 5%
21
1
2
3 522 CO2. Adhesion levels are expressed as the % of cfus adhered after three washes with
4
5
6 523 PBS at pH 7.1. Each adhesion assay was conducted in triplicate. Vertical bars represent
7
8 524 the standard deviations of three independent assays.
9
10
525
11
12
13 526 Figure 4. Detection of stained cells by confocal microscopy. Cell-milk suspensions of
14
15 527 the indicated bacteria untreated (left panel) or subjected to gastric treatment at pH 5.0
16
17
18
528 and further intestinal stress (right panel) were stained and analyzed under confocal
19
20 529 microscope as described in Materials and Methods for visualization of viable (green)
Fo
21
22 530 and non-viable (red). Bar = 20 µm
23
24
rP
25
26
27
28
29
ee
30
31
32
rR
33
34
35
36
37
ev
38
39
40
41
iew
42
43
44
45
46
47
48
49
50
51
52
53
54
55
56
57
58
59
60
22
1
2
3
4
5
6
7
8 Table 1. Comparative bacterial survival to gastric (G) and gastrointestinal (GI) stresses as detected by fluorescent staining and plate counting.
9
10
pH 5.0 pH 4.1 pH 3.0
11
12
13
Bacteria
Cell counts
(log cfu mL-1) Fo Fluorescence
(Green/Red)
Cell counts
(log cfu mL-1)
Fluorescence
(Green/Red)
Cell counts
(log cfu mL-1)
Fluorescence
(Green/Red)
14
15
16
G3-
stress*
GI3-stress* G3-stress
rP
GI3-stress G4-stress GI4-stress G-4stress GI4-stress G5-stress GI5-stress G5-stress GI5-stress
ee
9.06 5.77 6.69 1.97 9.07 5.62 7.85 1.77 9.02 2.37 8.30 1.07
17 L. paracasei
(0.03) (0.67) (0.51) (0.16) (0.02) (0.39) (1.32) (0.39) (0.02) (0.35) (0.90) (0.57)
18
19 7.58 7.56 8.72 1.66 7.54 7.37 8.12 1.54 7.31 7.33 8.74 1.61
20
21
22
L. acidophilus
B. lactis
(0.01)
9.14
(0.01)
4.75
(1.73)
9.15
(0.47)
1.10
(0.13)
9.12 rR (0.33)
4.52
(1.15)
8.92
(0.11)
1.09
(0.29)
9.02
(0.21)
3.83
(0.18)
7.80
(0.10)
0.88
ev
(0.11) (0.14) (0.73) (0.17) (0.09) (0.16) (2.21) (0.03) (0.05) (0.08) (1.84) (0.10)
23
24 8.29 5.44 7.88 1.80 8.29 5.79 7.45 1.66 8.11 2.93 6.53 1.55
L. delbrueckii
iew
25 (0.39) (0.42) (1.21) (0.16) (0.24) (0.30) (0.32) (0.12) (0.20) (0.25) (3.06) (0.17)
26 8.52 7.43 5.43 1.08 8.07 6.41 4.83 1.10 8.78 6.38 5.01 1.14
27 S. thermophilus
(0.01) (0.09) (0.18) (0.17) (0.01) (0.16) (0.62) (0.20) (0.12) (0.51) (1.4) (0.20)
28
29 *Bacteria were analyzed after incubation for 20 min at pHs 5.0, 4.1 and 3.0 (G-stress) and after a further incubation with 0.45% bile salts and 0.1% pancreatin (GI-stress; see
30
31 materials and methods and Fig. 1). The results are the mean of three independent experiments (SD in parenthesis). The plate counts obtained for untreated cultures of L. paracasei
32 subps. paracasei, L. acidophilus, B. lactis, L. delbrueckii subsp. bulgaricus and S. thermophilus were respectively 9.03, 7.78, 9.14, 8.47 and 8.60 log cfu mL-1 . The values of Green
33
34 /Red fluorescence ratio obtained for untreated cultures for L. paracasei subps. paracasei, L. acidophilus, B. lactis, L. delbrueckii subsp. bulgaricus and S. thermophilus were
35 respectively 8.65, 9.87, 10.80, 9.88 and 9.04.
36
37
38
39
40
41
42
43
44 23
45 http://mc.manuscriptcentral.com/efrt
46
47
1
2
3
4
5 Table 2. IL-6 and IL-8 secretion by Caco-2 cells after exposure to bacterial strains.
6
7 Bacteria IL-8 (pg mL-1) Il-6 (pg mL-1)
8 8h 24h
9 L. paracasei 34.91 ± 1.06 0.12 ± 0.07
10
11 L. acidophilus 38.13 ± 3.57 0.03 ± 0.02
12 B. lactis 35.94 ± 12.51 0.04 ± 0.01
13
L. delbrueckii 25.21 ± 9.25 0.06 ± 0.02
14
15 S. thermophilus 25.23 ± 2.87 0.05 ± 0.01
16
Caco-2 cells 41.11 ± 21.04 0.09 ± 0.01
17
18
19
20 Cytokine production was assayed by ELISA titration on supernatants from differentiated Caco-2
Fo
21
22 cells after incubation with the indicated bacterial strains for 8 and 24 h at 37ºC in an atmosphere
23
24 containing 5% CO2. Values indicate the mean ± standard deviation of three independent
rP
25
26 experiments measured in triplicate.
27
28
29
ee
30
31
32
rR
33
34
35
36
37
ev
38
39
40
41
iew
42
43
44
45
46
47
48
49
50
51
52
53
54
55
56
57
58
59
60
24
1
2
3
4
5
6
7
8
9
10
11
12
13
14 In vitro gastrointestinal tract model
15
16 L. paracasei L. acidophilus B. lactis L. delbrueckii S. thermophilus
17
18
milk
Fo
19
20 pH 4.6
21 samples
22
G1 1
r
23
+ lysozyme and pepsin, pH 5.0
24 G2 2
Pe
25
26 20 min, 37º C
+ bile salts and pancreatin, pH 8.0
27 G3 3
28 pH
20 min 37º C
er
29 4.1 + bile salts and pancreatin, pH 8.0

30 Gastric stress G4 4
31 pH 3.0
32 20 min, 37º C 120 min
+ bile salts and pancreatin, pH 8.0
Re
33 5 37º C
34 G5 pH 2.1
35 20 min, 37º C 120 min
36 6 37º C
G6 GI3
pH 1.8
vi
37 120 min
38 20 min, 37º C 37º C
39
G7 7 GI4
ew
40
41
42
43 GI5
44
45 Gastrointestinal stress
46
47
48
49
50
51
52
53
54
55
56
57
58
59
60 Figure 1
L. paracasei
100
1
Cell survival (%)

2 80
3 60
4
5 40
6
7 20
8 0
9 G1 G2 G3 G4 G5 G6 G7
10 Samples
11
L. acidophilus
12
13 100
14
Cell survival (%)
80
15
16 60
17
18 40
Fo
19
20
20
21 0
22 G1 G2 G3 G4 G5 G6 G7
r
23 Samples
24 B. lactis
Pe
25
26 100
27
Cell survival (%)
80
28
er
29 60
30
31 40
32
20
Re
33
34 0
35 G1 G2 G3 G4 G5 G6 G7
36 Samples
vi
37
L. bulgaricus
38
39 100
ew
40
Cell survival (%)
41 80
42
60
43
44 40
45
46 20
47 0
48 G1 G2 G3 G4 G5 G6 G7
49 Samples
50
S. thermophilus
51
52 100
53
Cell survival (%)
54 80
55
60
56
57 40
58
59 20
60 0
G1 G2
G6 G3 G4 G5 G7
Samples
Figure 2
1
2
3
4
5
6
7
8
9
10
11
12
13
14
15
16
17
18
Fo
19 12
20
21
22 10
r
23
24
Pe
25 8
26
adhesion (%)
27
28 6
er
29
30
31 4
32
Re
33
34 2
35
36
vi
37 0
38 L. paracasei L. acidophilus B. lactis L. delbrueckii S. thermophilus
39
ew
40
41
42
43
44
45
46
47
48
49
50
51
52
53
54
55
56
57
58
59
60
Figure 3
1
2
3
4
5
6 L. paracasei
7
8
9
10
11
12
13
14
15 L. acidophilus
16
17
18
Fo
19
20
21
22
r
23
24
Pe
25 B. lactis
26
27
28
er
29
30
31
32
Re
33
34
35 L. delbrueckii
36
vi
37
38
39
ew
40
41
42
43
44
S. thermophilus
45
46
47
48
49
50
51
52
53
54
55
56
57 Figure 4
58
59
60
1
2
3
4
5 Table S1. Comparative detection of bacterial resistance to gastric (G-stress) and gastrointestinal (GI-stress) by fluorescent staining and plate counting.
6
7 pH 5.0 pH 4.1 pH 3.0
8
9 Cell counts Fluorescence Cell counts Fluorescence Cell counts Fluorescence
10
Bacteria (% cfu mL-1) (% Green/Red) (% cfu mL-1) (%Green/Red) (% cfu mL-1) (% Green/Red)
11
12
13 Fo
G3-stress* GI3-stress* G3-stress GI3-stress G4-stress GI4-stress G4-stress GI4-stress G5-stress GI5-stress G5-stress GI5-stress
14
15
16
L. paracasei 100.33 0.13 77.34
rP29.44 108.04 0.06 90.75 22.54 97.24 <0.01 95.95 12.89
17
18
19
L. acidophilus
97.42 96.50 84.34 19.03
ee88.51 91.85 82.26 18.96 64.49 114 88.55 18.42
20
21
22
B. lactis
100.00
97.87
<0.01
0.31
84.72
79.75
12.02
22.84
84.96
47.65
rR <0.01
0.32
82.59
75.4
12.21
22.28
66.66
33.32
<0.01
<0.01
72.22
66.09
11.28
23.73
23
24
L. delbrueckii
99.06 6.99 60.06 19.88 31.84 2.67 ev 53.4 22.77 16.90 2.66 55.42 22.75
iew
25 S. thermophilus
26
27
*Bacteria were analyzed after incubation for 20 min at pHs 5.0, 4.1 and 3.0 (G-stress) and after a further incubation with 0.45% bile salts and 0.1% pancreatin (GI-stress; see materials and
28
methods). The values of cfu mL-1 obtained for untreated cultures of L. paracasei subps. paracasei, L. acidophilus, B. lactis, L. delbrueckii subsp. bulgaricus and S. thermophilus were
29
respectively 1.1x109, 6.0x107, 1.4x109, 3.0x108 and 4.0x108. The values of Green /Red fluorescence ratio obtained for untreated cultures for L. paracasei subps. paracasei, L. acidophilus,
30
B. lactis, L. delbrueckii subsp. bulgaricus and S. thermophilus were respectively 8.65, 9.87, 10.80, 9.88 and 9.04. In this table the results are represented as % of survival. For estimation of
31
G-stress, percentages were calculated using as 100% initial values before treatment. For estimation of GI-stress, percentages were calculated using as 100% the corresponding values
32
obtained with cells subjected to G-stress.
33
34
35
36
37
38
39
40
41
42
43
44
45 http://mc.manuscriptcentral.com/efrt
46
47
1
2
3
4
5
6
7
8
9
10
11
12
13
14
15
16
17
18
Fo
19
20
21
R2 = 0.99
22 10
r
23
24 8
Pe
25
Ratio Gree/Red
26 6
27
28 4
er
29
2
30
31 0
32
0 20 40 60 80 100
Re
33
34 proportion of added live cells (%)
35
36
vi
37
38
Fig S1. Standard curve for L. paracasei. Relationship between
39
proportion of live bacteria (prepared by mixing live and heat-
ew
40
41 killed bacterial suspensions) and Green/Red fluorescence ratio
42 (Ratio Green/Red).
43
44
45
46
47
48
49
50
51
52
53
54
55
56
57
58
59
60

Fernández de Palencia, 2008

Uploaded by

Document Information

Copyright

Available Formats

Share this document

Share or Embed Document

Sharing Options

Did you find this document useful?

Is this content inappropriate?

Copyright:

Available Formats

Fernández de Palencia, 2008

Uploaded by

Copyright:

Available Formats

Page 1 of 31 European Food Research and Technology

57 (LAB) [15] in a food matrix.

140 survival as detailed below.

331 the penetrability of the fluorescent stains.

369 The combination of S. thermophilus and L. delbrueckii subsp. bulgaricus is widely

374 paracasei strains ACA-DC221,333 and 3335 [43].

409 Consejo Superior de Investigaciones Científicas (grant 2006 7 0I026) and by

421 5. CODEX Alimentarius Commission (2003) Codex Stan 243–2003

426 Mätto J (2005) J Appl Microbiol 99:1330-1339

447 Isolauri E, Moreau M-C, Roberfroid M, Rowland I (1998) Br J Nutr 80:S147-S171

461 Microbiol 102:221–230

473 (2002) Antonie van Leeuwenhoek 82:187–216

478 38. Mättö K, Fondén R, Tolvanen T, von Wright A, Vilpponen-Salmela T, Satokari R,

487 Tsakalidou E (2006) Int Dairy J 16:189-199

521 per epithelial cell, followed by 1 h incubation at 37 ºC in an atmosphere containing 5%

29 4.1 + bile salts and pancreatin, pH 8.0

Cell survival (%)

You might also like