Abstract
Adhesion of the new Lactobacillus isolates, L. casei D12, L. casei Q85, L. casei Z11 and L. plantarum Q47, to the porcine intestinal cell line IPEC-J2 was investigated and compared to the recovery of the same bacterial strains from colon biopsies and faeces obtained from human intervention studies. Probiotic bacteria L. rhamnosus 19070, L. reuteri 12246 and L. casei F19 were used as reference strains. The new isolates exhibited low to moderate adhesion to IPEC-J2 cells in the range of 7–26%. A large variation in the recovery of strains was observed between the persons, suggesting host specificity of intestinal colonization. High correlation was shown between recovery from the different sections of the colon of the same subject, indicating consistency of bacterial colonization of the epithelium. The recovery of L. casei Z11 and L. casei Q85 was highest and comparable to the reference strains of L. rhamnosus 19070 and L. casei F19, indicating their potential to colonize the human intestine. Analysis of linear regression demonstrated poor correlation between in vitro and in vivo results, emphasizing the importance of critical evaluation of in vitro adhesion data for prediction of bacterial colonization of the gut.
Introduction
In recent years the beneficial effects of probiotic bacteria on human health have been demonstrated in various human intervention studies Citation[1]. As the functional properties of probiotic bacteria are known to be strain-specific, selection and assessment of potential probiotic isolates is important for development of new efficient probiotic preparations Citation[2]. It is believed that to be effective, the probiotic bacteria must posses a number of functional characteristics, including the ability to survive transit through the gastrointestinal (GI) tract and adhere to the epithelium. Assessment of bacterial adhesion is conventionally performed by using in vitro models, based on tissue-cultured cells and intestinal mucus preparations. However, these models only partially account for the variety of factors affecting probiotic colonization of the gut, and to date the relationship between in vitro and in vivo adhesion has not been clarified.
Examination of adhesive properties of selected probiotic and non-probiotic strains of lactobacilli has shown that bacterial binding to the colon adenocarcinoma cell line Caco-2 correlated with their recovery from faeces Citation[3]. In the present study we used porcine epithelial cell line IPEC-J2 as a model for assessment of bacterial adhesion because it is non-tumorigenic and exhibits enterocytic features such as microvilli, tight junctions and glycocalyx-bound mucin Citation[4]. The purpose was to investigate the ability of potential probiotic strains, Lactobacillus casei D12 (D12), L. casei Q85 (Q85), L. casei Z11 (Z11) and Lactobacillus plantarum Q47 (Q47), isolated from human biopsies and faeces, to survive in the human GI tract and adhere to the epithelium. Lactobacilli with documented probiotic properties, including Lactobacillus rhamnosus 19070, Lactobacillus reuteri 12246 and L. casei F19, were used as reference strains Citation[3], Citation[5–9]. In this study we applied a comparative approach, combining in vitro adhesion assay to IPEC-J2 cells and the analysis of colon biopsies and faecal samples obtained in the human intervention studies.
Materials and methods
Probiotic preparations
Two probiotic preparations of Lactobacillus strains were used for the human intervention studies. Preparation A included strains with proven probiotic effects: L. rhamnosus 19070 (University Hospital of Copenhagen, Denmark), L. reuteri 12246 (DSM, Deutsche Sammlung von mikroorganismen und Zellkulturen (DSM, Braunschweig, Germany) and L.casei F19 (Arla Foods, Denmark). Preparation B consisted of L. plantarum Q47, L. casei Q85 and L. casei Z11, isolated from adult human intestinal biopsies (University of Copenhagen, Denmark), and L. casei D12, isolated from the faeces of a 1-year-old Danish child (University of Copenhagen, Denmark). Bacterial cells were freeze-dried (Kemikalia, Lund, Sweden) and mixed together to produce the final amount of 1010 cfu of each strain per preparation. The preparations were aseptically packed in sealed bags and stored at −20°C until use.
Adhesion assay to IPEC-J2 cells
The piglet jejunal epithelial cell line IPEC-J2 was kindly provided by Professor Anthony Blikslager (North Carolina State University, USA). Cells were cultivated in Dulbecco's modified Eagle medium (DMEM) as described previously Citation[5]. Monolayers of IPEC-J2 cells were seeded at a concentration of 5×105 cells ml−1, dispensed into 12-well tissue culture plates (Nunc, Roskilde, Denmark) and incubated until confluence. Overnight cultures of bacteria were diluted with DMEM to a concentration of approximately 108 cells ml−1, added to monolayers of IPEC-J2 cells in the amount of 1 ml and incubated for 1 h at 37°C. Bacterial suspensions were discarded and the dishes were washed three times with DMEM to remove unbound bacteria. The IPEC-J2 cells were lysed by the addition of 0.1% v/v Triton-X100 (Merck), and the number of viable adherent bacteria was determined by plating serial dilutions onto MRS agar. The cfu were enumerated after incubation for 24 h at 30°C. Adhesion was expressed as the percentage of bacterial counts adhered to IPEC-J2 cells compared with the total number of bacteria added. The assays were performed twice with triplicate determinations.
Human intervention studies and sample collection
The intervention study included eight adults who underwent colonoscopy due to the previous removal of benign colonic polyps or a family history of polyps. The study was approved by the Ethical Committee of the Municipality of Copenhagen (protocol no. KF-11-094/03). The subjects were given one of the two probiotic preparations described above, A (n=4) or B (n=4), twice a day for 12 days. Before consumption the preparations were resuspended in drinking water. Faecal samples and biopsies from ascending, transverse and descending colon were taken on the last day of the intervention. Biopsies were washed twice in 3 ml of isotonic saline to remove faecal debris and free bacterial cells in the epithelium. The faecal samples and biopsies were processed within 24 h.
Isolation and identification of bacteria
The colon biopsy samples were added to 1.5 ml of saline peptone solution (SPS) and vigorously shaken in a FastPrep® FP120 instrument (Bio101 Savant Instruments, Inc., Holbrook, NY, USA) at the setting of 4.5 ms−1 for 40 s. Faecal samples were homogenized in SPS at a ratio of 1 g faeces (wet weight) per 9 ml. Tests for antibiotic resistance performed before the intervention studies showed that the ingested Lactobacillus strains were able to grow on vancomycin (data not shown). Consequently, suspensions from biopsies and stool samples were spread in serial dilution on Man Rogosa Sharpe (MRS) agar (Merck, Darmstadt, Germany) with added vancomycin (50 µg/ml). The plates were incubated anaerobically (Gas Pak system; Microbiology Systems, Cockeysville, MD, USA) at 37°C for 24 h. The single colonies were cultured in MRS broth under the same conditions as described above. Chromosomal DNA was extracted from bacterial cells using a GenElute Bacterial Genomic DNA kit (Sigma-Aldrich, Germany) according to the manufacturer's instructions. The isolates were differentiated by Rep-PCR with the use of primers REP1R-Dt (5′-IIINCGNCGNCATCNGGC-3′) and REP2R-Dt (5′-NCGNCTTATCNGGGCCTAC-3′). The Rep-PCR reaction mixture in a volume of 25 µl contained 2.5 µl 10× PCR buffer, 0.2 µM of each primer, 0.2 mM of each deoxynucleotide triphosphate, 4.0 mM magnesium chloride, 2.5 U Taq polymerase (Invitrogen A/S, Denmark) and 3 µl of template DNA (5–25 ng). The PCR cycling program included initial denaturation at 95°C for 4 min, followed by 30 cycles of denaturation at 90°C for 30 s, annealing at 45°C for 1 min and extension at 68°C for 8 min; and final extension at 68°C for 16 min. The PCR products were analysed by agarose gel electrophoresis (2% w/v agarose) at 120 V for 5 h and stained with ethidium bromide.
Statistical analysis
The correlation coefficients were computed by Pearson correlation calculations using GraphPad InStat version 3 for Windows (GraphPad Software, San Diego, CA, USA).
Results and discussion
Adhesion to IPEC-J2 cells
Lactobacillus strains, used in the intervention studies, showed large variation in their attachment to IPEC-J2 cells (). The strongest adhesion of 60% and 67% was determined for the probiotic strains L. rhamnosus 19070 and L. reuteri 12246, respectively. The same strains were also found to be highly adhesive to the human colon adenocarcinoma cell line Caco-2 in previous studies Citation[3]. Commercial probiotic L.casei F19 exhibited the weakest binding to IPEC-J2 cells: 6% on average. Low adhesion of this strain is in accordance with the results obtained with the use of immobilized human intestinal mucus Citation[10], although considerably lower than reported for the macrophages 3D4/21 derived cell line Citation[6]. Similarity between adhesion of L.casei F19 to IPEC-J2 cells in this study and to the immobilized mucus Citation[10] indicates that mucins produced by IPEC-J2 cells are probably the most important factor affecting bacterial attachment. The new Lactobacillus isolates demonstrated low to moderate binding, between 7 and 26%. Adhesion of L. plantarum Q47 of about 13% was slightly lower than previously found for the same cell line Citation[5], most probably due to the differences in the assays.
Table I. Adhesion of Lactobacillus strains used in this study to IPEC-J2 cells.
Recovery of lactobacilli from human colon biopsies and faeces
Ingested strains of lactobacilli were re-isolated from the biopsies of three of the four subjects who consumed preparation A (A1, A2 and A3). There was no growth of colonies from biopsies of the fourth person (A4) (). Reference strains, L. rhamnosus 19070 and L.casei F19, were repeatedly re-isolated from the biopsies, indicating their colonization of the colon. Strain 19070 was recovered at high frequencies of 40–96% from all the biopsies of two subjects (A1 and A3) and from the ascending and transverse colon of one subject (A2), in accordance with in vitro results on IPEC-J2 cells. Strain F19 was recovered from the biopsies of two subjects (A1 and A3) at variable frequencies between 4 and 50%. These results are supported by the other in vivo studies, showing high survival rates of L. rhamnosus 19070 and L.casei F19 during their passage through the GI tract Citation[3], Citation[11]. Although L. reuteri 12246 was found to be highly adhesive to IPEC-J2 cells, it was re-isolated from the biopsies of only one subject (A1). In contrast, this strain was predominantly isolated from faecal samples (subject A1), suggesting that it is able to survive transit through the GI tract, apparently without intimate contact with the mucosal surface.
Table II. Recovery of ingested bacterial preparations A and B from the colon biopsies and faecal samples collected in the human intervention studies.
Positive identification of ingested bacteria in biopsies and faecal samples was obtained for three of four persons who were administered preparation B (, subjects B1, B2 and B3). Strains L. casei Z11 and L. casei Q85 were re-isolated from the biopsies of three and two subjects, respectively. The frequencies of re-isolation were in the range of 4–90%, comparable to those of L. rhamnosus 19070 and L.casei F19. The same bacterial strains as in biopsies were also identified in faecal samples. The frequencies of re-isolation of ingested bacteria from faecal samples were generally lower than from biopsies, probably due to the high counts of vancomycin-resistant background flora. Recovery of L. casei Z11 and L. casei Q85 from both biopsies and faeces indicated the potential of these strains to colonize the intestinal epithelium, which is considered a prerequisite of probiotic activity. Strains L. casei D12 and L. plantarum Q47 were not detected among the analysed colonies, which possibly signified their poor survival and binding to the intestinal mucosa. Additionally, a low number of colonies obtained from a part of biopsies in this study might have reduced the possibility of re-isolation of consumed bacterial strains. The finding of two subjects who were negative for detection of test strains could either indicate the lack of survival through the GI tract due to individual variation or compliance problems in respect to ingestion of the probiotic preparations.
Generally, bacterial recovery was consistent between the colon sections of the same subject, with a Pearson coefficient of linear correlation of >0.9, suggesting that frequency of colonization was similar along the colon epithelium. At the same time, poor correlation in recovery data was shown between biopsies and faeces (Pearson coefficient < 0.4), which makes it doubtful that colonization abilities of the strains can be predicted from the analysis of their faecal content. However, taking into account the low number of persons which faecal samples were analysed, these results might be considered as preliminary and further work should be undertaken to substantiate the relationship between bacterial colonization of the gut and their recovery from faeces.
Correlation between in vitro and in vivo assays
Analysis of linear regression showed no correlation between bacterial adhesion to IPEC-J2 cells and their recovery from the biopsies (Pearson's coefficient < 0.4). These results are not unexpected considering significant differences between the IPEC-J2 assay used in this study and the actual physiological situation in the human gut. Although correlation between in vitro and in vivo estimates of bacterial adhesion has been demonstrated in a number of studies Citation[3], Citation[12], we believe that in vitro cell models are mostly applicable for preliminary estimation of adhesion abilities, as they do not account for all the factors involved in probiotic adhesion to the human intestine.
In conclusion, in this study we carried out in vitro and in vivo assays to evaluate adhesive abilities of the new Lactobacillus isolates compared to commercial probiotics. Among the new isolated strains, L. casei Z11 and L. casei Q85 were both adherent to IPEC-J2 cells and exhibited high recovery rates in vivo, indicating their potential to colonize the human intestine. Hence, these strains might be considered as possible candidates of probiotic strains worth further investigation. Furthermore, the absence of correlation between bacterial adhesion assessed by IPEC-J2 assay and their in vivo recovery in this study suggests that adhesion data obtained on cell models should be critically interpreted when used for prediction of bacterial colonization in the gut.
Acknowledgements
This work was financially supported by the Danish Dairy Board. Declaration of interest: The authors report no conflicts of interest. The authors alone are responsible for the content and writing of the paper.
References
- Montrose DC, Floch MH. Probiotics used in human studies. J Clin Gastroenterol 2005; 39: 469–84
- Morelli L. In vitro assessment of probiotic bacteria: from survival to functionality. Int Dairy J 2007; 17: 1278–83
- Jacobsen CN, Nielsen VR, Hayford AE, Møller PL, Michaelsen KF, Pærregaard A, et al. Screening of probiotic activities of forty-seven strains of Lactobacillus spp. by in vitro techniques and evaluation of the colonization ability of five selected strains in humans. Appl Environ Microbiol 1999; 65: 4949–56
- Berschneider M. Development of a normal cultured small intestinal epithelial cell line which transports Na and Cl. Gastroenterology 1989; 96(Suppl Pt 2)A41
- Larsen N, Nissen P, Willats WGT. The effect of calcium ions on adhesion and competitive exclusion of Lactobacillus ssp. and E. coli O138. Int J Food Microbiol 2007; 114: 113–19
- Ivec M, Botic T, Koren S, Jakobsen M, Weingartl H, Cencic A. Interactions of macrophages with probiotic bacteria lead to increased antiviral response against vesicular stomatitis virus. Antiviral Res 2007; 75: 266–74
- Rosenfeldt V, Michaelsen KF, Jakobsen M, Larsen CN, Møller PL, Tvede M, et al. Effect of probiotic Lactobacillus strains on acute diarrhea in a cohort of non-hospitalized children attending day care centers. Pediatr Inf Dis J 2002; 21: 417–19
- Rosenfeldt V, Benfeldt E, Nielsen SD, Michaelsen KF, Jeppesen DL, Valerius NH, et al. Effect of probiotic Lactobacillus strains in children with atopic dermatitis. J Allergy Clin Immunol 2003; 111: 389–95
- Møller PL, Pærregaard A, Gad M, Kristensen NN, Claesson MH. Colitic scid mice fed Lactobacillus spp. show an ameliorated gut histopathology and an altered cytokine profile by local T cells. Inflamm Bowel Dis 2005; 11: 814–19
- Juntunen M, Kirjavainen PV, Ouwehand AC, Salminen SJ, Isolauri E. Adherence of probiotic bacteria to human intestinal mucus in healthy infants and during rotavirus infection. Clin Diagn Lab Immunol 2001; 8: 293–6
- Heilig HGHJ, Zoetendal EG, Vaughan EE, Marteau P, Akkermans ADL, de Vos WM. Molecular diversity of Lactobacillus spp. and other lactic acid bacteria in the human intestine as determined by specific amplification of 16S ribosomal DNA. Appl Environ Microbiol 2002; 68: 114–23
- Muñoz-Provencio D, Lopis M, Antolín M, de Torres I, Guarner F, Pérez-Martínez G, et al. Adhesion properties of Lactobacillus casei strains to resected intestinal fragments and components of the extracellular matrix. Arch Microbiol 2009; 191: 153–61