L. Morelli
Istituto di Microbiologia, U. C. S. C., Via Emilia Parmense 84, 29100, Piacenza, Italy
The problem of choosing criteria for the in vitro selection of lactobacilli to be used as health-promoting, probiotic ingredients, in food and pharmaceutical preparations, was apparent even in the original works of Metchnikoff, who first proposed the therapeutic use of these bacteria. In the last decade, a consensus has been reached by scientists on some criteria, with particular attention being paid to the ecological origin of the bacteria, their tolerance to the hostile conditions of the stomach and the small intestine, and their ability to adhere to intestinal surfaces. Even though these criteria have been used to select probiotic lactobacilli, some doubts still remain about the real value of these criteria. A critical rethinking of selection criteria seems necessary in order to improve the process of developing better probiotics.
Time for Some New Thinking
The possibility that the ingestion of some selected bacteria may beneficially influence the gastrointestinal tract of humans was proposed by Elie Metchnikoff, the Russian-born Nobel Prize winner, working at the Pasteur Institute at the beginning of the 20th century, who noticed: " the different susceptibilities of people to the harmful action of microbes and their products. Some can swallow without any evil result a quantity of microbes which in the case of other individuals would produce a fatal attack of cholera. Everything depends upon the resistance offered to the microbes by the invaded organism. " (Metchnikoff, 1907, p. 164). He also stated that "The dependence of the intestinal microbes on the food makes it possible to adopt measures to modify the flora in our bodies and to replace the harmful microbes by useful microbes" (Metchnikoff, 1907, p. 162).
This sentence describes in a clear way the "probiotic concept"; the use of health-promoting bacteria able to exert a positive impact on the intestinal microflora. The word 'probiotics' was coined in the 1960s to name substances produced by microorganisms which promoted the growth of other microorganisms (Lilley and Stillwell, 1965). More recently, the meaning of this word has been refined several times and today a widely accepted definition of probiotics is: 'live microorganisms, which when consumed in adequate amounts, confer a health effect on the host' (Guarner and Schaafsma, 1998). If the complexity of the intestinal ecosystem, inhabited by more than 400 bacterial species and divided into several ecological niches, is taken in consideration, it becomes clear that the selection of the bacteria to be used as a food additive or a biotherapeutic agent (Elmer et al., 1996) is not a simple task (for a review on the complexity of the intestinal flora and recent advances in this field see Tannock, 1999). It is also unlikely that each single strain belonging to one species possesses all of the characteristics that will make it a suitable probiotic. Thus strain-specific criteria have been designed, with special attention to in vitro assays, which have been used to perform a preliminary selection. This is followed by in vivo studies in which healthy volunteers are dosed with strain(s) that performed well in vitro.
The in vitro selection studies are therefore relevant to the entire selection process that follows and obviously need to be based on a sound scientific foundation. In 1991, a project focused on the establishment of selection criteria for bacterial strains to be incorporated into dairy foods was funded by the European Union, and it was, as far as I am aware, the first European attempt to obtain a consensus on procedures to be used for such a selection (Morelli, 1994). In the meantime, several authors have also published their own list of criteria to be used for a preliminary screening of potential probiotic lactobacilli (Marteau and Rambaud, 1993; Huius int'Veld and Shortt, 1996; Salminen et al., 1996a; Tannock, 1997; Brassart et al., 1998; Charteris et al., 1998a; Klaenhammer and Kullen, 1999; Dunne et al., 1999) From these publications, it appears that a kind of general agreement among scientist has been reached, at least in general terms, on the properties that a strain must have in order to be further tested for human probiotic use:
· it must be of human origin
· it must survive during gastric transit
· it has to tolerate bile salts
· it has to adhere to gut epithelial tissue.
The first criterion is based on ecological reasons, and takes into consideration the original habitat of the bacteria to be selected for probiotic use. Strains belonging to bacterial species which are generally present into the intestinal flora of the animal species (i.e. humans) which is to be targeted have been generally selected by authors, who assume that these bacteria have a better chance of out-competing resident bacteria and of establishing at a numerically significant level in their new host. The following two criteria are focused on the assessment of the potential of one putative probiotic strain to overcome barriers which are present in the upper part of the gut, namely the gastric environment and the action of bile salts in the upper part of the intestine. The final criterion has its rationale in the need for bacteria to counter the flow of digesta, by sticking to intestinal surfaces. As it is difficult to obtain samples of human intestinal epithelia, this criterion has been translated, for in vitro testing, into assessment of adhesion to human derived, cultured cell lines.
These criteria have been used to select strains that have been successfully tested in vivo and are nowadays the active ingredients of probiotic products that have been proven to perform well for the well-being of consumers. However, I feel that problems caused by the selection procedure of probiotic lactobacilli, that were present from the very beginning of the history of probiotics (see the following section), are not yet fully resolved and more research effort is needed. This paper is intended to describe the need to critically rethink the criteria for selecting probiotic strains.
In the Beginning There Was Great Intuition, but Also Confusion
The problem of strain selection was apparent even in the work of Metchnikoff. In chapter five (Part III) of his frequently cited, but probably rarely read, book, Metchnikoff reviewed the pros and cons of various fermented dairy products. Then he drew attention to a bacterium previously isolated from Bulgarian yoghurts by a Swiss scientist who later studied at the Pasteur Institute. This strain was the most active bacillus in causing souring of milk and, even before Metchnikoff's studies, was already used in experiments with human volunteers. This bacterium was named the 'Bulgarian Bacillus'.
It is probably a mistake to identify this bacterium with the species named, in modern times, L. bulgaricus or L. delbrueckii subsp. bulgaricus and used for yogurt production. The original description by Metchnikoff deals with a bacterium able to produce 25 g/l of lactic acid when grown in milk. This high level of acidity is not typical of the current 'bulgaricus' strains and more closely resembles the values recorded for L. helveticus strains (Kandler and Weiss, 1986).
In addition, in international culture collections (ATCC cat. N. 521 = CIP 76. 19 =JCM 1003 = NCFB 87 = NCIB 2889= DSM 11445 ), there is only one strain identified as "one of the original Metchnikoff strains". It was deposited as Lactobacillus jugurti, a species nowadays recognised as a biotype of L. helveticus (Dellaglio et al., 1973; Kandler and Weiss, 1986). This strain however, "also gives a disagreeable taste of tallow" and Metchnikoff suggested associating it with another lactic acid bacterium, the so called 'paralactic bacillus' (whose real taxonomic position is unclear), to obtain a more pleasant flavour. Metchnikoff supported the assumption of Dr. Cohendy, another scientist at the Pasteur Institute, who stated that the Bulgarian Bacillus was "able to take its place in the intestinal flora of man". It is now clear, however, that species of lactobacilli inhabiting the human intestine are different from those used to produce fermented milk (Klein et al., 1998). These historical observations show that difficulties in selecting probiotic bacteria were present from the origin of the probiotic concept.
These historically based problems could also, at least partially, explain the inconsistent results often reported by scientific studies and the scepticism about probiotics showed by many members of the scientific community (Tannock, 1990; O'Sullivan et al., 1992).
Selecting the Right Probiotic Bacterium Remains a Difficult Task
Human Origin of Strains
The misidentification of the Bulgarian Bacillus as an intestinal species probably provides a good explanation for using true intestinal isolates in products. It was already attempted by Metchnikoff who suggested that a starch-digesting microorganism could be used instead of dairy bacteria. Unfortunately, this bacterium was isolated from a dog and this origin was regarded unfavourably by the public (Bibel, 1988). Studies carried out at Yale University in the 1930's showed that the Bulgarian Bacillus was killed when it passed through the stomach (Rettger, 1935). A suitable substitute was considered to be Lactobacillus acidophilus which was believed to be an inhabitant of the human gut and was found to tolerate gastric conditions.
Strains described as "acidophilus" have been isolated not only from the intestinal tract of humans, but also from other animals, including rodents and birds. In vitro assays (Morishita et al., 1971; Suegara et al., 1975; Barrow et al., 1980; Tannock et al., 1982; Mayra-Makinen et al., 1983), suggested that adhesion to epithelial tissues could be 'host-specific'; lactobacilli isolated from mammals adhered only to cells obtained from mammals. In vivo experiments however, did not produce clear-cut results. Dosing germfree mice with 19 Lactobacillus isolates demonstrated that, at least in these special animals, all of the strains were able to maintain substantial gastrointestinal populations for at least three weeks. Only Lactobacillus isolates from mice and rats, however, were able to form a continuous layer on the non-secretory epithelium of the forestomach (Lin and Savage, 1984). However, it must be noted that among these biofilm-forming strains, there was also one L. reuteri strain, isolated in my laboratory from calf faeces.
Further evidence of host-specificity was provided in the 1980s by taxonomic studies. The use of a genetic approach to taxonomy revealed that the "L. acidophilus" strains belonged to at least six different species, as determined by DNA-DNA hybridisation (Johnson, 1980; Lauer et al., 1980). These species are now named L. acidophilus, Lactobacillus crispatus, Lactobacillus gallinarum, Lactobacillus amylovorus, Lactobacillus johnsonii, Lactobacillus gasseri (Lauer and Kandler, 1980; Cato et al., 1983; Fujisawa et al., 1992). As a consequence of the taxonomic reorganisation of the acidophilus group of species, attempts were made to correlate the new taxonomic position of the strains with their original habitat (i.e. the intestinal tract of humans or other animals) (Sarra et al., 1980; Sarra et al., 1985; Axelsson and Lingren, 1987; Pryde et al., 1999; our unpublished results). The first results were obtained for strains isolated from animals (Table 1) and, even if further evidence is needed, they suggested that a host-related distribution of these six species did occur and could be used to support (at least for animals other than humans) the concept of host-specific colonisation properties. These data suggested that L. amylovorus and L. crispatus are the species-specific homofermentative lactobacilli in pigs and poultry respectively, while L. johnsonii seems to be widespread in the intestinal tract of animals.
Lactobacilli isolated from human faecal samples and identified by molecular genetical methods (Morelli et al., 1998; Dunne et al., 1999; Song et al., 1999 and 2000; Tannock et al., 2000) suggested that the most frequently encountered lactobacilli belonged to species that are not commonly isolated from other animals. The Lactobacillus casei group of species (formed by L. casei, Lactobacillus paracasei, Lactobacillus rhamnosus, Lactobacillus zeae), and the obligatory homofermentative species L. gasseri and, to a lesser extent, by L. crispatus and L. johnsonii were represented. Lactobacillus reuteri seemed to be the dominant heterofermentative species present in the human gut (Table 2). Comparing Tables 1 and 2, it appears that Lactobacillus species that are frequently encountered in the human gut are different from those isolated from other animals. If confirmed, this observation could provide scientific support to the selection criterion: "use a strain isolated from humans as a probiotic for human beings".
Further reflection, however, suggests more prudence and some doubts. If we consider L. johnsonii strains, it appears that they have been found in humans, poultry and calves. How can we discriminate the human biotype (if one exists) of this species from the others? As far as I know there are not any tests able to provide an answer. L. gasseri is present, not only in the human gut, but also in the urogenital tract (Giorgi et al., 1987; Boris et al., 1997). Is it possible to distinguish the strains of these bacteria according to their original site of isolation? Is it worthwhile undertaking this scientific effort, or has the 'human origin' to be taken in a broad sense? Again, in this case, there is not an answer. If we move our attention to heterofermentative lactobacilli, L. reuteri strains have been isolated from humans, poultry, calves and pigs (Sarra et al., 1980 and 1985; Stahl et al., 1994; Pryde et al., 1999). A recent paper (Jacobsen et al., 1999) has shown that one L. reuteri strain isolated from a pig was one the best performing strains during a human colonisation trial, together with two lactobacilli isolated from human sources but belonging to the L. rhamnosus species. Unfortunately, there was not a comparison with a human isolate of L. reuteri. The question arises if it is really necessary to use 'human' L. reuteri isolates as probiotic additives. The L. casei group is widely present in dairy foods and in the human gut. From the taxonomic point of view, strains isolated from dairy products are indistinguishable from those of human origin (Ferrero et al., 1996; Fitzsimons, 1999). It could be of interest, though, to investigate if it is possible to cluster these isolates according to their sources.
There have not been any other studies comparing probiotic effects of strains of lactobacilli belonging to the same species but isolated from different ecological habitats. Due to the significant presence of lactobacilli forming the non-starter flora of several dairy products (e.g. fresh cheese, mozzarella) it could be of relevance to compare the in vivo behaviour of a Lactobacillus strain isolated from cheese and another strain of the same species obtained from human faeces. This comparison could also lead to a better understanding of the phenotypes that allow a strain to survive and remain in a complex environment such as the gut.
Genetic tools have also allowed us to find out that there are humans harbouring a relatively simple Lactobacillus population, in which a very few strains were numerically predominant isolates for several days or weeks. This observation, originally made with infants by Reniero et al., (1991), was recently confirmed by Tannock and co-workers (Kimura et al., 1997; Tannock et al., 2000). It could open the way to a new and interesting method of selection: to select predominant strains, naturally occurring in human beings. The rationale behind this new criterion is, as we lack a complete understanding of the mechanisms which allow a strain to persist in the gut, to exploit a strain which has been shown to remain in this environment without continuous administration of the bacteria. What is clear, is that more attention must be paid to the correct taxonomy of the strains. It is surprising that even today (Chou and Weimer, 1999) strains that have been classified in the 1980s as L. johnsonii and L. helveticus (Dellaglio et al., 1973; Johnson et al., 1980; Lauer et al., 1980) are still presented in scientific papers as L. acidophilus, causing confusion and possibly mistakes.
In conclusion, it is my opinion that the criterion of human origin still holds true, but more for 'historical reasons' (the need to avoid strains similar to those used by Metchnikoff) and 'common sense' (as we are lacking knowledge about the real mechanisms which make a probiotic bacterium able to persist in the gut, it seems reasonable to use a strain naturally present in the selected environment) than for scientifically-based reasons which are still not evident.
Survival During Gastric Transit
To reach the intestine, strains must first pass through the stomach, which secretes hydrochloric acid and enzyme. More than two litres of gastric juice is secreted each day, with a pH as low as 1.5 providing a barrier to the entrance into the gut of bacteria. A clear quantitative measurement of the destructive action of this environment was provided in 1987 by Conway et al. They showed, using gastric juice obtained from human volunteers, that strains belonging to species used to produce yoghurt were extremely sensitive to killing by gastric juice while enteric species of lactobacilli were more resistant, with significant strain to strain variations. Interestingly, the best performing between the two L. acidophilus strains used in that study (strain ADH) has been reclassified as L. gasseri, the species of homofermentative lactobacilli which is thought to be typical of the human gut (Raya and Klaenhammer, 1992). Is it possible that the observed differences could be explained in terms of species and host-specificity? Another study conducted using human gastric juice (Goldin et al., 1992) showed that L. rhamnosus GG was unable to survive when challenged at pH 1 but there was no loss of viability at pH 3 or higher. In vitro results (Hood and Zottola, 1988; Charteris et al., 1998b) showed that enteric lactobacilli could tolerate exposure to pH 2 for several minutes, while higher pH slightly affected their viable counts and pH 1 was destructive for all of the lactobacilli that were tested. During the FLAIR project, the Irish group co-ordinated by K. Collins, confirmed that human gastric juice at very low pH (about 1-1.2) was a potent killing agent for lactobacilli, as three out six strains tested were killed after 30 min of exposure, while the other three were able to survive, but with a decrease in viability of 4-5 logs (Collins and Thorton report of the UCC activities, included in Morelli, 1994). The puzzling point is that L. paracasei F19, one of the strains with was found to be 'sensitive' when challenged either with human (K. Collins and G. Thorton, included into: FLAIR final report, Morelli, 1994) or simulated gastric juice (Charteris et al., 1998b) was recently shown to be an excellent coloniser of the human gut during in vivo trials (Fonden et al., 2000). The same remark could be made concerning data published recently by a Danish group (Jacobsen, 1999). Strains of lactobacilli which have a documented ability to survive and reproduce in the human gut scored poorly when challenged in vitro for 4 hours at pH 2.5. Furthermore, when data obtained by in vitro experiments are compared, it appears the lactobacilli of the "casei" group of species are the most sensitive. This observation does not correlate with the high rate of isolation from human faeces of this group.
A further issue to be considered is the food matrix or the other ingredients which are used to incorporate or protect cells of lactobacilli that are to be ingested by consumers. It was specifically demonstrated for probiotic bacteria that food intake could protect bacteria during gastric passage (Conway et al., 1987, Charteris et al., 1998b). This is a factor that should be taken into consideration during in vitro evaluation. Testing tolerance to gastric juice using pellets of centrifuged cells or freeze dried preparations, and not the final product in which these cells have to be incorporated, could lead to misleading results. While this criterion remains valid when used to eliminate extremely sensitive bacteria (such as S. thermophilus and L. delbrueckii subsp. bulgaricus) from the number of potential probiotic species, it is doubtful, in my opinion, whether it can predict that a strain is able to tolerate gastric transit in real life. A critical evaluation of results obtained by in vitro assays of tolerance of lactobacilli does not indicate, in my opinion, that scores obtained using these tests could be really predictive of in vivo behaviour of these strains. Intestinal strains of lactobacilli have never been found to be extremely sensitive in these assays.
Tolerance to Bile Salts
The ability to survive the action of bile salts is an absolute need of probiotic bacteria, and it is generally included among the criteria used to select potential probiotic strains. The sensitivity of dairy isolates of lactobacilli to bile salts was demonstrated in the 1970s by the Gilliland group (Gilliland and Speck, 1977). They also suggested the importance of assessing bile tolerance in the selection of lactobacilli for probiotic use (Gilliland 1979). In a further study (Gilliland et al., 1984), it was shown that bile resistance could differ among members of the same species of enteric lactobacilli and that this difference could account for differences in the ability of strains to colonise the intestinal tract of calves. Two strains of L. acidophilus, isolated from calves, exhibiting a different degree of bile resistance when assayed in vitro, were used in a feeding trial with calves. Results indicated that both strains were able to increase the total counts of lactobacilli in treated animals. Comparison with the control group of animals, however, revealed that statistically significant differences in the number of lactobacilli were achieved only in the jejunum of calves administered the bile resistant strain. In the ileum, numbers of lactobacilli were significantly higher for groups of calves receiving both strains of lactobacilli and the difference between the treated and untreated groups was not significant. In the large intestine, the treated animals did not show an increased number of lactobacilli compared to the control group. These results are a good example of how it is possible to demonstrate the in vivo relevance of a phenotypic trait associated with specific strains.
The bile preparation used by Gilliland to assay strains in vitro was 'oxgall', a product derived from bovine bile, at a concentration of 0.3% (w\v). The core of the assay was the measurement of the lag phase in the growth curve caused by the presence, in liquid medium, of oxgall. Similar methodologies have been used by several authors (see Table 3) to assess the bile resistance of potential, or already commercialised, probiotic lactobacilli. Results reported in all of these papers showed that the amount of delay detected in the growth curve of lactobacilli challenged with oxgall was strain, and not species, dependent. This observation was confirmed in other species of lactobacilli isolated from humans, or intended to be used for human consumption (Table 3), and can be considered a well established observation.
In these papers, bile preparations derived from bovines was always used to assay lactobacilli selected for human use. The composition of bile preparations was not taken into consideration. When this problem was addressed during the FLAIR project (Reports of The University College Cork Ireland and of NIZO-The Netherlands, cited in the Project Final Report, Morelli 1994), it was observed that commercially available bile preparations contained variable proportions of conjugated and deconjugated bile salts and that assessment of sensitivity gave different results according to the commercial product used. Two commercial preparations were compared for their bile salt composition and their ability to inhibit the growth of lactobacilli. One was found to contain 97.2% of conjugated bile salts whereas the latter was composed of only deconjugated bile acids. Results showed that all tested strains exhibited resistance to the preparation containing conjugated bile salts and were more sensitive to the deconjugated bile salts. It was also shown that porcine bile was more inhibitory than bovine bile, but that "regardless of the resistance patterns observed in the presence of either bovine or porcine bile, all the assayed bacteria (all of human origin) were capable of growth in a physiologically relevant concentration of human bile (approx. 0.3%)" (Dunne at al., 1999).
It was also recently shown that bile is a potent inhibitor of strains used in probiotic preparations and belonging to the genus Bacillus (Table 4). These results prompted the authors to suggest that any claimed probiotic effect of these preparations must be due to spores and not to vegetative cells (Spinosa et al., 2000). All of the above observations suggest that this selection criterion is capable of discriminating the enteric origin of the strains, but not to predict the real in vivo behaviour of strains. However, a recent paper (Kimoto et al., 1999) has reported that some isolates of dairy lactococci have a tolerance to bile salts comparable to that of enteric lactobacilli. The same authors conclude that this in vitro observation needs an in vivo confirmation.
An interesting development in the evaluation of the role played by bile resistance in defining the probiotic capabilities of strains is seen in the study of bile salt hydrolases at the genetic level. Deconjugation of bile acids has been studied in lactobacilli since the 1960s (Hill and Drasar, 1968) and the gene encoding a bile salt hydrolase has been cloned and sequenced (Christiaensen et al., 1992). Disruption of the gene was accomplished (Leer et al., 1993) and the implication for bacterial survival in the gut has been addressed in one paper (De Boever and Verstraete, 1999). Data obtained from the comparison of the wild type, deconjugating strain with the mutant clone, suggested that bile tolerance in vivo could be the result of complex interactions of lactobacilli with other members of the intestinal microflora. No doubt this kind of study will lead to a better understanding of the real behaviour of lactobacilli in the gut. They are also, however, reducing the relevance of assaying bile tolerance in vitro with procedures that seem to be too simplified and still need improved standardisation.
Adhesion to Intestinal Cell Lines
The gastrointestinal tract, especially the small intestine, is a dynamic environment and the flow of digesta washes out any bacterium unable to counter the flow either by rapidly multiplying or by attaching itself to intestinal surfaces. It is generally agreed that adhering probiotic strains are more likely to have an increased opportunity to colonise the intestine. Results obtained with probiotics used in farm animals have shown that some strains of Lactobacillus are able to adhere, in vitro, to small pieces of intestinal tissues in a species-dependent way and that it is possible to observe a positive correlation between results obtained in vitro and colonisation potential tested in vivo (Fuller, 1975 and 1978; Barrow et al., 1980; Savage, 1984). However, these studies were performed using freshly harvested tissues, and raised some questions. Savage (1984) concluded that ". . . the capacity to adhere to the surface is undoubtedly insufficient by itself to ensure that the microorganisms can colonise the epithelial habitat". He showed that one Lactobacillus strain, isolated from a pig, and another strain derived from a mouse, adhered in vitro to mouse tissues in equally high number, but only the latter was able to colonise the same kind of tissue surface in vivo. Fuller (1978) noticed that strains that did not adhere in vitro to chicken crop cells scored negative results during an in vivo feeding trial. Of the two strains that scored positively both in vitro and in vivo, the best performing in vitro (adhesion index of 22) was less able to colonise in vivo than the strain which scored an adhesion index of 2 in vitro. If in vitro assays for probiotic bacteria to be used in animal feeds are not fully reliable, the problem of assessing, under laboratory conditions, the potential for adhesion in vivo of probiotic strains intended for human use is, in my opinion, far from resolution.
Two approaches have been followed in adherence assays. The first was based on intestinal cells collected from ileostomic patients (Conway et al., 1987) whilst the second used a human foetal intestinal cell line (Kleeman and Klaenhammer, 1982; Hood and Zottola, 1987, 1988, 1989). Adhesion to human ileal cells gave striking results as it was possible to establish clear differences between strains. Two human isolates scored much better than two dairy isolates that were tested. The best adhering human isolate (L. gasseri ADH) was able to attach at a density of 120-205 bacterial cells per ileal cell, while the other strain (L. acidophilus N2) ranged from 51 to 69 bacteria per intestinal cell (Conway et al., 1987). Adhesion assayed by means of a human foetal intestinal (Kleeman and Klaenhammer, 1982) cell line was also able to establish differences between human and non-human isolates and also between strains. Unfortunately, in this paper, adhesion was measured subjectively and it was not possible to make a comparison with results reported by other authors. These methodologies suffer severe limitations since the use of freshly isolated intestinal cells yields fluctuating results because of the use of different donors, and the foetal cell line was poorly differentiated with ill-defined brush borders.
At the beginning of 1990s, some groups turned their attention to two tumor cell lines, Caco2 and HT29, in order to have a better model system to assay adhesion of lactobacilli in vitro.
These cells show morphological and functional differentiation, having the characteristics of mature enterocytes. They have been extensively used for assessing attachment of pathogenic bacteria. The use of these cell lines, especially the Caco-2 cells, resulted in the publication of a number of papers aimed at assessing the adhesion potential of lactobacilli belonging to a range of species that were isolated from different sources. Methodologies used for this assessment were not standardised so it is difficult to compare the results obtained in the different studies. In at least four of these papers, it was possible to calculate the ratio of bacterial cells adhering to one Caco-2 cell (Table 5). For the best scoring strains, the ratio was about two bacterial cells per Caco-2 cell. This figure is in striking contrast with results obtained by Conway et al., (1987) where lactobacilli adhering to ileal cells ranged from 51 to 205 bacteria per cell. The difference in adhesion levels of Lactobacillus rhamnosus GG detected with Caco-2 cells compared to that obtained with ileal cells (Gorbach and Goldin 1989) also raise doubts about the reliability of this model. We might assume that freshly harvested intestinal cells would be a closer model to the in vivo situation than a tumor cell line. It must be added that these results are probably overestimates because the procedure included the use of acid, culture supernatants. The results in Table 5 were obtained without the use of chelating agents. These chemicals have been shown to artificially increase the adhesion potential of lactobacilli. Even if non-enteric lactobacilli seem to be influenced more by these parameters than intestinal isolates, it seems clear that reported differences among strains of lactobacilli are very small (1 vs. 1.5 bacteria per Caco-2 cell) which are unlikely to be able to predict significant behavioural differences in vivo. Furthermore, the in vitro adhesion index system has been shown (Greene and Klaenhameer, 1994, Lehto and Salminen, 1997, Tuomola et al., 1998, Blum and Reniero, 2000) to be extremely sensitive to factors such as (Blum and Reniero, 2000):
· pH
· the presence of calcium ions
· the number of lactobacilli
· the presence of culture supernatant
· the growth phase in which the bacteria were harvested
Observations that the adhesion of lactobacilli to cell lines is concentration-dependent (Greene and Klaenhameer, 1994, Lehto and Salminen 1997, Tuomola et al., 1999) and that a saturation end-point was never reached, suggested that non-specific adhesion is measured by this in vitro procedure. High adhesion indices scored by dairy isolates (see Table 5) further strengthen this suggestion. Studies using subjective evaluation of adhesion to Caco-2 cells by lactobacilli report high levels of adhesion for non-enteric lactobacilli (Lehto and Salminen, 1997, Jacobsen et al., 1999). It is surprising to note that in one of these papers Lactobacillus rhamnosus GG, whose in vivo colonising ability has been reported (Alander et al., 1999), was found to be less adherent than a dairy Lactococcus (Lehto and Salminen, 1997).
Jacobsen et al., (1999) using a subjective index, have scored three well studied probiotic strains in the following way: L. johnsonii LA 1=17, L. rhamnosus GG= 630, L. plantarum 299v=355. Despite these different in vitro scorings, all of these strains have been shown to persist in the human gut using doses of equal magnitude (reviewed by Salminen et al., 1996b). It seems to me that cell line models are unlikely to be able to provide an absolute differentiation between adhering and non-adhering strains. Available data suggest that a strain, of human enteric origin, scoring a positive result in the in vitro system, is likely to perform well in vivo. But it cannot be assumed that a negative result in this model means a lack of colonisation potential in vitro. Some new approaches have been introduced, such as the use either of human intestinal tissue, maintained viable for short periods in vitro (Sarem-Damerdij et al., 1995), or human mucus glycoproteins (Tuomola et al., 1999 ). Even if some of the results seem to be reliable, the lack of knowledge about the molecular adhesion mechanisms of lactobacilli does not allow firm conclusions about the significance of these observations to be made. New data, obtained at the genetic level, as regards proteins secreted by lactobacilli and able to adhere to mucus (for a review see Vaughan et al., 1999) may provide a breakthrough in this field in the future.
Conclusions
Interest in the field of probiotics has boomed in recent years, paralleling the renewed interest in studies focusing on microbial ecology of the gut and powered by the use of randomised, blind or double-blind human trials. Molecular biology has provided good tools to assess the real behaviour of a specific strain in vivo and to learn more about the composition of the intestinal microflora. The development of new probiotic products has produced new scientific achievements and a strong demand for improved and scientifically-based selection criteria (Tannock, 1997). The ability of a specific probiotic strain to survive and reproduce in the hostile environment of the gut is the most relevant feature to be checked during the selection procedures. But in vivo testing is time-consuming and expensive. In vitro selection is therefore the first approach used to select a few strains that can be further evaluated in vivo. Criteria used at the moment have been defined without a clear knowledge of the bacterial characteristics that are important in the proliferation of probiotic bacteria in the intestinal tract. Nevertheless, results achieved are positive and promising, but clearly show the scientific limitations of current selection criteria. Further development of probiotic products requires a refinement of these criteria. A deeper understanding of the molecular mechanisms that are used by bacteria to tolerate and persist in the harsh environments of the upper part of our gastrointestinal tract is required to achieve new and improved criteria.
Acknowledgements
The author is indebted to all partners of the EU FLAIR Project AGRF-CT91-0053 and the EU FAIR Project CT96-1028, whose excellent research work has provided the basis for this paper. I am particularly in debt to the late A. Weerkamp, who introduced me to a critical assessment of criteria used for strain selection. I wish also to thank my former PhD student R. Reniero, now at the Nestlè Research Centre, who shared with me many years of fruitful research in this field.
Alander, M. Satokari, R., Korpela, R., Saxelin, M., Vilpponen-Salmela, T., Mattila-Sandholm, T., and von Wright A. 1999. Persistence of colonization of human colonic mucosa by a probiotic strain, Lactobacillus rhamnosus GG, after oral consumption. Appl. Environ. Microbiol. 65: 351-354.
Axelsson, L. and Lindgren, S. 1987. Characterization and DNA homology of Lactobacillus strains isolated from pig intestine. J. Appl. Bacteriol. 62: 433-440.
Barrow, P.A., Brooker, B.E, Fuller, R. and Newport, M.J. 1980. The attachment of bacteria to the gastric epithelium of pig and its importance in the microecology of the intestine. J. Appl. Microbiol. 48: 147-154.
Bernet, M.F., Brassart D., Neeser J.R., and Servin A.L. 1994. Lactobacillus acidophilus LA 1 binds to cultured human intestinal cell lines and inhibits cell attachment and cell invasion by enterovirulent bacteria. Gut. 35: 483-489.
Bibel, D.J. 1988. Elie Metchnikoff's bacillus of long life. ASM News. 54: 661-665. Blum, S., and Reniero, R. 2000. Industrial panel statements: adhesion of selected Lactobacillus strains to enterocyte-like Caco-2 cells in vitro: a critical evaluation of the reliability of in vitro adhesion assays. Proceedings of at the 4th Workshop of the PROBEMO- FAIR CT 96-1028 project "Functional Foods for EU health in 2000 ".
Boris, S., Suárez, J. E., and Barbés, C. 1997. Characterization of the aggregation promoting factor from Lactobacillus gasseri, a vaginal isolate. J. Appl. Microbiol. 83: 413-420.
Brassart, D., Schiffrin, E., Rochat, F., Offord, E.A., Macè, C., and Neeser, J-R., 1998. The future of functional food: Scientific basis and future requirements. Lebensmittel Technol. 7-8: 258-266.
Cato, E.P., Moore, W.E.C., and Johnson, J.L. 1983. Synonimy of strains of "Lactobacillus acidophilus" group A2 (Johnson et al., 1980) with the type strain of Lactobacillus crispatus (Brygoo and Aladame 1953) Moore and Holdeman 1970. Int. J. Syst. Bacteriol. 33: 426-428.
Charteris, W.P., Kelly, P.M., Morelli, L., and Collins, J.K. 1998a. Ingredient selection criteria for probiotic micro-organisms in functional dairy foods. Int. J. Dairy Technol. 51: 123-136.
Charteris, W.P., Kelly, P.M., Morelli, L., and Collins, J.K. 1998b. Development and application of an in vitro methodology to determine the transit tolerance of potentially probiotic Lactobacillus and Bifidobacterium species in the upper human gastrointestinal tract. J. Appl. Microbiol. 84: 759-768.
Chateau, N., Deschamps, A.M., and Hadj-Sassi, A. 1994. Heterogeneity of bile salts resistance in the Lactobacillus isolates from a probiotic consortium. Lett. Appl. Microbiol. 18: 42-44.
Chauvière, G., Coconnier, M., Kernéis, S., Fourniat, J., and Servin A.L. 1992. Adhesion of human Lactobacillus acidophilus strain LB to human enterocyte-like Caco-2 cells. J. Gen. Microbiol 138: 1689-96.
Chou, L.S., and Weimer, B. 1999. Isolation and characterization of acid- and bile-tolerant isolates from strains of Lactobacillus acidophilus. J. Dairy Sci. 82: 23-31.
Christiaensen, H., Leer, J.R., Pouwells, P.H., and Verstraete, W. 1992. Cloning and expression of a conjugated bile acid hydrolase gene from Lactobacillus plantarum by using a direct plate assay. Appl. Environ. Microbiol. 58: 3792-3798.
Coconnier, M.H., Klaenhammer, T.R., Kernéis, S., Bernet, M.F., and Servin, A.L. 1992. Protein-mediated adhesion of Lactobacillus acidophilus BG2FO4 on human enterocyte and mucus-secreting cell lines in culture. Appl. Environ. Microbiol. 58: 2034-2039.
Conway, P.L., Gorbach, S.L., and Goldin, B.R. 1987. Survival of lactic acid bacteria in the human stomach and adhesion to intestinal cells. J. Dairy Sci. 70: 1-12.
De Boever, P., and Verstraete, W. 1999. Bile salt deconjugation by Lactobacillus plantarum 80 and its implication for bacterial toxicity. J. Appl. Microbiol. 87: 345-352. Dellaglio, F., Bottazzi, V., and Trovatelli, L.D. 1973. Deoxyribonucleic acid homology and base composition in some thermophilic lactobacilli. J. Gen. Microbiol. 74: 289-297.
Dunne, C., Murphy, L., Flynn, S., O'Mahony, L., O'Halloran, S., Feeney, M., Morrissey, D., Thornton, G., Fitzgerald, G., Daly, C., Kiely, B., Quigley, E.M., O'Sullivan, G.C., Shanahan, F., and Collins, J.K. 1999. Probiotics: from myth to reality. Demonstration of functionality on animal models of disease and in human clinical trials. Antonie van Leeuwenhoek. 76: 279-292.
Elmer, G.W., Surawicz, C.M., and McFarland, L.V. 1996. Biotherapeutic agents. A neglected modality for the treatment and prevention of selected intestinal and vaginal infections. JAMA. 275: 870-876.
Ferrero, M., Cesena, C., Morelli, L., Scolari, G., and Vescovo, M. 1996. Molecular characterization of Lactobacillus casei strains. FEMS Microbiol. Lett. 140: 215-219.
Fitzsimons, N.A., Cogan, T.M., Condon, S., and Beresford, T. 1999. Phenotypic and genotypic characterisation of non-starter lactic acid bacteria in mature cheddar cheese. Appl. Environ. Microbiol. 65: 3418-3426. Fonden, R., Bjorneholm, R., and Ohlson, K. 2000. Lactobacillus F19 - a new probiotic strain. Poster presented at the 4th Workshop of the PROBEMO- FAIR CT 96-1028 project "Functional Foods for EU health in 2000".
Fujisawa, T., Benno, Y., Yaeshima, T., and Mitsuoka, T. 1992. Taxonomic study of the Lactobacillus acidophilus Group, with recognition of Lactobacillus gallinarum sp. nov. and Lactobacillus johnsonii sp. nov. and synonymy of Lactobacillus acidophilus group A3 (Johnson et al., 1980) with the type strain of Lactobacillus amylovorus (Nakamura, 1981). Int. J. Syst. Bacteriol. 42: 487-491. Fuller, R. 1975. Nature of the determinant responsible for the adhesion of lactobacilli to chicken crop epithelial cells. J. Gen. Microbiol. 87: 245-250.
Fuller, R. 1978. Epithelial attachment and others factor controlling the colonization of the intestine of the gnotobiotic chicken by lactobacilli. J. Appl. Bacteriol. 45: 147-154.
Gilliland, S.E., and Walker, D.K. 1989. Factors to consider when selecting a culture of Lactobacillus acidophilus as a dietary adjunct to produce a hypocholesteremic effect in humans. J. Dairy Sci. 73: 905-911
Gilliland, S.E, Staley, T.E., and Bush, L.J. 1984. Importance of bile tolerance of Lactobacillus acidophilus used as dietary adjunct. J. Dairy Sci. 67: 3045-3051.
Gilliland, S.E. 1979. Beneficial interrelationships between certain microorganisms and humans: candidate organisms for use as dietary adjuncts. J. Food Protec. 42: 164-167.
Gilliland, S.E. and Speck, M.L. 1977. Deconjugation of bile acids by intestinal lactobacilli. Appl. Environ. Microbiol. 33: 15-18.
Giorgi, A., Torriani, S., Dellaglio, F., et al., 1987. Identification of vaginal lactobacilli from asymptomatic women. Microbiologica (Italy), 10: 377-384.
Goldin, B.R., Gorbach, S.L., Saxelin, M., Barakat, S., Gualtieri, L., and Salminen, S. 1992. Survical of Lactobacillus species (strain GG) in human gastrointestinal tract. Dig. Dis. Sci, 173: 121-128.
Gorbach, S.L., and Goldin, B.R. 1989. Lactobacillus strains and methods of selection. U.S. Patent N. 4, 839, 281.
Greene, J. D., and Klaenhammer, T. R. 1994. Factors involved in adherence of lactobacilli to human Caco-2 cells. Appl. Environ. Microbiol. 60: 4487-4494.
Guarner, F., and Schaafsma, G.J. 1998. Probiotics. Int. J. Food Microbiol. 39: 237-238.
Hill, M.J., and Drasar, B.S. 1968. Degradation of bile salts by human intestinal bacteria. Gut. 9: 22-27
Hood, S.K., and Zottola, A. 1987. An electron microscopic study of the adherence properties of Lactobacillus acidophilus. Food Sci. 52: 791-792.
Hood, S.K., and Zottola, A. 1988. Effect of low pH on the ability of Lactobacillus acidophilus to survive and adhere to human intestinal cells. J. Food. Sci. 53: 1514-1516.
Hood, S.K., and Zottola, A. 1989. An electron microscopic study of the adherence of Lactobacillus acidophilus to human intestinal cells in vitro. Food Microstructure 8: 91-97.
Huius in't Veld J., and Shortt, C. 1996. Selection criteria for probiotic micro-organisms. In: Leeds A. R. and Rowland, I. R. (eds) Gut flora and Health - Past, Present and Future pp 19-26. The Royal Society Of Medicine Press Ltd., London.
Jacobsen, C.N., Rosenfeldt Nielsen, V., Hayford, A.E., Moller, P.L., Michaelsen, K.F., Paerregaard, A., Sandstrom, B., Tvede, M., and Jakobsen, M. 1999. 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. 65: 4949-4956.
Johnson, J.L., Phelps, C.F., Cummins, C.S., London, J., and Gasser, F. 1980. Taxonomy of the Lactobacillus acidophilus group. Int. J. System. Bacteriol. 30: 53-68.
Kandler, O., and Weiss, N. 1986. Genus Lactobacillus Beijerinck 1901. In Sneath, P. H. A., Mair, N. S., Sharpe, M. E., Holt. J., G. eds. Bergey's Manual of Determinative Bacteriology. Vol. 2. The Williams and Wilkins Co. Baltimore, MD, pp 1209-1234.
Kimoto, H., Kurisaki, J., Tsuji, N.M., Ohmomo, S., and Okamoto T. 1999. Lactococci as probiotic strains: adhesion to human enterocyte-like Caco-2 cells and tolerance to low pH and bile. Lett. Appl. Microbiol. 29: 313-316.
Kimura, K., McCartney, A.L., McConnel, M.A., and Tannock, G.W. 1997. Analysis of fecal populations of bifidobacteria and lactobacilli and investigation of the immunological responses of their human hosts to the predominant strains. App. Environ. Microbiol. 63: 3394-3398.
Klaenhammer, T.R., and Kullen, M.J. 1999. Selection and design of probiotics. Int. J. Food Microbiol. 50: 45-57.
Kleeman, E.G., and Klaenhammer, T.R. 1982. Adherence of Lactobacillus species to human fetal intestinal cells. J. Dairy Sci. 65: 2063-2069.
Klein, G., Pack, A., Bonaparte, C., and Reuter, G. 1998. Taxonomy and physiology of probiotic lactic acid bacteria. Internat. J. Food Microbiol. 41: 103-125.
Lauer, E., Helming, C., and Kandler, O. 1980. Heterogeneity of the species Lactobacillus acidophilus (Moro) Hansen and Mocquot as revealed by biochemical characteristics and DNA-DNA hybridization. Zbl. Bakt. Hyg., I. Abst. Orig. C1, 150-168.
Lauer, E., and Kandler, O. 1980. Lactobacillus gasseri sp. nov., a new species of the subgenus Thermobacterium. Zbl. Bact., I Abt Orig C1: 75-78.
Leer, R.J., Christiaensen, H., Verstraete, W., Peters, L., Posno, M., and Pouwels, P.H. 1993. Gene disruption in Lactobacillus plantarum strain 80 by site-specific recombination: isolation of a mutant stain deficient in conjugated bile salt hydrolase activity. Mol. Gen. Genet. 239: 269-272.
Lehto, E.M., and Salminen, S. 1997. Adhesion of two Lactobacillus strains, one Lactococcus and one Propionibacterium strains to cultured human intestinal Caco-2 cell line. Biosci. Microflora. 16: 15-17
Lilly, D.M., and Stillwell, R.H 1965. Probiotics: growth promoting factors produced by microorganisms. Science. 147: 747-748.
Lin, H-C.J., and Savage, D.C. 1984. Host specificity of the colonization of murine gastric epithelium by lactobacilli. FEMS Microbiol. Lett. 24: 67-71.
Marteau, P., and Rambaud, M. 1993. Potential of using lactic acid bacteria for therapy and immunomodulation in man. FEMS Microbiol. Rev. 12: 207-220.
Mayra-Makinen, A., Manninen, M., and Gyllenberg, M. 1983. The adherence of lactic acid bacteria to the columnar epithelium cells of pigs and calves. J. Appl. Bacteriol. 55: 241-245.
Metchnikoff, E. 1907. Lactic acid as inhibiting intestinal putrefaction. In: The prolongation of life: optimistic studies. pp. 161-183. W. Heinemann, London.
Morelli, L. 1994. EU FLAIR Project AGRF-CT91-0053 The selection and charcterization of human probiotic strains, the process technology necessary for their incorporation into fermented foods and evaluation in selected animal models. Final report.
Morelli, L., Cesena, C., de Haen, C., and Gozzini, L. 1998. Taxonomic Lactobacillus composition of feces from human newborns during the first few days. Microb. Ecol. 35: 205-212.
Morishita, Y. Mitsuoka, T. Kaneuchi, C. Yamamoto, S., and Ogata, M. 1971. Specific establishment of lactobacilli in the digestive tract of germ-free chickens. Japanese Journal of Microbiology. 15: 531-538.
O'Sullivan, M.G., Thornton, G., O'Sullivan, G.C., and Collins, J.K. 1992. Probiotic bacteria: myth or reality? Trends Food Sci. Technol. 3: 309-314.
Pryde, S.E., Richardson, A.J., Stewart, C.S., and Flint, J. 1999. Molecular analysis of the microbial diversity present in the colonic wall, colonic lumen, and cecal lumen of a pig. Appl. Environ. Microbiol. 65: 5372-5377
Raya-R.R., and Klaenhammer, T.R. 1992 High-frequency plasmid transduction by Lactobacillus gasseri bacteriophage adh. Appl. Environ. Microbiol. 58: 187-193.
Reniero, R., Morelli, L., De Haen, C., and Bottazzi, V. 1991. Detection of permanent Lactobacillus casei subsp. casei strains in weaned infants' gut. Lett. Appl. Microbiol. 13: 3-6.
Rettger, L.F., Levy, M.N., Weinstein, L., and Weiss, J.E. 1935. Lactobacillus acidophilus and its therapeutic action. Yale University Press.
Salminen, S., Laine, M., von Wright, A., Vuopio-Varkila, J., Korhonen, T., and Mattila-Sandholm, T. 1996a. Development of selection criteria for probiotic strains to assess their potential in functional foods: a Nordic and European approach. Biosci. Microflora 15: 61-67.
Salminen, S., Isolauri, E., and Salminen, E. 1996b. Clinical uses of probiotic for stabilizing the gut mucosal barrier: successful strains and future challenges. Antonie van Leeuwenhoek 70: 347-358.
Sarem, F., Sarem-Damerdij, L.O., and Nicolas, J.P. 1996. Comparison of the adherence of three Lactobacillus strains to Caco-2 and Int-407 human intestinal cell lines. Lett. Appl. Microbiol. 22: 439-442.
Sarem-Damerdij, L.O., Sarem, F., Marchal, L., and Nicolas, J.P. 1995. In vitro colonization ability of human colon mucosa by exogeneous Lactobacillus strains. FEMS Microbiol. Lett. 131: 133-137.
Sarra, P.G., Dellaglio, F., and Bottazzi, V. 1985. Taxonomy of lactobacilli isolated from the alimentary tract of chickens. System. Appl. Microbiol. 6: 86-89.
Sarra, P.G., Magri, M., Bottazzi, V., and Dellaglio, F. 1980. Genetic heterogeneity among Lactobacillus acidophilus strains. Antonie van Leewenhoek 46: 169-176. Savage, D.C. 1984. Adherence of the normal flora. in: Attachment of organisms to the gut mucosa Vol. I. Boedeker, E. C. ed., Boca Raton CRC Press. pp. 3-10.
Song, Y.L., Kato, N., Liu, C. X., Kato, H., and Watanabe, K. 1999. Identification of and hydrogen peroxide production by fecal and vaginal lactobacilli isolated from Japanese women and newborn infants. J. Clin. Microbiol. 37: 3062-3064.
Song, Y.L., Kato, N., Liu, C.X., Matsumiya, Y., Kato, H., and Watanabe, K. 2000. Rapid identification of 11 human intestinal Lactobacillus species by multiplex PCR assays using group- and species-specific primers derived from the 16S-23S rRNA intergenic spacer region and its flanking 23S rRNA. FEMS Microbiol. Lett. 187: 167-173.
Spinosa, M.R., Braccini, T., Ricca, E., Morelli, L., Pozzi, G., and Oggioni, M.R 2000. Fate of ingested Bacillus spores. Res. Microbiol. 151: 361-368. Stahl, M., Pettersson, B., Molin, G., Uhlen, M., and Ahrne, S. 1994 Restriction fragment length polymorphism of Lactobacillus reuteri and Lactobacillus fermentum, originating from intestinal mucosa, based on 16S rRNA genes. Syst. Appl. Microbiol. 17: 108-115.
Suegara, N.M., Morotomi, T., Watanabe, Y., Kawai, Y., and Mutai, M. 1975. Behaviour of microflora in the rat stomach: adhesion of lactobacilli to keratinized epithelial cells of the rat stomach in vitro. Infect. Immun. 12: 173-179.
Tannock, G.W., Szylit, O., Duval, Y., and Raibaud, P. 1982 Colonization of tissue surfaces in the gastrointestinal tract of gnotobiotic animals by Lactobacillus strains. Can. J. Microbiol. 28: 1196-1198.
Tannock, G.W., Munro, K., Harmsen, H.J.M., Welling, G.W., Smart, J., and Gopal, P.K. 2000. Analysis of the fecal microflora of human subjects consuming a probiotic product containing Lactobacillus rhamnosus DR20. Appl. Environ. Microbiol. 66: 2578-2588.
Tannock, G.W. 1990. The microecology of lactobacilli inhabiting the gastrointestinal tract. In Advances in microbial ecology ed. K. C. Marshall. Plenum Press, N.Y. p. 147-171.
Tannock, G.W. 1997. Probiotic properties of lactic acid bacteria: plenty of scope for fundamental R & D. Trends Biotechnol. 15: 270-274.
Tannock, G.W. 1999. Analysis of the intestinal microflora: a renaissance. Antonie van Leeuwenhoek. 76: 265-278.
Tuomola, E.M., and Salminen, S.J. 1998. Adhesion of some probiotic and dairy Lactobacillus strains to Caco-2 cell cultures. Int. J. Food Microbiol. 41: 45-51.
Tuomola, E.M., Owwehand, A.C., and Salminen, S.J. 1999. Human ileostomy glycoproteins as a model for small intestinal mucus to investigate adhesion of probiotics. Lett. Appl. Microbiol. 28: 159-163.
Usman, and Hosono, A. 1999. Bile tolerance, taurocholate deconjugation, and binding of cholesterol by Lactobacillus gasseri strains. J. Dairy Sci. 82: 243-248.
Vaughan, E.E., Mollet, B., and deVos, W.M. 1999. Functionality of probiotics and intestinal lactobacilli: light in the intestinal tract tunnel. Curr. Opinion. Biotechnol. 10: 605-610.
Walker, D.K., and Gilliland, S.E. 1993. Relationships among bile tolerance, bile salts deconjugation and assimilation of cholesterol by Lactobacillus acidophilus. J. Dairy Sci. 76: 956-961.