Anti-Infective Activities of Lactobacillus Strains in the Human Intestinal Microbiota: from Probiotics to Gastrointestinal Anti-Infectious Biotherapeutic Agents

Abstract

SUMMARY

A vast and diverse array of microbial species displaying great phylogenic, genomic, and metabolic diversity have colonized the gastrointestinal tract. Resident microbes play a beneficial role by regulating the intestinal immune system, stimulating the maturation of host tissues, and playing a variety of roles in nutrition and in host resistance to gastric and enteric bacterial pathogens. The mechanisms by which the resident microbial species combat gastrointestinal pathogens are complex and include competitive metabolic interactions and the production of antimicrobial molecules. The human intestinal microbiota is a source from which Lactobacillus probiotic strains have often been isolated. Only six probiotic Lactobacillus strains isolated from human intestinal microbiota, i.e., L. rhamnosus GG, L. casei Shirota YIT9029, L. casei DN-114 001, L. johnsonii NCC 533, L. acidophilus LB, and L. reuteri DSM 17938, have been well characterized with regard to their potential antimicrobial effects against the major gastric and enteric bacterial pathogens and rotavirus. In this review, we describe the current knowledge concerning the experimental antibacterial activities, including antibiotic-like and cell-regulating activities, and therapeutic effects demonstrated in well-conducted, placebo-controlled, randomized clinical trials of these probiotic Lactobacillus strains. What is known about the antimicrobial activities supported by the molecules secreted by such probiotic Lactobacillus strains suggests that they constitute a promising new source for the development of innovative anti-infectious agents that act luminally and intracellularly in the gastrointestinal tract.

INTRODUCTION

The gastrointestinal (GI) tract forms complex ecosystem that functions in concert with the resident microbiota as a structural and functional barrier that protects the host from attack by unwanted, harmful enterovirulent microorganisms (1). The mucosal surface of the gastrointestinal tract faces the external environment (2, 3). The stomach is a muscular organ that secretes the acid and enzymes involved in digesting food. Histologically, the human stomach can be divided into three regions: the cardia, the fundus/corpus, and the antrum (3). Specialized secretory cell phenotypes are present: acid-secreting parietal cells, mucus neck cells, and pepsinogen-secreting zymogenic cells in the fundus and corpus and gastrin-secreting cells and gland cells in the antrum (4). These different cell types are localized in glandular invaginations, which are known as oxyntic glands in the corpus/fundus region and as pyloric glands in the antrum. Members of the first line of defense of the gastrointestinal tract against the unwanted intrusion of pathogenic bacteria are present in the stomach (3). Mucus-producing surface mucus cells cover the whole gastric mucosa, creating a mucus barrier composed of mucin glycoproteins: the membrane-associated MUC1 and the secreted MUC5AC and MUC6. A large number of bacteria are present in the outer mucus layer, whereas the inner mucus layer is virtually free of bacteria. In the gastric epithelium several antimicrobial peptides (AMPs) are present, such as β-defensin 1, β-defensin 2, and cathelicidin LL37. In addition, hepcidin has been identified as a major regulator of iron homeostasis and links the iron metabolism and host response to infection (5). One of the main functions of the stomach is to achieve digestion, and the harsh gastric environment contributes to inactivating ingested microorganisms, including pathogens, to prevent them from reaching the intestine. The pyloric sphincter connects the stomach with the intestine and controls passage of the digested food into the intestine.

Anatomically, the intestine is formed by four segments: the duodenum, jejunum, ileum, and colon. The gastrointestinal epithelium consists of a single layer of fully differentiated, polarized epithelial cells of various phenotypes that create an impermeable, regulated epithelial barrier separating the external and internal environments. Four highly specialized cell phenotypes compose the intestinal epithelium: enterocytes (also known as fluid-transporting cells), neuroendocrine cells, mucin-secreting cells (also known as goblet cells), and Paneth cells (6). Tight junctions positioned most apically in the junctional domain of intestinal epithelial cells are the primary cellular determinant ensuring the closure of the intestinal epithelial barrier. There are several lines of chemical defenses in the intestine that function to prevent the passage of luminal enteric bacterial pathogens across the epithelium. The intestinal mucosa is coated by secreted mucus delivered by mucin-secreting cells (7, 8). The outer mucus layer favors the growth of the mucosa-associated resident microbiota and of enterovirulent bacteria by providing nutrients, and the density of the inner mucus layer limits the contact between luminal enterovirulent bacteria and the intestinal epithelial cells. The thinness of the mucus layer in parts of the small intestine renders efficient specialized immune anti-infective mechanisms. In contrast, the roles of membrane-bound mucins are poorly documented. The Paneth cells are pyramid-shaped, columnar, exocrine cells located in the crypts of the epithelium that provide AMPs, including defensins, C-type lectins, and cathelicidins (9, 10). AMPs are retained by the mucus overlying the intestinal epithelium and act rapidly to kill or inactivate pathogenic microorganisms. For example, α-defensins and cathelicidins kill both Gram-negative and Gram-position pathogens, whereas C-type lectins kill only Gram-positive bacteria. Nevertheless, several enterovirulent bacteria can partially evade the action of AMPs by altering the anionic charge of their own surface molecules. Intestinal host defense systems against pathogenic enteric bacteria include both adaptive and innate immunity. Adaptive immune responses are induced mainly in the follicle-associated epithelium that overlies the organized gut-associated lymphoid tissue through the interaction of intestinal epithelial cells with antigen-presenting cells and lymphoid cells. The intestinal epithelium senses the microbial environment and produces strong cellular defense responses, including the release of cytokines and chemokines which trigger the recruitment of leukocytes and others immune cells (11–13). Pathogen recognition receptors (PRRs), including Toll-like receptors, retinoic acid-inducible gene 1-like receptors, NOD-like receptors, and DNA receptors, recognize pathogens expressing various signature molecules called pathogen-associated molecular patterns (PAMPs) and rapidly trigger a panel of antimicrobial immune responses and also some adaptive immune responses. In addition, autophagy, which is an evolutionarily conserved process by which cell constituents are recycled, acts as a cell defense mechanism against intracellular pathogenic bacteria (14).

GASTROINTESTINAL MICROBIOTA

Knowledge about the composition of the microbiota of the human stomach is currently increasing (3). A low level of resident bacteria (10¹ to 10³/ml) colonize the human stomach. The five main phyla that have been identified are the Firmicutes, Actinobacteria, Proteobacteria, Fusobacteria, and Bacteroidetes. The dominant genera in Firmicutes are Lactobacillus, Streptococcus, Veillonella, Staphylococcus, and Bacillus. Lactobacillus species that have been identified in the gastric microbiota are L. antri, L. gastricus, L. kalixensis, L. reuteri, L. ultunensis, L. plantarum, L. salivarius, L. fermentum, and L. gasseri.

The adult human intestine has been estimated to contain trillions of microbes, including hundreds of species and thousands of subspecies, which display age- and geography-related differences and are distributed as a function of the intestinal site; they have a predominantly symbiotic relationship with their host (15, 16). The intestinal microbiota of healthy adults expresses two dominant phyla, the Gram-negative Bacteroidetes and the Gram-positive Firmicutes. Less abundant are the Proteobacteria, the Verrucomicrobia, the Tenericutes, the Deferribacteres, and the Fusobacteria. Microbial colonization in the human GI tract evolves along the length of the intestinal tract. A low level of 10¹ to 10³ bacteria per gram of content is present in the duodenum, progressing to 10⁴ to 10⁷ bacteria per gram of content in the jejunum and ileum and reaching 10¹¹ to 10¹² bacteria per gram of content in the colon. In the colon, most of the bacteria present are anaerobes, with about 1,000-fold fewer facultative anaerobes. Infants acquire their commensal bacteria from other human beings, in particular from the mother. Microbial colonization of the gut begins during birth and early infancy and then proceeds under the influence of breastfeeding and skin-to-skin contact with the mother and other people (17, 18). Many of these resident intestinal bacteria are adapted to the intestinal environment and develop complex ecological networks with other bacteria to acquire nutrients. Supported by both phylogenetic and functional analyses conducted by the international MetaHIT consortium, data obtained from examining different metagenomes indicate that the microbial communities include three distinct clusters known as “enterotypes” (19). Each enterotype displays specific species and functional compositions that differ from those of the other two.

The intestinal microbiota fulfills three functions (20). First, it assists host nutrition. The species that make up the resident microbiota ferment components of the host's diet that the host is unable to digest and synthesize low-molecular-weight metabolites, amino acids, and vitamins. After being absorbed from the intestinal lumen, these low-molecular-weight metabolites are transported into the systemic circulation, where they play roles in both health and disease. The functions of low-molecular-weight metabolites produced by the intestinal microbiota that constitute the intestinal metabolome remain largely unknown, although the roles of some of them, such as short-chain fatty acids and polyamines, have been investigated (21). It is noteworthy that the antibiotic treatment induces modifications in the composition of the intestinal microbiota that have an impact on the intestinal homeostasis as a result of changes in the intestinal metabolome (22). Second, the intestinal microbiota also participates in the structural and functional maturation of the epithelial cells lining the epithelial barrier and influences intestinal immune development (23). The third function relates to the intestinal host defense mechanisms, since several of the species that make up the resident microbiota inhibit the unwanted intrusion of harmful microorganisms by competing for nutrients and blocking the deleterious effects of bacterial virulence factors on the host cell by enhancing the mechanisms of the host epithelial and immune defenses and producing organic acids and antibacterial compounds (24, 25). In addition, the regulation of several host defense mechanisms results in activation of innate immune receptors by microbial molecules produced by the resident intestinal microbiota (26).

The resident intestinal microbiota has recently been shown to constitute an innovative therapeutic strategy in the treatment of several intestinal disorders. Indeed, fecal bacteriotherapy, also called fecal microbiota transplantation, has recently been successfully used as a therapy to correct the dysbiosis that characterizes chronic Clostridium difficile infection (27). Moreover, owing to the role of resident intestinal microbiota in inflammatory bowel disease (IBD) (28) and the presence of an “activated intestinal microbiota” containing an elevated number of normally underrepresented potentially harmful bacteria resulting from dysbiosis (20, 29), microbiota transplantation also appears to be potentially useful to treat IBD and irritable bowel syndrome (27, 30, 31). Interestingly, an emerging recently described role of members of the microbiota is to provide protection against pathobionts, i.e., normally harmless microorganisms that can become pathogens under certain environmental conditions (32, 33).

Pathogenic bacteria infect specific regions of the gastrointestinal tract. Helicobacter pylori, which is never found in the small or large intestine, infects the stomach mucosa and to do this has developed special adaptation properties that enable this conditional pathogen to survive in the harsh gastric environment (34, 35). Vibrio cholerae and enterotoxigenic Escherichia coli (ETEC) target the polarized epithelial cells lining the small intestine, whereas C. difficile, Shigella spp., enterohemorrhagic E. coli (EHEC), and Campylobacter spp. target the cells lining the colon epithelium, and Yersinia spp. and Salmonella spp. can affect both the small intestine and the colon (36–40). The high level of interdependency within intestinal microbiota bacterial communities makes it particularly difficult to study the cooperation between the various intestinal bacterial species in producing the chemical barrier effect of the intestinal microbiota against the unwanted intrusion of bacterial pathogens. Various intestinal microbiota bacterial species that produce antimicrobial molecules triggering a pivotal role in the chemical barrier effect of the intestinal microbiota against the unwanted intrusion of bacterial pathogens have been identified. By producing bacteriocins, resident Gram-positive bacteria in the intestinal microbiota have bacteriostatic or bactericidal effects against closely related Gram-positive species (41–43). Based on biochemical characteristics, bacteriocins of intestinal microbiota Gram-positive bacteria are classified in two groups, i.e., lantibiotics and heat-stable proteins not containing lanthionine residues, each of which is subclassified into multiple subgroups (44–46). Bacteriocin-associated antibacterial activities have been observed both in vitro and in vivo at concentrations in the nanomolar range and develop against Gram-positive bacteria closely related to the producing strain. Several bacteriocins act upon the cell envelopes of target pathogens, and others are active within the cell and affect gene expression (47, 48). Pores and ion channels formed in the cytoplasmic membrane of the target microbial cell and the subsequent leakage of intracellular components and protein production are the typical modes of bacteriostatic and bactericidal activities, despite the structural and physicochemical differences observed between the different classes of bacteriocins. In addition, resident Escherichia coli strains exert potent bactericidal activity against closely related Gram-negative species by also producing bacteriocins (43, 49), known as the low-molecular-weight microcins (50) and the larger colicins (51), which are typically plasmid encoded. Microcins range from 1 to 10 kDa, are subclassified based on the presence and localization of posttranslational modifications and the organization of the gene cluster and the leader sequence, and have MICs in the nanomolar range. They disrupt a wide range of functions in the target cell, including ATP synthetase and DNA gyrase. Colicins range in size from 30 to 80 kDa and have MICs in the picomolar to nanomolar range. They have varied killing mechanisms, which include pore formation, DNase or RNase activity, or inhibition of peptidoglycan biosynthesis. In bacteria, an intercellular communication process called quorum sensing (QS) is based on the synthesis and secretion of small hormone-like molecules, termed autoinducers, coordinated mainly in response to the bacterial population density (52, 53). In bacterial pathogens, QS molecules bind to cognate receptors which, after activation, directly or indirectly control expression of target genes coding for virulence factors (54). A QS mechanism(s) regulates the production of bacteriocins by lactic acid bacteria via secreted bacteriocin-like peptide pheromones (55–57). Interestingly, Lactobacillus QS molecules controlling bacteriocin production have been found to be activated in response to infection (58). Moreover, it has been reported that Lactobacillus QS-quenching compounds have been found to be involved in the control of virulence of bacterial pathogens in vitro and in vivo (59–65).

Elie Metchnikoff, the Russian-born Nobel Prize winner who worked at the Pasteur Institute in Paris, reported the first observation of a beneficial role played by some bacteria of the human intestinal flora and suggested that “the dependence of the intestinal microbes on food makes it possible to adopt measures to modify the flora in our bodies and to replace the harmful microbes by useful microbes” (66). Henry Tissier (67), observing that “bifid” bacteria are abundant in the stools of healthy children but present at only low levels in the stools of children with diarrhea, postulated that these bacteria administered to patients with diarrhea could restore a healthy gut flora. For Fuller (68), the natural balance of the gut microflora could be restored by administering probiotics. In this review, we highlight the key traits of six human intestinal probiotic strains of Lactobacillus that have displayed experimentally demonstrated antibacterial activities against gastric or enterovirulent bacterial pathogens and have clinically demonstrated therapeutic efficacy against enterovirulent bacterium- and rotavirus-induced diarrhea and H. pylori-induced gastritis. Experimental and clinical data indicate that the Lactobacillus-produced molecules underpin activities that could be a potential source of new anti-infectious molecules offering alternatives to antibiotics to treat gastrointestinal infections.

PROBIOTIC LACTOBACILLUS STRAINS ISOLATED FROM THE HUMAN INTESTINAL MICROBIOTA

The FAO/WHO Expert Committee has defined probiotic strains as “live microorganisms which, when consumed in appropriate amounts in food, confer a health benefit on the host” (69, 70). In addition, a set of major criteria has been defined for probiotic species, among which may be cited proper taxonomic identification by molecular techniques, deposition in an internationally recognized culture collection, a lack of transmissible antibiotic resistance genes, persistence in a viable state in the gastrointestinal tract, experimentally and clinically demonstrated health benefits, safety for human use, and resistance to technological processes as shown by, e.g., the preservation of cell viability and probiotic activities throughout the processing, handling, and storage of the food product containing the probiotic strain (69). The probiotic activity that has undergone the most experimental and clinical investigation is that exerted against gastric or enterovirulent bacterial pathogens and rotaviruses, both of which constitute major human health problems. Lactobacillus strains isolated from the human intestinal microbiota display antibacterial activities as a result of producing metabolites such as lactic acid, bacteriocins, nonbacteriocin compounds (defined as sensitive to proteolytic enzymes but unable to be precipitated with 80% saturated ammonium sulfate), and nonproteinaceous molecules that exercise a direct bactericidal effect (41–43, 71) and/or downregulate the expression of the virulence factors of enterovirulent pathogens or modulate the deleterious effects of these factors on the host's intestinal cell structure, machinery, and functions (72–75). It is noteworthy that a given effect of a probiotic Lactobacillus is strain specific and cannot be extrapolated to other strains of the Lactobacillus genus or even to other strains belonging to the same species and subspecies. In addition, substances produced by probiotic Lactobacillus strains, including cell wall components and secreted molecules, have been reported to display immunomodulatory activities, mainly in in vitro experiments (76, 77). The mechanistic studies showing the immunomodulatory effects of probiotics are based principally on in vitro cell culture models (78). Positive results have recently been reported in in vivo models, but currently there has been no convincing clinical demonstration of a probiotic-induced immunostimulatory effect in human patients.

A large number of Lactobacillus strains have been isolated from human and animal intestinal microbiota, and the properties of these strains, including their adhesion to cultured epithelial cells or mucus and their inhibitory activities, have been reported mainly from in vitro experiments. Only six Lactobacillus strains isolated from the human intestinal microbiota have been clearly demonstrated to have probiotic antimicrobial and antirotavirus properties in a comprehensive set of in vitro and in vivo experiments and randomized controlled trials (RCTs). These probiotic strains are L. rhamnosus strain GG (ATCC 53103) (Valio Ltd., Finland) (79–82), L. casei strain Shirota YIT9029 (Yakult Honsha Co., Ltd., Japan) (83, 84), L. acidophilus strain LB (Forest Laboratories, Inc., New York, NY) (85), L. johnsonii NCC 533 (first designed La1) (CNCM I-1225) (Nestlé, Switzerland) (86), L. casei DN-114 001 (CNCM I-1518) (Danone, France) (87), and L. reuteri DSM 17938 (ATCC 55730 cured of the pLR581 and pLR585 antibiotic resistance plasmids, also designed SD2112, ING1, and MM53) (BioGaia AB, Sweden) (88–90). (L. acidophilus strain LB was historically identified as such on the basis of its biochemical and metabolic activities; subsequent molecular investigation reclassified the identity of this strain as a symbiotic culture of L. fermentum [L. fermentum LB-f] [CNCM I-2998] [91, 92] and L. delbrueckii strains in a 95:5 ratio.)

Other probiotic Lactobacillus strains have been isolated from the human intestinal microbiota; these include L. acidophilus strain NCFM (also designated RL8K/NCK45/NCK56/N2) (ATCC 700396) (Danisco A/S, Denmark) (93), L. plantarum 299v (DSM 6595) (Probi AB, Sweden), and L. fermentum ME-3 (DSM 14241) (University of Tartu and Tere AS, Estonia) (94). It is noteworthy that these probiotic strains have been experimentally tested only in vitro against gastrointestinal pathogens or in animal infection models or have been experimentally tested for other probiotic effects. It was noted that there have been no RCTs for these strains in human infection situations, but several of these strains have been therapeutically tested in humans for other probiotic effects. In addition, it was noted that the probiotic L. rhamnosus R0011 (CNCM I-1720) and L. helveticus R0052 (CNCM I-1722) strains isolated from dairy cultures (Institut Rosell-Lallemand Inc., Canada) (95, 96) have been shown to exhibit both experimental and clinical anti-infectious effects (97).

According to a U.S. Food and Drug Administration (FDA) working definition, probiotics are classified as “live biotherapeutics”: “live microorganisms with an intended therapeutic effect in humans” (98, 99). Guidelines for the clinical use of probiotic strains were published after a Yale University workshop in 2005 and were updated in 2007 (100). The advice is graded as “A,” “B,” “C,” or no category. Classified in the A-grade advice are probiotic Lactobacillus strains used to treat acute childhood diarrhea (101) and C. difficile-associated diarrhea (102). Probiotic Lactobacillus strains used to treat chronic disorders of the gut, including IBD, are classified as having B-grade advice because there have been some negative studies (103). It was noted that the C grade relates to results that were significant but unable to receive stronger grades because of the numbers of patients enrolled in studies. The probiotic strains are considered to be safe, or “generally recognized as safe” (GRAS) (104–106). However, as Sanders et al. (105) point out and as stated in the report of the Agency for Healthcare Research and Quality (AHRQ) (104), there was a small number of studies specifically designed to assess probiotic safety, contrasting with the long history of safe use of foods containing probiotic bacteria. The information that is required for the development of probiotic products containing live or lyophilized dormant probiotic Lactobacillus in food or dietary supplements includes the step-by-step description of the manufacturing process, which is carried out under aseptic conditions, and the quality control process, including the process-input parameters and the expected output results that ensure that the product remains stable and pharmacologically effective over the indicated conservation time (69, 99). Indeed, Fayol-Messaoudi et al. (107) have found that the bactericidal activity against Salmonella enterica serovar Typhimurium of 24 h-cultures of L. rhamnosus GG, L. johnsonii NCC 533, L. casei Shirota, and L. casei DN-114 001 rapidly decreased by 4 log CFU/ml after 1 day in storage at 4°C. Grzeskowiak et al. (80) have investigated the probiotic antibacterial effect of L. rhamnosus GG isolated from 10 probiotic products, including capsules, commercial infant foods, and freeze-dried powders from several different countries. Compared to parental L. rhamnosus GG, the antibacterial activities of the product-isolated strains varied significantly. Moreover, the resuscitation of dormant probiotic strains (108, 109) needs particular attention during processing. Indeed, as demonstrated by Muller et al. (110), the resuscitation of dried probiotic strains, including L. johnsonii NCC 533, is a critical aspect for obtaining active and effective probiotic strains, as multivariate factors, including the pH, the diluent, and the reconstitution time, can all have a strong influence on the content, viability, and activity of the restituted probiotic bacteria. As a consequence, the authors stress that it is necessary for the resuscitation conditions to be optimized for each of the lyophilized probiotic strains used. These experimental observations are in line with what was previously defined by the FAO/WHO expert group (69) and by Sanders (111), i.e., the need to conduct appropriate quality control measures for all foods containing probiotic strains to ascertain whether the original properties of each of the probiotic strains have been preserved after the manufacturing process undergone by the food.

IN VITRO ANTIBACTERIAL ACTIVITIES

Direct Activities against Enterovirulent Bacteria

Bacteriostatic activity.

Lactobacillus strains exert direct antagonistic activities by a bacteriostatic activity that blocks the growth of enterovirulent bacteria as a result of the strain-specific production of bacteriocins. Several bacteriocins are used as biopreservatives in food and food products to inhibit or control the growth of food-borne bacterial pathogens (41–43). The generally held view is that bacteriocins exhibit less potential as chemotherapeutics for infections with Gram-negative pathogens. However, it should be noted that several bacteriocins and bacteriocin-like molecules produced by probiotic Lactobacillus strains have been also found to be active against the growth of several Gram-negative gastric or enterovirulent bacterial pathogens, including H. pylori, EHEC, Shigella, Salmonella, and Campylobacter (112–116).

Bactericidal activity.

Certain probiotic Lactobacillus strains display bactericidal activity against Gram-negative or Gram-positive gastric or enterovirulent bacteria after direct contact in vitro. Bactericidal activities of probiotic Lactobacillus strains have been generally explored using probiotic cultures. For some of the six probiotic Lactobacillus strains examined here, it has been shown that bacteria isolated from a culture do not have a bactericidal effect after direct contact with pathogens and that the bactericidal activities of cultures are reproduced with the cell-free spent culture supernatants (CFCSs) (Fig. 1 to 3 and Table 1). It is important to note that this activity develops within the specified limits for an antibiotic acting against a pathogenic microorganism, i.e., the bactericidal activity producing greater than a 3-log reduction of a viable cell count of a test microorganism after incubation for a fixed length of time under controlled conditions (117). A loss of ∼4 log CFU/ml of Shigella viability has been observed to be triggered by the L. rhamnosus GG (118, 119), L. johnsonii NCC 533 (120), L. reuteri ATCC 55730 (121), and L. acidophilus LB (122) strains after 4 h of direct contact. The viability of Listeria was affected by 3 to 4 log CFU/ml after 4 h of exposure to the L. johnsonii NCC 533 (120) and L. acidophilus LB (122) strains. The L. rhamnosus GG (119, 123), L. casei Shirota (124), L. reuteri ATCC 55730 (121), and L. acidophilus LB (122, 125) strains decreased the viability of enterovirulent E. coli by 3 to 4 log CFU/ml after 4 h of direct contact. The viability of S. Typhimurium was dramatically lowered, by ∼5 log CFU/ml, after 4 h of exposure to the L. rhamnosus GG (107, 118, 119, 126–131), L. johnsonii NCC 533 (107, 120, 128, 129, 132–134), L. casei Shirota (107, 128, 129, 135), L. casei DN-114 001 (107), L. reuteri ATCC 55730 (121), and L. acidophilus LB (122, 136, 137) strains. Antagonistic activity against Vibrio cholerae has been reported only for L. reuteri ATCC 55730 (121). It is noteworthy that the bactericidal activity must develop when a bacterial pathogen attaches to human intestinal cells, since diffusely adhering E. coli strains (DEAC) expressing Afa/Dr adhesins (Afa/Dr DAEC) attached within brush borders of enterocyte-like Caco-2/TC7 cells are killed after treatment of the infected cells with L. acidophilus LB CFCS (125) (Fig. 2). A problem in intestinal antibiotic therapy is the ineffectiveness of antibiotics against enteroinvasive bacterial pathogens that have already entered host intestinal cells and taken up residence within the cell cytoplasm or intracellular vacuoles. Consequently, the observation that a compound(s) secreted by L. acidophilus LB was able to kill S. Typhimurium resident in the intracellular vacuoles located within the enterocyte-like Caco-2/TC7 cells is particularly interesting (136).

FIG 1.

Bactericidal effect of probiotic L. acidophilus strain LB against gastric or enterovirulent bacterial pathogens. (A) Time course of the bactericidal effect of L. acidophilus strain LB against wild-type ETEC H10407 expressing colonization factor CFA/I, EPEC E2348/69, Afa/Dr DAEC C1845, S. Typhimurium SL1344, and H. pylori 1101 after direct contact. (B) Scanning electron microscopy micrographs showing the transformation of the helical form of H. pylori to the U-shaped form after treatment with L. acidophilus strain LB. For both panels A and B, pathogens were placed in direct contact with L. acidophilus LB CFCS. (The time course of the bactericidal effect in panel A is based on data extracted from reference 182 with permission of the publisher and from references 122 and 139. The two micrographs in panel B are reprinted from reference 139.)

FIG 3.

Bactericidal effect of probiotic L. acidophilus strain LB against wild-type S. Typhimurium SL1344 residing within intracellularly localized vacuoles in preinfected cultured human intestinal Caco-2/TC7 cells. (A) Time course of the bacterial effect. (B) Confocal laser microscopy scanning examination of fluorescein-labeled S. Typhimurium SL1344 cells showing the altered morphology of bacteria residing within a typical intracellular vacuole in L. acidophilus strain LB-treated cells compared to untreated cells. For panels A and B, preinfected cells were treated with L. acidophilus LB CFCS. The drawing on the right indicates the localization in enterocytes of the effects reported in panels A and B. (The time course of the bactericidal effects in panel A and the two micrographs in panel B are reproduced from reference 136.)

TABLE 1.

Overviews of in vitro antibacterial effects of probiotic Lactobacillus strains isolated from human intestinal microbiota against gastric or enterovirulent bacterial pathogens

Pathogen	Probiotic strain	Experimental conditions	Observed effect(s)	Reference(s)
Shigella	L. rhamnosus GG	Direct contact	Bactericidal	118, 119
	L. johnsonii NCC 533	Direct contact with culture or CFCS	Bactericidal	120
	L. casei DN-114 001	Direct contact	Inhibition of upregulation of proinflammatory genes	224
	L. reuteri ATCC 55730	Direct contact	Bactericidal	121
	L. acidophilus LB	Direct contact with culture or CFCS	Bactericidal	122, 125
Listeria	L. johnsonii NCC 533	Direct contact with culture or CFCS	Bactericidal	120
	L. acidophilus LB	Direct contact with culture or CFCS	Bactericidal	122, 125
Enterovirulent E. coli	L. rhamnosus GG	Direct contact	Bactericidal	119, 123
		Peptides with NPSRQERR and PDENK sequences isolated from CFCS	Bactericidal	141
		Direct contact	Decrease of Shiga toxin Stx2A mRNA	145
		Direct contact	Inhibition of adhesion onto cultured epithelial cells	177, 178, 179
		Direct contact	Inhibition of TJ lesions in cultured enterocyte-like cells	215.
		Direct contact	Inhibition of IL-8, CCL, and CXCL production	218, 219, 220, 222, 223
		Direct contact	Increased MUC2 and MUC3 mRNA, which in turn inhibited the adhesion of EPEC and EHEC	228
	L. johnsonii NCC 533	Direct contact	Inhibition of adhesion onto cultured epithelial cells	179, 180
	L. casei Shirota	Direct contact	Bactericidal	124
		Direct contact	Inhibition of adhesion onto cultured epithelial cells	177, 178
	L. reuteri ATCC 55730	Direct contact	Bactericidal	121
		Direct contact	Repression of A/E locus	146
		Direct contact	Inhibition of adhesion onto cultured epithelial cells	179
	L. acidophilus LB	Direct contact with culture or CFCS	Bactericidal	122, 125
		Direct contact with heat-treated CFCS	Conservation of bactericidal	122, 125
		Direct contact with culture or CFCS	Inhibition of adhesion onto cultured epithelial cells	125, 181, 182, 183, 187
		Direct contact with heat-treated CFCS	Conservation of inhibitory effect against adhesion onto cultured epithelial cells	125, 181, 182, 183, 187, 188
		Direct contact with untreated or heat-treated CFCS	Inhibition of structural and functional injuries at the brush borders of cultured enterocyte-like cells	125
		Direct contact with CFCS	Inhibition of TJ lesions in cultured enterocyte-like cells	207
	L. casei DN-114 001	Direct contact	Inhibition of adhesion onto cultured epithelial cells	185
		Direct contact	Inhibition of TJ lesions in cultured enterocyte-like cells	186
Vibrio cholerae	L. reuteri ATCC 55730	Direct contact	Bactericidal	121
S. Typhimurium	L. rhamnosus GG	Direct contact with culture or CFCS	Bactericidal	107, 118, 119, 126–131
		Peptides with NPSRQERR and PDENK sequences isolated from CFCS	Bactericidal	141
		Direct contact	Inhibition of adhesion onto cultured epithelial cells	178
		Direct contact with culture or CFCS	Inhibition of cell-entry into cultured enterocyte-like cells	126, 128–130
		Direct contact	Inhibition of IL-8 production	219, 220
	L. johnsonii NCC 533	Direct contact with culture or CFCS	Bactericidal	107, 120, 128, 129, 132–134
		Direct contact with culture or CFCS	Bactericidal effect of the produced hydrogen peroxide and cooperation with lactic acid	132, 133
		Direct contact with culture or CFCS	Inhibition of cell entry into cultured enterocyte-like cells	128–130, 180
	L. casei Shirota	Direct contact with culture or CFCS	Bactericidal	107, 128, 129, 135
		Direct contact	Inhibition of adhesion onto cultured epithelial cells	178
		Direct contact with culture or CFCS	Inhibition of cell-entry into cultured enterocyte-like cells	128, 129
		Direct contact with CFCS	Inhibition of flagellum swimming motility	148
		Direct contact with CFCS	Inhibition of IL-8 production	221
	L. casei DN-114 001	Direct contact with culture or CFCS	Bactericidal	107
	L. acidophilus LB	Direct contact with culture or CFCS	Bactericidal	122, 125
		Direct contact with CFCS	Inhibition of flagellum swimming motility	147
		Direct contact with culture or CFCS	Inhibition of cell entry into cultured enterocyte-like cells	136, 137, 139, 147, 182, 183
		Direct contact with CFCS	Bactericidal effect against intracellular vacuole-localized S. Typhimurium	136
		Direct contact with heat-treated CFCS	Conservation of the bactericidal effect, inhibitory effect against flagellum motility, inhibitory effect against cell entry into cultured enterocyte-like cells and against intracellular vacuole-localized S. Typhimurium	136, 137, 139, 147, 182, 183
H. pylori	L. rhamnosus GG	Direct contact with CFCS	Low bactericidal activity and loss of activity by CFCS heat treatment	139
		Direct contact with a produced bacteriocin	Bactericidal activity	151
		Direct contact	Inhibition of adhesion onto gastric cells	235
		Direct contact with CFCS	Absence of inhibitory effect against adhesion onto mucus-secreting cells	139
		Direct contact	Inhibition of IL-8 production	235, 236
	L. johnsonii NCC 533	Direct contact with a produced bacteriocin	Bactericidal activity	139, 150, 151
		GroEL protein	GroEL protein-dependent aggregation of bound cells onto cultured epithelial cells	160, 234
		Direct contact with CFCS	Inhibition of IL-8 production	149
		Direct contact with CFCS	Inhibition of flagellum swimming motility	160
	L. casei Shirota	Direct contact with CFCS	Bactericidal activity with formation of coccoid forms	149
		Direct contact with CFCS	Inhibition of urease activity	149
		Direct contact with CFCS	Inhibition of flagellum swimming motility with formation of U-shaped forms	148
	L. acidophilus LB	Direct contact with culture or CFCS	Bactericidal with formation of coccoid forms	139
		Direct contact with culture or CFCS	Inhibition of urease activity	139
		Direct contact with CFCS	Bactericidal effect against preadhering cells onto mucus-secreting cells	139
		Direct contact with CFCS	Inhibition of adhesion onto mucus-secreting cells	139
		Direct contact with heat-treated CFCS	Conservation of bactericidal activity and inhibition of urease activity	139

FIG 2.

Bactericidal effect of the probiotic L. acidophilus strain LB against gastric or enterovirulent bacterial pathogens adhering to infected cultured human intestinal cells. (A) Time course of the bactericidal effect of L. acidophilus strain LB against the wild-type Afa/Dr DAEC C1845 strain adhering to the brush borders of enterocyte-like Caco-2/TC7 cells. (B and C) Scanning electron micrographs showing the L. acidophilus strain LB-induced changes in the wild-type Afa/Dr DAEC C1845 strain adhering to the brush borders of enterocyte-like Caco-2/TC7 cells and in H. pylori strain 1101 adhering to cultured human mucus-secreting HT29-MTX cells, respectively. For panels A to C, preinfected cells were treated with L. acidophilus LB CFCS. The drawing on the right indicates the localization in enterocytes of the effects reported in panels A to C. (The time course of bactericidal effects in panel A and the two micrographs in panel B are reproduced from reference 125 with permission from BMJ Publishing Group, Ltd. The two micrographs in panel C are reproduced from reference 139.)

The bactericidal activity of L. rhamnosus GG, L. johnsonii NCC 533, L. casei Shirota, L. casei DN-114 001, and L. acidophilus LB has been shown to be entirely triggered by secreted compounds present in CFCSs (107, 122, 125, 129, 130, 138). The activity exerted seems to result from various molecules acting alone or synergistically (Table 1). In vitro, the bactericidal activity of probiotic Lactobacillus cultures has been proposed to result from the acidic pH. However, Fayol-Messaoudi et al. (107), investigating how the in vitro bactericidal activity of L. rhamnosus GG, L. casei Shirota, L. casei DN-114-001, and L. johnsonii NCC 533 CFCSs against S. Typhimurium develops, observed that the acidic pH contributed only a small part. During fermentative metabolism, lactobacilli produce organic acids as terminal products, triggering antagonistic activities against bacterial pathogens through intracellular acidification and membrane permeabilization. The main metabolic compound secreted, i.e., lactic acid, has been suspected to play a pivotal role in the bactericidal activity of probiotic Lactobacillus strains. It has been observed that the in vitro bactericidal activity of lactic acid against S. Typhimurium increased linearly with increasing lactic acid concentrations (107, 119, 129), but at the concentrations present in the probiotic Lactobacillus cultures (50 to 80 mM), it displayed no bactericidal activity (107, 122, 136, 139). The antimicrobial effect of lactic acid is not just due to the lowering of the intracellular pH. Indeed, hydrogen peroxide produced by L. johnsonii NCC 533 and other L. johnsonii strains in vitro kills S. Typhimurium (133), an effect enhanced in the presence of the membrane permeabilizer lactic acid (132). As deduced from in vitro experiments that have tested the sensitivities of secreted compounds present in L. rhamnosus GG, L. johnsonii NCC 533, and L. acidophilus LB CFCSs to a set of physical and chemical treatments and from partial isolation experiments, bactericidal activity against S. Typhimurium results from small (dialysis cutoff, 1,000 Da), nonproteinaceous compounds (120, 122, 123). The compound(s) supporting the bactericidal activity present in L. acidophilus LB CFCS is heat resistant (122). Recently, five low-molecular-weight, nonproteinaceous bactericidal compounds that are heat stable and active at acidic pH values have been found in L. rhamnosus GG CFCS, which either with or without lactic acid display antimicrobial activity against S. Typhimurium (140). Moreover, Lu et al. (141) have reported that seven heat-resistant small peptides, two of which have NPSRQERR and PDENK sequences, are present in L. rhamnosus GG CFCS and display antibacterial activity against enteroaggregative E. coli (EAEC) strain 042 and Salmonella enterica serovar Typhi. It was noted that Lactobacillus strains that produce bacteriocins that are active to kill Salmonella, Campylobacter, and E. coli have been found (142–144), but none of the six Lactobacillus strains described here seems to produce bacteriocin.

The bacterial membrane damage that accompanies the bactericidal effects produced by the CFCSs of several human intestinal microbiota Lactobacillus strains resembles the bacterial cell damage produced by antibiotics and bacteriocins (41–43) and by several intestinal epithelial AMPs (9). For example, before L. acidophilus LB CFCS-induced S. Typhimurium cell death is achieved, there is a change in the cell membrane (122) accompanied by the release of lipopolysaccharide, an increase in membrane permeability, and a loss of intracellular ATP (137).

Activity on the expression or functionality of virulence factors.

Molecules produced by probiotic Lactobacillus strains can also directly affect the expression or functionality of virulence factors of enterovirulent bacteria without affecting cell viability (Table 1). In Shiga toxin-producing E. coli O157:H7, L. rhamnosus GG has been shown to reduce the levels of Shiga toxin stx_2A mRNA (145). L. reuteri ATCC 55730 repressed the expression of the locus of the enterocyte effacement-encoded regulator involved in the attachment/effacement (A/E) lesion of enterocyte microvilli by EHEC (146). Impairment of S. Typhimurium swimming motility has been observed after treatment with CFCSs of L. acidophilus LB and L. casei Shirota as a result of membrane depolarization that affects the functionality of the flagellar motor but not flagellum expression (147, 148). The inhibition of the swimming motility results from the secretion of a small compound(s) (dialysis cutoff, 1,000 Da), which in the case of the L. acidophilus LB compound(s) was heat and trypsin resistant whereas in that of the L. casei Shirota compound(s) was heat and trypsin sensitive (147, 148).

Direct Activities against H. pylori

Bactericidal activity.

H. pylori is a member of the class Epsilonproteobacteria, composed almost exclusively of helical and curved organisms (34, 35). H. pylori colonizes the gastric tissue and causes serious disorders, ranging from mild gastritis to the onset of chronic gastric inflammation that can lead to ulcers and gastric cancer, although most infected individuals are asymptomatic. It has been documented that the human probiotic intestinal microbiota L. johnsonii NCC 533, L. casei Shirota, and L. acidophilus LB display direct antagonistic activity in vitro against the gastritis-associated bacterium H. pylori (Fig. 1 and Table 1). pH-dependent bactericidal activity against H. pylori was exhibited by L. casei Shirota CFCS (149). Similarly, direct contact with the CFCSs of L. johnsonii NCC 533 (150) or L. acidophilus LB (139) resulted in the rapid and dramatic loss (∼6 log) of H. pylori viability. Moreover, H. pylori cells were morphologically affected by treatment with L. johnsonii NCC 533 (149), L. acidophilus (139), or L. casei Shirota (148) CFCS, shifting from their characteristic helical form to U-shaped and coccoid forms (Fig. 1). Avonts and De Vuyst (151) have also shown that L. johnsonii NCC 533 and L. casei Shirota can produce bacteriocins which are active against H. pylori. It is noteworthy that other Lactobacillus strains producing bacteriocins that are also active against H. pylori have been identified (113, 152–154). The molecule(s) present in L. acidophilus LB CFCS that exerts bactericidal activity against H. pylori remain to be identified. The appearance of coccoid forms after treatment with the above-mentioned Lactobacillus strains is important since it resembles the morphological changes produced in H. pylori by antibiotics (155, 156) or colloidal bismuth subcitrate (157). Moreover, the transformation into coccoid forms is interesting in terms of pathogenesis, since even though these forms conserve the genes that code for virulence factors found in the spiral form, they are known to be less likely to colonize and induce inflammation than the corresponding spiral forms (158).

Activity against the expression and functionality of virulence factors.

Secreted compounds produced by human probiotic intestinal microbiota Lactobacillus strains have a direct effect on the expression or function of H. pylori virulence factors (Table 1). Urease is a surface protein component of H. pylori which allows it to survive within the stomach by neutralizing the gastric acidic environment (35). Both L. acidophilus LB (139) and L. casei Shirota (149) CFCSs have been shown to induce a dramatic loss of H. pylori urease activity. In contrast, L. rhamnosus GG CFCS has a lower effect on H. pylori urease activity (139). Flagellar motility together with the helical cell shape of the pathogen has been shown to be essential for the ability of H. pylori to colonize the stomach (35). H. pylori cells swim within the gastric mucus layer, propelled by a bundle of rotating flagella (159). L. johnsonii NCC 533 secretes nonproteinaceous compounds of >1,000 Da that inhibit the swimming motility of H. pylori (160). L. casei Shirota CFCS irreversibly inhibits the swimming motility of H. pylori by changing the morphology of the pathogen from its characteristic helical form with polar flagella to U-shaped and coccoid forms devoid of flagella (148). Irreversible inhibition of H. pylori swimming motility by L. casei Shirota results from the secretion of a small compound(s) (dialysis cutoff, 1,000 Da) that are heat and trypsin sensitive (148).

ACTIVITIES AGAINST THE DELETERIOUS EFFECTS INDUCED BY INFECTIOUS AGENTS AT THE INTESTINAL EPITHELIAL BARRIER

Activities against Enterovirulent Bacteria

Enterovirulent bacteria initiate their infectious process by attaching themselves to the target epithelial cells that line the intestinal epithelium by use of specialized adhesive factors (36, 161–164). For example, for interacting with the polarized host epithelial cells that line the intestinal barrier (165–167), many bacterial species move in the luminal intestinal compartment by rotating their flagella (168, 169). The adhesion of enterovirulent bacteria to the brush border of the enterocyte involves much more than just simple attachment (161). In the case of ETEC, it permits the optimal delivery of cytotoxic toxins in the vicinity of their membrane-associated receptors, followed by signaling events affecting electrolytes and fluid secretion (164). For EAEC, attachment optimizes the delivery of autotransporter toxins at the brush border (170). Intestinal bacterial pathogens are equipped with a variety of weapons that provide them with a variety of mechanisms for subverting the cellular machinery and circumventing host defenses. Pathogens have developed sophisticated ways to secrete proteins from the cytoplasmic compartment outside the bacterial cell. This allows the type III secretion systems (T3SS) of enteropathogenic E. coli (EPEC) and EHEC to insert a translocated intimin receptor into the host cell membrane, thus triggering the recruitment of actin directly underneath the attached bacteria to form pedestal structures leading to intimate attachment of the bacterium, resulting in characteristic A/E lesions on the brush border microvilli and in dramatic defects in the absorption/secretion functions (36, 162, 171, 172). Some enterovirulent bacteria have developed sophisticated strategies for altering and opening the junctional domain of the intestinal epithelial barrier (6, 173). For enteroinvasive pathogens such as Shigella (40) and Salmonella (163), adhesion initiates an orderly series of T3SS-dependent, bacterial effector-controlled molecular events within a defined area on the host cell membrane, which facilitate the formation of the dramatic actin-rich cell surface ruffles that are pivotal to the successful completion of bacterial invasion, after which the internalized bacteria adopt specific intracellular lifestyles. Some of these pathogens, such as Shigella (163) and Listeria (174), live in the cell cytoplasm, within which the bacteria move by means of actin-based motility, and thence spread into the neighboring cells via cellular membrane invaginations known as transpodia. Once other invasive pathogens, such as Salmonella, have been internalized, they take up residence within the cell cytoplasm inside large vesicles where they replicate (163, 175). In addition, emerging evidence indicates that these effectors are mimetic proteins expressing functional domains or motifs by which bacteria activate cell signaling pathways that modify signaling-regulated cell functions. Antagonistic activities of probiotic Lactobacillus strains against the structural and functional cell injuries promoted by enterovirulent pathogens at the intestinal barrier have been investigated mainly using cultured fully differentiated colon carcinoma cells that structurally and functionally mimic the human intestinal barrier and are used extensively to dissect the mechanisms of virulence of the major enterovirulent pathogens (176). Data have demonstrated that regulatory molecules produced by human intestinal microbiota Lactobacillus strains act by modulating several receptor signaling cascades that are known to have a pivotal role in the deleterious effects of enterovirulent bacteria on the structural organization and functionality of cell types lining the intestinal epithelial barrier (Table 1).

Effects at the brush border.

Inhibition of the intestinal cell association of enterovirulent bacteria involved in acute infantile and traveler's diarrhea by human probiotic Lactobacillus strains has been reported (Table 1). Inhibition (∼5 to 6 log CFU/ml) of the adhesion of ETEC expressing colonization factors CFA/I and CFA-II, EPEC, EHEC, and Afa/Dr DAEC to the brush border of cultured human enterocyte-like Caco-2 cells (176) develops in the presence of adhering L. rhamnosus GG (177–179), L. johnsonii NCC 533 (179, 180), L. casei Shirota (177, 178), L. reuteri ATCC 55730 (179), and L. acidophilus LB (125, 181–183). The interaction of S. Typhimurium with enterocyte-like Caco-2 cells was inhibited by ∼6 to 7 log CFU/ml in the presence of L. rhamnosus GG (128–130, 178), L. johnsonii NCC 533 (128, 129, 180), and L. casei Shirota (128, 129, 178) cultures and CFCSs. In contrast, L. rhamnosus GG did not affect the adhesion of S. Typhimurium to human colonic tissue specimens (184). L. casei DN-114 001 inhibited by ∼5 log CFU/ml the interaction of adherent-invasive E. coli isolated from Crohn's disease patients with cultured human intestinal epithelial cells (185) but failed to block the adhesion of EPEC (186). It is noteworthy that heat-treated L. acidophilus LB culture or CFCS conserved the antagonistic activity of the live strain against the attachment at the brush borders of Caco-2 cells of ETEC (181, 183, 187, 188), EPEC (182, 183), and Afa/Dr DAEC (125). Moreover, L. rhamnosus GG and L. casei Shirota inhibited the adhesion of enterovirulent bacteria onto mucus (177, 178, 189). The presence of carbohydrate-binding specificities in L. johnsonii NCC 533 mimicking cell surface adhesins of enteric bacterial pathogens (190) and the recently evidenced presence of pili involved in adhesion of L. rhamnosus GG (191–197) may explain the competitive inhibition exerted by these two probiotic strains against adhesion of enterovirulent bacteria.

The entry of S. Typhimurium into enterocyte-like Caco-2 cells (198, 199) was entirely abolished (decrease of ∼7 to 8 log CFU/ml) in the presence of cultures or CFCSs of L. rhamnosus GG, L. casei Shirota, L. johnsonii NCC 533, L. acidophilus LB, and L. casei DN-114 001 (120, 126, 128–130, 136, 137, 139, 182, 183). For Lehto and Salminen (131), the inhibition of internalization of S. Typhimurium into Caco-2 cells by L. rhamnosus GG resulted from the acidic pH. In contrast, experiments using noncultivated Lactobacillus culture medium acidified to pH 4.5 have demonstrated that the inhibitory activity of CFCSs of L. rhamnosus GG, L. johnsonii NCC 533, L. casei Shirota, L. acidophilus LB, and L. casei DN-114 001 did not result solely from the acidic pH (128, 129, 136, 137, 139, 182, 183). It has recently been demonstrated that a compound(s) secreted by Lactobacillus strains blocked the swimming motility of S. Typhimurium, which plays a pivotal role in enabling the pathogen to swim into the intestinal contents and interact with the intestinal epithelial cells (165, 167, 200). A heat-stable and trypsin-resistant secreted compound(s) present in the L. acidophilus LB CFCS (147) and a heat- and trypsin-sensitive compound(s) present in the L. casei Shirota CFCS (148) by the blockade of S. Typhimurium delayed the penetration of the invasive bacteria into enterocyte-like Caco-2/TC7 cells.

The entry of Salmonella into intestinal cells is followed by a characteristic intracellular lifestyle, including the presence of live bacteria within large intracellular vacuoles (201). S. Typhimurium is a typical invasive enteropathogen that develops a mechanism of microbial “nutritional virulence” which, through gaining access to host nutrients in infected tissues, controls its growth, virulence, disease progression, and infection (202). Internalized S. Typhimurium cells live in intracellular vacuoles where they have adapted to the host cell environment by expressing versatile catabolic pathways to exploit multiple host nutrients for bacterial growth (175). In Caco-2/TC7 cells preinfected with S. Typhimurium, the treatment of the cells with L. acidophilus LB CFCS resulted in a dramatic and irreversible decrease in the intracellular level of S. Typhimurium (a decrease of ∼4 to 5 log CFU/ml) (136) (Fig. 3). Moreover, transmission electron microscopy observation of the S. Typhimurium cells that remained in the intracellular vacuoles revealed cells displaying the morphology typical of dead cells (136) (Fig. 3). This observation is of interest, because intracellular enterovirulent bacteria are known to be resistant to antibiotics and can constitute a population of persistent and dormant pathogens (203).

The brush border of enterocytes is constituted by a regular array of microvilli, the membrane of which is endowed with hydrolases, such as sucrase-isomaltase (SI), alkaline phosphatase (AP), lactase-phloridzin hydrolase, maltase-glucoamylase aminopeptidase N, and dipeptidylpeptidase IV (DPP IV), and transporters, such as sodium/glucose cotransporter 1, GLUT1, GLUT2, GLUT3, and GLUT5 hexose transporters, peptide transporter 1, H⁺-coupled dipeptide transporter, cholesteryl ester transfer protein, and Na⁺/H⁺ exchanger isoforms (204). The protection of brush border-associated intestinal functions has been demonstrated only for L. acidophilus LB. The enterovirulent Afa/Dr DAEC, which promotes the destruction of the enterocyte brush border and, in turn, a dramatic loss of brush border-associated functions (205), has been used to examine the protective effect of L. acidophilus strain LB. The killing of Afa/Dr DAEC adhering to enterocyte-like Caco-2/TC7 cells, seen after treatment with L. acidophilus LB CFCS, resulted in the maintenance of a normal F-actin brush border cytoskeleton (125). Moreover, when Afa/Dr DAEC-infected intestinal cells were treated with L. acidophilus LB CFCS at a concentration that did not affect the viability of Afa/Dr DAEC, the expression of brush border-associated functional hydrolases SI, DPP IV, and AP and fructose transporter GLUT5 was normal, in contrast to the dramatic loss of expression observed in untreated Afa/Dr DAEC-infected cells (125) (Fig. 4). The secreted autotransporter toxin, Sat, belonging to the subfamily of SPATE toxins (170), expressed by Afa/Dr DAEC strain C1845 (206) promotes an increase of fluid dome formation in Caco-2/TC7 cell monolayers by modifying the transcellular passage of fluids. L. acidophilus LB CFCS treatment of Caco-2 cell monolayers results in the disappearance of the Sat-induced fluid dome formation, indicating a regulatory effect on the intestinal transcellular pathway of fluids (207).

FIG 4.

Normal expression of structural or functional brush border-associated proteins in L. acidophilus LB-treated, Afa/Dr DAEC strain C1845-infected cultured human enterocyte-like Caco-2/TC7 cells. (A) Transmission electron micrographs show the well-ordered microvilli in uninfected cells and the disappearance of the brush border in C1845-infected cells. (B) Confocal laser microscopy scanning examination of immunolabeling of brush border-associated structural F actin and functional sucrase-isomaltase (SI) in uninfected, C1845-infected, and L. acidophilus strain LB-treated C1845-infected Caco-2/TC7 cells. Micrographs on the left show the normal mosaic pattern of distribution of fluorescein isothiocyanate (FITC)-labeled F actin (green) and rhodamine isothiocyanate (RITC)-labeled SI (red) in uninfected Caco-2/TC7 cells. Central micrographs show the disappearance of FITC-labeled F actin (green) and RITC-labeled SI (red) located centrally in the cells and the persistence of the proteins at the cell-to-cell contacts. The micrographs on the right show that L. acidophilus strain LB-treated C1845-infected cells conserve the normal mosaic pattern of the FITC-labeled F-actin (green) and RITC-labeled SI (red) distribution. For panel B, cells were apically infected with C1845 alone or in the presence of L. acidophilus strain LB CFCS (0.5-fold concentrate, which does not induce a bactericidal effect). (The micrographs in panels A and B are reproduced from reference 125 with permission from BMJ Publishing Group, Ltd.)

It has been clearly demonstrated that the structural and functional lesions triggered by some enterovirulent bacteria result from virulence factors, including bacterial effectors secreted by the T3SS, and toxins, which hijack the cell signaling pathways that control the polarized organization of the cell cytoskeleton and deregulate the activities of functional membrane-associated proteins that have specific intestinal functions (162–164, 205, 208). Lactobacillus secreted compounds have been found to downregulate the expression of virulence factors acting at the brush border (145, 146). Other antagonistic activities observed using enterocyte-like cellular models also certainly result from regulatory effects by these secreted compounds at the cellular signaling pathway level. It is noteworthy that similar regulatory effects have been observed experimentally for several probiotic Lactobacillus strains when intestinal functions were impaired in a noninfectious context (209–213).

Effects at the epithelial junctional domain.

The intestinal barrier is kept closed by three intercellular junctional complexes: tight junctions (TJs), adherent junctions, and desmosomes (214). In particular, the TJs form a highly regulated structure that acts as a “fence” separating the apical and basolateral membrane domains of polarized cells, thereby segregating the various functional proteins in each of the domains. The TJs also function as a “gate” which allows paracellular vectorial transport to occur across the epithelial cell barrier. The junctional domain of the intestinal epithelium is targeted by enterovirulent bacteria (173). Only L. rhamnosus GG and L. casei DN-114 001 have been shown to have cytoprotective effects against the enterovirulent bacterium-induced structural and functional injuries at the intestinal junctional domain produced by enteric pathogens (Table 1). In T84 cell monolayers, Johnson-Henry et al. (215) observed that the EHEC-induced changes in electrical resistance, dextran permeability, and distribution and expression of claudin 1 and zonula occludens 1 (ZO-1) are antagonized by live L. rhamnosus GG but not by the heat-inactivated bacteria. In T84 cells, L. casei DN-114 001 abrogates the EPEC strain E2348/69-induced increase in paracellular permeability and rearrangements of ZO-1 in a dose-dependent manner (186). The antagonistic activity of L. acidophilus LB CFCS against the Sat-induced increase in paracellular permeability and formation of fluid-formed domes in enterocyte-like Caco-2/TC7 cell monolayers (207) probably includes an effect at the TJs, since the toxin affects the structural TJ organization (206). Consistent with this hypothesis, it has been observed that L. acidophilus LB culture protects TJs of cultured human intestinal HT-29 cells by counteracting the aspirin-induced delocalization of structural TJ-associated ZO-1 protein (211).

Activation of host epithelial defense responses.

Enterovirulent bacteria have the capacity to induce a host-controlled inflammatory response which includes release of inflammatory cytokines such as interleukin-6 (IL-6), IL-8, IL-1β, tumor necrosis factor alpha (TNF-α), and TNF-β. IL-8 is involved in the transmigration of polymorphonuclear leukocytes across the intestinal barrier, which initiates deleterious inflammatory cellular lesions (216). Other pathogen-induced dramatic proinflammatory responses are more deleterious for the host; for example, the highly harmful Shigella, causing bacillary dysentery in humans, has the capacity to cause the inflammatory destruction of the intestinal epithelium (40). Probiotic Lactobacillus strains have been shown to antagonize the pathogen-induced production of proinflammatory cytokines (75). In addition, probiotic Lactobacillus strains are able to increase the production of molecules generated by host intestinal epithelial cells which are major players in the first line of host defenses against enterovirulent bacteria, such as AMPs and mucins (217).

In enterocyte-like Caco-2 cells, the NF-κB-dependent flagellin-induced increase in IL-8 production was antagonized by both live and UV-inactivated L. rhamnosus GG (218) (Table 1). In cultured human enterocyte-like HT-29 and crypt colonic T84 cells, L. rhamnosus GG decreased the production of IL-8 after V. cholerae, Salmonella, and EHEC infection but, surprisingly, did not significantly alter the Shigella-induced IL-8 production (219, 220). L. casei Shirota CFCS decreased Salmonella enterica serovar Enteritidis-induced IL-8 production and, in addition, promoted the expression of cytoprotective heat shock protein (Hsp) 70 in Caco-2 cells (221). L. rhamnosus GG suppressed the expression of CCL20 and CXCL10 triggered in Caco-2 cells by effector molecules, peptidoglycan or flagellin, of enterovirulent E. coli (222). In Caco-2BBe cells, in which EHEC infection induces the upregulation of proinflammatory genes, preincubation of the pathogen with L. rhamnosus GG prior to infection reduced the EHEC-induced upregulation of the CXCL1, CXCL8, and NF-κB1A genes (223). Tao et al. (210) identified a heat-stable, low-molecular-weight peptide as the L. rhamnosus GG factor inducing the p38 and Jun N-terminal protein kinase (JNK) mitogen-activated protein kinase (MAPK)-dependent expression of cell regulatory Hsps 25 and 72 in cultured mouse colonic YAMC cells. L. casei DN-114 001 triggered an NF-κ-dependent downregulation of the transcription of genes encoding proinflammatory effectors and adherence molecules in Shigella flexneri-infected Caco-2 cells (224). Moreover, a DNA microarray analysis conducted with distal duodenal mucosa biopsy samples from human patients receiving L. rhamnosus GG revealed that 334 genes were upregulated, including genes that are involved in immune and infectious responses, and that 92 genes were downregulated (225). In mice, L. casei Shirota enhanced the expression of genes acting in defense and immune responses (226). Altogether, these results indicated that these Lactobacillus strains produced effectors having regulatory activities on the host signaling pathways activated in response to enterovirulent infection and on the host signaling pathways involved in intestinal mucosal defenses.

Only L. rhamnosus GG has been shown to promote the production of intestinal mucus. L. rhamnosus GG mediates the upregulation of epithelial mucin MUC2 and MUC3 mRNAs or proteins in Caco-2 cells and HT-29 cells (227, 228), which is accompanied by a concomitant inhibition of adhesion of EPEC and EHEC (228).

Effects against Rotavirus

Human rotavirus is associated with 450,000 deaths each year, mainly in developing countries and among children under 5 years of age (229). Infection is characterized by a spectrum of responses that can vary from asymptomatic to mild or severe symptoms and can result in a lethal dehydrating illness. Via either the entire virus or specific structural and nonstructural proteins, including the enterotoxin NSP4, the rotavirus induces substantial structural and functional intestinal mucosa lesions that lead to secretory diarrhea (229). Antirotavirus activities of probiotic Lactobacillus strains have been poorly investigated in vitro. Secreted compounds produced by the probiotic strains L. rhamnosus GG, L. johnsonii NCC 533, and L. acidophilus LB do not directly affect the rotavirus, since the pretreated rhesus rotavirus (RRV) strain replicates normally, in the same way as untreated RRV, in the African green monkey MA104 epithelial cells used classically to maintain infective rotavirus strains in culture (unpublished data). It has been reported that soluble factors released by L. casei DN-114 001 block the infection of cultured human mucus-secreting HT29-MTX cells by rotavirus RF and WA strains (230). In cultured pig and human epithelial cells infected with rotavirus, Maragkoudakis et al. (231) observed that the rotavirus-induced release of reactive oxygen species was decreased in the presence of either L. rhamnosus GG or L. casei Shirota. In a nontransformed porcine jejunum epithelial cell line, the presence of L. rhamnosus GG resulted in a reduced rotavirus-induced IL-6 response (232). Rotavirus induces signaling-dependent structural and functional lesions at the brush border and junctional domains of human enterocyte-like Caco-2 cells similar to those observed in intestinal biopsy specimens from children with rotavirus-associated acute diarrhea (233). It has been difficult to investigate experimentally the impact of probiotic Lactobacillus strains with known antirotavirus therapeutic effects on the rotavirus-induced structural and functional lesions in Caco-2 cells because of the cellular toxicity of Lactobacillus CFCSs when the cells are infected long enough for all the replications and stages of the intracellular lifestyle of the rotavirus to occur (unpublished data). To overcome this problem, further investigations can be conducted using fractionated Lactobacillus CFCSs.

Effects against H. pylori

Probiotic Lactobacillus strains display antagonistic activities against the H. pylori cell association or cell responses accompanying H. pylori infection (Table 1). Secreted GroEL of L. johnsonii NCC 533 has the capacity to generate the aggregation of H. pylori (234), a phenomenon observed on the surface of H. pylori-infected cultured human gastric epithelial cells (160). L. rhamnosus GG inhibits the adhesion of H. pylori onto AGS gastric cells, an effect that is abolished by heat treatment of the Lactobacillus strain (235). When human mucus-secreting HT29-MTX cells were infected with H. pylori, the adhesion of the pathogen was inhibited in a dose-dependent manner by L. acidophilus LB CFCS (139). Moreover, L. acidophilus LB CFCS treatment resulted in the death of the adhering H. pylori, and any remaining adhering H. pylori cells displayed lower urease activity and, when observed by scanning electron microscopy, appeared to have undergone lysis as determined by cell morphology (139) (Fig. 2). In contrast, L. rhamnosus GG CFCS-treated H. pylori adheres normally to HT29-MTX cells and displays normal urease activity (139). Conversely, L. rhamnosus GG enhanced the H. pylori-induced barrier injury following prolonged incubation (236). In human adenocarcinoma AGS cells, L. johnsonii NCC 533 (149) and L. rhamnosus GG (235, 236) decreased the H. pylori-induced IL-8 production.

ACTIVITIES IN ANIMAL INFECTION MODELS

Several antibacterial activities of probiotic Lactobacillus strains observed in vitro have been also observed in animal infectious models (Table 2). In addition, activities of certain probiotic Lactobacillus strains against rotavirus infection have been examined in rotavirus-infected animals. These studies have been conducted in classical rodent infectious models, including conventional or axenic rodents. It was interesting to note that some other animal models are promising to study probiotic activities. Citrobacter rodentium-infected conventional mice mimicked the deleterious structural and functional cellular lesions of the human diarrhea-associated EPEC and EHEC (237, 238). The conventional streptomycin-pretreated mice infected with S. Typhimurium established by Hardt and coworkers (239, 240) mimicked the S. Typhimurium-induced cell deleterious effects. The impact of probiotics on the pathogen-induced deleterious effects on intestinal epithelial cells and immune responses has rarely been investigated in these two models (241, 242). Caenorhabditis elegans is an emerging model to study microbial pathogenesis (243) and also for studying microorganisms that have implications for human health (244). This model seems useful for the elucidation of the mechanisms by which probiotics combat enteric pathogens (245–248) and may influence quality of life.

TABLE 2.

Overview of antibacterial effects of probiotic Lactobacillus strains isolated from human intestinal microbiota against gastric or enterovirulent bacterial pathogens in animal infection models

Pathogen	Probiotic strain	Animal model	Observed effect(s)	Reference
Listeria	L. casei Shirota	Conventional rats	Decrease of translocation from the intestine	254
Enterovirulent E. coli	L. acidophilus LB	Suckling mice	Increase of survival rate	249
Campylobacter jejuni	L. acidophilus LB	Axenic mice	Decrease of mucosa-associated bacteria and bacterial translocation	253
S. Typhimurium	L. rhamnosus GG	Conventional mice	Decrease of viable bacteria in feces	130
		Axenic mice	Decrease of intestinal colonization and increase of survival rate	130
		Newborn rat pups	Decrease of intestinal colonization	250
	L. johnsonii NCC 533	Conventional mice	Decrease of viable bacteria in feces	120
		Axenic mice	Decrease of intestinal colonization and increase of survival rate	120
		Conventional mice	Decrease of bacteria associated with the intestinal tissues or present in the intestinal contents	128
	L. casei Shirota	Conventional mice	Decrease of bacteria associated with the intestinal tissues or present in the intestinal contents	128
		Fosfomycin-treated mice	Decrease of intestinal colonization and translocation of a multiresistant strain	135
	L. acidophilus LB	Conventional mice	Decrease of fecal excretion	122
H. pylori	L. johnsonii NCC 533	Conventional mice	Decrease of stomach colonization	149
		Mongolian gerbils	Decrease of stomach colonization and gastritis	160
	L. casei Shirota	Conventional mice	Decrease of stomach colonization and gastritis	149
	L. acidophilus LB	Conventional mice	Decrease of stomach colonization, urease activity in stomach, and gastritis	139

Bacterium-Infected Animals

Once human intestinal microbiota Lactobacillus probiotic strains are established in the gastrointestinal tracts of axenic animals, they reduce the association of gastric or enterovirulent bacterial pathogens with the gastric and intestinal epithelia, inhibit the translocation of pathogens, lower cytopathic effects on epithelial cells, stimulate cellular defense responses, and increase the survival of infected animals. Administration of heat-treated L. acidophilus LB cultures increases the survival of ETEC-infected suckling mice compared to that of their untreated counterparts (249). In newborn rat pups, L. rhamnosus GG decreases intestinal colonization by enteroinvasive E. coli (250). In rabbits, administration of L. casei Shirota reduces the severity of diarrhea, lowers Shiga toxin-producing E. coli intestinal colonization, and decreases the intestinal concentrations of the Stx1 and Stx2 toxins and lowers the associated histological damage (251). In pigs infected with an ETEC strain, administration of L. rhamnosus GG reduces the duration of diarrhea (252).

Establishment of L. rhamnosus GG in the guts of axenic mice challenged with S. Typhimurium strain C5 results in lower cecal colonization levels, a reduced translocation rate of the pathogen, and survival rates higher than those of noncolonized mice (130). L. johnsonii NCC 533 displayed its antibacterial activity in axenic mice orally infected by S. Typhimurium, since the level of Salmonella was lower in the feces of the treated conventional mice, and the establishment of L. johnsonii NCC 533 in the guts of axenic mice resulted in improved survival (120). Moreover, decreased intestinal colonization and translocation of C. jejuni caused by a heat-treated L. acidophilus LB culture has been observed in axenic mice (253). It is possible but not demonstrated that when probiotic Lactobacillus cells colonize the intestinal epithelia of axenic mice, the observed in vivo antagonistic effect against Salmonella results in a local production of antibacterial molecules having a bactericidal activity observed in vitro (120, 130).

In conventional mice orally infected with S. Typhimurium strain C5, oral administration of L. rhamnosus GG leads to a lower level of the pathogen in the feces than in that of untreated mice (130). In conventional rats orally receiving L. casei Shirota, there is a reduction in the number of L. monocytogenes organisms found in the stomach, cecum, and feces, accompanied by reduced translocation from the intestine (254). In conventional mice, L. casei Shirota colonizing epithelial cells lining the stomach, small intestine, and colon and present in the intestinal contents reduced the levels of S. Typhimurium associated with the epithelial cells of the duodenum and jejunum or present in the intestinal contents (128). It can be deduced from these experiments that when probiotic Lactobacillus cells colonize epithelial cells in different parts of the conventional mouse intestine (128), the adherent bacteria create a barrier effect against Salmonella similar to what is observed in vitro when probiotic bacteria attached to the brush borders of enterocyte-like Caco-2 cells in culture inhibit attachment and entry of Salmonella into cells (107). L. acidophilus LB CFCS treatment displayed antibacterial activity in the conventional C3H/He/Oujco mouse infected with S. Typhimurium strain C5, since lower levels of fecal excretion of S. Typhimurium were observed than in untreated mice (122). This suggests that molecules present in L. acidophilus LB CFCS that exert in vitro bactericidal activity against this pathogen (122, 136, 137) are able to produce their bactericidal effect in vivo. It is noteworthy that in conventional mice infected with S. Typhimurium there was an absence of an antagonistic effect when the infected mice were treated with uncultivated deMan-Rogosa-Sharpe broth acidified at pH 4.5 (107, 122), demonstrating that the pH effect observed in vitro (131) is irrelevant for probiotic activity in vivo.

L. casei Shirota displays an antagonistic activity against the multidrug-resistant S. Typhimurium strain DT104 in a fosfomycin-treated murine model, since the intestinal growth of the pathogen and its subsequent lethal extraintestinal translocation were inhibited in a manner correlated with intestinal colonization by the Lactobacillus strain (135). This observation is interesting in the light of the currently increasing public health problem of multidrug-resistant enterovirulent bacteria (255, 256) and persister cells (257, 258).

Administration of L. johnsonii NCC 533 to H. pylori-infected mice reduced the inflammatory infiltration into the stomach lamina propria, although the level of colonizing H. pylori was not decreased (149). Similarly, H. pylori colonization and gastritis were significantly less intense in L. johnsonii NCC 533-treated Mongolian gerbils than in untreated animals (160). Oral administration of L. casei Shirota in mice previously infected with H. pylori resulted in significantly lower levels of H. pylori colonization in the antrum and body mucosa and a reduction in the gastric mucosal inflammation compared with those in the H. pylori-infected untreated group (149). Oral treatment of conventional mice with concentrated L. acidophilus LB CFCS antagonized Helicobacter felis infection in mice by inhibiting gastric colonization and decreasing H. felis urease activity, which in turn prevented the development of gastric inflammation. These effects were not suppressed by subjecting L. acidophilus LB CFCS to heat treatment (139). L. rhamnosus GG has been found to enhance gastric ulcer healing in a rat model by decreasing cell apoptosis and increasing angiogenesis (259).

The exact mechanism(s) by which the probiotic Lactobacillus strains exerted their antagonistic activities in vivo against infecting gastric or enterovirulent pathogens remains to be identified. Two reports have presented evidence for specific mechanisms developed by a probiotic Lactobacillus strain and a probiotic E. coli strain. It has been postulated that probiotic strains may compete against bacterial pathogens in intestinal ecological niches. It is well demonstrated that bacteriocins produced by Gram-positive and Gram-negative probiotic bacteria inhibit strains closely related to the producer strain (43). Consequently, it is thought that these peptides may assist the producers to compete within their specific intestinal ecological niches. This has recently been demonstrated for a Gram-negative probiotic strain. Deriu et al. (260) have shown that when the probiotic Gram-negative E. coli Nissle strain (261) was administered to mice with S. Typhimurium-induced colitis, there was an impressive reduction in S. Typhimurium colonization as a result of successful competition of the probiotic strain with Salmonella for iron acquisition via non-enterobactin-mediated pathways. Moreover, the production of a bacteriocin in vivo by L. salivarius UCC118 can efficiently protect mice against infection by the invasive L. monocytogenes (262). Such ecological competition between the six Lactobacillus probiotic strains examined and Gram-negative enteric bacterial pathogens remains to be demonstrated.

Rotavirus-Infected Animals

There are few studies that have analyzed the protective effect of the six probiotic Lactobacillus strains against rotavirus infection in animal models. Neonatal mice and rats provide reliable animal models for studying the kinetics of viremia, spread, and pathology of rotavirus and also immune responses during a primary rotavirus infection (229). In rotavirus-infected rodent models, L. rhamnosus GG reversed the rotavirus-induced increase in intestinal barrier permeability (263), reduced both the duration and the severity of the resulting diarrhea and the histopathological changes and virus load in the intestine in combination with antirotavirus antibodies (264), and shortened the duration of diarrhea and decreased epithelium vacuolation in the jejunum (265). In germfree suckling rats receiving milk fermented by the L. casei strain DN-114 001 and infected with rotavirus, the frequency of stools and severity of diarrhea were reduced, as were intestinal cell lesions, including cell vacuolation and changes in the morphology of the intestinal villi (266). In rotavirus-infected mice, oral administration of L. reuteri DSM 17938 reduced the duration of diarrhea and decreased the accompanying intestinal cell lesions (267).

CLINICAL STUDIES

The major forms of the six probiotic Lactobacillus strains available to consumers consist principally of yogurt or bottled fermented milk, but capsules or sachets containing the lyophilized strain in powder form for rehydration prior to consumption and chewable tablets are also available as dietary supplements (268). In addition, several probiotic strains are now included in dietetic products for children (269). The RCTs conducted in children and adults using human probiotic Lactobacillus strains have been intervention studies in which various vehicles containing the probiotic Lactobacillus strains have been used (Tables 3 and 4). Those conducted to investigate the anti-infectious therapeutic efficacy of L. rhamnosus GG have been carried out mainly using lyophilized powders of Lactobacillus cells or fermented milk containing the probiotic strain. RCTs designed to evaluate the anti-infectious clinical efficacy of the probiotic strains L. casei Shirota and L. casei DN-114 001 have been conducted using Lactobacillus-containing milk drinks. Drinkable whey-based or commercial dairy products containing L. johnsonii NCC 533 have been used to evaluate in RCTs the anti-H. pylori therapeutic efficacy of this probiotic strain. For the L. acidophilus LB, the forms used in RCTs are sachets or capsules containing a combination of 10 billion heat-treated and lyophilized L. acidophilus LB cells and 160 mg of 2-fold-concentrated CFCS (Lacteol).

TABLE 3.

Overview of therapeutic effects against acute or persistent diarrhea in RCTs of probiotic Lactobacillus strains isolated from human intestinal microbiota

Probiotic strain	Cause of infection	No. (age group) of patients	Treatment	Reported clinical effect(s) (control/treated)a	Reference
L. rhamnosus GG	Not documented	71 (children)	Fermented milk, 109 CFU/ml for 5 days	Shortened duration of acute diarrhea (2.4/1.4 days)	296
	Rotavirus	42 (children)	Freeze-dried powder, 1010 CFU/ml twice daily for 5 days	Lowering of no. of patients with acute diarrhea (43/10 at day 3 of treatment)	297
	Rotavirus	49 (children)	Freeze-dried powder, 1010-1011 CFU/ml twice daily for 5 days	Shortened duration of acute diarrhea (2.7/1.8 days) accompanied by stimulation of rotavirus-specific IgA antibody responses	299
	Rotavirus	40 (children)	Freeze-dried powder, 1010-1011 CFU/ml twice daily for 2 days	Lowering of no. of patients with acute diarrhea (75/31% at day 2 of treatment)	300
	Rotavirus	26 (children)	Freeze-dried powder, 1010-1011 CFU/ml twice daily for 2 days	Shortened duration of acute diarrhea (3.3/1.9 days)	301
	Rotavirus (28%) and enterovirulent bacteria (21%)	123 (children)	Freeze-dried powder, 5 × 109 CFU/ml twice daily for 5 days with ORS	Shortened duration of acute diarrhea (3.7/2.7 days with a decrease of frequency of stools	302
	Rotavirus	123 (children)	Freeze-dried powder, 5 × 109 CFU/ml twice daily with ORS	Shortened duration of acute diarrhea (30.4/17.7 h)	303
	Rotavirus and enterovirulent bacteria	204 (children)	3.7 × 1010 CFU/ml once daily for 6 days/wk for 15 mo	Prophylactic effect on the incidence of diarrhea	304
	Rotavirus	287 (children)	Freeze-dried powder, 1010 CFU/ml with ORS until diarrhea stopped	Shortened duration of acute diarrhea (76.6/57.2 h)	305
	Rotavirus	81 (children)	6 × 109 CFU/ml twice daily for duration of hospital stay	Prophylactic effect on the incidence of diarrhea	331
	Rotavirus (27.5%) and enterovirulent bacteria (36%)	179 (infants)	Fermented milk, 109 CFU/ml per day with ORS	No significant differences in duration of diarrhea, rate of treatment failure, and proportion of unresolved diarrhea	314
	Not documented	192 (children)	6 × 109 CFU/ml per day with ORS	Shortened duration of acute diarrhea (115.5/78.5 h)	306
	Not documented	662 (children)	6 × 1010 CFU/ml per day with ORS	No effect on duration of diarrhea (6.8/6.6 days)	313
	Rotavirus and enterovirulent bacteria	235 (children)	6 × 1010 CFU/ml per day with ORS	Shortened duration of persistent diarrhea (9.2/5.3 days)	307
	Not documented	559 (children)	6 × 1010-1012 CFU/ml per day with ORS	Decrease of frequency of stools and duration of diarrhea	308
	Not documented	229 (infants)	109 CFU/ml per day for 10 days	No effect on duration of diarrhea or numbers of stools	315
	Rotavirus and enterovirulent bacteria	64 (children)	5 × 109 CFU/ml three times per day for 3 days with ORS	No change in duration of diarrhea, total stools, or diarrhea score	316
L. casei DN-114 001	Not documented	287 (children)	Fermented milk, 108 CFU/ml daily	Shortened duration of acute diarrhea (8.0/4.3 days)	338
	Not documented	928 (children)	Fermented milk, 108 CFU/ml daily	Decrease of no. of patients with acute diarrhea (22.0/15.9%)	337
L. reuteri DSM 17938	Rotavirus	66 (children)	1010-1011 CFU/ml daily for up 5 days	Shortened duration of acute diarrhea (2.9/1.7 days)	317
	Rotavirus	40 (children)	107-1010 CFU/ml daily for up to 5 days	Shortened duration of acute diarrhea (2.5/1.7 days) and no. of patients with acute diarrhea (80/48%)	282
	Rotavirus	74 (children)	4 × 108 CFU/ml daily	Decrease of no. of patients with acute diarrhea (82/45%)	319
Heat-treated L. acidophilus LB cultureb	Rotavirus	73 (children)	6 sachets with ORSc	Shortened duration of acute diarrhea (74.0/42.9 h)	325
	Enterovirulent bacterium-induced acute diarrhea	80 (children)	6 sachets during 35 h with ORSc	Shortened duration of acute diarrhea (30.4/8.2 h)	327
	Enterovirulent bacterium-induced acute diarrhea	80 (children)	8 sachets during 96 h with ORSc	Shortened duration of acute diarrhea (63.4/39.5 h)	147

^aNumber of days of acute diarrhea or duration of diarrhea or number or percentage of patients with acute diarrhea.

^bLacteol.

^cOne sachet consisted of 10 billion heat-treated and lyophilized L. acidophilus LB cells and 160 mg of 2-fold-concentrated neutralized CFCS.

TABLE 4.

Overview of therapeutic effects against H. pylori infection in RCTs of probiotic Lactobacillus strains isolated from human intestinal microbiota

Probiotic strain	No. (age group) of patients	Treatment	Associated treatment	Reported clinical effects (control/treated)	Reference
L. rhamnosus GG	120 (adults)	109 CFU/ml twice daily for 14 days in conjunction with triple therapy	Pantoprazole, 40 mg; clarithromycin, 500 mg; tinidazole, 500 mg	No improvement of H. pylori eradication rate	382
	43 (adults)	6 × 109 CFU/ml twice daily for 7 days after 1 wk of triple therapy	Rabeprazole, 20 mg; clarithromycin, 500 mg; tinidazole, 500 mg (daily)	No improvement of H. pylori eradication rate	383
	47 (adults)	Probiotic prepn including L. rhamnosus GG (109 CFU/ml) twice daily during triple therapy and once a day during 3 wk after cessation of triple therapy	Lansoprazole, 30 mg; clarithromycin, 500 mg; amoxicillin, 1,000 mg (daily)	No improvement of H. pylori eradication rate	384
	83 (children)	109 CFU/ml twice daily for 7 days in conjunction with triple therapy	Amoxicillin, 25 mg; clarithromycin, 10 mg; omeprazole, 0.5 mg (twice daily)	No improvement of H. pylori eradication rate	385
L. johnsonii NCC 533 (La1)	20 (adults)	50 ml of drinkable, whey-based, CFCS four times daily for 14 days in conjunction with monotherapy	Omeprazole, 20 mg four times daily	Decrease in breath test values	150
	53 (adults)	180 ml fermented milk twice daily for 3 wk	Clarithromycin, 500 mg during the last 2 wk of milk therapy	Improvement of clarithromycin-induced H. pylori eradication by decreasing H. pylori density and gastritis in stomach.	373
	50 (adults)	Fermented milk (LC1a) twice daily for 16 weeks	No anti-H. pylori therapy	No cure of H. pylori infection, decrease of inflammatory score (6.3/5.3)	374
	12 (adults)	80 ml fermented milk (LC1) (107 CFU/ml) 8 times daily for 2 weeks	No anti-H. pylori therapy	Decrease in breath test values	375
	100 (children)	2 × 80 ml fermented milk (LC1) (107 CFU/ml) for 4 wk	No anti-H. pylori therapy	Decrease in breath test values	376
	136 (children)	80 ml fermented milk (LC1) (107 CFU/ml) daily for 2 weeks	No anti-H. pylori therapy	Decrease in breath test values	377
L. casei Shirota	14 (adults)	Drinkable fermented milkb (108 CFU/ml) for 3 wk	No anti-H. pylori therapy	Decrease in breath test values	378
	64 (adults)	Drinkable fermented milkb (108 CFU/ml) for 8 wk	Clarithromycin, amoxicillin, and omeprazole	Improvement of triple-therapy-induced H. pylori eradication	379
L. casei DN-114 001	86 (children)	Drinkable fermented milkc for 14 days	Omeprazole, amoxicillin, and clarithromycin	Improvement of triple-therapy-induced H. pylori eradication (57.5/84.6%)	380
L. reuteri DSM 17938	40 (adults)	Chewable tabletsd (108 CFU/ml) daily for 28 days	After probiotic treatment: rabeprazole (20 mg twice a day) plus amoxicillin (1 g, twice daily) for 5 days followed by rabeprazole (20 mg twice a day), clarithromycin (500 mg, twice daily), and tinidazole (500 mg, twice daily) for the next 5 days	Decrease in breath test values at the end of probiotic treatment, no difference in H. pylori eradication rates after triple therapy or not	386
Heat-treated L. acidophilus LB culturee	120 (adults)	1 sachetf daily for 7 days and 3 days after cessation of triple therapy	Rabeprazole (20 mg), clarithromycin (250 mg), and amoxicillin (500 mg) for 7 days	Improvement of triple-therapy-induced H. pylori eradication (72/88%)	381
	84 (adults)	2 sachets daily for 4 wk	Omeprazole (20 mg) and amoxicillin (1,000 mg)	No improvement of the low H. pylori eradication rate of double therapy	387
	120 (children)	1 sachet daily for 8 wk	No anti-H. pylori therapy	No decrease in breath test values	388

^aFermented milk LC1 (Nestlé Company).

^bYakult drink (Yakult Company).

^cActimel drink (Danone Company).

^dBioGaia AB.

^eLacteol.

^fOne sachet consisted of 10 billion heat-treated and lyophilized L. acidophilus LB cells and 160 mg of 2-fold-concentrated L. acidophilus LB spent culture medium.

Human probiotic Lactobacillus strains survived after oral administration in laboratory animals and were present at appreciable levels in the gastrointestinal tract. For example, in conventional or germfree mice, L. rhamnosus GG cells have been found in the different segments of the gut (130), and L. johnsonii NCC 553 and L. casei Shirota cells efficiently colonize the different parts of the conventional mouse intestine (128). In human adults and infants receiving L. rhamnosus GG, the strain has been found in fecal and/or intestinal mucosa biopsy samples (270–275). L. casei Shirota cells have been found in the feces of human volunteers receiving fermented milk containing L. casei Shirota (276–278). L. casei DN-114 001 cells have been found in ileal and fecal samples of human volunteers receiving fermented milk containing L. casei DN-114 001 or the derived L. casei strain DN-114 001Rif (279, 280). In human adult volunteers and infants, L. reuteri ATCC 55730 cells have been found in fecal samples (275, 281, 282). It is noteworthy that probiotic Lactobacillus strains do not colonize the gut and that the persistence of a strain is limited to the duration of the administration of the probiotic preparation. Indeed, as soon as the probiotic strain is no longer consumed, probiotic cells disappear from the intestinal tract of the consumer.

Therapeutic Effects against Various Forms of Acute Diarrhea

The World Health Organization (WHO) has defined diarrhea as the occurrence of three or more loose or watery stools within a 24-h period. Diarrhea is described as acute when it began less than 14 days before and as persistent when its duration is 14 days or more. Rotavirus is worldwide a major cause of severe diarrhea and of diarrhea leading to mortality in children. Other diarrhea-associated viruses include astroviruses, noroviruses, sapoviruses, and enteric adenoviruses. The major bacterial pathogens involved in diarrhea are ETEC, Salmonella, Shigella, Yersinia, Campylobacter, and Vibrio cholerae. The major causes of acute infectious diarrhea due to enteric viruses or enterovirulent bacteria vary from place to place, and there are relatively few vaccines available to prevent human mucosal infections (283). Rotavirus vaccines (284–286) are effective in preventing rotavirus-associated acute diarrhea in all countries (287–289) and particularly in those in which high mortality is attributable to rotavirus-induced diarrhea (290, 291). Several interventions, including improved sanitation and hand hygiene and the promotion of breast-feeding, have been clinically used for the prevention of rotavirus-induced acute diarrhea in children (292). Such treatment is intended to prevent the dehydration and nutritional damage and to reduce the length and severity of acute diarrhea. To prevent dehydration and nutritional damage in children and infants with acute diarrhea, the therapeutic regimen recommended by the WHO, the European Society for Pediatric Infectious Diseases (ESPID), and the European Society for Pediatric Gastroenterology, Hepatology, and Nutrition (ESPGHAN) is to provide an oral rehydration solution (ORS) and to continue feeding (293). Although ORS administration effectively mitigates dehydration, it has no effect on the length and severity of diarrheal episodes. A variety of adjuvant therapies have been used to treat the infectious origin of acute diarrhea (294). Luminally acting antidiarrheic drugs are confined mainly to the gastrointestinal tract, but several of these agents can cause adverse effects outside the gastrointestinal tract. Moreover, several problems resulting from the use of antibiotic therapy to treat diarrhea associated with enterovirulent bacteria have been observed, including the emergence of resistance to single antibiotics, followed by the emergence of multiple resistance and, if broad-spectrum antibiotics are used, a adverse effect on the intestinal microbiota, which can in turn favor the emergence of C. difficile-associated diarrhea (295).

Acute infectious diarrhea.

Human probiotic Lactobacillus strains and drugs derived from several of these strains have been clinically investigated to treat acute infectious diarrhea (Table 3). In RCTs investigating their therapeutic efficacy, they have been generally used in association with ORS (293). In RTCs, L. rhamnosus GG alone or in combination with ORS has demonstrated therapeutic efficacy in significantly reducing by at least half the duration of acute diarrhea mainly resulting from rotavirus infection in children (296–310). Two meta-analyses have been conducted to specifically examine the RCTs conducted with L. rhamnosus GG (311, 312). The conclusion is that L. rhamnosus GG had no impact on the total stool volume but did reduce the duration of diarrhea, particularly in the context of a rotavirus etiology. In contrast, four trials in children and infants with enterovirulent bacterium- or rotavirus- and enterovirulent bacterium-induced diarrhea showed no decrease in the daily number of stools or duration of diarrhea after treatment with L rhamnosus GG (313–316). Multicenter and other RCTs in children and infants have shown that the administration of L. reuteri DSM 17938 (L. reuteri ATCC 55730) shortened the duration of acute diarrhea and lowered the diarrhea relapse rate (282, 317–319).

Heat treatment of probiotic Lactobacillus strains results in a marked decrease of antimicrobial activity (320) compared to that of the parental live strains (135, 139, 215, 298). Heat-treated Lactobacillus strains cannot be considered to be probiotics (321), because the definition of probiotics states that the strains are alive (69). However, a heat-treated Lactobacillus strain may retain the pharmacological “probiotic” activities of the parental live strain when the substances produced by the strain that underpin the pharmacological activities are able to resist the treatments to which they are subjected during the pharmaceutical manufacturing process, including heat treatment (322). L. acidophilus strain LB is one of the few probiotic Lactobacillus strains which display this characteristic since, as we have already mentioned, it produces heat-stable antimicrobial compounds. Lacteol has been authorized for use as an antidiarrheal agent by the French Safety of Medicines Agency (marketing authorizations AMM 3400933073602, AMM 3400934847837, and AMM 3400934840224) (Forest Laboratories, Inc., New York, NY). In three RCTs conducted with Lacteol in combination with ORS in infants and young children with rotavirus-associated diarrhea, the treatment shortened by at least half the duration of diarrhea and accelerated the reappearance of the first normal consistent and formed stool compared to the placebo combined with ORS (323–325).

While the numerous experimental in vitro and in vivo data reported above demonstrated the bactericidal activity of L. rhamnosus GG against enterovirulent bacteria, it is both surprising and intriguing to note that therapeutic efficacy was absent in RCTs conducted with this probiotic strain in children with established bacterially induced diarrhea (302, 305, 314, 326). Consistent with the in vitro and in vivo antibacterial activities reported above for live and heat-treated L. acidophilus strain LB against enteroadherent and enteroinvasive bacteria (122, 137, 147, 181, 182), two RCTs have demonstrated the therapeutic efficacy of Lacteol in combination with ORS in decreasing the frequency and duration (reduced by at least half) of enterovirulent bacterium-associated acute watery diarrhea in children between the ages of 1 and 4 years (207, 327). Moreover, in adults over 16 years of age with chronic bacterium- or parasite-associated diarrhea, a prospective, randomized, multicenter clinical trial has demonstrated that administering Lacteol led to the recovery of consistent stools in 81% of the treated patients (328).

Nosocomial infections.

It is noteworthy that the few prevention clinical studies conducted with probiotic Lactobacillus strains have given disappointing results, even though the regular consumption of probiotic food products is often recommended on health grounds, including to prevent intestinal infections and to improve quality of life (69, 111, 329). Contrasting results have been obtained in trials investigating the effect of probiotic Lactobacillus strains in preventing nosocomial infections. Hojsak et al. (330) observed a reduced risk of gastrointestinal infections in children after L. rhamnosus GG treatment. L. rhamnosus GG significantly reduced the occurrence of nosocomial diarrhea in infants, particularly in cases resulting from rotavirus infection (331). In contrast, two RCTs in children have shown that the consumption of L. rhamnosus GG was ineffective in preventing nosocomial diarrheal episodes and also failed to reduce the number of days with acute diarrhea (332, 333). The occurrence of the primary episodes of acute diarrhea was significantly retarded in children receiving a probiotic drink containing L. casei Shirota compared to untreated control children (334). In two multicenter RCTs investigating the effect of L. casei DN-114 001 on overall common infectious diseases (CIDs), including gastrointestinal tract infections, it has been observed that the consumption of a fermented dairy product containing L. casei DN-114 001 was associated with a shorter duration of CIDs in aging patients (335, 336). In contrast, a trial in healthy children receiving yogurt containing L. casei strain DN-114 001 found that there was no reduction of the incidence of acute diarrhea but that the period of acute diarrhea was lower than that in children receiving traditional yogurt (337, 338). A study in children found no difference between the probiotic L. reuteri DSM 17938 group and the placebo group with regard to the incidence of rotavirus infection, the incidence and duration of acute diarrhea, the incidence of chronic diarrhea, the duration of hospital stay in days, and the frequency of the need for rehydration (339).

Traveler's diarrhea.

Acute diarrhea induced by enterovirulent bacteria is common among travelers and tourists. Three RCTs investigating the efficacy of human probiotic Lactobacillus strains in preventing intestinal infectious episodes in travelers gave negative results. Two trials were conducted in adults and infants traveling in southern Turkey (given 2 × 10⁹ L. rhamnosus GG cells daily for 1 to 2 weeks of travel) (340) and in adult American travelers visiting South America, the subcontinent of India, Central America, the Middle East, West Africa, and North Africa (given one capsule [2 × 10⁹ L. rhamnosus GG cells] once daily, starting 2 days prior to departure and continuing throughout the travel) (273). The results show that the occurrence of acute diarrhea was quite similar in the placebo group and the L. rhamnosus GG group (273, 340). Similarly, another study demonstrated a lack of prevention of infectious diarrhea by Lacteol (one capsule consisting of 10 billion heat-treated and lyophilized L. acidophilus LB cells and 160 mg of 2-fold-concentrated L. acidophilus LB spent culture medium twice daily from 1 day before their departure to 3 days after their return) compared to placebo in adult travelers visiting West, East, Central and North Africa, Oceania, South America, Asia, Central America/the Caribbean, and the Middle East from 2001 to 2004 (341). These studies do not prejudge the efficacy of other probiotic Lactobacillus strains to prevent traveler's diarrhea.

Therapeutic Effects against C. difficile-Associated Diarrhea

C. difficile is a Gram-positive anaerobic bacterium that is one of the most important causes of antibiotic-associated diarrhea in the developed world, and discontinuation of antibiotics generally promotes a cure of mild infections (342). The incidence and severity of C. difficile infection appear to be on the increase, leading to significant morbidity and mortality and placing a considerable economic burden on health care systems, particularly in Europe and North America (343, 344). Recent severe C. difficile disease and higher mortality rates have been associated with the emergence of strains of increased virulence, or “hypervirulent” isolates, belonging to the BI/NAP1/027 category and which are fluoroquinolone resistant (345). C. difficile produces a number of putative virulence factors, including two members of the large clostridial cytotoxin family, known as toxin A and toxin B, both of which induce intense colonic inflammation and dramatic structural and functional epithelial tissue damage followed by a rapid fluid loss into the intestinal lumen, leading to diarrhea and an adaptive immune response (342, 346).

The antagonistic activity exerted by probiotic Lactobacillus strains against C. difficile has received little experimental investigation (118, 347). Despite this lack of experimental evidence, the use of probiotic Lactobacillus strains to treat C. difficile-induced diarrhea has been extensively promoted in a high number of reviews (348–352), but it remains controversial (353). Meta-analyses have evidenced that probiotic strains are associated with a reduction in C. difficile-associated diarrhea (102, 354–357). There were a small number of reports showing persuasive therapeutic effects against C. difficile-induced diarrhea after the administration of a single probiotic Lactobacillus strain. RCTs have shown therapeutic efficacy of L. rhamnosus GG to treat C. difficile-induced diarrhea (102, 358–361). However, several reports have shown a lack of therapeutic efficacy against C. difficile-induced diarrhea for L. rhamnosus GG (362). There is a lack of reports documenting the experimental and therapeutic effects of the probiotic L. casei Shirota YIT9029, L. acidophilus LB, L. johnsonii NCC 533, L. casei DN-114 001, and L. reuteri DSM 17938 strains against C. difficile infection.

Hell et al. (363) have suggested that a combination of different probiotic strains may be more useful to reduce C. difficile infection. It is noteworthy that fecal microbiota transplantation (FMT) is an emerging treatment to restore the normal microbial homeostasis after antibiotic agents have disrupted the intestinal microbiota, leading to C. difficile-associated diarrhea. Interestingly, Russell et al. (364) recently reported the success of FMT in a young child showing infection caused by C. difficile strain BI/NAP1/O27. This pathogenic strain is refractory to metronidazole, vancomycin, rifaximin, and nitazoxanide antibiotic treatments or treatment with L. rhamnosus GG. Kassam et al. (365) conducted a meta-analysis of 11 studies concerning patients with C. difficile infection and treated with FMT and concluded that “FMT holds considerable promise as a therapy for recurrent C. difficile infection, but well-designed, RCTs and long-term follow-up registries are still required.”

Meta-Analyses

The clinical trials discussed above all have methodological flaws, including having groups of patients that were extremely different in size and with regard to age, differences in the concentrations of probiotic bacteria administered and in the duration of administration, and differences in the vehicles containing the probiotic strains. Moreover, different outcomes to evaluate or quantitatively measure the therapeutic efficacy have been used, such as the period of acute diarrhea or the time to the reappearance of a consistent stool. To mitigate these problems, meta-analyses have attempted to examine RCTs conducted with a wide variety of probiotic lactic acid-producing strains, such as L. rhamnosus strain GG, strains of L. acidophilus, L. casei, L. plantarum, L. fermentum, and L. bulgaricus, Bifidobacterium strains BB12 and BB536, B. infantis and B. lactis strains, and VSL3 (viable lyophilized bacteria of L. paracasei, L. plantarum, L. acidophilus, L. delbrueckii subsp. bulgaricus, B. longum, B. breve, B. infantis, and a Streptococcus salivarius subsp.), and other, non-lactic acid-producing probiotic strains (355, 366–368). Moreover, Allen et al. (369) have examined the Cochrane Infectious Diseases Group's trials register (2010), the Cochrane Controlled Trials Register (2010), MEDLINE (1966 to 2010), EMBASE (1988 to 2010), and lists of references in reviews for analyzing RCTs and quasi-RCTs that compare probiotic treatment with no probiotic administration or treatments with or without placebo administration in patients with acute diarrhea. The authors of these meta-analyses conclude that “probiotic strains used alongside rehydration therapy appear to have clear beneficial effects in shortening the duration and reducing stool frequency in acute infectious diarrhea, although the size of the effect varied considerably between studies.” Other meta-analyses concluded that probiotic Lactobacillus strains significantly reduce C. difficile-associated diarrhea and acute diarrhea of diverse causes (102, 357, 370). In contrast, two meta-analyses showed that probiotic Lactobacillus strains have no effect on traveler's diarrhea (355, 369). Moreover, there is an absence of convincing evidence that probiotics are effective for treating persistent diarrhea in children (371).

Therapeutic Effects against Infectious Gastritis

Combinations of drugs, including a proton pump inhibitor (PPI) plus three antibiotics or a PPI plus bismuth plus two antibiotics, provide the best therapeutic efficacy for treating gastric H. pylori infection (372). However, the effectiveness of many of these treatments has been compromised by an increase in the resistance of H. pylori to antibiotic treatment. Several human probiotic Lactobacillus strains have been shown to improve the efficacy of the anti-H. pylori triple therapy in well-conducted RCTs (Table 4). Oral administration of L. johnsonii NCC 533 CFCS in omeprazole-treated patients induced a decrease in breath test values (150). Oral administration of acidified milk containing L. johnsonii NCC 533 in clarithromycin-treated patients induced a decrease in H. pylori levels and reduced inflammation in the antrum and corpus (373). The intake of fermented milk containing L. johnsonii NCC 533 without taking antibiotics resulted in the persistence of H. pylori but with a decreased inflammatory score (374) or breath test values (375). It was noted that heat treatment of L. johnsonii NCC 533 affects its therapeutic efficacy (376, 377). Patients receiving a fermented milk drink containing L. casei Shirota displayed lower urease activity (378). Eradication of H. pylori after triple treatment with clarithromycin, amoxicillin, and omeprazole was better when a fermented milk drink containing L. casei Shirota was combined than with the triple therapy (379). A multicenter RCT showed that with adjunct of a drinkable fermented milk containing L. casei DN-114 001 improved the rate of eradication of H. pylori in children treated with amoxicillin, clarithromycin, and omeprazole (380). An RCT carried out by Canducci et al. (381) has shown that in patients with H. pylori infection and receiving a triple therapy based on rabeprazole, clarithromycin, and amoxicillin for 7 days, supplementation with Lacteol significantly increased the eradication of H. pylori (87% of patients) compared to the triple therapy alone (72% of patients).

Negative results have been reported in four RCTs investigating the therapeutic efficacy of L. rhamnosus GG against H. pylori-induced infection in conjunction with pantoprazole, clarithromycin, and tinidazole (382), rabeprazole, clarithromycin, and tinidazole (383), lanzoprazole, clarithromycin, and amoxicillin (384), or amoxicillin, clarithromycin, and omeprazole (385). An RCT in adults has shown that the administration of tablets containing L. reuteri DSM 17938 reduced the occurrence of dyspeptic symptoms but did not improve H. pylori eradication (386). De Francesco et al. (387) showed that Lacteol did not ameliorate the low eradication rate of an anti-H. pylori therapy composed of a proton pump inhibitor and amoxicillin. Gotteland et al. (388) evaluated Lacteol treatment alone versus triple therapy in an open, prospective, randomized trial and observed that the treatment was relatively ineffective in eradicating H. pylori in contrast to the triple therapy (lanzoprazole, clarithromycin, and amoxicillin).

CONCLUDING REMARKS

Despite the fact that a large number of Lactobacillus strains have been isolated from human microbiota and tested in vitro for their probiotic activities, such as their adhesiveness to intestinal cells and mucus and antagonistic activities against gastroenteropathogens, only a few of these strains have displayed the properties needed for industrial development to generate biotherapeutic anti-infectious agents (389). The experimental data and RCTs analyzed above provide evidence of the following: (i) L. rhamnosus strain GG in fermented milk and lyophilized forms, fermented milk containing L. casei strain DN-114 001, the lyophilized form of L. reuteri DSM 17938, and Lacteol containing lyophilized, heat-treated L. acidophilus LB cells and concentrated spent culture medium are therapeutically effective against rotavirus-associated acute diarrhea when used in addition to ORS; (ii) only Lacteol has been shown in RCTs to demonstrate therapeutic efficacy against enterovirulent bacterium-induced acute diarrhea when used in addition to ORS; and (iii) fermented milk containing L. johnsonii NCC 533, L. casei Shirota, or L. casei DN-114 001 and Lacteol are therapeutically effective in antagonizing gastritis-associated H. pylori or improving the eradication rate of the triple therapy. In contrast, there is no clinical evidence that consumption of these probiotic strains prevents the occurrence of intestinal infectious episodes. The mechanism underlying these anti-infectious effects appears to be multifaceted. Indeed, a large set of mechanisms have been identified, including (i) inhibition of the growth of pathogens or a direct bactericidal effect exerted by secreted molecules, (ii) inhibition of expression of virulence genes coding for virulence factors or interference with the cell membrane expression or secretion of virulence factors, including toxins, (iii) competitive exclusion of pathogenic bacteria by competition for binding sites, (iv) inhibition of the signaling-dependent structural and functional deleterious effects triggered by virulence factors in host intestinal cells, and (v) stimulation of antimicrobial host intestinal cell responses. It was noticed, interestingly, that experimental data and RCTs have provided evidence that Lactobacillus species isolated from the human vaginal microbiota developed anti-infectious properties against vaginosis-associated bacterial pathogens, including Prevotella bivia and Gardnerella vaginalis, by mechanisms similar to that of human intestinal microbiota Lactobacillus strains (390). On the basis of the experimental and clinical evidence of the production of strain-specific derived bioactive molecules by human microbiota Lactobacillus strains having concentration-dependent pharmacological activities such as antibiotic-like activities and host cell regulatory activities, Shahanan et al. (391) have proposed that these bacterial molecules can be collectively termed “pharmabiotics.”

The increasing occurrence of drug-resistant bacterial pathogens is currently one of the major public health challenges. Although medical practice has limited the development and spread of pathogens, the rapid global emergence of pathogens resistant to most conventional antibiotics and the recent emergence of multidrug-resistant pathogens constitute an increasing major public health threat (256, 392). Moreover, antibiotic resistance results also from the presence of dormant bacteria and persisters exhibiting an extraordinary tolerance to antibiotics resulting from transient growth inhibition and inactivity of essential cell functions controlled by a multiplicity of bacterially regulated mechanisms (258, 393). It is becoming increasingly evident that in order to reduce the rate of appearance of antibiotic-resistant strains, better management and a more reasonable use of antibiotics both in humans and in animal husbandry are necessary. However, the lack of new antibiotics coming onto the market highlights the need to discover new antimicrobial agents and, to do this, to develop innovative strategies. Host intestinal antimicrobial molecules, including epithelial cell-produced AMPs and bacteriocin and nonbacteriocin molecules generated by microbiota Gram-positive and Gram-negative bacteria, are potential sources of new antimicrobial molecules. AMPs such as cationic AMPs, which have the advantages of broad-spectrum activity and a lower propensity to select for drug resistance phenotypes, have been intensely investigated as a potential source of complementary antimicrobial agents (9, 10, 394, 395). However, many pathogens are able to overcome host AMPs. They often kill bacteria by membrane disruption, which is a highly effective and rapid mechanism (396). Importantly, since human and bacterial membranes are similar, it is often difficult to target the bacteria and avoid damaging the patient's cells. To make it possible to use membrane-disrupting AMPs in antibacterial chemotherapy in a viable fashion, their safety needs to be investigated. In a recent commentary on the development and spread of antibiotic resistance in bacteria, Bush et al. (397) said, “The use of probiotics is likely to become more important in years to come as microbiological studies of the roles of gastrointestinal bacterial populations (the human microbiome) lead to the identification of those bacterial genera and species that have key roles in human health and disease. Such advances may well lead to the use of bacteria and their products as specific therapeutics.” It has recently been convincingly argued that bacteriocins are potential alternatives to traditional antibiotics; however, the anti-infectious therapeutic efficacy of bacteriocins in human beings remains to be demonstrated in properly designed clinical studies (41–43). Moreover, combinations of bacteriocins with old antibiotics, which have unfortunately displayed toxic side effects in human beings, are currently being investigated with a view to reducing their toxicity by lowering the concentrations generally used, accompanied by an enhancement of their therapeutic efficacies.

So far, attempts to purify the antimicrobial nonbacteriocin compounds secreted by most probiotic Lactobacillus strains have failed, and their structures remain unknown. It is essential to purify these compounds in order to identify their exact mechanisms of action and to resolve the structures of these new, natural antimicrobial agents before there can be any hope of developing an innovative drug design strategy involving them. However, this does not seem to be an easy task, because the antimicrobial activity is exercised by compounds that become very unstable once they are extracted from the culture medium. In addition, it has been difficult in some cases to purify these molecules because bactericidal activity disappears in the final stages of the purification process, as the active molecules probably act synergistically with lactic acid exerting a bacterial membrane permeabilization activity (unpublished data). Genomic analyses (398, 399) have recently been conducted in order to understand the core mechanisms that control and regulate probiotic Lactobacillus bacterial growth, survival, signaling, and fermentative processes and, in some cases, potentially underlying probiotic activities (74, 400, 401). Genomic analyses have been published for only four of the six probiotic strains examined here: L. rhamnosus GG (402, 403), L. johnsonii NCC 533 (404, 405), L. casei Shirota (403), and L. casei DN-114 001 (403). Genes coding for several strain-specific properties have been identified. Genes of L. johnsonii NCC 533 code for the cell factors that account for the particular properties of adhesion/colonization (406, 407). A cluster of genes in L. casei Shirota code for the synthesis of the high-molecular-mass components that can be relevant for the bacterial whole-cell-induced, cell contact-dependent immune modulation (408). Genes of L. rhamnosus GG code for the pili involved in adhesion (191–194), for the formation of biofilm (409), and for enzymes involved in the biosynthesis of extracellular polysaccharides (410). Genes of L. casei DN-114 001 code for adaptation of metabolic properties involved in intestinal colonization (411). Currently, the comparative genomic studies have not given pertinent information about the antibacterial molecules produced by the probiotic Lactobacillus strains. A strategy involving mutant strains has been successfully developed to identify structures, antibacterial activities, genetic systems, and biosyntheses, as well as the mechanisms of action, of bacteriocins of Gram-negative E. coli, i.e., microcins and colicins (49–51). The establishment of mutant strains of Lactobacillus in order to identify the genes coding for anti-infectious molecules has rarely been attempted. Mutants of L. reuteri strains have been used for examining the activity of the antibiotic 3-hydroxypropionaldehyde, also known as reuterin (412), against enteric bacterial pathogens (146). Recently, using wild-type L. salivarius strain UCC118 and a stable mutant in which the gene coding for the production of the broad-spectrum class IIb bacteriocin Abp118 had been deleted, the probiotic activities of the strain have been characterized, including the protection of mice against L. monocytogenes infection (262) and changes in the composition of mouse and pig intestinal microbiota (413). A major shortcoming in the field of anti-infectious Lactobacillus molecules is the absence of comparative genomic analyses coupled with a strategy involving mutants.

Biographies

Vanessa Liévin-Le Moal is a microbiologist who trained at the Faculty of Pharmacy, University Paris-Sud at Châtenay-Malabry, France, and received her master's degree in 2001 and her Ph.D. in microbiology in 2005. She studied the role of intestinal host defense mechanisms against enterovirulent bacterial infection in INSERM Units 510 and 756. She is currently an associate professor at the Faculty of Pharmacy at Châtenay-Malabry, working on antiparasitic therapeutics at CNRS Unit 8076 BioCis.

Alain L. Servin received his Doctorat d'Université from University Orsay Paris-Sud, France, in 1973 and his Doctorat d'Etat es Sciences in Molecular Pharmacology from University Paris VI in 1987. In 1980 he joined the French National Institute of Health and Medical Research (INSERM), working on biophysics and cellular pharmacology at INSERM Institute Saint Antoine Hospital, Paris, and INSERM Unit 178 at Paul Brousse Hospital at Villejuif. He was research director at INSERM and head of INSERM units between 1990 and 2010 at the Faculty of Pharmacy Châtenay-Malabry, University Paris-Sud. He began his interest in cellular microbiology in 1990. He and his colleagues focus their research on the molecular and cellular mechanisms of Afa/Dr DAEC and rotavirus pathogenesis. He also studies the role of intestinal and vaginal microbiota in controlling enteric and vaginal infections and holds patents in this area. He is currently an associate-researcher at CNRS Unit 8076 BioCis, Faculty of Pharmacy Châtenay-Malabry. Dr. Servin's homepage can be found at http://cvscience.aviesan.fr/cv/827/alain-l-servin.