Normal ora: diversity and functions Lynne V. McFarland
From the Department of Medicinal Chemistry, University of Washington, and Biocodex, Inc. Seattle, WA, USA
Correspondence to: Lynne V. McFarland, Ph.D., Biocodex, Inc., 1910 Fairview Avenue E, Suite #208, Seattle, WA 98102, USA. Fax: 206-323-2968; E-mail: email@example.com
Microbial Ecology in Health and Disease 2000; 12: 193–207
Recent research into the therapeutic use of living organisms has focused attention on the impact of various disruptive factors (antibiotics, surgery, immunosuppression) and their impact on the host’s normal ora. This review covers what is meant by ‘normal ora’, how the microecology differs by the niche in the body, type of diet, age and health status. In addition, the functions and tools used to investigate normal ora will be explored. The functions of the normal ora include digestion of substrates, production of vitamins, stimulation of cell maturation, stimulation of the immune system, aid in intestinal transit and colonization resistance. A variety of factors can disrupt the normal ora including age, diet, stress, illness and exposure to antibiotics. Research involving microecologic populations is difŽcult due to the challenge of unraveling the complex dynamics within a usually inaccessible niche, but progress is being made.
In the past, the role that microorganisms played in the normal functioning of the body was not appreciated. In the early 1900’s when Dr. Metchnikoff was credited with the discovery of the importance of intestinal ora, other physicians felt that the colon was totally unnecessary and often surgically removed them from their patients (1). The colon was described as ‘a poisonous cess-pit infecting the body with rheumatism, tuberculosis, cancer and other diseases’. Today, we know that normal ora is a dynamic and complex mixture of microbes that have diverse func- tions including digestion of essential nutrients, maturation of intestinal physiology, stimulation of immune system, systemic effects on blood lipids and the inhibition of harmful bacteria. Current research techniques allow better evaluation of the speciŽc bacterial and fungal microbes within various body sites. This paper revisits some popular conceptions about normal ora and updates them with regard to recent scientiŽc Žndings.
WHAT IS ‘NORMAL FLORA’?
Although the term ‘normal ora’ is commonly used, it is really a misnomer. Microbial ora has spatial and tempo- ral complexity that differs by individual, body niche, age, geographic location, health status, diet and type of host (2, 3). Even within the same individual, the composition of the microbial ora can vary according to changes in diet, stress, sexual behavior, medication, hormonal changes and other host-related factors (3–7). With this caveat in mind, the Želd of ‘normal ora’ can be examined for common
predominant types of ora present within body niches and shared functional traits.
The adult human body contains 1014 cells, of which only 10% compose the body proper and 90% are accounted for by members of the microora (8). The predominating types of species in humans differ according to the body niche (oral cavity, skin, vagina, stomach, ileum, colon or urinary tract), as shown in Fig. 1 (3, 9–13). Normal ora found in the oral cavity has been found to vary by the area sampled (tooth enamel, tongue, gingivital surface, saliva) and the state of periodontal health (14, 15). The oral cavity contains a wide mixture of microbes, which are mainly anaerobic bacteria. Gagliardi et al. sampled normal ora in the healthy esophagus during upper endoscopy proce- dures in 30 patients and the predominant ora was found to be Streptococcus viridans (16). Lactobacilli and alpha- hemolytic Streptococcus species are frequently isolated on tonsils of healthy children (15, 17). Lactobacillus species that have the ability to adhere to mannose-containing receptors, such as L. plantarum, have a distinct advantage in surviving in the oral cavity (18). Results from different studies proŽling predominant ora may be difŽcult to compare as subject age, sampling techniques (washing of the surface, aspirates or biopsies), diet, sampled location and microbiological assay techniques may produce signiŽ- cantly different results. Fewer bacteria exist in the stomach (usually below 103:g due to acidic lumen). Helicobacter pylori has been found in patients with peptic ulcers and gastric neoplasia, but is also found in 60% of healthy hosts, which casts suspicion that this microbe is always a cause for gastric disease (19–21). The concentration of
© Taylor & Francis 2000. ISSN 0891-060X Microbial Ecology in Health and Disease
L. V. McFarland194
microbes increases as progression is made down the intesti- nal tract: small intestine (º104: ml contents), to 106–107: ml at the ileocecal region and 1011–1012:g in the colon. The intestinal microora consists of 1011 organisms:gram of feces with over 500 different species, ranging in concen- trations from 102–1011:ml luminal contents (22). Although the variety of organisms is complex, generally there are more anaerobic microbes than aerobes (9). The develop- ment of new techniques and genetic probes has allowed better characterization of the types of organisms that comprise the normal intestinal ora. Franks et al. devel- oped six 16S rRNA-targeted oligonucleotide probes that can detect at least 66% of the anaerobic fecal ora in humans (23). When these probes were used to characterize the ora in nine healthy human volunteers, Bacteroides species accounted for 20% of the total fecal population, Clostridium coccides and Eubacterium rectale accounted for 29%, Gram-positive bacteria accounted for 12% and BiŽdobacterium species accounted for 3% of the fecal ora. These probes may prove very valuable in the characteriza- tion of the microecologic proŽles, but more research is needed (as discussed later).
Previous studies have reported that ora in the healthy vagina is typically a mixture of aerobic Lactobacillus spe- cies, including L. jensenii, L. acidophilus or L. rhamnosus (9). Two strains of lactobacilli (L. crispatus and L. jensenii ) protect vaginal surfaces by producing H2O2, which inhibits the colonization of pathogenic anaerobes and mycoplas- mas associated with bacterial vaginosis, Neisseria gonor- rhoeae or other sexually transmitted diseases (5, 13).
Vaginal ora has been shown to change over the menstrual cycle (4), sexual activity and hygiene habits (7, 24), and use of intravaginal microbicides (such as nonoxynol-4) (25).
However, studies show most healthy women (52–78%) have transient changes in vaginal ora (4, 7, 24). Some more recent prospective studies have shown only a minor- ity (22–26%) of healthy women had a lactobacilli-predom- inant ora (5, 24). Thus, the characterization of vaginal ora is open to debate and requires additional prospective studies in well-deŽned populations of healthy women.
The process of the development of normal ora starts at birth. It is thought that colonization begins during parturi- tion when the neonate’s intestine is seeded with mostly Gram-positive facultative anaerobes from the vaginal mi- croora during delivery (22, 26, 27). Whether the vaginal ora in the last trimester is similar to vaginal ora when the woman is not pregnant is not known. However, Kar- voven et al. documented that the vaginal ora collected from mothers after delivery was the same as ora found in the stools of neonates (27). Neonates born by caesarian section usually acquire their Žrst microbes from the envi- ronment of the hospital nursery (26). Neonates are quickly colonized by facultative anaerobes (E. coli and Streptococ- cus), reaching concentrations of 108 to 1010:g feces within 1–2 days (9, 26). Previous studies reported that anaerobic ora do not become established until the second month of life (9). The hypothesis for this observation was that when the newly seeded facultative microbes grew and produced a more anaerobic environment, this established an anaero- bic environment suitable for anaerobes (9). However, these
Fig. 1. Predominant ora in different niches of the human body. Compiled from references: (3), (9–13).
Normal ora description 195
Microbial-associated characteristics (MAC) relating to different functions of the normal intestinal microecology
1. Digestion of metabolizable substrates 2. Colonization resistance 3. Production of vitamins 4. Development of attachment sites 5. Induces development of the immune system 6. Production of exogenous enzymes 7. Stimulation of intestinal transit 8. Maturation and turn-over of intestinal cells
Compiled from references: (22), (36–49).
FUNCTIONS OF THE NORMAL FLORA
The functions of the normal ora have been called ‘mi- croora-associated characteristics’ (MAC) by several re- searchers (22, 36–38). These MAC (Table I) include digestion of metabolizable substrates, colonization resis- tance, vitamin production, mucosal cell development, im- mune system stimulation and intestinal transit regulation (36–49). An important role of the intestinal ora is the digestion of metabolizable substrates. A major source of nutrients is the upper intestinal tract and the available substrates may include dietary Žbers, starches, oligosac- charides, sugars, some lipids and proteins. Another source of nutrients is within the colon itself and includes endoge- nous mucins, sloughed epithelial and enterocyte tissues, bacterial debris, bile acids and cholesterol. The types of metabolizable substrates are key in determining the type of ora present in the colon. The main products of bacte- rial digestion of non-absorbed dietary carbohydrates are short chain fatty acids (SCFAs). The SCFAs produced in the largest quantities by the normal ora include acetic, propionic and butyric acids (50). Acetic and propionic acids are rapidly absorbed and are a major source of energy, in addition to stimulating salt and water absorp- tion (22, 51). Butyric acid has several functions including the maintenance of the integrity of the colonic epithelial layer, as a chief energy source for these cells and regulat- ing cell growth and differentiation (22, 51). Treem et al. studied fecal SCFA in patients with Inammatory Bowel Disease (IBD) (52). Fecal homogenates from 10 patients with IBD and 10 age-matched controls were compared as to the ability of the fecal homogenate to produce SCFA. Patients with IBD produced less total SCFA, less acetate acid and less propionic acid. Since normal ora is respon- sible for the production of SCFA, this study may indi- rectly link normal ora disruption with IBD. However, Treem et al. did not characterize the ora responsible, nor compare the microbial proŽles of patients with IBD and controls. In some rare instances, the role for a member of the normal ora has been identiŽed for a speciŽc func- tion. Oxalobacter formigenes is a normal anaerobe in the colon responsible for regulating the breakdown of oxalic acid and it has been demonstrated that patients with recurrent calcium oxalate kidney stone formation prob- lems do not harbor this useful bacterium (53). Normal ora are also involved in the conversion of primary bile salts. The identiŽcation of the genes encoding conjugated bile salt hydrolases has been identiŽed in a strain of L. johnsonii (54). Usually the role for a speciŽc member of the normal ora is not limited to one function, rather the role is usually to interact with other members of the ora for more complex functions (such as colonization resis- tance).
conclusions were based on standard culturing techniques (9, 28, 29). Harmsen et al. compared newer techniques (FISH, 16S rRNA probes) with culturing techniques and found high counts were found by culturing only after 8–9 days, but FISH detected anaerobes by the second day (30). In addition, another study indicated that neonates can acquire strict anaerobes, such as C. difŽcile, within the Žrst 2 days of stay on a neonatal ward (31). Therefore, the theory that colonization of the neonatal intestine by anaer- obes is dependent upon facultative bacteria producing an anaerobic environment may be erroneous.
The type of diet largely inuences the types of ora in pre-weaning infants. Several studies have reported that infants who are breast-fed have higher concentrations of biŽdobacteria as compared to formula-fed infants (28, 29, 32, 33). Breast milk contains low protein content and high levels of oligosaccharides and glycoproteins, which are considered to be growth factors for biŽdobacteria (26). Formula-fed infants have a more complex microbiota con- sisting of BiŽdobacterium, Bacteroides, Clostridium and Streptococcus species (22, 28). However, other researchers have not found a signiŽcant difference in breast-fed and formula-fed infants (34).
Historically, the characterization of neonatal fecal ora has relied on culturing techniques, which are not able to detect non-culturable species and may not be able to distinguish different microbial populations. More recent techniques using 16S rRNA probes, which are able to amplify the bacterial genes and are followed by sequence analysis, are more sensitive than the traditional culturing techniques (23). Harmsen et al. used 16S rRNA probes and conŽrmed previous Žndings that breast-fed infants have predominant populations of biŽdobacteria in their stools and formula-fed neonates had a more complex mixture of organisms (32). By 2 years of age, children have a similar complexity and range of microbes as adults (35). Differences in breast-fed and formula-fed microbial proŽ- les found by different studies may be due to sampling techniques, time of collection, assay method, hospital prac- tices or by the buffering capacity of different formulas (34).
L. V. McFarland196
Colonization resistance is the Žrst line of defense against invasion by exogenous, pathogenic organisms or indige- nous opportunistic organisms and the normal ora is responsible for this formidable task (8, 55). Even though the focus here is the intestinal tract, it should be remem- bered that colonization resistance plays an important role at other body sites (oral, skin, vagina, etc.). Colonization resistance is a dynamic phenomenon that may differ dra- matically by microbial species, type of host, diet and other host factors. Colonization resistance has been found to be an extremely effective natural barrier against such patho- gens as C. difŽcile, Salmonella, Shigella, Pseudomonas, pathogenic E. coli strains, Candida albicans and others (8, 56–58).
Autochthonous ora (indigenous species that normally inhabit a given ecologic niche) may become disrupted and allochthonous species (not normally present) are then able to colonize the site. Colonization with allochthonous bac- teria may or may not result in disease, as the colonizing organism may not be a pathogenic species. Early evidence for colonization resistance arose from the observation that, in germ-free animals (animals with no ora), animals were extremely sensitive to colonization with pathogens and subsequently developed disease at higher rates than ani- mals with intestinal ora (57). This increased susceptibility could be corrected if fecal ora or mixtures of bacteria and yeasts were administered (38). Germ-free animals and ani- mals that have been given certain antibiotics are able to be colonized by exogenous bacteria at doses 1000 to 100 000 fold less than in animals with normal ora present (11). Freter and Abrams found the concentration of Shigella exneri depends upon the degree of normal ora present in
the gut. In germ-free mice, S. exneri is present at 109:g, but if the mice are pre-colonized with a mixture of anaero- bic bacteria, S. exneri is held to 105:g and when E. coli was added to the mixture, the levels dropped to 103:g (59). Romond et al. showed that biŽdobacteria, given to gnoto- biotic mice, prevented the colonization by E. coli (60).
The role of colonization resistance has been well studied in the case of the resistance to C. difŽcile. C. difŽcile readily colonizes neonatal animals and human babies at the time when there is scarce ora established, but is cleared from the intestines once a more mature microecol- ogy develops (55). In the adult host, colonization resis- tance adequately prevents infection by C. difŽcile unless the microbial barrier is disrupted (61). Once the ora is disrupted, C. difŽcile colonizes the intestines, produces two major toxins and may cause overt disease including di- arrhea, colitis, pseudomembranous colitis or toxic mega- colon (62, 63). If protective ora is seeded back into the intestines (either by fecal infusion of normal stools or selected biotherapeutic organisms), colonization of C. difŽcile may be prevented and the disease does not develop (64).
The mechanism of colonization resistance is dynamic and complex (Table II). Bacteria comprising the normal ora are capable of producing substances and antimicro- bial peptides that are inhibitory against colonizing bacteria (40, 65). Bacteriocins are a group of anti-bacterial proteins produced by intestinal ora that have a broad inhibitory spectrum including many Gram-positive and Gram-nega- tive bacteria. Lactobacilli bacteria have been shown to produce a bacteriocin called reuterin, which is inhibitory in vitro for Salmonella, Shigella, Clostridium and Listeria species (66). Although an intriguing in vitro Žnding, it has not been shown that these bacteriocins reach concentra- tions inhibitory to pathogens in the intestinal lumen, thus the clinical signiŽcance of bacteriocins is not known.
Normal ora may also produce other metabolic end- products that are inhibitory to other microbes. Most nota- ble is hydrogen peroxide (H2O2) produced under anaerobic conditions by several strains of normal ora (66). The presence of H2O2 results in peroxidation of lipid mem- branes, increased bacterial membrane permeability, de- struction of bacterial nuclear acids in bacterial strains that do not possess catalase. The vaginal tract is usually pre- dominantly colonized with lactobacilli (Fig. 1) and H2O2 producing Lactobacillus strains have been found in 75% of vagina samples from healthy women. Vaginal colonization with Lactobacillus strains that produce H2O2 has been shown to be protective of infections caused by Chlamydia trachomatis, Gardnerella vaginalis, Ureaplasma urealyticum and the development of bacterial vaginosis (13). In women with bacterial vaginosis, these strains of H2O2 producing lactobacilli are absent and, instead, high concentrations of Gardnerella vaginalis and anaerobes are present (66). An- other protective mechanism is the production of a low pH
Mechanisms of action for colonization resistance
Factors activeFactorMechanism against
Production of Salmonella,Bacteriocins Shigellainhibitory
substances Toxic metabolic Chlamydia,Hydrogen peroxide
endproducts Gardnerella Acidic endproducts, S. aureus, E. coliAdverse microenvi- short chain fattyroments acids Monomeric glucoseNutrient or sub- Clostridium
difŽcilestrate depletion Non-toxigenicAttachment enterotoxigenic
E. colistrains of E. coliinterference Immune system rotavirus,Secretory IgA
Compiled from references: (11), (13), (41), (55), (59), (60), (63), (67), (68), (75), (80).
Normal ora description 197
environment, which may be inhibitory for certain patho- gens. The production of acids as an end product of carbohydrate metabolism is common in many species of the normal ora and is inhibitory against Gram-positive and Gram-negative bacteria. Several pathogens, including Staphylococcus aureus, Salmonella, E. coli, and Bacillus cereus, are inhibited by acids produced by normal ora such as lactobacilli and biŽdobacteria. A common end product of microbial fermentation is short chain fatty acids (SCFA). The presence of these SCFA have been shown to be inhibitory to nonindigenous bacteria (40). Rolfe showed in a hamster model of C. difŽcile disease that SCFA levels inhibited C. difŽcile growth (40). As newborn hamsters age, they start to produce high levels of acetic, butyric and propionic acids by day 16–19. Growth of C. difŽcile was signiŽcantly reduced when levels of these SCFA increased (40). If normal ora are disrupted, de- creased levels of SCFA result and pathogenic microbes may take advantage of this decrease and reproduce to levels that induce disease. To date however, the identiŽca- tion of which speciŽc SCFA or mixtures of SCFAs are responsible for the inhibition of pathogens has not been demonstrated.
Competition for nutrients may be another mechanism for colonization resistance. As it is extremely difŽcult to assess the levels of speciŽc nutrients in the interior of the colon, most of the research on nutrient depletion has been done using continuous ow culture techniques. Wilson et al. inoculated normal ora from a mouse into a continu- ous ow culture and found one or more of the ora competed more successfully for monomeric glucose, N- acetylglucosamine and sialic acid, resulting in signiŽcantly reduced levels of C. difŽcile (67). Sweeney et al. found that even small numbers of an ingested E. coli strain F-18 could supplant established ora, as this strain utilized an avail- able nutrient (gluconate) more efŽciently than the other microbes present in the system (68). However, continuous ow cultures are extremely dependent on culturing and incubation parameters and the applicability of these results is unclear.
Normal ora may also produce extracellular enzymes that are inhibitory or interfere with pathogen attachment. A yeast (Saccharomyces boulardii ) has been shown to produce a protease that destroys toxin A and toxin B receptor sites in rabbit ileal models for C. difŽcile disease (69). The toxins of C. difŽcile act by inactivating Rho proteins that keep the cytoskeleton of the intestinal entero- cyte intact, thereby distorting the cellular morphology leading to uid loss and diarrhea (70). Rho proteins are also involved in yeast budding processes (reproduction), thus this yeast may produce the protease to protect itself against soil Clostridia that may produce similar toxins to C. difŽcile, but in the human, the protease may coinciden- tally protect the host against infection with C. difŽcile.
Survival in various body niches necessitates the attach- ment to receptor sites, especially in the colonic lumen where peristalsis sweeps unattached microbes and undi- gested debris away. Competition for attachment sites is a successful mechanism to inhibit colonization of pathogenic microbes. Colonization with non-enterotoxin producing E. coli has been reported to prevent the subsequent infection with enterotoxin strains of E. coli in several mouse and pig animal models (40). Fuller reviews several studies in ani- mals showing positive interference with attachment of normal ora when an non-indigenous microbe is ingested (71). A variety of host factors are also involved with colonization resistance including secretory IgA levels, peri- staltic movement, production of mucus, epithelial or ente- rocyte turnover (40).
An area of intense research is to determine which mem- ber or members of the numerous species of microbes present in the intestine are responsible for colonization resistance. Most studies support the role of anaerobic ora as the major microbes responsible for colonization resis- tance (11, 56). Hazenberg et al. found a mixture of human anaerobic ora restored colonization resistance against P. aeruginosa in the germ-free mice model (57). Giuliano et al. showed cefoxitin increased the numbers of fecal enter- obacteriaceae, enterococci and yeasts in human volunteers (72). The increase in all three groups of these aerobic organisms may be due to the suppression of selected anaerobic bacteria by cefoxitin. Leonard et al. reported that ceftriaxone decreased anaerobic ora in mice and human volunteers and colonization resistance was also depressed in ceftriaxone treated animals (73). Clindamycin is effective against anaerobes and has been shown to impair colonization resistance (58). Some antibiotics that are effective against anaerobes, but are only present at low levels in the gut (tinidazole, cephradine), do not impair colonization resistance (58). SpeciŽc members of the nor- mal ora have also been studied including BiŽdobacteria longum, peptostreptococci and Clostridium cocleatum. Romond et al. tested BiŽdobacterium longum in germ-free mice model (60). Pre-colonization of germ-free mice with E. coli delayed colonization by B. longum for one month, compared to a colonization time of only 24 hours in the totally germ-free animals. Herias et al. observed that pep- tostreptococci reduced translocation of E. coli in germ-free rat model by priming the immune system (74). Boureau et al. showed that a normal member of the mouse intestinal ora, C. cocleatum, inhibited the growth of pathogenic C. difŽcile (75). C. cocleatum was found to produce a variety of glycosidases that attacked intestinal oligosaccharides. As the peptide core of the mucin layer in the intestine is protected by oligosaccharides, this bacteria may interfere with the attachment of C. difŽcile but further study is needed. Bourlioux et al. found a Ruminococcus species that prevented C. perfringens colonization in a gnotobiotic animal model (76). Although these studies have identiŽed
L. V. McFarland198
potential candidates for species which may be responsible for colonization resistance; it is more likely that a complex mixture of many microbes is responsible.
A note of caution must be exercised with these animal studies, as the behavior of microbial species has been found to differ depending upon which animal is studied (including human). Wong et al. evaluated the mouse as a model for studying microbial ecology interactions that may occur in humans (77). Human strains fed to mice usually survived, with the exception of Bacillus and Lacto- bacillus species. They also found that in the mouse, human organisms only produced 25% of the expected organic acids, which may suggest a change in metabolism when human strains colonize the mouse. Resistance to a chal- lenge strain (Salmonella ) was preserved, indicating that the strains maintain their ability to produce colonization resis- tance (77).
Production of vitamins
Intestinal ora are also involved in the production of vitamins including panthothenic acid (B5), biotin (vitamin H), pyridoxine (vitamin B6) and menaquinone (Vitamin K2) (66, 78, 79). Without the intestinal ora, these vita- mins would not be produced or in some cases, not broken down into an absorbable form. Intestinal ora can usually provide the daily minimum requirement for many of the vitamins in humans. Microbes that are added to foods (such as fermented diary products) may also increase the dietary levels of these vitamins (66). Vitamins B12, niacin, riboavin and thiamin are also made by intestinal ora, but are not absorbed in the colon. The value of micro- bially supplemented foods is that the Vitamin B12 is ingested and it can be absorbed in the small intestine and utilized.
Attachment to the intestinal mucosa is an important sur- vival trait for organisms in the intestinal tract. The pres- ence of intestinal ora has been shown to stimulate the production of epithelial glycoconjugates, which may be receptors for some pathogenic bacteria (22). Umesaki et al. documented that the presence of a strict anaerobe, B. thetaiotaomicron and a segmented Žlamentous bacterium were associated with fucosylated glycoconjugates in the small intestinal tract (80). Colonic mucins are a class of high molecular weight glycoproteins, which are secreted by the mucosa and exocrine glands. Mucin is present in the lumen and functions as a lubricant, a modulator of water and electrolyte absorption, may aid in attachment of mi- crobes and protects the mucosa from injury (81). A large number of oligosaccharide side-chains of mucin glyco- proteins aids in the stability of the mucin layer. Carlstedt- Duke et al. found antibiotics including bacitracin, clindamycin and vancomycin resulted in altered mucin degradation in healthy human volunteers (82). The effect
on mucin required 5 weeks before it was restored to pre-antibiotic levels, which may indicate the length of time that it takes for normal ora to recover from antibiotic exposure. The most abundant intestinal species (Bac- teroides) does not possess the glycosidases necessary for mucin degradation and only 1% of normal ora species can degrade mucin. Hoskins et al. identiŽed Žve strains of mucin oligosaccharide chain-degrading bacteria from healthy humans, of which three were Ruminococcus strains and two were BiŽdobacterium strains (83). Mucin-degrad- ing bacteria represent a distinct subset of normal ora with a speciŽc function.
Stimulation of the immune system
Normal ora also induces the maturation of the gut-asso- ciated lymphoid system (GALT). The intestinal ora pro- vides an array of antigenic stimulants to the GALT cells, affecting both local and systemic levels (22). Gnotobiotic mice have been shown to have fewer intraepithelial lymphocytes, plasma cells and Peyer’s patches than mice with intact intestinal ora (78). When gnotobiotic mice are immunized, the only local immune response is secretory IgA, however, mice with intact intestinal ora also respond with IgM and IgG (40, 41). Intestinal ora may also be involved in the development of tolerance to antigens (42). Herias et al. gave germ-free rats E. coli alone or a mixture of E. coli, Lactobacillus acidophilus and a strain of an obligate anaerobe (Peptostreptococcus) and observed two effects (74). The peptostreptococci reduced translocation rates of E. coli and increased serum anti-E. coli antibodies. It may be that peptostreptococci act as an immune system primer to other bacterial antigens, thereby leading to a decrease in translocation. Further evidence that normal ora may act as an immune primer was found in a study using BALB:c mice. Pulverer et al. found that normal ora releases low molecular weight substances that interact with MALT (mucosa associated lymphoid tissue) and these substances appear to be essential for an adequate immune response (43). Antibiotic decontamination in a mouse model resulted in a decreased immune response. Thus, normal ora may have an important role as an immune system primer.
The presence of intestinal ora has been shown to stimu- late peristalsis or otherwise involve the enteric nervous system (44, 45, 78). In gnotobiotic mice models, the small intestine has thinner walls and is smaller than in mice with intestinal ora (78). Husebye et al. studied germ-free rats (who have enlarged caeca) and found a slower progress of chyme through the intestinal tract when compared to rats with a normal microbiota (46). He found that the mecha- nism may be through the enteric nervous system, rather than stimulating motility by a direct action on intestinal smooth muscle. This may explain why colonic bacteria are
Normal ora description 199
Geographical differences in intestinal microora in humans
English:mixed Western Japanese:vege-Ugandan vegetarianMicroorganisms or groups of Americans: Japanese:mixed tarian dietdiet (in London) Western dietmicroorganisms mixed Westerndiet (London)
10.1Total anaerobes 9.3 10.2 9.9 11.5 8.2 7.5 9.4–9.8 9.6Total aerobes 8.0
7.2À 4.8ÀFacultative anaerobes 5.7ÀEnterococci 7.0À 5.5 8.4 8.4
8.2À 9.8 10.1 11.1Bacteroides 9.7À 9.3À 10.0 8.29.9À 9.5BiŽdobacteria
6.0ÀLactobacilli 7.2À 7.3 5.7 4.0 4.0–4.6 4.4 5.1–9.7 9.5Clostridia 4.4–5.0 3.1À1.3Yeasts
À Indicates a signiŽcant difference between groups as reported in the original publication (Adapted from reference 3).
able to inuence transit through the small intestine in gnotobiotic mice. Pothoulakis et al. found that one intesti- nal response to C. difŽcile involved events relating to the neural cascade (84). The role of normal ora on the enteric nervous system requires further study and may have promise for therapeutic intervention by medications that inhibit the neural response.
Colonic physiology and maturation
It has been shown that the presence of intestinal ora stimulates the maturation and turnover rates in colonic epithelial cells. The surface layer of the mucosa in the intestines is replaced every 2–3 days, which allows basal stem cells to migrate up the crypt, presenting differentiated mature cells that are involved in nutrient absorption and mucin secretion. In germ-free mice (gnotobiotic), the rate of cell turnover was found to be signiŽcantly slower than mice with established microora (78, 80). Bry et al. also found that normal ora induced enterocyte cell turnover (85).
FACTORS THAT INFLUENCE NORMAL FLORA