Australasian Biotechnology (backfiles) AusBiotech ISSN: 1036-7128 Vol. 8, Num. 4, 1998

Measuring Health Efficacy Of Probiotic And Prebiotic Foods

Martin J. Playne and Ross G. Crittenden

Cooperative Research Centre for Food Industry Innovation, Food Science Australia, PO Box 20, Highett, Victoria, 3190, Australia
e-mail: martin.playne@dfst.csiro.au

Code Number:AU98032
Sizes of Files:
      Text: 24K
      Graphics: Photographs (jpg) - 127K

This paper is based on presentations made to the 6th ASEAN Food Conference, Singapore, November 1997, and to the 14th Australasian Biotechnology Conference, Glenelg, South Australia, April 1998

Introduction

Most probiotic and prebiotic food products released on the market are not supported by strong evidence of their ability to improve human health. This is partly because their effect on disease conditions is indirect, and of a preventative nature rather than a therapeutic nature, and consequently difficult to demonstrate quantitatively. Additionally, promotion of the so-called "health foods" has tended to rely on image of the product rather than on its actual ability to improve health. This has occurred because of regulatory restrictions in most countries which make it difficult to include health claims on food labels, as well as the time and cost of conducting animal experiments and human trials to prove health efficacy. Furthermore, these health foods usually contain a number of ingredients, many of which can each contribute to modification of the health of the host. For example, a bio-yoghurt can contain two strains of probiotic bacteria (e.g., strains of Lactobacillus acidophilus and Bifidobacterium spp), a strain of Streptococcus thermophilus, fructo-oligosaccharide as a prebiotic, and all the components of skim milk powder, including possibly bio-active peptides.

Figure 1. Typical yoghurts containing probiotics and prebiotics.

Consequently, health efficacy tests have to be conducted on the three individual bacterial strains, on the prebiotic, and on the whole probiotic food to isolate the effects. Furthermore, the probiotic food will form only one part of the daily diet, and there will be an interaction with the other dietary components consumed. It is against this background that researchers must devise suitable tests and model systems to screen new strains of bacteria for their ability to act as probiotics and provide a health benefit to the host. This paper describes the types of tests and model systems that are currently available, including laboratory tests, fermentor systems, animal models, bioassays, clinical trials and epidemiological studies with humans.

Laboratory Tests

Laboratory tests are quite limited in their ability to measure the effects of consumed probiotic bacteria on the the health of the host. Laboratory test systems are more suited for the assessment of the ability of the bacteria to utilise various substrates, to survive shelf storage, and to transit the stomach and small intestine. However, some in vitro tests can be indicative of the health-promoting properties of the bacterium. These include:

• Beta-galactosidase activity (as an indicator of ability to reduce lactose intolerance)
• Ability to inhibit the growth of pathogens (e.g., Salmonella spp., Escherichia coli, Candida spp, Clostridium difficile, Clostridium perfringens)
• Antibiotic sensitivity (against a range of medically-important antibiotics, including fungicides)
• Bile salt deconjugation activity (as an indicator of ability to reduce cholesterol)
• Inability to form secondary bile acids from primary bile acids
• Adhesion to human cell lines (such as Caco2 and HT29) or to pieces of epithelial tissue
• Production or utilisation of vitamins (e.g., folate)
• Toxicity tests (e.g., platelet aggregation)

Appropriate laboratory tests for prebiotics include:

• Digestibility of the prebiotic by pancreatic and brush border enzymes
• Preferential utilisation of the prebiotic by probiotic bacteria

Fermentor Models

Fermentors can range from simple sealed 50 ml test tubes, fitted with gas relief valves with no pH control to perform batch fermentations (eg., Miller and Wolin, 1974 ), through to sophisticated , fully-automated, multi-compartment fermentors designed to simulate the human gut, which operate in continuous flow with differentiated flow rates for the solid and liquid feed components. A range of fermentor designs used in studies of human and animal metabolism and function have been reviewed by Rumney and Rowland (1992).

The uses of fermentors in probiotic research have been to optimise media for growth to high cell numbers, to examine the utilisation of different carbohydrates, to simulate environments in the gut, to examine interactions of probiotic and gut microorganisms, and to mimic the gastrointestinal tract. Fermentor systems have a particular role to play in the assessment of different prebiotics (e.g., oligosaccharides, resistant starches). The advantage here is that the quantities required for tests can be quite small. Consequently, the normal food-grade carbohydrate can be purified to remove residual simple sugars, which are normally present in commercial samples. The performance of the pure carbohydrate can then be tested in the fermentor.

A sophisticated multi-stage continuous fermentor has been developed by researchers at the TNO Nutrition and Food Research Institute in the Netherlands. It incorporates peristaltic action, addition of solids and liquids at different flow rates, and removal of solutes by dialysis ( Minekus et al., 1995). Simpler multi-stage systems (see Fig. 2) are those of groups at Gent University (Molly et al., 1993 ), and at the Dunn Clinical Nutrition Laboratory (McBain and Macfarlane, 1997).

Figure 2. Multi-stage fermentor to simulate the human colon at Ghent University.

Animal Models

Mouse, rat and pig models have been used to test probiotic strains, and to understand mechanisms of function of probiotics. While intact animals have mostly been used, caecotomised animals, and fistulated animals, to enable ileal, caecal and colonic sampling, have been developed. Gnotobiotic and specific pathogen free (SPF) animals have also been used, as have animals treated with antibiotics prior to experimental use ( Tannock, 1995a). All the above animals differ from humans in the relative sizes of their gut compartments (e.g., size of caecum), in the presence or absence of bile secretion, in coprophagic habits, and in the nature of their microbial population in the gut (Tannock, 1995b; Guillot & Ruckebusch, 1995). Additionally, many host / microbe interactions may be species- specific (e.g., colonization, immune responses). Consequently, whilst animal models are extremely valuable for probiotic research, data on health efficacy in humans also requires scientifically valid human clinical studies.

Human Trials

Human studies can be divided into: clinical trials with volunteers, epidemiological studies and anecdotal studies. Clinics may be conducted in-house, where subjects are housed for the entire period of the experiment, and externally, where subjects visit on a frequent basis, or by home visitation, where a nurse visits the subject at their home. In addition to these systems, which are used mainly with healthy volunteers, a number of other systems have been employed e.g., volunteers in the military forces, and volunteers amongst prisoners in jails. Patient groups have also been employed, in which the patients are suffering from a pre-existing medical condition, and have voluntarily agreed to take part in trials where a probiotic or prebiotic treatment is included in the design of the experiment. Ileostomy patients have also been used in probiotic studies, as in these patients, it is possible to achieve easy microbial sampling of digested foods passing the small intestine.

Anecdotal records and epidemiological studies can provide limited information on the effect of probiotics on human health conditions. Anecdotal records are those gathered by medical practitioners or paramedics in the course of patient interviews of their medical history, symptoms, and reaction to various medications, including probiotics, but no control or placebo patients exist. This anecdotal data is useful as an indicator of potentially useful treatment methods, because it is derived from a very large informal database held by general medical practitioners, medical specialists, dieteticians, and pharmacists on a large proportion of the population. However, it does not provide the "hard" evidence needed to make health claims for foods. Epidemiological studies are of limited value for studies on probiotics, because of the complex interactions between gut microflora and dietary ingredients, and the lack of control groups. Large numbers of patient records and long timespans are involved in such studies, limiting their use for the commercial development of new probiotic strains.

Well-designed clinical trials are of most use for probiotic research. Randomization, crossover designs, double blind treatment allocations, and placebo controls are required. A sufficient number of subjects must be included to allow statistical evaluation of the data. Allocation of patients to treatments must be balanced. Sufficiently long changeover periods must be allowed when patients are changed to new treatments. Adequate and well-designed sampling procedures of patients must be adopted. Procedures intrusive to patients are generally not accepted by ethics committees. Most countries require assessment of proposed trials prior to commencement by the relevant ethics committees. This also applies to experimentation using animals.

Measurement of Samples from Animals and Humans

In both animal and human studies, it is important to ensure adequate sampling of gut and faecal samples and other tissues, including blood. Breath hydrogen can also be a useful measurement with humans, to detect microbial utilisation of carbohydrates. In probiotic research, examples of such measures are: bile acids in faecal water; short chain fatty acids (SCFA), pH, lactate, and ammonia in colonic and faecal samples; blood insulin; aberrant crypt assay, and DNA comet assays on cell wall tissues of the intestine; and the enzymes- glucuronidase, azoreductase and nitro-reductase in colonic samples (Table 1).

Analysis of the microbial composition of gut samples and faeces is crucial to most probiotic studies. Because of the complexity of the gut microflora (perhaps 400 species present, but with only 10 to 15 species in high numbers), analysis is difficult and time consuming, especially as quantitative results are usually required. Particularly important measurements are those which identify the added probiotic strain, and are able to distinguish that strain from other strains of the same species which may be naturally resident in the gut. It is important that this be done quantitatively. Probe specificity may not be adequate to allow simple DNA probing of a whole faecal sample. It is usually necessary to quantitatively enumerate the genus with conventional microbiological plating methods, and then to determine DNA patterns on a random selection of colonies of the genus using a technique such as pulsed field gel electrophoresis (see Fig. 3). From this analysis, one can determine the ratio of the added probiotic bacteria to the naturally-resident bacteria of that genus. This enables one to quantify the survival and colonisation of the probiotic bacteria in the gut.

Figure 3. Pilse field gel electrophoresis patterns of genomic DNA fragments from different strains of bifidobacteria.

Safety Assessment

Most probiotic bacteria used and under development belong to the genera, Lactobacillus and Bifidobacterium. Both these genera are generally regarded as safe for human and animal consumption. There are a few strains of three species in these genera that have been implicated as demonstrating pathogenicity under certain circumstances. These species are: Bifidobacterium dentium, Lactobacillus rhamnosus, and Gardnerella vaginalis (closely related to Bifidobacterium spp). This pathogenicity is rare and has been reviewed thoroughly (Aguirre and Collins, 1993; Gasser, 1994).

If a patient is suffering open wounds , or has recently undergone surgery, orally or in the gastrointestinal tract, then translocation into the bloodstream of almost any bacteria present in large numbers in the gut may lead to infection. Under such circumstances, dosing with probiotics should not be advised. Treatment with probiotics of persons who are immunocompromised (including AIDS patients) is also a questionable practice.

New isolates of potential probiotic bacteria obtained from animal or human sources such as from gut or faecal samples should not be consumed until some basic tests of toxicity have been completed. A full complement of tests has been described by Donohue and Salminen (1996), and include virulence factors; antibiotic resistance; plasmid transfer; assessment of the bacteria's metabolic products; assessment of the acute and subacute toxicity of ingestion of large amounts of the bacteria; assessment of infective properties, using cell lines and immunocompromised animal models; determination of efficacy with dose / response studies; assessment of side effects in human volunteers; and monitoring the effects on the population of a newly introduced probiotic in the market. Additionally, platelet aggregation tests can also be useful indicators of bacteria with potential problems (Harty et al.,1993; Korpela et al.,1997).

The genera, Lactobacillus and Bifidobacterium, both have excellent safety records over a long period of time, and one would anticipate that not all of the above tests would be needed before new isolates of those species are consumed by humans. However, where some other genera are used as probiotics (e.g., Enterococcus, Bacillus, Clostridium), then more stringent safety testing should be performed.

Taxonomy of Probiotic Bacteria

Knowledge of the identity of probiotic strains is extremely important. This requires access to molecular techniques (e.g., DNA fingerprinting), as well as to traditional phenotypic measurements of morphology and biochemical reactions. Such knowledge provides a manufacturer with quality assurance on its products, and information on its competitor's products. It also assures researchers that they are performing tests over time on the same defined strain. Over the last 10 years, there have been a number of taxonomic reclassifications which have affected the naming of common probiotic species. These are described below.

Species which were previously known as "L. acidophilus" are now divided into six species, based on genotypic as well as phenotypic evidence (Fujisawa et al., 1992) . These species are: L. acidophilus, L. amylovorus, L. crispatus, L gallinarum, L.gasseri, and L. johnsonii. Species previously named L. casei subsp rhamnosus have been re-classified into L. rhamnosus, with a few being placed in L. paracasei subsp paracasei. The only other species of Lactobacillus likely to be involved in probiotic foods are L.plantarum, L.fermentum, and L. reuterii. Of these, L. fermentum and L. reuterii are closely related and will easily be mis-identified.

Labels on commercial products containing Bifidobacterium species are usually labelled "Bifidobacterium spp.", "Bifidobacterium bifidum", "Bifidobacterium longum", "Bifidobacterium infantis" or "Bifidobacterium breve". Sometimes, they are inaccurately named as "bifidus" or "Lactobacillus bifidus".

Conclusions

Until this time, most manufacturers of probiotic cultures and most food manufacturers using these cultures have supplied strains of lactobacilli and bifidobacteria that grow well to high numbers, are robust in a range of manufacturing and storage conditions, and which are resistant to acid conditions found in the stomach. Only a limited number of currently available commercial probiotics have been studied in well-controlled animal and human trials, such that their efficacy has been demonstrated conclusively for a number of disease conditions. Some such health efficacy data exists for Lactobacillus rhamnosus (GG, Valio) and Lactobacillus johnsonii (LC1, Nestle).

A combination of methods have to be used to assess a new probiotic strain. It is not possible to rely only on in vitro methods, however, they can be used to eliminate strains not worthy of further study. In probiotic research,it is necessary to have access to in vitro test systems, fermentation models, animal models and human trials to successfully bring forward to commercialisation new and improved probiotic strains with proven health efficacy. Such proof will be needed with the new generation of probiotics in order to meet consumer expectations. Research groups need to collaborate with commercial companies producing probiotic cultures and food ingredients which can act as prebiotics (such as oligosaccharides and resistant starches), and with health care specialists. The research group has to ensure that its culture collection is maintained at the highest standards, that rigid adherence to quality control is followed, and that it has access to DNA fingerprinting and allied techniques to support phenotypic methods to identify bacteria at both the species and the strain level. The probiotic research programme of the Cooperative Research Centre (CRC) for Food Industry Innovation in Melbourne and Sydney has been structured so that it meets these requirements.

Acknowledgements

We thank our CSIRO colleagues Clinton Grant, Leone Morris, Ian Haynes and Kathryn Binding for their contributions to the research, and our colleagues in Sydney and Adelaide in the CRC for Food Industry Innovation. We acknowledge funding provided by CSIRO, the CRC, and the Dairy Research and Development Corporation.

References

Aguirre, M and Collins, K. (1993) Lactic acid bacteria and human clinical infection. J.Appl. Bacteriol., 75, 95-107

Donohue, D.C. and Salminen, S. (1996) Safety of probiotic bacteria. Asia Pacific J. Clin. Nutr. 5., 25-28

Fujisawa, T., Benno, Y., Yaeshima, T. And Mitsuoka, T. (1992) Taxonomic study of the Lactobacillus acidophilus group, with recognition of Lactobacillus gallinarum sp. Nov. And Lactobacillus johnsonii sp. Nov. and synonymy of Lactobacillus acidophilus group A3 (Johnson et al. 1980) with the type strain of Lactobacillus amylovorus (Nakamura 1981). Int. J. Syst. Bacteriol., 42, 487-491

Gasser, F. (1994) Safety of lactic acid bacteria and their occurrence in human clinical infections. Bull. Inst. Pasteur 92, 45-67

Guillot, J.F. and Ruckebusch, Y. (1995) Microflore digestive des animaux IN "Bacteries Lactiques" Vol 2 (eds: H de Roissart and F.M. Luquet) (Lorica, Paris) p343-367

Harty, D.W.S., Patrikakis, M., Hume, E.B.H., Oakey, H.J. and Knox, K.W. (1993) The aggregation of human platelets by Lactobacillus species. J Gen. Microbiol., 139, 2945-2951

Kordias, N., Davidson, B.E., Playne, M.J., and Hillier, A.J. (1995) Molecular techniques for the identification of strains and species of Bifidobacterium . Proc. Fed. Asian Oceanian Molec. Biol. Biochem. Soc., Sydney, Sept 1995

Korpela, R., Moilanen, E., Saxelin, M., and Vapaatalo, H.(1997) Lactobacillus rhamnosus GG (ATCC 53103) and platelet aggregation in vitro. Int. J. Food Microbiol., 37, 83-86

McBain, A.J. and Macfarlane, G.T. (1997) Investigations of bifidobacterial ecology and oligosaccharide metabolism in a three stage compound continuous culture system. Scand. J. Gastroenterol., 32, 32-40

Miller, T.L. and Wolin,M.J. (1974) A serum bottle modification of the Hungate technique for cultivating obligate anaerobes. Appl. Microbiol., 27, 985-987

Minekus, M., Marteau, P., Havenaar, R., and Huis in `t Veld, J. (1995) A multicompartmental dynamic computer-controlled model simulating the stomach and small intestine. ATLA : alternatives to laboratory animals 23, 197-209

Mitsuoka, T. (1969) Zentrl. Bakt. Parasitkde Orig. Abt 1, 210, 52-64

Molly, K., Van de Woestyne, M., and Verstraete, W. (1993) Development of a 5-step multi-chamber reactor as a simulation of the human intestinal microbial ecosystem. Appl. Microbiol Biotechnol., 39, 254-258

Rumney, C.J. and Rowland, I.R. (1992) In vivo and in vitro models of the human colonic flora. Crit. Rev. Food Sci Nutr., 31, 299-331

Tannock, G.W. (1995a) Normal Microflora (Chapman and Hall, London) 115p

Tannock, G.W. (1995b) Microecology of the gastrointestinal tract in relation to lactic acid bacteria. Int. Dairy J., 5, 1059-1070

Copyright 1998 Australian Biotechnology Association Ltd.

The following images related to this document are available:

Photo images