Gnotobiotic Animals In Nutrition Research Paper


Gnotobiotic (germfree, defined colonized) rodents have become powerful tools to advance our understanding of the host–microbiome relationship. However, the maintenance and ultimately the monitoring of gnotobiotic rodents is a critical, labor-intensive, and costly process (e.g., sterility, not absence of specific pathogens, must be demonstrated in germfree animals). Here, we provide information on the housing and maintenance of gnotobiotic animals, elucidate prophylactic measurements to avoid contamination, and make specific recommendations for sampling procedures, sampling frequencies, and test methods.

germfree, gnotobiotic, monitoring


Gnotobiotic animals (or gnotobiotes) are typically derived from aseptic hysterectomy or embryo transfer (ET) using axenic or gnotobiotic foster mothers. Originally, these animals were Cesarean-derived and had to be hand reared (Pleasants 1959). For decades, many populations of gnotobiotic mice have been established through hysterectomy. However, this procedure is potentially associated with the vertical transmission of harmful agents. This risk might be reduced (e.g., for Helicobacter spp., Mycoplasma spp., or Pneumocystis spp.) through extensive and adequate testing of the donors via molecular and serological methods and/or treatment with suitable broad-spectrum antibiotics prior to hysterectomy derivation. The risk of vertical agent transmission may be substantially diminished using ET to establish a new gnotobiotic population (Reetz et al. 1988; Van Keuren and Saunders 2004). Compared with hysterectomy derivation, ET has the advantage of avoiding the post-implantation vertical transmission of an infection, which has previously been described during Pasteurellaceae infections (Reetz et al. 1988; Ward et al. 1978). Strains or stocks of mice and rats can be converted to a germfree status through ET rederivation using germfree foster mothers maintained in isolators under germfree conditions (Inzunza et al. 2005). The generation of germfree mice through ET has been described in detail in several studies (Giraud 2008; Inzunza et al. 2005; Okamoto and Matsumoto 1999). However, ET performed with germfree recipients might be complicated, as the enlarged cecum might hinder surgery.

The necessary gnotobiology techniques and equipment were developed more than 50 years ago (Gustafsson 1959; Reyniers and Sacksteder 1958a; Trexler and Orcutt 1999; Trexler and Reynolds 1957; van der Waaij and Andreas 1971). The historical development of gnotobiotic mice and the explanation of various terms and definitions has been previously described (Trexler 1983). However, during the last 10 years, the demand for gnotobiotic animals has undergone a rapid expansion for various reasons. Metagenomic techniques have been developed to study the autochthonous flora, and these techniques not only provide insight into the microbiome of animals but also provide a foundation to study interactions and the relationship between the microflora and the host. Many characteristics of the host, regarding immunology, physiology, metabolism, body weight, behavior, etc., are indeed microflora-associated properties (Bleich and Hansen 2012; Nicholson et al. 2012). Undoubtedly, genetically defined germfree animals associated with a defined microflora are irreplaceable for progress in microbiome research and advancement of the current understanding of the microbiota and host relationship (Faith et al. 2014; Nature Biotechnology 2013; Wos-Oxley et al. 2012; Yi and Li 2012). This idea has been emphasized through recent studies in which defined bacteria and bacterial communities of human origin were established in germfree mice and used to identify gut microbe–host phenotype relationships (Eun et al. 2014; Faith et al. 2014). Even phenotypic effects can be vertically transmitted through the microbiome and thus mimic alterations of host genes (Moon et al. 2015).

The term gnotobiotic refers to all life forms in a host being fully defined and includes both germfree and defined flora animals. Germfree or axenic animals are theoretically free of all life forms (e.g., bacteria and viruses). This status is hypothetical because endogenous or previously uncharacterized viruses might be present or integrated into the host genome. Defined flora animals are intentionally associated with one or more known life forms (“associated animals”). The association of germfree mice can be performed through gavaging pure cultures of the organisms of interest. Recently, a simplified human intestinal microbiota (SIHUMI) was transferred to germfree rats and mice through oral gavage, and this microbiota was stably established and transmitted to the progeny of the animals (Becker et al. 2011; Eun et al. 2014). More commonly, association is achieved through the addition of these microorganisms to the drinking water or food (Schaedler et al. 1965b). In specific situations (e.g., when an existing flora is transmitted to other animals), it is also possible to add fresh feces or cecal contents to the drinking water, to directly inoculate the animals with cecal or fecal contents, or through cohousing and direct contact between associated and germfree animals (Gillilland et al. 2012; Seedorf et al. 2014). Similar to germfree animals, associated animals must be maintained in an otherwise sterile environment to avoid contact and colonization with additional and unknown microorganisms.

In contrast, “normal” mice are colonized by various bacterial species. In particular, the gastrointestinal tract is a complex microbial ecosystem with different habitats (characterized by regional differences of pH, ambient oxygen, etc.) of which each is colonized with different microbes. These autochthonous or indigenous organisms live at high concentrations in the host, and some organisms exist under these conditions during the host's entire lifetime (Cerf-Bensussan and Gaboriau-Routhiau 2010). Most of these bacteria are obligate anaerobes that are extremely sensitive to oxygen. These bacteria are typically not detected during routine testing. In the last decade, however, several methods have been developed to detect this important part of the intestinal flora, representing more than 99.9% of the total microflora. During traditional bacterial testing, only facultative anaerobic bacteria are typically cultured, comprising a small portion of the total bacterial flora, such as Enterobacteriaceae as well as members of the Lactobacillus, Streptococcus, Enterococcus, and Staphylococcus species, to name a few. While strictly anaerobic bacteria are detected in extremely high numbers in the large intestine and feces (1011 bacteria/g feces), the bacterial flora that can easily be cultured are detected at low levels (106 bacteria/g feces or less). Naturally, the composition of the bacterial flora also affects various physiological parameters, including immune mechanisms and drug metabolism. These complex interactions between different bacteria in the intestinal tract have been known for many decades (Dubos et al. 1965; Schaedler et al. 1965a). Details regarding the gastrointestinal microflora of mice have been previously reviewed (Giraud 2008; Sartor 2008; Schaedler and Orcutt 1983).

Thus, it is obvious that germfree animals differ from animals with a conventional microflora in various aspects, including anatomical, immunological, and physiological parameters (examples provided in Figure 1). The most obvious anomaly is the enlarged size and liquid content of the cecum of germfree animals. In addition, the fecal pellets of these animals exhibit increased water content and are softer than normal. Further, the bowel motility of germfree animals is significantly reduced. The enlarged cecum is a consequence of both osmosis, because mucopolysaccharides are not degraded and bind sodium, and intestinal atonia, resulting from the accumulation of muscle depressant substances that are normally degraded by bacteria. Although mice with a conventional microflora are independent of different vitamins, such as vitamin K, because these are synthesized by the microflora, vitamins must be added to the food to avoid deficiencies in germfree mice. However, currently available autoclavable (fortified) diets have sufficient vitamin levels to compensate for deficiencies in this context. A detailed description of the morphological, metabolic, and nutritional aspects was previously reviewed (Wostmann 1981).

Figure 1

Characteristics of germfree mice. Cecal wall thickness of germfree animals (A) and strict (B) and low (C) barrier specific-pathogen-free (SPF) mice. Small intestinal lymphoid structures in germfree (D) and SPF (E) mice. Physiological parameters that are altered in germfree mice (F). SCFA: short chain fatty acid.

Figure 1

Characteristics of germfree mice. Cecal wall thickness of germfree animals (A) and strict (B) and low (C) barrier specific-pathogen-free (SPF) mice. Small intestinal lymphoid structures in germfree (D) and SPF (E) mice. Physiological parameters that are altered in germfree mice (F). SCFA: short chain fatty acid.

Another hallmark of the germfree status is the lack of typical mouse and ammonia odors that are produced when urea is hydrolyzed through bacterial ureases. The smell of ammonia could therefore represent an initial indicator of bacterial contamination. Germfree rodents have an underdeveloped immune system, resulting in smaller lymph nodes and lower immunoglobulin levels as a result of decreased antigenic stimulation. The intestinal walls of these animals are thinner due to the absence of or a reduced gut-associated immune system (Round and Mazmanian 2009). Furthermore, circulating leukocyte levels are reduced. As a consequence, these animals exhibit increased or modified susceptibility to various infectious agents and environmental organisms (Reyniers and Sacksteder 1958b; Stecher and Hardt 2008). In addition, germfree animals differ with respect to other anatomical characteristics and have smaller hearts, livers, and lungs compared with conventional mice, resulting for example in a lower cardiac output and blood flow to organs. Germfree females have a prolonged diestrus period that reduces the frequency of the estrus and therefore copulation and implantation rates. Differences between germfree and colonized animals with regard to the role of the intestinal microflora have previously been reviewed (Giraud 2008; Gordon and Wostmann 1960; Turnbaugh et al. 2007). As a consequence of the increased cecal volume of germfree rodents, cecal torsion and volvulus, ischemia, and obstruction might occur, eventually resulting in death (Figure 2). Some strains or stocks might therefore not be easily maintained in the germfree status. Furthermore, it is likely that the enlarged cecum might also impact on the reproductive performance, reflecting reduced abdominal space.

Housing and Maintenance

Gnotobiotic animals are at an increased risk for contamination through handling or technical failures or inappropriate sterilization measures. When handled with sufficient experience and appropriate maintenance procedures, rodents can be maintained germfree for years or decades (Hardy 2012). Long-term housing of gnotobiotic animals requires sterile conditions, which are easiest fulfilled in positive pressure isolators. These techniques were established more than 100 years ago. Nuttall and Thierfelder raised guinea pigs and subsequently chickens in a germfree environment (Nuttall and Thierfelder 1895–96; 1897). Küster developed a prototype of the modern isolator and raised germfree goats as early as 1915 (Heine 1968). Isolators comprise a chamber with built-in gloves, a port, and a ventilation system. The chambers were originally constructed using stainless steel and therefore were heavy, expensive, and difficult to handle. For example, the Reyniers system was initially constructed to serve as an autoclave, and the Gustafsson system was constructed with locks functioning as steam sterilizers. These systems have been replaced with plastic isolators, typically flexible film isolators, which can be made from different materials, most commonly polyvinylchloride (PVC). All handling can be facilitated on tables adjusted to the proper height. For ventilation, isolators are equipped with a germtight air inlet filter and a liquid air outlet trap or a germfree outlet filter (HEPA filter) to avoid the introduction of microorganisms into the isolator. These chambers are further supplied with long-arm gloves that should be resistant to chemicals and are protected from the inside by wearing cloth gloves. Furthermore, isolators feature a chemically sterilizable lock for supply via a sterilizing cylinder (Figure 3) or supply isolators. Certain systems facilitate sterile supply, even without chemical disinfectants. Although housing in positive-pressure isolators is considered safe, breeding populations of gnotobiotic mice should be housed in at least two separate isolators to avoid losses due to isolator contamination. Additionally, animals from one isolator should not be transferred to animals housed in another isolator, whenever avoidable.

Figure 3

Loading of an isolator (A–C). The sterilizing cylinder into which the supplies are transferred is autoclaved, connected to a port on the isolator using a flexible plastic sleeve, and the inside of the sleeve is sterilized using peracetic acid (personnel are protected by wearing powered air purifying respirators). After disinfection, the internal door on the isolator and the external seal on the transport cylinder are opened, and the sterile materials are transferred into the isolator. A static microisolator (gnotocage) can be used for the temporary maintenance of germfree mice (D).

Figure 3

Loading of an isolator (A–C). The sterilizing cylinder into which the supplies are transferred is autoclaved, connected to a port on the isolator using a flexible plastic sleeve, and the inside of the sleeve is sterilized using peracetic acid (personnel are protected by wearing powered air purifying respirators). After disinfection, the internal door on the isolator and the external seal on the transport cylinder are opened, and the sterile materials are transferred into the isolator. A static microisolator (gnotocage) can be used for the temporary maintenance of germfree mice (D).

Housing for shorter periods is possible in certain types of individually ventilated cages or sterile static micro-isolators (Gnotocages, Han-Gnotocages, see Figure 3), but the risk of contamination significantly increases (Hecht et al. 2014; Hedrich and Nicklas 2012). However, with sufficient care, the housing of gnotobiotic animals is possible without contamination for longer periods and even for a few years (Stehr et al. 2009). Maintaining animals under these conditions might be considered for experiments with gnotobiotic animals, as access to animals is much easier. However, the germfree status must be assessed and reassured on a regular basis, independent from the maintenance conditions.

The interior and all inner surfaces of newly assembled isolators, including the entry lock, must be sterilized using chemical sterilization agents. We prefer buffered peracetic acid (diluted at 5%), which might also be used for all subsequent transfer procedures. Alternatively, other fast-acting and reliable sterilizing agents can be used (e.g., chlorine dioxide, hydrogen peroxide). All supplies including food, water, and bedding introduced into an isolator must be sterilized. The most frequently used sterilization method is autoclaving for 30 min at 134°C. However, autoclaving parameters might vary depending on the situation and should be established for each potential indication by experienced personnel possessing the required skills (e.g., special attention should be given to the packing of food and bedding in small perforated bags to guarantee optimal stem penetration during autoclaving).

In addition, the validation of the autoclaving process is an important factor (e.g., the addition of bioindicators to water bottles and the inclusion of control cylinders loaded with bioindicators when autoclaving supply cylinders). As the diet might harden and become difficult to gnaw after autoclaving, the use of gamma-irradiated food (e.g., 50 kGY) is occasionally preferred. However, even irradiated diets might not be considered sterile due to the presence of radioresistant bacteria (e.g., Deinococcus radiodurans can survive 17.5 kGy) (Daly and Minton 1997). However, the risk of introducing bacteria via this route is considered relatively low, particularly because all diet batches are sampled and tested for bacterial contamination. Here, the management of a separate isolator maintained exclusively to introduce and monitor each new diet charge is recommended. The introduction of chemicals (e.g., drugs) may be necessary for experimental purposes. In cases in which they are sensitive to heat and irradiation, sterile filtration and introduction in a sterile tube is also possible. The docking of the sterilizing cylinder is a critical step, because each opening of the isolator and connection with the sluice poses a threat to sterility. It is of crucial importance that the surfaces of all goods introduced are accessible to/wetted with the sterilization agent. Details regarding the types and supply of isolators, including disinfection procedures and the introduction and removal of materials, are described in detail elsewhere (Arvidsson et al. 2011; Giraud 2008; Trexler 1983).

Prophylactic Measures to Avoid Contamination

Extreme precaution must be taken to avoid isolator contamination. It is of vital importance that only experienced and trained personnel maintain gnotobiotic animals. These individuals must carefully assess the animals and all equipment. All components of an isolator (e.g., gloves, transfer ports, caps) that are subject to deterioration must be replaced early, before deterioration, to avoid contamination. All working procedures must be carefully planned in advance to avoid the unnecessary opening of an isolator. All transfers (in and out) and sterilization steps must be performed without any compromises via well-trained, experienced, and competent personnel. Only sterilization agents that reliably inactivate all microorganisms within a short time must be used. Experiments should never be performed in a breeding isolator, because that approach would require an increase in the frequency of transfers in and out of an isolator.

Potential Indicators of a Break

Animals must be regularly monitored for any changes in behavior or appearance as an indicator of a potential contamination. Furthermore, a change in the smell of the exhaust air might indicate further testing. However, the lack of ammonia odor does not reliably exclude bacterial contamination, because the typical smell appears only when bacteria hydrolyze urea. For example, the Schaedler flora comprises urease-negative bacteria (Dewhirst et al. 1999). Changes in the appearance or the consistency of fecal pellets can also be indicators of contamination. Morphological changes in the intestinal tract and other organs should prompt further examinations. However, as long as food is not chemically defined or otherwise systematically standardized for content of microbe-associated molecular patterns, molecules such as lipopolysaccharide (LPS), might also induce morphological and immunological changes (Enss et al. 1996).

Monitoring Germfree Rodents

General Remarks

Laboratory rodents with a “normal” microflora are typically monitored in accordance with the recommendations of the Federation of Laboratory Animal Science Associations (FELASA) and regularly tested for microorganisms that might affect animal health or the outcome of animal experiments (Mähler et al. 2014; Pritchett-Corning et al. 2014). Attempts to determine the autochthonous flora are typically not undertaken when defining the hygienic status. The agents tested are primarily species-specific viruses, bacteria, or parasites typically transmitted through animals carrying such agents or via contaminated biological materials. These agents do not play a significant role as contaminants in gnotobiotic animals. Agents most easily introduced include environmental bacteria or fungi, which can survive outside an animal under environmental conditions (e.g., at low humidity). These agents, primarily spore-forming bacteria, micrococci, or molds, are easily detectable. Testing for more fastidious bacteria, such as mycobacteria or mycoplasma, is less important given that it is highly unlikely that these organisms gain access to an isolator-housed population through handling. However, these microbes could be carried over through hysterectomy derivation; thus, animals should be tested early after conversion to the germfree status. Protozoans might theoretically be transferred during cesarean derivation. It is extremely unlikely that intestinal worms or arthropods gain entry to a gnotobiotic colony; therefore, respective testing is less important.

The risk of introducing rodent viruses into a running gnotobiotic colony is relatively low, but contamination could occur during the rederivation process. Therefore, extended screening for infectious agents (e.g., as recommended by FELASA) might be performed only after conversion of the animals to the germfree status and (depending on the likelihood of contamination) subsequently at low frequency. These points should be considered when establishing a surveillance program for gnotobiotic animals. Nevertheless, the testing of germfree animals is aimed at demonstrating freedom from all agents, as the animals must be microbiologically sterile. Testing animals with a defined flora will focus on the presence of the agent(s) with which these animals are intentionally associated and will also demonstrate the absence of additional organisms. Testing procedures are therefore dependent on the type(s) of organism(s) present in the colony.

However, testing should not be restricted to laboratory testing. It is important that experienced personnel regularly assess germfree isolators (at least weekly) and that all abnormalities in materials (e.g., gloves, plastic isolator, transfer cylinder, transfer sleeves) and animals (e.g., behavior, consistency of fecal pellets, odor of the exhaust air) are recorded and reported.

It is reasonable to use different techniques to warrant the germfree status, because all approaches have drawbacks. The first description of determining a germfree status was published in the 1950s (Wagner 1955). Trexler defined minimum routine examination procedures, which should include the microscopic examination of feces as wet mounts and stained preparations, aerobic and anaerobic culture at room temperature and at 37°C, and careful observation of the animals and their environments (Trexler 1983). However, many additional and more sophisticated surveillance methods have been developed during the last decades, which increase the reliability of testing. Because testing in the last decades was restricted to microscopic and culture methods, previously undetected contaminants might currently be detected using new techniques, such as quantitative polymerase chain reaction (PCR). The term sterile therefore only states that the presence of viable microorganisms has not been demonstrated.

When samples such as swabs (e.g., from inner surfaces of an isolator) or fecal pellets are collected, drying should be avoided because many agents do not survive desiccation. To avoid drying, the use of composite samples (immersing food trap material, feces, and interior swabs in used drinking water) is recommended. Certain microorganisms might not survive for long periods outside an animal. Thus, the time between the collection of samples and testing should be as short as possible. Transport media potentially maintain the viability of the organisms present in a sample. It is advantageous when the testing lab is near the sampling location, because long transport and storage might reduce the survival rate of certain microorganisms and increase the risk of contamination. Therefore, results suggesting contamination should always be confirmed through de novo sampling and testing, as the sample might have been contaminated outside the isolator.

Samples for Testing

An overview is provided in Table 1

Dr Elena Verdú’s lab seeks to understand the complex pathophysiology of gastrointestinal disease, with a focus on microbiota-diet interactions, to identify novel therapeutic targets for these disorders.



1/ What strikes you most in the evolution of research on gut microbiota and why?

One interesting aspect relates to the way we have approached the study of the microbiota. We started off with classical culture techniques. We realized that we were only culturing a fraction of the microbiota, and moved on to heavily rely on high throughput molecular techniques. These have greatly enhanced our capacity to know “who is there”.

Moreover, metagenomic and metabolomic techniques are allowing us to understand what the microbiota “is doing”. These concepts regarding the limitations of culture techniques are undergoing some revisiting, and we are beginning to understand that what was considered unculturable before, is in fact our lack of capacity –or effort!– to culture what is there.

The groups of Dr M Surette and Dr P Bercik at the Farncombe Institute in McMaster University are engaged in a clinical collaborative study that investigates the recovery of microbiota by culture techniques from fecal samples from patients with irritable bowel syndrome, and compares this with state of the art molecular approaches. In a recent study presented at the Canadian Digestive Disease Week (Lau et al. 2014, Characterization of the cultivable human gut microbiota by culture-enriched molecular profiling, CDDW2014 Abstract) Surette’s group has shown that by using more than 50 different culture media and conditions, they are able to culture between 90-100 % of the species detected by advanced pyrosequencing techniques.

We are beginning to realize that old methods will be invaluable for the isolation of species present in the microbiota, if sufficient effort is employed. When used together with molecular and metagenomic approaches and gnotobiotic technology, the renaissance of culture techniques will increase our understanding of the role of the microbiome in health and disease. We have come full circle.


2/ You are our nutrition expert on the platform. What is according to you the health potential of better understanding the gut microbiota in this area?

There is little doubt that diet and nutrition affect the microbiota. Several recent studies have demonstrated that long-term diet is an important determinant of the microbiota (Wu et al., Science 2011). On the other hand, the microbiota may be a disease-promoting factor in food intolerances because of the important role it plays in immune and functional gut homeostasis (Natividad et al., PlosOne 2009; Rodriguez, FEMS Microbiol Ecol. 2011; Noval Rivas, J Allergy Clin Immunol. 2013). The relationship is bi-directional. We are currently conducting studies that aim at understanding how the gut microbiota interacts with specific dietary components. Specifically we are interested in the dietary protein gluten, the trigger in celiac disease, and how the microbiota may modify disease severity and expression.

3/ How do you see the evolution of clinical applications and interventions in the field of gut microbiota?

There is evidence for an association between intestinal dysbiosis and disease states, including food intolerances like celiac disease. However, evidence demonstrating causality and mechanisms of action is lacking. The challenge will be to assign function to the microbial profiles associated with specific dietary components or interventions and how they affect expression of common diseases and their therapies. The continued use of gnotobiotic animal models in combination with clinical research will be necessary to explore this complex relationship. These studies will help develop targeted interventions directed at modulation of the microbiota to prevent or treat diseases, including food intolerance.


Leave a Reply

Your email address will not be published. Required fields are marked *