What we need to know about the germ-free animal models

1. Introduction

For a long time it was thought that the contents of the bowel are simply waste products, disregarding a huge and vital community inhabiting in different organs of the body [1]. The role of gut microbes to human health was first appreciated by Elie Metchnikov in the early 20^th century [2]. Now, after several decades, it seems that not only, can we not ignore our intestinal guests, but we also intensively need these microorganisms for a healthy life [3],[4]. This population of microorganisms is named the “gut microbiota” (GM). For a long time, it was assumed that the microorganisms could affect only the gastrointestinal tract. However, it is now well known that the GM profoundly display regulatory roles in different body systems [5].

The human GM contains approximately more than 100 trillion bacteria from more than 100 bacterial phyla and about 1000 different species. The GM encodes about 4,000,000 genes [6],[7], and the entire genome of the GM is called the “gut microbiome,” [8]. The GM also contains additional large numbers of some other microorganisms, including viruses, fungi, protozoa, and archaea [9]–[11]. Two phyla Firmicutes and Bacteroidetes constitute 70% of the total microbiota [12]. Other common bacteria in the human gut are proteobacteria, veromicrobiota, fusobacteria, cyanobacteria, actinobacteria, and spirochetes [13], which mostly reside in the colon. It is of noteworthy to point out that the microbial composition differs between different parts of the gut [14], as well as the lumen and the intestinal mucosa layer [15]. The human gut is extensively innervated, with neurons from the extrinsic and intrinsic plexuses [16],[17], which functionally have close relations to the GM.

Accumulating evidence suggests the GM plays a basic role in different physiological activities, such as the stimulation of the growth of microvilli, food digestion, maintenance of intestinal barrier integrity, improvement of the immune system, fermentation of dietary fibers, and inhibition of colonization of the digestive tract by harmful pathogens. Further, the GM plays a role in protecting against pathogenic organisms, metabolizing vital substances including (sterols, bile acids, and drugs), generating short-chain fatty acids (SCFAs), energy harvest and storage, synthesis of vitamins, neuronal activities, proliferation of neurons, brain functions, behaviors, social cognition, emotion, neurogenesis, neurotransmission, protection against oxidative stress, gastrointestinal motility, absorption of nutrients and the production of bioactive molecules [18]–[20]. Therefore, we should accept the opinion that, without the regulatory effects of GM on various systems, our body would be the target of many disorders [21]–[24].

2. Purpose of using germ-free animal models

In many fields of host-microbe, interactions investigating germ-free (GF) animal models are appropriate experimental subjects. The use of GF animals facilitates the assessment of the role of the GM in all aspects of physiology, normal aging, and the nervous, digestive, immune systems, and metabolic function [33].

GF animal models provide biological model systems for studying either the complete absence of microbes, or colonized selected, and known microbes [34]. Experimental models using GF animals are valuable subjects to assess how the GM may affect host physiology [35]–[38]. In this regard, early studies show that the GM affects vascular remodeling in the intestine and increases vascular endothelial growth factor receptor 1 expression and vessel density [39]. Also, GM suppresses tonic Hedgehog (Hh) signaling in the small intestine, thus regulating intestinal barrier function. Hh pathway activity is mainly suppressed through Toll-like receptor (TLR2/TLR6) signaling in the intestinal epithelium, identifying intestinal epithelial neuropilin-1 (NRP1) as a microbiota-dependent Hh regulator that contributes to the stabilization of the intestinal epithelium barrier [40].

GF animal models are considerably used in evaluating the mechanistic understanding of microbe-induced changes in disease models [41]–[50], linking the GM dysbiosis with intestinal (irritable bowel syndrome, inflammatory bowel disease, etc.) and non-intestinal (metabolic syndrome, cancers, brain diseases, etc.) disorders. The GF animals can help us to elucidate the role of commensal microbiota in the development and function of the organism.

It is worth noting that, since those early models, advances in the knowledge of nutritional differences, intestinal morphology, intestinal epithelial properties, intestinal function, metabolic characteristics, and mucosal immunity [51],[52] have significantly developed GF animal models.

3. History of GF animal modeling

To test whether GF life of an animal host is possible, Nuttall and Thierfelder (1896) were the first to generate and manage to have them survive for 13 days [53]. They raised the first GF animals (guinea pigs), which were generated by aseptic caesarean section at the University of Berlin and kept them for 2 weeks [53]. However, because of the lack of knowledge concerning appropriate nutrition and adequate equipment, rearing and maintenance of healthy GF animals was a challenging task due to technological constraints until the mid-1900s when the first GF rat colony was established [54]. Nevertheless, GF research programs were developed independently at 3 different institutions and proved conclusively that life without microbes is possible, though not desirable. Systematic studies with GF animals started when, in the mid-twentieth century, a group headed by James Reyniers at the University of Notre Dame was the first to rear successive generations of GF rodents [55],[56]. Reyniers and his colleagues established academic organizations in the mid-1940s [55] and early 1950s devoted to understanding host–microbial interactions [57]. Almost simultaneously, Bengt Gustafsson at Lund University in Sweden also produced GF animals with a new isolation breeding system [58],[59]. Then, a third GF program started at Nagoya University under the leadership of Masasumi Miyakawa [60]. Pleasants (1959) developed the first GF mouse colony in the United States [61]. By the late 1950s, researchers were successful in rearing GF guinea pigs, mice, and chickens [56]. In the 1960s, life without GM was featured prominently in the medical, scientific, and public press, often reported as a compound of fact and fiction in the future. Then, after World War II, following the appearance of antibiotics, GF living became an interesting topic [62]. The concept of humans living in sterile worlds was realized as early as 1971, perhaps most notably with David Vetter, a patient with severe combined immunodeficiency who grew up in GF conditions as an infant and became known as “Bubble Boy”[62].

In industrial agriculture, GF animals are bred to make pathogen-free animals to help in veterinary work [62],[63]. Also, GF technology has been used to protect GF immunocompromised newborn [62],[64]–[67]. In spite of many applications of GF technology, it has not been widely implemented outside the laboratory yet. The main reason is that GF modeling is a labor-intensive technology that requires constant control to manage the state of cleanliness [68]–[70]. However, the existing methodology underlying the production of GF animals has remained essentially unchanged.

With the advent of next generation sequencing and developments in microbial ecology, the use of gnotobiotic models is now a valuable resource for understanding host-microbe interactions in health and disease. Different methods are used in the study of GM, including isolated GF animals, antibiotic-treated animals, probiotic feeding, fecal microbiota transplantation, and mouse humanization. Methodologically, isolated GF and antibiotic-treated models are known as the main approaches by which exploring the effects of the microbiota on physiology and disease in mice is established (Figure 1). Both approaches have strengths and weaknesses [71],[72].

The colonization of GF animals with a minimal microbiome offers an attractive method to evaluate the etiology of disease-associated microbial changes [73]. In fact, the production of minimal microbiomes and their application in gnotobiotic models allow mechanistic studies of host-microbe interactions under controlled conditions [73].

4. Isolated GF animal model

Isolated GF animals are born, bred, and raised for their whole lifetime in sterile isolators to prevent their exposure to microorganisms including bacteria, viruses, fungi, parasites, and protozoa, throughout its lifetime [88]–[90]. The key principle supporting the creation of isolated GF animals is the fact that the environment of uterine is sterile and colonization of the GI tract takes place after birth in normal humans and rodents [91]. Nevertheless, it is worth noting that researches on the potential for bacterial transfer across the placenta have also detected bacteria in placental tissue [92], umbilical cord blood [93], amniotic fluid [94]–[96], and fetal membranes [96],[97].

The production and transformation of isolated GF animals is difficult and expensive; however, environmental circumstances, isolator technology, and necessary equipment including feed, water, cages, and bedding have been improved, resulting in cost-effective systems [98]–[100].

To establish an initial isolated GF colony, newborns are carefully delivered by caesarean section to avoid contamination with microbes residing in the mother's vagina and skin [56],[59],[101]–[105]. The fetus is then delivered from the uterus in the isolator, and the neonate can be reared by a GF foster mother or artificially fed with formula milk [98]. Over the postnatal period of life, colonies are maintained in aseptic isolators in a GF unit where the food, water, and bedding are sterile, therefore eradicating the opportunity for postnatal colonization of their GI tract, allowing a direct comparison with the normal colonized gut of their counterparts.

Early isolated GF animals were housed in stainless steel isolators [59], then replaced by light, cheaper, more flexible, plastic polyvinyl chloride isolators [98],[106]. Figure 2 show the different stages of isolated GF animal production.

Isolated GF mice can also be reproduced by embryo transfer to axenic mice and then be kept in a GF isolator. The embryos are then removed from the uterus in the GF isolator and place on heating pads. These pups will be adopted by a foster mother [107]. A substitute method could be the transfer of an embryo at the 2-cell stage into a pseudo pregnant GF mother [102]–[104].

Although the creation of new strains of isolated GF animals requires that the fetus remains sterile in the uterus, however, the later descendents of isolated GF animals produce through a much simpler process. Isolated GF animals are mated and mothers can give birth naturally in the isolator without exposure to any microorganisms in the new litter [56],[59],[101]–[103]. The majority of commercially present isolated GF animals produced with this method can be transported in a sterile container to local GF facilities.

Methodologically, the use of first-generation isolated GF animals in experiments is imprudent because their mothers are not GF and some microbes or bacterial metabolites may be transferred from mother to the fetus through the placenta [52].

5. Antibiotic-treated animal model

Whereas isolated GF mouse models are mostly considered the gold standard for microbiota studies, the production and working with isolated GF animals, with strictly controlled identified microbes, is very difficult and expensive due to the use of isolators and rigid gnotobiotic working procedures.

As a result, depletion of the GM by antibiotic treatment, as a rapid, inexpensive, and accessible alternative has been considered for GF animals [108]. Actually, the antibiotic-induced short-term disruption of the GM in this model has been thought to be a less intrusive model to probe causality in microbiota-dependent effects [80],[109],[110]. As an alternative to isolated GF technology, treatment of mice with antibiotics is considered to be more practical and a potentially relevant alternative to GF mice [111].

6. Validation of bacterial depletion

An important aspect of working with GF animals is to confirm that the animals are without any microbes. Three screening methods are commonly used to access GF status. These include anaerobic/aerobic liquid culture, Gram-stain, and polymerase chain reaction (PCR) using universal and specific 16S rRNA bacterial primers [36].

In fact, the fastest way to confirm the presence of any microbes is culture-based methods by evaluating colony-forming units (CFUs) from fecal samples placed in aerobic and/or anaerobic conditions on non-selective media, or alternatively by making Gram-stained fecal smears and assessing them under a light microscope [71]. The technician swabs the cages and, using bacterial culture methods, analyzes stool samples to confirm that the GF unit is truly sterile [104],[128]. These methods ensure that the GF mice do not have a contact with any bacteria in the gut and on any other surface. Furthermore, they only consider cultivable microbes.

Recent studies show that, in practice, bacterial culture and Gram-stain are sufficient to screen for GF status because both have high sensitivity and specificity, while PCR shows high specificity but less sensitivity [129]. However, each technique has limitations. Cultivating GM outside the intestine requires an anaerobic environment, such as an anaerobic chamber, which is costly. Hence, the GF status of mice must be frequently monitored by fecal sample culturing for aerobic/anaerobic bacteria and fungi. Here, performing molecular techniques, such as PCR amplification, is a good alternative for bacteria that cannot be cultured [52],[102]. Quantitative PCR of the gene encoding 16S rRNA also allows for culture-independent evaluation of the bacterial load of the gastrointestinal tract [71]. However, it should be noted that bacterial 16S rRNA gene contamination in the breeding diet can occur, thus, PCR-based control should not be considered as the only sterility test for PCR-based sterility controls.

7. Advantages of isolated GF animal model

Isolated GF animals seem to be the best controlled models for microbial transplantation, and this model has been subject of the most experimental research on the GM so far. As previously described, fecal microbiota transplantation is a method in which selected bacteria can be transfered from a donor subject to a recipient one. Because isolated GF animals are free of microorganisms, they are a good model system for the response to bacterial introduction; indicating that they are suitable for studying the effects of microbes on host development and function. On the other hand, in vivo experimental models show a reduction or absence of several inflammatory and complex diseases in isolated GF animals; suggesting that the GM is associated with the development of these diseases [130]. Therefore, it is not surprising that isolated GF mice live longer than normally colonized control animals [131]–[133]. It is probably due to the absence of pathological infections. Therefore, this animal model provides conditions through which the positive and negative role of the GM on lifespan can be evaluated.

8. Disadvantages of isolated GF animal model

Despite the many advantages that the isolated GF model has, some disadvantages can limit their use. First of all, since these animals are never exposed to microorganisms, they display impaired physiology and immune development from birth. Technically, the production and maintenance of isolated GF animals need particular facilities, and the cost, labor, and skills necessary to preserve them can make these models inaccessible to many investigators [71].

Isolated GF mice should be regularly monitored for contamination using a combination of culture, microscopy, serology, gross morphology, and sequence-based diagnostic techniques [34],[129] and this limits the number of different genotypes that can be studied. Additionally, keeping animals in isolators may make some studies (e.g., behavioral testing or pathogen infections) impractical or challenging [71].

Growing evidence shows that isolated GF animals have some biochemical and physiological abnormalities such as altered immune systems [43], mild chronic diarrhea [134] and impaired metabolism [135], and reduced reproduction [136]. Particularly, immune system is known to be primed by the GM in early life [42],[137]–[140]. The immune response to fecal microbiota transplantation in isolated GF mice, which have never previously encountered the bacteria, must be expected to be important for at least some disease models [108].

9. Advantages of antibiotic-treated animal model

The advantage of antibiotic-treated over isolated GF studies is the timing of the GM changes or reductions and translatability to humans [141]. Treatment with broad-spectrum antibiotics is commonly used to eliminate the GM in mice and can be easily applied to any mouse genotype or condition [71]. Because of differences in their action mechanism, antibiotics are able to selectively exhaust different types of microbes.

Individual antibiotics can be used to alter the GM composition in order to identify bacterial classes associated with different phenotypes [142],[143]. A cocktail of different classes of antibiotics can be used to broadly deplete the GM [71]. Antibiotics also have the advantage of allowing the examination of the consequences of intestinal microbial depletion at different stages of life [144]. In fact, by targeting different groups of bacteria through different classes of antibiotics, it is possible to develop hypotheses about which bacteria are responsible for disease manifestations. For instance, while both clindamycin and metronidazole target anaerobes, polymyxin B specifically targets Gram-negative bacteria and vancomycin is only effective against Gram-positive bacteria. [142],[145]. It is also possible to transfer host phenotypes with normal GM to antibiotic-treated animals through fecal microbiota transplantation, however problems related to reproducibility and antibiotic resistance genes must be considered [108] (Figure 3).

10. Disadvantages of antibiotic-treated animal model

Antibiotic-induced dysbiosis presents several challenges, especially when used for fecal microbiota transplantation studies. Although a broad-spectrum antibiotic approach significantly reduces most bacterial species, bacteria will still remain in the gut, as demonstrated by denaturing gradient gel electrophoresis [111] or cultivation [146]. It is difficult to precisely control the effect of an antibiotic administration in terms of species are completely eradicated and which species are only reduced, and the residual microbiota from antibiotic-treated mice may also influence colonization over time [108]. Because the immune system is primed by the GM at early postnatal age [42],[137]–[139], exposure to microbes prior to elimination with antibiotics can have long-term effects on the physiology of the host. An important potential drawback of eliminating microbiota with antibiotics using broad-spectrum antibiotics can cause the evolution or development of antibiotic-resistant bacteria and the selection and overgrowth of resistant bacterial species [147]–[149]. It may play a significant dominant role in the microbial profile after recolonization [146] or may be detrimental to animal health.

Although oral administration of antibiotics decreases the GM, other microbial communities, for instance the skin and lung microbiota, are not ever directly affected. It depends on the pharmacokinetics of the antibiotic substance and may also developmentally affect the immune system [150],[151]. Also, if antibiotics are administered through drinking water, the possible disadvantages of antibiotic-induced intestinal dysbiosis could be the systemic or even central effects of the antibiotics themselves as well as changes in consumption [152]. Antibiotics may even directly affect the brain. There is evidence that they can modulate the vagus nerve [153] and the enteric nervous system [154]. Mounting evidence indicates that bacteriophages, fungi, and eukaryotic viruses, which are not directly targeted by antibacterial antibiotics, cannot be discounted in GM homeostasis and immune priming [155],[156]. In addition, antibiotic therapy can allow the overgrowth of common fungal species, possibly confusing results because these organisms can modify immune function [157],[158] (Figure 3).

11. Effects of dysbiosis on other organs

Microbiota disruption affects the anatomy and function of various organs, such as the liver and gastrointestinal tract [159],[160]. One of the most obvious anatomical changes is the enlargement of the cecum, due to mucus and undigested fiber accumulation, which is observed in both isolated GF and antibiotic-treated mice [34],[122]. In addition, GF mice have elongated villus structures, reduced villus width, and weakened capillary networks in small intestinal villi [133],[161],[162]. The immune cell populations are also influenced by antibiotic treatment [163]–[165]. Oral antibiotic regimens have been shown to reduce cultivable bacteria in the respiratory system tract [114],[125],[166] and vagina of mice [167] with no effect on skin bacterial communities [168].

12. Conclusion

The importance of the GM in physiology of almost all body organs has encouraged research in this field. However, there are numerous problems in research strategies in the human GM. Therefore, animal models take a major role in many aspects of GM investigations. There are two very often used animal models: the isolated GF and antibiotic-treated models. Each of these models display advantages and disadvantages. Whereas both models are currently used by researchers, there are numerous differences between the two models such that, it is suggested that they should be viewed as distinct models for the GM manipulation. Consequently, it is very important to apply appropriate models when working on the GM. On one hand, because isolated GF animals are not exposed to bacteria from conception onwards, their use for experimental questions about the impact of altered microbiota composition in postnatal life may be limited [169]. On the other hand, given the experimental variables and several side effects of antibiotic- treated protocols, it is increasingly evident that the interpretation of data collected from experiments on microbiota disrupted by antibiotics should be approached with caution. A possible approach to circumvent the uncontrolled situation of antibiotic-treated animals and the effect of GF early in life is to use the generation of GF parents and use succeeding generations [108].

Accordingly, it is believed that findings obtained from GF animal models should be used with caution to develop strategies for the disease treatment and/or prevention [170]. Taken together, laboratory animals are currently the major models for the study of the GM, and each model has its own limitations; nevertheless, reproducibility must always be emphasized as an undisputed essential feature of the system.