In 2005 Sarkis Mazmanian and colleagues showed that a single polysaccharide from an intestinal commensal, Bacteroides fragilis, could largely correct the subnormal and functionally distorted development of the immune system that occurs in germ-free mice (Mazmanian et al. 2005). More recently they have shown, using three different models of intestinal inflammation, that the same polysaccharide, given by mouth, can turn on crucial immunoregulatory pathways (Mazmanian et al. 2008). In the discussion of the latter paper they state:-

• “We propose that the mammalian genome does not encode for all functions required for immunological development but rather that mammals depend on critical interactions with their microbiome (the collective genomes of the microbiota) for health.”

To put it even more simply, some genes needed by the immune system might have been “entrusted” to microorganisms. Clearly these have to be organisms with which mammals have co-evolved for a very long time, and that are always present. They will obviously not be organisms that merely cause sporadic infections. The latter can modify the human genome by elimination of susceptible genotypes, but they cannot be relied upon to supply genes and functions that we need.

This point should lead to a reappraisal of the “hygiene hypothesis”, which seeks to explain the increasing incidence of chronic inflammatory disorders in developed countries. The hygiene hypothesis has a long history, but interest in it grew in 1989 when it was observed that allergic disorders were less common in the younger siblings of large families (Strachan 1989). It was postulated that childhood infections passed on from older family members might in some way protect from allergies. To many allergologists this seemed a strange notion, because the allergies are commonest in the crowded, developed inner cities where the childhood infections are rife, and least common in isolated rural communities. However from a Darwinian point of view there are even stronger reasons for suspecting that sporadic childhood infections are not the crucial factor. Most of the human childhood viruses were picked up from animals during the close contact resulting from animal husbandry which began only about 10,000 years ago (~500 generations). Moreover, these viruses require large communities to sustain endemicity…… hundreds of thousands, perhaps millions… so endemicity is very recent (Armelagos and Harper 2005). Similarly the non-viral childhood infections are sporadic and could not have evolved the role of delivering essential genes. Perhaps it is not surprising then that several rigorous studies have failed to show that childhood infections protect from allergic disorders (Benn et al. 2004; Dunder et al. 2007; Bremner et al. 2008), despite reproducing the effect of family size and birth order. (The latter might be explained by a fascinating effect of hepatitis A virus, which binds to a receptor on T cells (TIM-1) that modulates their differentiation (McIntire et al. 2003). HAV seropositivity has declined from almost universal in about 1970, to rare or sporadic now).

So if the relevant environmental changes are not to be found amongst the childhood infections, can Mazmanian’s studies provide an alternative explanation for the increasing prevalence of chronic inflammatory disorders? This is far from clear because it is likely that in the developed countries Bacteroides fragilis is universally present in human intestines, just as it always was in the past. There is some variation in time of appearance of this organism in the microbiota (delayed by caesarian section, for example), but this does not relate to allergic manifestations (Adlerberth et al. 2007), and ultimately it is always present. I suggest that the answer to this dilemma lies in the existence of other “Old Friends” in addition to the gut microbiota, of which we have become aware via a rather different train of thought. Initially the hygiene hypothesis assumed that whatever was changing in the microbial environment was leading to an imbalance of effector mechanisms (more Th2 activity, which mediate allergies, because less Th1 was being driven by infections). However from about 1998 we started to point out that other chronic inflammatory disorders were increasing in parallel with Th2-mediated allergies, and that some of these, such as Crohn’s disease and multiple sclerosis, were mediated by Th1 cells (Rook and Stanford 1998; Rook 2000). The common factor was clearly a failure of immunoregulation that was allowing inappropriate, poorly controlled activity of both Th1 and Th2 responses, manifested as allergies, autoimmunity or inflammatory bowel disease. These ideas were amplified by others (Wills-Karp et al. 2001; Bach 2002; Yazdanbakhsh et al. 2002). Then in 2003 we suggested the “Old Friends” hypothesis. Our thinking was that if there are organisms that have an evolutionarily determined role in driving appropriate background levels of immunoregulation, they are likely to be those organisms that must be tolerated… a state dependent upon immunoregulation. These organisms fall into two classes in addition to the commensal microbiota considered by Mazmanian. First, there are helminths that although not always harmless, must be tolerated once they are established, because attempts to eliminate them are futile and merely cause immunopathology. (Helminth infections were universal, but are now rare in rich countries). Secondly, there are the organisms we have termed “pseudo-commensals” because although not replicating in or on the human subject, they were inevitably always present in the gut, because they were always present in mud and untreated water. (For instance there can be easily 10⁹ saprophytic mycobacteria in a litre of untreated water, and huge numbers of fermenting lactobacilli in non-refrigerated vegetable foods and drinks). So we introduced the “Old Friends” hypothesis (Rook et al. 2003; Rook 2007), and suggested that these organisms might have been entrusted by evolution with the role of setting up the regulatory circuits of the immune system. I suggest that these organisms are as relevant as the true commensal microbiota highlighted by the work of Mazmanian and colleagues (Mazmanian et al. 2005; Mazmanian et al. 2008). In fact it is interesting to speculate that an organism that is a “pseudo-commensal” is a risky one to use as the provider of crucial genes, because changing lifestyles can, (and indeed do), eliminate environmental organisms, while true commensals continue to be passed on through the generations. Importantly, these other “Old Friends” all satisfy the following criteria (Discussed in Rook 2007):-

1. Abundant during mammalian evolution (i.e. going back much further than the ~500 generations since the start of agriculture)
2. Virtually absent …and increasingly so over the last century….. from the modern environment (which is probably not true of Bacteroides fragilis)
3. proven to have therapeutic effects in animal models of chronic inflammatory disorders
4. some proven to have therapeutic effects in clinical trials  (Listed in Rook 2007)