Hello all, since my trip to the conference here in Keystone, CO, has been graciously funded, in part, by GrEBES, it is my duty to report back to you what I’ve learned each day. The conference (actually a joint symposium on Innate Immunity and the Microbiome) began last night with two keynote talks by Ruslan Medzhitov (immunity) and Jeffrey Gordon (microbiome). However, I was too tired from traveling (and perhaps an hour and a half of skiing) to write about it right afterwards. So here it is now
– Ruslan Medzhitov talk, titled “Damage control in host-pathogen interactions”, focused on the importance of considering host tolerance of pathogens when thinking about and studying how hosts protect themselves from the damage associated with pathogens.
Avoidance – behavior that limits exposure to pathogens, such as avoiding food that tastes or smells rottenResistance – reducing pathogen load via the actions of the immune systemTolerance – reducing the negative impact of pathogens on host fitness without affecting pathogen burden
The reason tolerance is important is because avoidance is not always possible, and resistance comes with a cost, namely, that many immune response damage the host as well as the pathogen. Each different immune response (referred to as effector mechanisms by Ruslan) has a different fitness cost associated with it. The effectors with the lowest cost tend to be those that are constitutive, e.g. physical barriers such as the skin, and production of IgA (Ruslan admitted determining *actual* fitness cost is very difficult, and often proxies are used that may or may not have any relevance to the evolution of these particular traits) whereas effectors with higher fitness costs are only expressed more conservatively. These fitness costs, then, construct the hierarchy of immune response.
For any effector mechanism, an increase in expression results in an increase of immunopathology (disease symptoms resulting from the immune system itself). However, coexpression of multiple effector mechanisms means that each one can operate below the level that causes immunopathology. And, indeed, inactivation of particular effector mechanisms often results in the hyperactivation of another that leads to immunopathology.
The general mechanism of this trade off is that mechanisms of tolerance come at a cost to normal function. Thus, many mechanisms of tolerance are inducible, and induction of tolerance to a particular pathogen can compromise a concurrent response to another infection. This results in situations such as increased mortality due to bacterial pneumonia following an influenza infection.
To illustrate this, Ruslan presented findings from coinfection of mice with influenza and Legionella. Separately, neither hardly ever causes mortality, but together, thee mortalities rates are incredibly high. What’s particularly interesting, though is that in both cases of Legionella-only infections and coinfection, the bacterial loads in the mice are very similar, indicating the mortality is caused not by increased pathogen loads, but a decrease in tolerance to the pathogen.
Unfortunately, attempts to reduce the damage to the host via inflammation via anti-inflammatory treatments did not rescue the mice, and defects in innate and adaptive immunity actually increased mortality. RIP3 (an enzyme in the apoptosis signaling pathway) knockout mice had prolonged survival during coinfection, but eventually succumbed, and blocking apoptosis actually resulted in more necrosis and and pathology in the lungs. So, attempts at boosting tolerance to reduce the fitness cost of pathogens seems to be a much trickier problem than the beginning of the talk foretold. One reason that might be in this case is that the influenza causes significant epithelial damage in the lungs (which the Legionella cannot do on its own), which is then exacerbated by coinfection with the bacterium. However, preventing apoptosis in this case to preserve the epithelium and prevent invasion by the bacterium may in fact only result in the death of the cells anyway from lysis by the virus.
So, there’s definitely a lot of work to be done here, but the idea of effectively reducing a pathogen’s virulence by increasing host tolerance is a very attractive one, especially because it could possibly avoid the evolutionary arms race that often results from increased resistance. And understanding the mechanisms of tolerance are important for determining how situations such as Typhoid Mary can arise.
– Jeffrey Gordon‘s talk, titled “Dining in with trillions of friends: exploring the gut microbiome”, focused on what we currently know and what we need to know about how the microbiota affects the nutritional value of our food, a topic of increasing importance as the human populations continues to grow at a rapid pace.
What we know already is there are significant diet x microbiota x microbiome interactions (one thing to point out here is that the microbiota is the collection of microbial taxa associated with the host while the microbiome is the collection of genes supplied by the microbiota). Various studies in mice and humans, particularly the Calorie Restriction Society whose members take meticulous notes on their eating habits, have illustrated these interactions very well.
Of particular interest to the Gordon lab is how nutrition, and specifically malnutrition, affect the assembly and composition of the microbiota, and vice versa. Malnutrition generally comes in two forms, marasmus and kwashiorkor. Marasmus is much more abrupt and the prognosis is generally worse, however both are generally treated using Ready to use therapeutic food (RUTF). To determine how malnutrition and the microbiota affect each other, they studied 317 twin pairs in southern Malawi that were either both healthy, discordant for malnutrition (only one had it), or both mal-/under-nurished.
They first looked at how the microbiota matured through time (i.e. it’s progression from an infant-like community to an adult-like community). They noticed malnurished children had a delayed maturation of their microbiota. Administration of RUTF to discordant malnurished children intiallly caused rapid maturation of their microbiota, but it did not continue the way it did in their health co-twins.
To do more manipulative studies, they collected samples of the microbiota from the twins from fecal samples, froze them and then used them to colonize germ-free mice (humanized mice). What they found was that mice given microbiota from kwashiorkor children with a nutrient deficient Malawian diet experience weight loss (this didn’t seem that surprising to me).
Furthermore, they created clonal arrays of cultured bacteria from donors (apparently the fraction of gut microbes that ad culturable is much higher than for other bacterial communities) which could be sequenced and used for testing their individual responses to different diets. They also used INseq, which uses transponson insertion to disrupt genes and figure out which ones are required for successful colonization and growth in the gut. This revealed putative essential genes in four Bacteroides species including seven genes in a single polysaccharide utilization locus.
Creating various diets to determine what was required for successful bacterial growth, they developed a four ingredient linear model that could explain 60% of the variation in species abundance without considering, for example, the host or any other factors.
Lastly, the used they model communities from the arrays to see how they respond to an introduction of a new strain, say from fermented milk products (yogurt). What they found was very little change in the community, but a significant change in the metatranscriptome.
Jeff Gordon concluded by saying he hopes these kind of studies will lead to (1) forecasting diet x microbiota interactions, especially by putting the nutrional value of food into the context of the microbiota, (2) the identification of next generation probiotics from various communities, and (3) proofs of concept, efficacy, efficiency, and safety for possible future therapeutics.
More to come soon