MAP causes Crohn's disease, but...

Mycobacterium avium subspecies paratuberculosis (MAP) is an important trigger for Crohn’s disease. However, most Crohn’s patients don’t have MAP, thus will not benefit from anti-MAP therapy.

How does MAP trigger Crohn’s?

MAP forces our immune system to fight the fungus Malassezia, a second microbe present in and on everyone. When our immune system decides to fight Malassezia in our gut, IBD symptoms appear. Picking a fight against Malassezia is like starting the Vietnam war: there is no way our immune system can win, and the collateral damage is devastating.

Genes controlling immunity against MAP and Malassezia

The difference between a healthy gut and IBD is mostly genetically determined. Genetic variants associated with Crohn’s disease increase immune sensitivity to both MAP and Malassezia. Crohn’s patients are—unfortunately for them—hypersensitive to MAP and Malassezia antigens. This hypersensitivity is what causes Crohn’s.

This is very similar to peanut allergies. Peanuts/Malassezia are ubiquitous and harmless—except if your immune system decides to fight the peanut/Malassezia invader! In both cases, the correct immune response is to leave peanuts/Malassezia in peace. Fortunately for those allergic to peanuts, avoiding this legume prevents peanut hypersensitivity symptoms. Unfortunately for those with IBD, Malassezia colonize our bodies from birth, for life.

Why do MAP and Malassezia hypersensitivity go hand-in-hand?

MAP and Malassezia are detected by white blood cells called phagocytes. Phagocytes find MAP and Malassezia using a receptor called Mincle. Mincle is a “hook” which attaches specifically to glycolipids on the surface of these two microbes.

Under normal circumstances, phagocytes do not express Mincle, so Malassezia remain largely invisible to them—despite being present on the skin, in the gut and in other internal organs. When the body suspects that a mycobacterial infection (either MAP or Mycobacterium tuberculosis) is present, phagocytes upregulate Mincle to find and clear the mycobacteria invader. By upregulating Mincle, they inadvertently find Malassezia too, because the Mincle hook binds efficiently to both mycobacteria and Malassezia!

Though Mycobacteria (MAP) and Malassezia look very different under the microscope, human phagocytes have trouble distinguishing them because these microbes bind to the same phagocyte receptors (Mincle and Dectin-2). From a phagocyte’s point of view, Mycobacteria (MAP) are disguised as Malassezia!

How do we know that MAP alone is not sufficient to cause Crohn’s?

We know this because colitis can be triggered in rats without exposing the rat’s gut to live MAP—we only need to inject a MAP antigen called “cord-factor” into one of the rat’s paws. Here are the details.

Crohn’s and ulcerative colitis are part of a group of diseases called spondyloarthritides. The main spondyloarthritide symptoms are colitis/IBD, arthritis, uveitis and psoriasis. Symptoms are caused by the immune system attacking Malassezia in each affected organ—respectively the gut, joints/spine, eyes and skin.

An excellent spondyloarthritide animal model was accidentally developed by Pearson in the late 1950s (Pearson et al 1960). Pearson's model is still widely used by researchers currently studying these diseases. Pearson noticed that when he injected dead mycobacteria into a rat’s paw (similar to MAP), the rat would develop all spondyloarthritide symptoms within 3 weeks. The trigger for this disease was the lipid extract of mycobacteria, not the whole microbe. In 2010, it was realized that the component of mycobacteria causing symptoms was a glycolipid called cord-factor, binding to phagocytes' Mincle hook (Schoenen et al 2010).

Pearson’s rat model clearly demonstrates that simulating Mincle is sufficient to activate T cells which cause spondyloarthritide symptoms, including colitis. However, T cells only detect proteins, and can’t detect cord-factor at all. So which proteins are these T cells going after? It was long thought that the T cells were going after rat proteins: inflammation was believed to be autoimmune. What researchers did not know is that the rat body is colonized by Malassezia from birth—just like ours!

This is what’s really happening in Pearson’s rat model: T cells are looking for Malassezia proteins, which had been present all along, but which the immune system was ignoring. Exposing rats to mycobacteria (like MAP) or cord-factor is sufficient to trigger immune hypersensitivity to Malassezia. And this hypersensitivity causes spondyloarthritides, including colitis. All these symptoms are the result of a failed attempt by the rat’s immune system to clear Malassezia from its body.

The same occurs in humans: Mycobacterium tuberculosis-induced spondyloarthritides are called Poncet’s disease (Dall et al 1989). Though MAP-induced spondyloarthritides do not have a name yet, they likely occur in the same way as Poncet’s disease and Pearson’s rat model.

Venn diagram of spondyloarthritide-associated symptoms. Having any one of these greatly increases the risk of having the others, mainly due to common susceptibility alleles (for example HLA-B*27 and CARD9). Pearson showed that injecting mycobacterial cord-factor into a rat’s paw causes all of these symptoms. AAU = acute anterior uveitis, ReA = reactive arthritis, EnA = enteropathic arthropathies, PsA = psoriatic arthritis.

The key players: Mincle, CARD9 and TNF-α

CARD9 is a protein produced by white blood cells. CARD9 monitors Mincle, and determines if the Mincle hook has caught a microbe. If so, CARD9 initiates production of TNF-α, which tells T cells to attack the microbe attached to the Mincle hook.

Anti-TNF-α drugs (ex: Humira or Remicade) prevent T cells from attacking the microbe attached to Mincle, for better or worse. When this microbe is Malassezia, these drugs prevent diseases previously thought to be autoimmune, but which we now know are “Malassezia-immune”. When the microbe attached to Mincle is Mycobacterium tuberculosis, anti-TNF-α drugs are actually preventing the immune system from fighting a dangerous mycobacterial infection (Ai et al 2016), and should be discontinued.

What’s more, the CARD9 protein comes in different variants (alleles) which are passed down from parents to their children. One of these variants (CARD9 rs4077515:A) greatly increases the risk of Crohn’s disease and ulcerative colitis by increasing TNF-α production when Malassezia is detected by the Mincle hook, resulting in immune hypersensitivity to Malassezia.

Phagocytes catch Mycobacteria (MAP) and Malassezia by expressing the Mincle “hook”, which binds to glycolipids on the surface these microbes. When a phagocyte catches either Mycobacteria (MAP) or Malassezia, it uses CARD9 to secrete TNF-α, which tells T cells to attack! Each time a phagocyte catches a microbe, T cells are sent to patrol the gut for identical microbes. T cells attacking Malassezia is the root cause of each spondyloarthritide-associated symptom (including Crohn’s disease).

Individuals who carry the CARD9 rs4077515:A allele are at higher risk of Crohn’s disease because phagocytes produce much more TNF-α when they catch Malassezia. This is illustrated by a larger and louder megaphone, which results in more T cells patrolling the gut for Malassezia. This figure is based on Limon et al 2019’s Figure 6D.

Anti-TNF-α drugs (ex: Humira and Remicade) neutralize TNF-α, thus preventing activation of T cells which would otherwise have attacked Malassezia in the gut. This is illustrated by a cork in the TNF-α megaphone.

How can I clear Malassezia from my body?

We don't know yet, but we’re eager to find out! We are setting up clinical trials to this effect right now. Candidate drugs include itraconazole and terbinafine, two inexpensive antifungals which are moderately effective against Malassezia in vitro (Leong et al 2017). One small retrospective study reported that itraconazole alone might be effective (Samuel et al 2010). No prospective studies have tested the efficacy of any drug to clear Malassezia from internal organs.

Notes

(1) This post is based on Limon et al 2019 and Laurence et al 2018.

(2) Many different types of hypersensitivity exist. Peanut allergies are a Th2-type hypersensitivity, whereas IBD is a Th1/Th17-type hypersensitivity.

(3) Though this post focuses on MAP and Malassezia, other bacterial species can force phagocytes to upregulate Mincle too (for example, Streptococcus and Klebsiella). This means dietary changes and antibacterial drugs can sometimes restore or break immune tolerance to Malassezia—respectively improving or worsening symptoms—by affecting the mix of bacteria in the gut, despite Malassezia's continued presence. It is very difficult to predict outcomes of dietary changes and antibacterial therapies in Crohn’s.

(4) Dectin-1 and Dectin-2 phagocyte receptors can also detect Malassezia (and other fungi), but Malassezia are very good at hiding antigens on which Dectin-1 and Dectin-2 attach. Malassezia cannot hide the glycolipid on which Mincle attaches, which is why Mincle is the most important receptor in Malassezia diseases.

(5) CARD9 mediates activation of NF-kB following binding of Mincle, Dectin-1 or Dectin-2 to their respective ligands (glycolipids, beta-glucan and mannan). This initiates transcription of various cytokines (including TNF-α) which together launch a Th1/Th17 immune response against fungi or mycobacteria.