24 February, 2009

Not-So-Mad Science: How EHEC Finds and Spoons Your Gut Epithelium

Figure A: This is a picture of mouse cecal epithelium infected with EHEC. EHEC is FITC green, actin is red (that's smooth muscle to the right of the picture, I didn't have the concentration of phalloidin high enough in this slide to also stain epithelial actin), and nuclei are blue with DAPI. It's at 40X, I think. I usually take prettier pictures than this.


NOTE: This post discusses some animal research. If this offends you, go somewhere else.

Enterohemorrhagic Escherichia coli (EHEC) is an enteric human pathogen that has been the cause of many food-borne disease outbreaks over the past 3 decades. One of the most famous outbreaks was borne by Jack-In-the-Box fast food restaurants in contaminated beef and the most recent outbreak was a spinach scare of 2007 when cow manure was used as fertilizer and then improperly washed. The primary manifestation of an EHEC infection is bloody diarrhea (hemorrhagic colitis), although intestinal cramping is apparently also common. In most cases in most adults, EHEC infection is just very unpleasant and complications are rare. But when complications do arise, they are nasty and life-threatening. The primary complications are sepsis and hemolytic uremic syndrome (HUS). HUS is characterized by thrombocytopenia (low platelet count), hemolytic anemia (exploding blood cells), and uremia (acute renal failure). Thrombosis can be widespread and lead to microangiopathy in the gut, kidneys, and CNS.

Doesn’t EHEC sound friendly? This is why you should always cook your beef very very well, because EHEC doesn’t cause pathology in cattle like it does in humans (we don’t know why yet, either), and also because EHEC can establish a productive infection with an inoculating dose as low as 25 CFUs (CFUs = microbiology term, approximately equivalent to 1 bacterial cell).

Epidemiologically, patients hospitalized with HUS usually have stool (poop!) samples positive for EHEC, specifically strains of EHEC that carry the genes for Shiga toxin. In vitro tests with Vero cells (a kidney cell line from monkeys, I think but am not sure that they’re from African green monkeys, also known as vervets, that are disappointingly not actually green) and human glomerular epithelial cells have shown that Shiga toxin is directly cytotoxic to these cells. There is not yet a good animal model for the pathophysiology of HUS. When mice are injected with Shiga toxin, they develop some kidney damage, but none of the wider complications. When rabbits are injected with Shiga toxin, they quickly develop neurological symptoms, but the other symptoms remain unreliable. And did I mention that the LD50 for Stxs are generally in the micrograms/kg body weight kind of range? Fun, huh?

Most EHEC strains make Shiga toxins (Stxs), which are AB5 toxins N-glycosidases that cleave specific bases on the rRNA, thereby shutting down protein synthesis and killing target cells. Not all EHEC makes Stx equally, and different strains of EHEC carry different types of Stx. From a diagnostic standpoint, this gets confusing.

Thankfully, EHEC isn’t invasive like Shigella, Yersinia, or Salmonella species. That is, it doesn’t disrupt the epithelial barrier in the gut and directly invade the underlying tissue. This isn’t to say that this won’t happen later on in infection when gut epithelial hemorrhage becomes more widespread. It’s more to say that EHEC doesn’t have the machinery to mount a direct invasion. However, EHEC does snuggle up to the gastrointestinal epithelium (GIE) quite nicely (not as cute as one might imagine). It even convinces the GIE to build it a nice little actin pedestal. Basically, EHEC cozies up to the GIE and assembles a Type III secretion system and then shoots a bunch of proteins into the GIE cells (see? I told you is isn’t cute), most prominently Tir (translocated intimin receptor). These proteins are then naturally inserted into the outer membrane of the GIE cells, and EHEC has the ligands for these proteins, mostly intimin, already displayed. So EHEC locks on and secretes still more proteins into the poor GIE cell it is humping and these proteins rearrange the host cell’s actin cytoskeleton to form up a pedestal. We don’t know exactly why EHEC forms these pedestals, and they’re really hard to find with fluorescent microscopy (I’ve been trying with infected mouse cecal epithelium), but it thought that the pedestals protect the EHEC from innate immune cells, give it a base from which to multiply and form tightly-adherent micro-colonies, and/or to prevent EHEC being swept out by peristalsis.

But this then brings up the question: how does EHEC find the GIE in the first place? Sure, there’s going to be some random collisions that might result in adherence, but in general this isn’t enough. So, scientists being scientists, some scientists set off in search of an answer to that very question. They’re cited down below.

Bansal et al did some cool in vitro experiments where they examined the chemotaxis, colonization, and gene expression responses of EHEC to epinephrine (EPI), norepinephrine (NOR), and indole. What they found was that EHEC moves towards both EPI and NOR and away from indole. EPI and NOR are produced by human hosts, although the exact in vivo gut levels thereof are currently unknown (any volunteers?). Indole is produced by bacteria. So the authors posit that the GIE is producing EPI and NOR naturally, and that the EHEC are sensing these chemicals and accordingly moving towards them.

These findings were borne out well. This group didn’t just measure the direction EHEC swam in, but they also looked at biofilm formation (EPI and NOR increased, indole decreased) in vitro, looked at in vitro attachment to cultured epithelial cells in the presence of EPI, NOR, and indole (same results), and also at gene expression. Gene expression analysis revealed that Tir and intimin et al were upregulated with EPI and NOR and decreased with indole.

They didn’t look at whether or not EPI, NOR, and/or indole modulated expression of Stx or other non-adherence-associated virulence factors, but this is something I would be interested to see them try. I did twice. An interesting note is that they co-cultured EHEC with HeLa cells (I know, not GIE, but still epithelial) for 3h to test whether or not adherence was modulated. I have done a similar experiment looking for Stx induction under epithelial co-culture conditions, and I found that 0.6% of EHEC were stuck to the epithelial layer within 15 minutes. That might not seem like much, but when you’ve got 10^6 EHEC cells floating around, that’s rather significant (still 6,000 CFUs!). And in 3h I saw that increase to 8.3% of all cells. I’m impressed, aren’t you? However, I never tested my epithelial monolayers (I did this with both HeLa and Caco-2) for EPI or NOR production, nor did I monitor indole levels in any of the cultures. Still, it’s cool that this group was able to begin answering the question of how EHEC finds the GIE without eyes.

I’m still not sure whether or not to regard them as competitors. Either way, they’re at Texas A&M, so maybe they’ve already got it bad enough.

Bansal, T., Englert, D., Lee, J., Hegde, M., Wood, T., & Jayaraman, A. (2007). Differential Effects of Epinephrine, Norepinephrine, and Indole on Escherichia coli O157:H7 Chemotaxis, Colonization, and Gene Expression Infection and Immunity, 75 (9), 4597-4607 DOI: 10.1128/IAI.00630-07

NOTE 2: I do not know what's going on with the font. I guess this is the price I pay for writing this post in MS Word instead of Notepad. There is no meaning to the font being different for separate paragraphs.

1 comment:

Kat said...

Two corrections I think are in order. EHEC does not actually produce Shigatoxin, but rather a Shiga-like toxin called verotoxin. Second is that verotoxin and other shiga-like toxins need certain receptors found in humans to do their nasty business, and apparently cattle don't have these receptors, so they carry the bacteria without getting sick.