Toll-like receptor signaling in the intestine: friend or foe (Part 2)

Toll-like receptor signaling in the intestine: friend or foe (Part 2)


– So we started with a very simple hypothesis, which was that TLR4 signaling is important for intestinal inflammation. And so we reasoned that because we were studying innate immunity, we wanted to study an animal model that really didn’t
necessarily require TNB cells, that was dependent on bacteria, and that was something early, not like one of these chronic models ’cause we reasoned that innate immunity maybe played a role in the early stages of IBD or colitis in a mouse,
and not the later stages. So for that reason, we decided to use DSS, which you know, at the time, we didn’t really know that much about animal models, so we started with that. And as you may know from other
lectures that you’ve been to, dextran sodium sulfate is, I don’t know, like a soap or something that they drink and for some reason it
disrupts the epithelial barrier just of the colon, I would say, for some reason, ’cause
it goes down their entire intestinal track, but it seems only the colon to be affected. It really does, wreaks havoc
on the epithelial lining and this results in colitis. And it’s dependent on bacteria, it’s ameliorated by
antibiotics, et cetera. So this is the model we chose to examine. And so I just took a long
time to explain to you that TLR4 recognizes LPS, there’s this adapter molecule, MyD88, that is important for lots of TLRs. When we started these experiments, and we used a knockout of these pathways, the prediction might have been that these animals would be protected
against colitis, right? They don’t recognize bacteria, so maybe they don’t know
there’s a problem, okay? The answer is both yes and no. The animals are actually sicker in the absence of TLR4 and MyD88. This is one representation of sicker that I can get away with
showing to a GI audience. The animals bleed significantly more when they’re knocked
out for four or MyD88, and you can see that
based on their hemoglobins that their mean hemoglobin
drops significantly. We did a lot of work to
understand why that was. So they have increased
rectal bleeding, fine. If you go a little too long with the DSS you kill them more often than
you would a wild-type mouse. But in terms of inflammation,
technically speaking, actually they have less inflammation. They’re not able to recruit
neutrophils to the intestine, they’re not able to recruit
macrophages, as well to the intestine, and as
a result of this inability to recruit these inflammatory cells, we believe that’s the reason why they have bacterial translocation. Bacteria escape the gut and are found in mesenteric lymph nodes
to a much higher extent than they are in the wild-type mice. In addition, they have a very clear defect in epithelial proliferation, so it’s a defect that spans
from the epithelial monolayer of the gut all the way
to the lamina propria infiltrate that occurs in this DSS model. So for simplicity, we can conceptualize that the TLR4 knockout phenotype, at least in the setting of DSS, shows defects in leukocyte recruitment and bacterial clearance, and
that maybe is to be expected if you don’t have a receptor
in the innate immune system, but what was surprising was this epithelial homeostasis
and repair defect. And I’ll actually focus the
majority of the talk on that. So we hypothesized that
proliferation apoptosis in intestinal epithelial cells was regulated by TLR4. This is a proof of just
the proliferation part, you’ll have to trust me
on the apoptosis part, and that’s published. This is an animal that’s been sacrificed after they’ve gotten a dose of BRDU, which is a nucleotide that is incorporated into their DNA if it’s replicating, and then you can easily come back and use an antibody against it to tag it, so a dark nucleus is the sign of a cell that was proliferating. You can see that before you do
anything to a wild-type mouse they’re of course always
proliferating cells at the base of the crypt. You give them seven days of DSS and then you can see that
there is an increase, but after they’ve been
through the injury of DSS, in the recovery phase,
you can see that there’s a very dramatic proliferative response to try to repair that defect
that’s been inflicted. But look at, by comparison,
the TLR4 knockout at, again, the comparable time point at the same magnification. I hope I don’t have to convince you they have very perturbed,
very decreased proliferation. At baseline, we don’t really see, if you don’t injure the mouse, the TLR4 knockout intestines look normal. I’m not entirely confident
that they don’t have some subtle abnormalities,
but it’s so subtle that at least using level 1 techniques we can’t detect any abnormality. So, you know, with this
slide, I will summarize a couple years worth of work. To try to understand why it is that these TLR4 knockout mice
have defective proliferation. And so what we found was
that in the absence of TLR4, when you injure the intestine, you cannot express COX-2. So no TLR4, you can’t make COX-2. And COX-2 is essential for the production of Prostaglandin E2. Prostaglandin E2 has many complicated things it can do, among
them it can activate epidermal growth factor receptor signaling in an indirect way by causing the release of EGF-like factors,
we happen to be looking at Amphiregulin but there are a couple others that it can do as well. We’re confident that this
is part of the mechanism because if we give TLR4
knockout mice back PGE2, we can restore them back
to a wild-type phenotype. They proliferate and we
can bypass the defect by giving them exogenous PGE2. Another way in which this
might be working, as well, is that maybe it’s also the
macrophages cooperating. You know, the lamina
propria’s rich in macrophages. These macrophages could
also be TLR4-expressing, respond to PAMPs, and they
can then activate COX-2 and make PGE2 and still
the epithelial cell could benefit from this
in a kind of a trans way, not in cis but in trans. Okay, so all of this
led us to hypothesize, well, if TLR4 is important
for all of these processes involved in expression of
COX-2 EGFR and proliferation, ergo this might also be
involved in the extreme of that, which is the development of cancer. And so we hypothesize that TLR4 signaling is important for
colitis-associated cancer, as distinct from sporadic cancer, which is far more common
in the population. And so this is kind of a,
you know, an example of that. Unlike the idea that in
sporadic colon cancer, when Tim spends the rest of his day in the endoscopy and
he’s looking for polyp, sometimes raised, sometimes
flat, that he’ll remove and for all intents and
purposes that person is cured. IBD must be different. It looks different. These areas that go on
to become cancerous, we believe are relatively flat, probably not completely flat, but they certainly don’t become broccoli sprouts and are easy to remove. And in fact, that has
been also demonstrated in the genetic events that occur with P53 and k-ras being
much earlier in the cascade in colitis-associated cancer
and APC being a late event. So we were interested whether
or not TLR4, TLRs in general are superimposed on this
colitis-associated cancer pathway. Are they involved in initiating events, like reactive oxygen generation and things that will mutate the DNA and make it more likely
that you’ll develop a monoclonal proliferating
population of cells? Or does it work later? Once you already have
dysplasia, maybe it allows, helps in the growth of
that already-present tumor. The first thing we did
was we asked the question, is TLR overexpressed in
colitis-associated cancers and in ulcerative colitis,
and so to answer this, I think it’s dark enough. At this time of the morning
with not enough coffee I’d be reluctant to turn
the lights down any further. But for the most part,
you can see on the left that normal colon from, this happens to be a
completely normal person, just has some TLR4-expressing cells at the base of the crypts,
and Scott Plevy has thought that these are maybe
entrochromaffin cells. Whereas you can see next to it is a colitis-associated cancer, so cancer from two different UC patients, the TLR4 expression
seems to be much higher. Obviously, I’m talking to
a sophisticated audience, and immunohistochemistry
and immunofluorescence can be problematic. How do you really know that this is what you’re staining for? And remember, this is archived tissue so there are all sorts
of things that go along with dealing with archived tissue. So at Mount Sinai we
were lucky enough to have frozen tissue from a colitis cancer and the surrounding mucosa, so this is a UC patient,
the surrounding mucosa is not cancerous, but it’s
also not completely normal. You can see on the western blot, and this is the right size for TLR4, that the tumor expressed much more TLR4 than the surrounding tissue. So at least, at the moment, I would say we’ve probably done about 50 UC cancers and it’s about 80% of them
have a high expression of TLR4, whereas it’s much,
much lower in sporadic cancers.

Leave a Reply

Your email address will not be published. Required fields are marked *