Endometriosis Foundation of America
Medical Conference 2019
From Biomarkers to Precision Surgery
March 8-9, 2019 - Lenox Hill Hospital, NYC
Thank you. I'd like to thank Dr. Seckin and the Endometriosis Foundation for inviting me and for putting on such a wonderful forum for us all to get together and discuss this very important topic.
I've entitled the talk today Endometriosis: Stem Cells, micro RNA, and Inflammation. I'd probably add in their retrograde menstruation as well. I did, on that slide just to throw something into the controversy here. I intentionally made the title a little bit long and complicated to make us think that it's not just one issue; it's not whether it is or is not retrograde menstruation. This is a very complex disease or even a syndrome and we need to think about it as systemic, complicated, multifaceted disease.
I think again, I don't have to explain to this audience what endometriosis is, but I think the point is we see many different forms. We've all talked about the different colored lesions, the endometriomas versus the peritoneal lesions. We've talked already at great length about the different clinical presentations. They may have pain, they may have infertility, they have neither. It clearly isn't behaving as one single entity.
We know that these symptoms aren't related to the stage of the disease. Very poorly correlated, so I think we really have to think of this not just as the lesions in the pelvis, but a more systemic disease that has multiple manifestations and I'll give you some evidence today about some of the mechanisms that drive that.
Let's talk about Sampson's Theory. I think we all are familiar with Sampson's Theory. We've already had a debate about that today and I am a believe still that most endometriosis arises through Sampson's Theory of retrograde menstruation and settles in those dependent portions of the pelvis. I'm going to put some data today and a paper that I just got news was accepted today in "Biology of Reproduction" where we mimic retrograde menstruation in a mouse model.
Here, we take mice. We create endometriosis, experimental endometriosis, by placing uterine tissue in the peritoneal cavity, and then we digest endometrium from GFP mice. They have that green tag signal that we can identify the endometrium that we are then injecting into the peritoneal cavity of these mice. It's dispersed, single-cell, injected into the peritoneal cavity, free-floating, and then we take out those endometriosis implants and we say, "Are these free-floating cells, what we postulate as the equivalent of retrograde menstruation, are they actually incorporated into the endometriosis lesions?"
We sort them for GFP and you probably can't see it from the back of the room. Whoops, sorry. Let me go back. That there are some GFP cells here, positive cells, in those endometriosis lesions. When we look at them to define them again, you probably can't see this well from the back of the room, but we see some of those cells injected and incorporated into blood vessels; some of them have CD45-positives, so there's some immune component that comes from, again, these injected free-floating cells and incorporate into the lesions; some of them are stromal cells stained with vimentin; and here, GFP vimentin. Here, some of these GFP cells also show expression of PCNA to show that they are proliferated. They proliferate, they expand, they contribute to those endometriosis lesions.
I think this tells us that those free-floating cells can adhere and attach and contribute to the continued growth of endometrial lesions.
But of course, I don't think all endometriosis comes from retrograde menstruation. I don't think it has to be an either/or. I don't think retrograde menstruation into the peritoneal cavity explains endometriosis that we see in the lung, the brain. It certainly doesn't explain the endometriosis that's been seen in men, especially back in the old days when men got high doses of estrogen for prostate cancer therapy. I haven't heard any case reports yet about transgender women with endometriosis, but I expect we'll be seeing some of those as well. I think there are other ways we can develop endometriosis other than just retrograde menstruation.
That led us to look at stem cells. talked to you about some of the progenitor cells and induced stem cells, but there are natural sources of multipotent stem cells that have been shown to contribute to other organ systems. Probably the richest source of multipotent stem cells is the bone marrow. The bone marrow contains these mesenchymal stem cells that can differentiate to multiple different mature cell types. We reasoned if they could become many types of other tissues that don't have the dramatic turnover that endometrium does, certainly they should be able to contribute to endometrium and endometriosis.
It's going back now for over 10 years. We did some simple experiments to prove that bone marrow cells could indeed become endometrium and endometriosis. This is where we took male bone marrow, we transplanted it into females, and we simply looked for a Y chromosome. Could we detect that in the endometrium?
Here's a male-to-male transparent. The red dots there, it's Y chromosome fish looking at Y chromosome. Sorry, this is a male, this is a positive control. This is the uterus of a female that did not get a bone marrow transplant as the negative control. This is the uterus of a female mouse that got a bone marrow transplant, and you can see, this is the uterine lumen, this is the epithelium on either side and here is an endometrial epithelial cell that has a Y chromosome. It's really coming from the bone marrow transplant from that male. I won't show you here, but we did multiple different types of characterization of these cells to really show they're truly epithelial cells, not migratory leukocytes. This is a stromal cell.
Then we looked at women who had bone marrow transplants. These are women who had had bone marrow transplants in the old days when they gave chemotherapy, radiation, and then actually, it took bone marrow biopsies, not the peripheral mobilized cells. And then looked at some endometrium biopsies to see if we could find cells in the endometrium that had the HLA type of the bone marrow done rather than the recipient. Could there be bone marrow-derived cells that are incorporated into the endometrium?
Here, the brown staining shows the mismatched HLA type, the donor HLA type, in the endometrium of women who had received a bone marrow transplant. Here, the brown staining here in the epithelium. Here are some stromal cells. The brown arrows point to the mismatched HLA type. The arrowheads point to the endogenous stromal cells. They do not stain for the mismatched HLA type. The red-colored cells are the leukocytes that also stain brown, as you imagine, after a bone marrow transplant. All the hematopoietic cells that might migrate in and out of the uterus would be of the bone marrow donor HLA type.
So in mice and humans, we've shown clearly that bone marrow cells, one type of stem cell, can go to the endometrium. This is in our endometriosis mouse model too. Not only endometrium but in endometriosis.
Here, we use the beta-galactosidase gene as a marker. You can see that these bone marrow-derived cells, after bone marrow transplant, incorporating into endometriosis lesions. This is by immunohistochemistry and this is looking for enzymatic activity. So two different ways of assuring that these cells really are bone marrow-derived cells incorporating into the endometriosis.
I think stem cells and bone marrow-derived and maybe even other endogenous stem cells may contribute to endometriosis. I think this may be a major source of the endometriosis that we see outside of the peritoneal cavity or in men that is not likely to be derived from retrograde menstruation, and certainly, in men, it's not hematogenous or lymphatic spread from their uterus. It really, in some sense, constitutes a novel mechanism of disease. Ectopic inappropriate differentiation of stem cells probably occurs in other diseases as well. We clearly see here that it can be a source of diseases such as endometriosis, but could there be other diseases that we haven't identified yet where ectopic inappropriate differentiation of the stem cells contributes to the disease? I think so. We'll be looking for more of those.
The other thing we found, and this work was done by somebody who was in my lab, is here in the audience, who did much of this work when she was with me a couple of years ago. We looked at, okay, we've got some stem cells, progenitive cells in endometriosis. These cells we now know that cells are migrating throughout the body and contributing to endometriosis, but could cells come from the endometriosis and actually travel to other distant or remote locations?
We showed we could detect these cells in the circulation. Lo and behold, when we started looking and sorting cells in different organs after transplanting tagged endometrium into mice, so our endometriosis model, we had various fluorescent tags that we could use to identify the endometriosis cells, and then we looked at all sorts of other organs throughout the body and asked, when we sorted these cells, could we find some of these tagged cells that originated in the endometrium in these remote locations?
We looked in the lung, the spleen, the liver, the brain. Here's a negative control and here's our endometriosis animals. You probably can't see it from the back of the room, but as we go to the right here, these are very positive cells for this fluorescent marker, and you can see that all of these different organs had a few, not many, but a few cells from the endometriosis. Endometriosis cells are traveling to, essentially, every organ we looked at and in every animal, we looked at, 100% of them had at least a few cells that we could detect. If we look at different organs, not as many in the spleens; quite a few in the brain; more in the liver and lung were the more common regions.
Again, these would be clinically unrecognizable, single or small clusters of cells. We looked at various markers to show that they were certainly derived from the endometriosis. These are not forming endometriomas, but there's certainly some sort of cell migration transplant of cells from endometriosis to multiple other organ systems. How that plays into symptomatology, how that plays into this syndrome of endometriosis, we're not sure yet, but it certainly makes one wonder if this indeed could have clinical manifestations that are yet unrecognized. We really have to start thinking about endometriosis as not just the lesions in the pelvis, but a disease that affects multiple organ systems.
Stem cells, certainly we know ... Well, I shouldn't say, "Certainly we know" to this audience. I suspect stem cells come from the uterus in a retrograde fashion and implant in the peritoneal cavity to create most endometriosis. This is a cartoon.
They can come from bone marrow and go to uterus or endometriosis. That's another source of this. But then they leave the endometriosis. I didn't show you, some of them go back to the uterus and can actually reincorporate in the endometrium. Some of them, in smaller numbers, go to these remote organs causing this systemic-type effects.
Then we asked could there really be effects of endometriosis remote from the peritoneal cavity? We wanted to look at the endometrium, which we know is affected in women with endometriosis and can be a cause of implantation failure, diminished endometrial receptivity. We took a mouse model, we implanted endometrium right near the uterus in the pelvis to mimic typical endometriosis. We also, in a separate set of mice, placed endometrium under the skin and the should blade, remote from the peritoneal cavity. We asked what happens to gene expression in the endometrium?
Indeed, some things were the same with the remote endometriosis, somethings were less affected. HOXA10, we've previously shown it's a gene important for implementation, is not as affected with the remote endometriosis, but progesterone receptor was affected even more dramatically with the remote endometriosis. So clearly, endometriosis, even remote from the peritoneal cavity, can have an effect on the uterus and other organ systems.
We also started looking at another signaling mechanism and that's microRNAs. Some of you are probably familiar with this, but these are short, non-coding RNAs. They're in the genome. They are transcribed, but they are not translated into proteins. They are processed into these short, approximately 22-nucleotide fragments that then combine the messenger RNAs of other genes and typically block the translation of those genes or may lead to degradation of that RNA. They generally block the expression of other genes.
These are secreted from cells and can be found in the circulation, typical in exosomes. We had done some work in endometriosis awhile back where we found several microRNAs have very differential expression in the endometriosis lesions. Here, we look at one called Let-7. Very low in endometriosis patients.
What does Let-7 do? Well again, microRNAs generally block gene expression. What genes do we know Let-7 blocks? Well, some such as RAS and MYC, which are known for important genes for mitotic signaling; cell cycle genes; genes involved in cell adhesion and angiogenesis. It stops the expression of these genes. In endometriosis, we show that it's low, so we'd expect all of these genes that are involved in proliferation, migration, adhesion all to be increased in women with endometriosis.
But then we asked well, if it was there in the endometriosis, like so many other microRNAs, could we find differences in microRNAs in the circulation? Was it not just localized at the lesion? Were these having some sort of widespread systemic effects?
We did a large-scale screen of the 2, 3,000 microRNAs that are known to exist and we found several of them in the circulation were dramatically increased or decreased in women with endometriosis with a potential to develop this as a biomarker of the disease. We've been working on that as well. We published some of that before and here's the most recent data we presented at ASRM and I have a publication now under review looking at developing a panel of circulating microRNAs as a non-invasive endometriosis biomarker. Again, two different independent datasets show that this is a very good receiver operator characteristic curve here. It turned out to be a very good test for endometriosis.
But our main intent was not to originally define biomarkers, but really to look at the systemic effects of endometriosis, really to look at the underlying biology of the disease. These microRNAs aren't floating around the circulation just waiting for us to measure them as biomarkers. We suspected that they had a functional role.
Here, we look at microRNAs that are differentially in abundance in endometriosis and their effect on various inflammatory cytokine productions by macrophages. In the black bars here, we mimicked what was going on in endometriosis in terms of the microRNAs. micro RNA 125 is one that's increased in endometriosis. Let-7, I showed you, is one that decreased, so in black here, we used the Let-7 inhibitor. You can see TNF-alpha, IL 1-beta, IL 6, IL 8, all inflammatory cytokines that are known to be increased in endometriosis are driven up by these changes in these microRNAs that are exactly what we see in endometriosis.
We think some of the systemic inflammatory phenotypes that we see in endometriosis is modulated not just by the lesions themselves making these microRNAs, but the effect on macrophages and other leukocytes and their effect on cytokine production. The microRNAs probably are having, again, a systemic effect. This is happening throughout the entire body as these microRNAs are in the circulation acting on immune cells also in the circulation.
We've looked in other models as well as some of the systemic effects of endometriosis. This is looking at weight and fat distribution. We know from epidemiologic studies that women with endometriosis tend to have a lower BMI. Is that a risk factor or is that cause and effect? Well, when we used our animal model and we induced experimental endometriosis in the animals, they lost weight or never gained the same amount of weight as those that had the sham surgery. We found that they had differences in body weight, fat content, and what I won't show you here is that many enzymes in liver were differentially expressed so there is a metabolic phenotype associated with endometriosis. Endometriosis is affecting the liver. Some of these are regulated by the microRNAs that I showed you earlier, but there clearly seems to be an endometriosis metabolic phenotype.
We also looked in the brain. We know women with endometriosis have central sensitization to pain. We know that they have a higher incidence of anxiety and depression, so we wanted to test these in the animal model to see if this was a cause and effect relationship.
We took animals. We had sham surgery or we induced endometriosis in half of them. When we looked in their brains and we could see differences in gene expression, multiple genes, we did an array but some of the ones that were most impressive, I'm showing here. We saw differences in electrophysiology. We did a patch clamp on the neurons from the brain. They are sending out different electrical signals. They are functioning differently.
Finally, we did behavioral testing on the mice. These were mice that weren't showing any outward symptoms of pain, the classic ways that you monitor pain in mice. But yet, when we looked at their movement throughout a cage which is a measure of anxiety in mice, you could see that the sham mice moved throughout the cage, whereas those with endometriosis tended to congregate more on the sides of the cage, the outer portions of the cage. That's a classic established test for anxiety in mice. You can see the endometriosis mice all spent less time in the center of the cage showing higher anxiety in mice caused by endometriosis.
We looked at the duration of putting the paw on the hotplate, which is a measure of pain sensitivity. The endometriosis mice had higher pain sensitivity. We looked at a test called tail suspension test where mice that are depressed will be sort of lethargic and you hold them up by your tail and they lay there for awhile limply, whereas mice that are not depressed, they're very vigorous. You can see in this mobility assay, the standard test for depression shows that the endometriosis mice had more depression.
Again, endometriosis is affecting the brain and some of our work that I haven't completed yet and we haven't published yet are showing some of these genes are affected directly by the microRNAs in the circulation that is entering the brain, causing the anxiety, depression, and central sensitization that we see in some women with endometriosis.
It really is a systemic disease that affects a large number of organ systems. I think we have to think about endometriosis as more of a syndrome as a widespread, whole body, systemic disease. While we can target visible lesions in the pelvis, I think that's very important and I think that's what's initiating a lot of these systemic effects, we have to remember that some of these effects may be lasting. When we treated the mice after inducing these brain effects, the behaviors did not always reverse. So important that we need to recognize, even when we remove the lesions from the pelvis, some of the systemic manifestations may still be present and we still might need medical therapy, and early recognition of the disease before you have all of these widespread effects are important. So a movement towards early recognition, clinically, surgically, and by non-invasive biomarkers.
Of course, we don't have great medications that work. Progestin resistance is common. Progestins, in our studies, women, many of them had lower progesterone receptor expression levels that we show in this paper and other studies, including Dr. Boland's, have shown progesterone receptor epigenetic modifications. We know that the progesterone receptor's defective in many patients.
Indeed, progesterone, we look at a number of stem cells recruited. Progesterone does a lousy job at blocking stem cell recruitment to endometriosis lesions. Lowering estrogen has a better effect. Here's medroxyprogesterone. Didn't lower them, but when we used Letrozole or here, a GnRH agonist, you can see you get a nice reduction in the number of stem cells recruited to these lesions. I think that's probably a better approaching.
I'm running out of time now, so I'll just not linger on this, but I at least wanted to say a couple of words about the new GnRH antagonist, Elagolix, that's out in the market that's proving to be effective. I was involved in some of the clinical trials here. They're proving to be quite effective and I think the lowering of estrogen levels seems to, in our models, also improve recruitment of stem cells or reduce recruitment of stem cells to these lesions, corrects some of the micro RNA imbalances that we're seeing. Probably a better approach than the progestin therapy, at least for a large number of women who have progestin resistance.
But I'll end here by just saying that endometriosis is certainly a biologically complex and systemic disease. We need to not just focus on the lesions in the pelvis but think about this as a wider, whole body systemic disease. Thank you.