Asgi Fazleabas, PhD
The role of microRNAs in the pathophysiology of endometriosis
Advancing the Science and Surgery of Endometriosis
Monday and Tuesday, April 18-19, 2016
The Union Club, New York
Thank you so much and again Tamer, thank you for the invitation. This is my, I think, the fifth time I have been here and I really do enjoy coming here catching up with old friends and talking about the work we do.
What I thought I would like to do with you today is bring you up to date with some of the work. I am a basic scientist so I really ask fundamental questions related to this disease that we are all so very interested in and this very enigmatic disease. What I am going to do is take you through a little journey and hopefully I can get that done in 15 minutes about some of our more recent work and trying to understand the implications of microRNAs and their role in the pathogenesis of this disease.
The question of course is what kind of model can we use? We all know that by the time the diagnosis is made in women with disease it has been there for an extended period of time so it is very difficult to ask the primary biological questions about what initiates this disease. How does this disease impact both pain and infertility? We have to come up with appropriate model systems and several years ago my laboratory, along with Thomas D’Hooghe’s laboratory in Belgium, developed a baboon as a non-human primate model. This is a ___ monkey that gets spontaneous disease, their physiology, their anatomy and endocrinology is very, very similar to the human female. Because of the pathogenetic relationship and the (mic cuts out briefly) women it allows us to utilize this model to try to understand the fundamental aspect of this disease. Basically we use this animal model to essentially create a peritoneal disease with the implications of retrograde menstruation. Essentially what we do with these animals is once we evaluate them laparoscopically because they do get spontaneous disease and make sure they have no spontaneous disease we inoculate the peritoneal cavity with menstrual tissue on two consecutive menstrual cycles. And with that these animals will develop full-blown disease that is very, very similar to what you see in the human situation. This is a baboon that has got endometriosis three months after we have inoculated these animals with menstrual tissue and you can see how similar the peritoneal disease is to the human female.
The advantage of this animal model is that we can then sample the same animal prior to and after the induction of the disease so we know from day one when we inoculated these animals when the disease began and then we can follow the progression of the disease in these animals. We utilize this model for several years and essentially what we have been able to show is that since our interest is really the impact of this disease on infertility is that progesterone resistance, which is now considered one of the hallmarks of this disease occur sometime about six to nine months after the disease is initiated. What you see basically is that you see a very early response once you inoculate these animals in terms of the gene expression of the endometrium, which has a very estrogenic response in the mid-secretory phase. We always focus on the mid-secretory phase because that is the window of implantation, the time at which the embryo attaches to the endometrium. Subsequently somewhere between six and nine months the endometrium changes and becomes what we call progesterone resistant in that the normal genes that are regulated by progesterone during the mid-secretory phase that we think might be important for the initiation of pregnancy are all affected.
In addition when we look at the global changes, so using alphametrics and various genorays what we were also able to show is that one month after we initiate the disease in these animals and we look at their endometrium using global gene expression technologies that 40 percent of the global changes that we see in these animals at one month are compatible to what is being reported in women who had the disease for the extended period of time. Essentially what this is telling you is that the early insult that you see in the peritoneum is having a marked effect on the endometrium.
We have done a lot of this work and what I want to really focus on today is trying to ask the questions in terms of what is the upstream mechanisms that might be involved in some of these processes that are responsible for contributing to infertility that is associated with this disease. Based on all these questions we went back then to try to understand whether microRNAs play a role in this process.
Essentially what microRNAs are is they are small 17 to 24 base, single-stranded RNA molecules that have a profound effect on gene expression. Essentially, just to give you a very small prima on what microRNAs are basically pre-microRNAs that are processed within the nucleus I exported through very unique mechanisms into the cytoplasm and they undergo specific processes where the mature microRNA complexes with this RNA silencing complex. Essentially what this does is binds generally to three prime regions of the RNA and essentially acts as a break. So unlike we think about normal transcription where either if genetic transcription or regulation brings about the results of microRNAs which then translate into protein that have function, which is the classic way of looking at what happens during gene regulation micro-RNA regulation actually acts more as a break. Essentially it does the opposite of what we think of normal transcriptional regulation in that microRNAs in general tend to suppress protein expression and thereby function.
Essentially with micro-RNAs, one of the breakthroughs of the year in 2002 and just to give you a feel for how this field has just exploded this is just a publication using simply the title microRNA using PubMed. You can see from its initial discovery, which was 2002 when it was the molecule of the year, you can see this precipitous increase in the number of papers that have been published in the last ten to 12 years with regard to trying to understand the role of microRNAs as it relates to biological function.
When it comes to endometriosis then there have been a number of papers that have described micro-RNAs that have been isolated, described and identified with regard to micro-RNAs and endometriosis. Since the first report related to micro-RNAs and endometriosis in 2007 PubMed now shows between about 92 and 100 papers that are associated with micro-RNAs and endometriosis. But the question still remains whether this is a cause or consequence of this disease that we see these changes in micro-RNAs which is always a question we always have in endometriosis because it takes so long to diagnose the disease. The second question is what do these micro-RNAs do if they are being regulated, how do they work, and how do we identify target genes that we can potentially target or suppress depending on what we want to do with regard to treatment and therapy of this disease.
What I am going to take you through is how we have approached these questions and how we have been trying to in a very short period of time tried to address these issues with regard to cause and consequence and function of these micro-RNAs. Going back then is how are these being regulated? So again we have the advantage of our baboon model where essentially we can use human platform, so that is the advantage again because of close phylogenetic relationship we can do direct comparisons between human and non-human primates. Essentially what we did then is an initial experiment was take endometrium from animals prior to the induction of the disease and three months after the induction of the disease and using Agilent array tried to identify the micro-RNAs that were altered in response to the disease doing a direct comparison in the same animal model. What we basically find is that there are a number of micro-RNAs that are significantly altered, the ones in bold are the ones that have also been reported to be altered in women. So here again you can see within three months of the induction of the disease using peritoneal inoculation of menstrual tissue you can see the micro-RNAS that are very, very similar to what has already been reported in women.
For the purpose of today’s talk I am going to focus on three of these, micro-RNA 451 which is highly downregulated in response to disease, and 29C and 21 which are upregulated in response to the disease. What I am going to do is try to take you through how we tried to approach what the potential mechanisms and functions of these micro-RNAs are.
Looking through the literature the decrease in micro-RNA 451 which has been extensively studied in breast cancer, in hormone resistant breast cancers the decreases seen in breast cancers is associated with the enhanced proliferation and invasion. Again, you know this is a characteristic of endometriosis. Micro-RNA 29C which is upregulated targets one of the genes it targets a number of genes. That is the one thing with the RNAs, they are complex. They target multiple, sometimes thousands of genes so one micro-RNA can have lots of regulatory mechanisms. But the one we are interested in is specific immuno fill in called FKB 4 which is a target of micro-RNA 29C which we think contributes to the progesterone resistance of the disease. Very recent work in looking at micro-RNA 21 which is also highly upregulated in breast cancer and it promotes fibrosis, inflammation and EMT, again mechanisms that we know are associated with this disease.
I am going to take you through very quickly in terms of how we try to approach these and try to see whether there are functional implications associated with this disease. First and foremost when you do any kind of array you want to make sure that that array validates what you are seeing by looking at the tissue itself. This is the 451 expression which is the one that is markedly downregulated and then you can see then in the baboon model progressively as once the disease is induced you see a dramatic down regulation across the timeline in the same animal, the same group of animals between one and 15 months along with the lesions that are then harvested at 15 months. If you then look at women with this disease you see the same thing. Eutopic endometrium in women with endometriosis compared to controls, again, all in secretory phase. You can see the dramatic down regulation of 451 which complements what we see in our induced model.
Then of course looking at targets One target we have been really interested in looking at is YWHAZ at 14.3.3 zeta which is a gene that is involved in regulating proliferation. And what you can see is just as you would expect when a micro-RNA goes down its ___ gene goes up and you can see that the expression of this gene both in the eutopic endometrium and the ectopic endometrium is markedly increased. One thing you have to do with all micro-RNA studies is to validate the ___. Essentially what you do is you can then use your target gene with a luciferase construct, which just gives you a read out and essentially if you over-express 451 you can see that the gene expression based on luciferase exactly ____ read out light activity goes down but if you essentially mutate that region that 451 binds to on the eutopic gene you do not see any effect. So that tells you it is a bona fide target of the micro-RNAs. We think that this gene based on the literature is also associated with proliferation and in women with this disease Juan Garcia in Madrid has shown that this particular protein also is upregulated in the uterine flushings of women with this disease and it increases as the progression and the severity of the disease increases.
Basically based on the breast cancer literature and also work that is being reported in women we wanted to see whether this micro-RNA plays any role in proliferation. We first went back and looked at our baboon endometrium and so these are the control endometrium looking at the in situ hybridization of micro-RNA 451. You can see in the control animal you see very beautiful localization of the micro-RNA by in-situ hybridization in the epithelial cells which is completely eliminated in the animals with this disease. Then you could look at the target which is YWHAZ which interacts with beta-catenin and YAP which are basic regulatory genes that are involved in proliferation. These genes are not expressed in the endometrium of animals that do not have the disease. However, in contrast, if you look at the expression of YWHAZ, beta-catenin and YAP they all are markedly upregulated in the eutopic endometrium and also in the ectopic endometrium that I am not going to get into today, when 451 is downregulated. This is correlated with a significant increase in proliferation which many use phosphohistone.3, which is a mark of mitotic activity. You can see that there is a significant increase in the expression of proliferation in the mid-secretory phase when epithelial cells should not proliferate in the endometrium suggesting that this change in micro-RNAs is involved in this aberrant proliferation that we see in the eutopic endometrium.
You can go ahead and then confirm this by simply taking cell lines, and in this case what we have used is an epithelial cell line, essentially when you over-express 451 you down regulate both messenger RNA and protein for YWHAZ and this is collated then if you do an empty T assay and look at proliferation there is significant decrease in the proliferative activity of these cells when you over-express 451 suggesting that this could potentially be an important mechanism by which the proliferation takes place aberrantly in the eutopic and ectopic endometrium associated with this disease.
I am going to summarize a huge amount of work here in this one diagram. Essentially what we see is that micro-RNA 451 has two major targets that we really would be interested in. In the cell cycle targets, the cell cycle genes and YWHAZ and essentially the regulation of these contributes through YWHAZ and beta-catenin and interaction of beta-catenin in YAP ____ to the nucleus enhances cell survival, promotes invasion, and c-FOS. If you remember from the first slide I showed you which is highly upregulated in the eutopic endometrium interacts with YAP to enhance EMT. In addition YWHAZ together with beta-catenin and the other cell cycle genes enhances the proliferating index associated with this tissue. Essentially what you are seeing is an aberrant change in micro-RNA 451 which then enhances cell cycle activity and proliferation activity.
A lot of our work right now is focused on targetting 451 since we know that if we over-express it we can suppress proliferation and we are using nano particle technologies and a number of other methodologies to try target 451 and see if we target lesions in various animal models and cell in vitro experiments whether we can suppress this proliferation response.
Now switching very quickly to this idea of progesterone resistance; if we look at FKBP4 you can see now both in the baboon and the human in the normal eutopic endometrium you see a significant increase in immunolocalization of this protein. It is important for progesterone action. And again, in the mid-secretory phase of both baboons and women with this disease you see that it is dramatically downregulated.
Again confirming that there is an association and in this case 29c goes up dramatically in response to the disease. It is increased more markedly when progesterone resistance is dominant in this tissue. This is associated again with a significant decrease in FKBP4 again later on in the disease process when progesterone resistance is markedly evident in this tissue. Again you can predict that by now using a 29c mimic using an FKBP4 luciferase component and showing that it is a bona fide carbon. Here again since this is required for progesterone signaling we can go back and use now stromal cells and these are stromal fibroblasts which we know do not undergo decidualization appropriately in response to endometriosis, that if you over-express 29c you down regulate FKBP4 and if you look at progesterone target genes such as Decorin and Hox A10, which have all be shown very clearly to be downregulated in endometriosis you can suppress this along with the decidualization markers IGFBP1 and prolactin suggesting that basically the upregulation 29c perhaps through FKBP4 and other mechanisms suppresses the ability of these cells to respond to progesterone and undergo decidualization.
The working hypothesis right now what we have is that normally what you need is FKBP4 which complexes with each protein 90 the progesterone receptor and the ligan forms this complex in the cytoplasm which is required for translocation in the nucleus to initiate transcription of regulation of progesterone responsive genes. However, in the presence of endometriosis when FKBP4 is low because of increased levels of 29c this complex does not form appropriately so as a result you are not able to initiate progesterone regulated transcription and as a result the endometrium is no longer responsive appropriately to progesterone signaling which is so critical for the establishment of pregnancy.
Again, what I have tried to show you is that in the absence of endometriosis when we look at these two micro-RNAs that I focused on initially the increase in 451 and the decrease in 29c is associated in a situation where there is no proliferation and apoptosis and progesterone responsiveness. As a result there is a decrease in potential lesion development, inappropriate proliferation of the uterus which then, because the progesterone then can be responsive to the endometrium that pregnancy can ensue in the situation where we are looking at infertility.
However, in the presence of endometriosis now you have the opposite effect where 451 is significantly downregulated and 29c is upregulated. As a result their target genes, and we are just focusing on two of the target genes, there are multiple target genes associated with that. Altered proliferation decreases apoptosis in response to the interaction beta-catenin and YAP and in this case the ability of the progesterone receptor to be translocated to the nucleus. This results in the ability of these lesions to develop because you now have the ability of these cells to undergo hypo-proliferation the eutopic endometrium is in an estrogenic state which also has aberrant proliferation and because progesterone is not able to signal appropriately you potentially have the situation of infertility.
Now, is this relevant to the human situation? In collaboration with our colleagues in Brazil we have been able to get some samples from women that have deep infiltrating disease and essentially looked at the controlled eutopic endometrium and in the same group of women before and after surgery for deep infiltrating disease. You can see here and these were taken in the mid-secretory phase and you can see in the women pre-operatively that you see an increase in 29c and a decrease in FKBP4 and five of these women got pregnant with assisted reproductive therapies after surgery. The four of them that did not in the six months after the surgery that we were then able to get endometrial tissue you can see that post-surgery in these women that now 29c has gone down and FKBP4 has gone up.
We followed these four women that did not get pregnant in the six months and now three of them have also got pregnant and only one of these nine women have not got pregnant after the surgical responsiveness. This suggests that again perhaps surgical intervention is responsible for removing the insult of the lesions and there is a short window of time that these women can get pregnant.
Now I am going to switch to 21 and I know I am going through this pretty quickly there is a lot of data here that I want to get through but I just want to give you an idea of how we approached this process. Micro-RNA 21 again is highly upregulated in the eutopic and ectopic tissue of baboons with this disease and you again can see by in cycle hybridization here are lesions taken from baboons at 15 months. You can see the beautiful in cycle hybridization of micro-RNA 21 both in the glands and also in the subset of immune cells that are found associated with these lesions.
Micro-RNA 21 is really a double-edged sword. It has been extensively studied like I said in the breast cancer literature and essentially it is regulated by inflammation, a response that is critical for endometriosis and essentially contributes to fibrosis, it contributes to angiogenesis and it contributes to reduced apoptosis. All of which are critical for maintaining the lesions and things that we need to target if we are going to be able to suppress lesion development. Then, if you look to see whether there is a correlation because IL-6 through STAT3 regulates micro-RNA 21 you can see that if you look in lesions for IL-6 you see a significant increase in the messenger RNA for IL-6 and if you take cell lines whether they are stromal cell lines or epithelial cell lines and you now treat them with IL-6 you see also a significant increase in the expression of micro-RNA 21. This suggests perhaps that the inflammatory response you see in the peritoneal cavity in the lesions are important for driving the upregulation of micro-RNA 21.
So to ask whether IL-6 has a direct regulatory mechanism in terms of regulating lesions development we have developed a mouse model and investigate what we use is that this mouse model has a progesterone receptor cre with a Rosa mT/mG locus. So we have both coloring we can red and green fluorescence that we can then monitor using in vivo imaging to see what happens with lesion development. In this case all we have simply done is created lesions in these animals using their own endometrium and as you can see when we then treat them with IL-6 there is a significant increase in response to IL-6 treatment in development. This is a very, very nice model system. We have to use model systems. The mouse does not get spontaneous disease but being able to engineer these animals with ___ technologies we can then utilize various color mechanisms to try to see whether we can do in vivo imaging. This really helps us in terms at trying to look at lesion development and trying to understand.
One of the nice things about this animal model is when we looked at them three months after we created the lesions, independent of treating them with IL-6, they become infertile. Basically what you are seeing is an effect that is associated with this.
Very quickly then going back and then asking the question is 21c also involved in both EMT and fibrosis if you now look at EMT in our baboon model at six and 15 months you can see that there is a beautiful upregulation of EMT going on and these are ___ positive cells within the epithelial cells of the lesion and you can see that these cells are trying to transform into stromal fibroblasts and that is associated with an increase in TGF-beta and phosphorous Smad, which is involved with the process of EMT.
In addition, if you look again to the development of processes of lesion development in our baboon model you can see that if you look at the extent of fibrosis there is a significant increase in alpha-SMA, which is a myofibroblastic ____ associated fibrosis and then if you stain with Masson trichrome you can also see the significant increase in the amount of collagen deposition associated with that, which is also associated with the increase in micro-RNA 21.
Quickly, since I need to finish, basically what we see is that the increased expression of 29c and 21 through progesterone resistance and inflammation plays a significant role in the development of fibrosis that is associated with lesion development and also downstream with a regulation of phosphorous STAT3 enhances both growth and angiogenesis associated with the development of this disease.
Finally, just to give you a snapshot of what we think might be happening, the very, very early insult of the presence of lesions that we see in our baboon model alters the steroid hormone receptor profile and again, I have not got into the epigenetic changes that are involved in response to this, which also epigenetically modifies the genome and results in altered expression of micro-RNAs. This then results in aberrant gene expression which can feed back to bring about aberrant proliferation EMT and fibrosis to sustain lesion development in these animals. Downstream the progesterone resistance inhibits the initiation of decidualization, impairs the ability of these cells to decidualize and as a result of the decidualization effects this endometrium and the progesterone resistance this endometrium is no longer responsive to the embryonic signals that are so critical for the establishment of pregnancy.
I have just given you a snapshot and a lot of our work now is focused on trying to look at the mechanisms associated with this and whether we can target these to try to reverse some of these processes using both of our animal models .
With that, first and foremost we really need to thank the people that did the work. All the micro-RNA work is being done by Niraj Joshi who is standing somewhere right there in the back. I have not talked about any of the work that Mike Strugg and Ren-Wei Su have done with regard to any of the notch regulation. Mark Olsen is really responsible for all the animal work we do and Sharra and Samantha are involved with all of the histology and all of the immunohisto chemistry that we do. I have a wonderful group of collaborators all over the world that have really helped us. This is not a one-person job it really takes a global village to try to do this.