Dr. Michael R. Freeman Video (Text Version)
Cholesterol and Prostate Cancer Natasha Kyprianou
Next we will discuss cholesterol metabolism. It is my pleasure to introduce Dr. Michael Freeman. Dr. Freeman is a David Retik Chair of Urologic Research at Children's Hospital in Boston and the Professor of Surgery at Harvard Medical School. Dr. Freeman’s research for a number of years has focused on cholesterol and ion channels as regulators of signaling pathways in prostate cancer. Dr. Freeman made groundbreaking discoveries on prostate cholesterol metabolism that have been supported by three Idea Development Awards and several Postdoctoral Fellowship Awards from the Prostate Cancer Research Program with the Department of Defense. He has also been the recipient of an NIH Merit Award. Today Dr. Freeman will be presenting his innovative findings that are of major mechanistic significance on the role for circulating cellular cholesterol in prostate cancer progression; Dr. Freeman.
Michael R. Freeman, Ph.D.; David Retik Chair of Urologic Research, Children's Hospital, Boston and Professor of Surgery, Harvard Medical School
Thank you very much Natasha for the invitation to give this lecture. I also want to particularly thank the DoD; we have been well-supported by the DoD for this work for a number of years and it has made an enormous difference in my laboratory. A lot of what we have done in the past 8 or 9 years has related to the problem of cholesterol metabolism and prostate cancer and in other contexts. Cholesterol is a neutral lipid that most people are familiar with. When it is elevated in the blood it is a risk factor for cardiovascular disease. It is a very prominent plasma membrane lipid; it makes up about 30% of the plasma membrane lipids and mostly associated with the plasma membrane in cells. This slide summarizes a lot of work from my laboratory which in the general conclusion is that circulating cholesterol activates oncogenic pathways in prostate cancer cells and cholesterol accumulation in the membrane we believe is oncogenic by a variety of routes and we have worked out some of the mechanisms of this. This is illustrated in this slide showing that, in animals where we’ve raised cholesterol using standard diet strategy, the very potent oncogenic kinase AKT is activated and these are xenograft tumors and tumor apoptosis is inhibited. We think from these results and a variety of other data from other groups that circulating cholesterol actually is very potently oncogenic in prostate cancer.
The implication from this is that there should be a chemopreventive effect from taking cholesterol-lowering drugs which are widely used and are considered very safe. If what I am saying is true here, then there should be evidence that cholesterol-lowering therapy actually has some impact on prostate cancer and in fact, there is. Elizabeth Platz’s group at Johns Hopkins had made major contributions here, but starting with this paper in JNCI in 2006 and then a series of subsequent papers from the Platz’s Group and from other groups, we now know that statin use results in a decreased risk of prostate cancer progression particularly in men who have taken statins continually for over 5 years and similarly high-circulating cholesterol in humans represents an increased risk for prostate cancer. I think this data is now very solid and this conclusion is not going to go away despite some confusion in the literature.
HMG-CoA reductase inhibitors or statins actually do not just inhibit cholesterol synthesis though; cholesterol is way down in this mevalonate acid pathway and statins inhibit up near the front of the pathway. So another thing statins do is inhibit isoprenoid synthesis. Isoprenoids are precursors for lipid anchors on a variety of oncogenic proteins, and you can see from this diagram that there is quite a bit of different signaling strategies that could be altered by statin therapy. I think that this is the reason there is confusion in the literature; also statins don’t have identical potency and identical ability to penetrate tissues.
Cholesterol is a major mediator of changes in membrane structure. The fused ring structure with the polar head group causes reordering of phospholipids in the plasma membrane, and this has very significant consequences because it alters the configuration of signaling proteins in the plasma membrane, many of which play oncogenic roles as you all know. This is a nice demonstration of the effect of disrupting cholesterol at the level of the plasma membrane in looking at two downstream targets, EGF receptor and AKT, both of these can be inhibited by disrupting cholesterol in the membrane indicating that these are cholesterol-sensitive targets.
It turns out that my laboratory has contributed a lot of literature to this idea. There are quite a few cholesterol-sensitive targets, but we’ve focused heavily on AKT which seems to be very sensitive to cholesterol manipulations of several kinds but there are many others—RB receptors, receptors that are involved in inflammatory pathways, and of course downstream targets—are also cholesterol sensitive indirectly if the upstream molecules are cholesterol sensitive. So this represents a very large signaling network that can be affected potentially by changes in cholesterol metabolism.
In animal studies where we have looked at cholesterol—changes in cholesterol levels—doing so in a way that doesn’t affect energy, androgen, circulating androgen levels; it doesn’t affect IGF or insulin signaling, we’ve seen very potent increases in xenograft tumor growth—prostate cancer growth. The comment about androgen actually there is an exception. In the castrate condition, castration results in increases in circulating cholesterol. This is seen in animals and also in humans. We see very potent effects of high-cholesterol, high-cholesterol diet on tumor xenografts suggesting that the castrate condition actually is oncogenic potentially because of changes in cholesterol metabolism and specifically elevation of circulating cholesterol. So as Jim Mohler indicated, cholesterol is also a potential source of intertumoral androgen; virtually all of these enzymes are actually present in tumor cells and some of them are upregulated with aggressive disease. Cholesterol can feed into these metabolic pathways and potentially downstream androgenic steroid hormones can accumulate in the tumors.
This is some data from Keith Solomon showing that in a high-cholesterol condition we actually have accumulation of DHT and T, and in this experiment we have higher levels of DHT than T, suggesting the possibility that circulating cholesterol can be a significant source of intercrine androgen synthesis. The problem with carrying out experiments on statins in preclinical models in the rodent is that the statins are not targeting cholesterol specifically as I said. They are also not very effective at lowering cholesterol in the mouse, which is the standard preclinical model. We have used an alternative strategy; in recent experiments, a compound ezetimibe which is FDA-approved; it localizes too the intestinal brush border and it inhibits cholesterol absorption. The target is thought to be specific, this NPC1-1L1 cholesterol transporter and there is a downstream effect of inhibiting cholesterol absorption from the intestine in which LDL receptors are upregulated. So circulating cholesterol is lowered.
This is a series of experiments that we published a couple of years ago showing that ezetimibe is a potent inhibitor of tumor-androgenesis in a human prostate cancer xenograft model and by using a combination of diet and this drug, we can actually get these very nice step-wise changes in cholesterol levels and we can actually show dose-dependent effects. We believe these are cholesterol-specific effects and not effects on isoprenoids or other types of targets which statins might affect. In this case, this is thrombospondin which is a potent androgenesis inhibitor and in a high cholesterol condition you have low levels of the androgenesis inhibitor, so high levels of androgenesis in this experiment.
What is the mechanism of this; I mean if cholesterol is a potential support of androgen then maybe the whole phenomenon is dependent on androgen. We just published some experiments which indicate that isn’t the case. Again using ezetimibe in a strain of hamster that has spontaneously enlarged prostate that is age dependent, we actually see a very potent suppressive effect on prostatic enlargement. This is actually in a situation where the prostate is already enlarged, so this is regression of the prostate and we see in effects it is comparable to finasteride which is a standard therapy for prostate enlargement. When we look at the tissue we see that finasteride evokes prostatic epithelial atrophy so it is affecting androgen metabolism but ezetimibe does not. It is altering prostate structure by a mechanism that seems to be independent of androgen.
This leads us to the idea that the normal prostate can sense changes in circulating cholesterol, and we have identified at one very interesting molecule, the transcription factor ATF3 which we believe is a cholesterol sensor in a high-cholesterol condition. The levels of the gene and protein go down in culture using prostate epithelial cells in cholesterol reduced medium—the gene is induced, and so we are using this transcription factor to try to work out a wiring diagram for cholesterol sensitivity in prostate cells. If cholesterol is so abundant, then why is it very limiting for these processes that I have talked about? That could be a feature of lipogenic metabolism that happens in tumor cells partly as a phenomenon known as the Warburg effect in which the synthesis of macro-molecules and membranes is upregulated. We think that alterations in cholesterol metabolism are a component of this. The Warburg effect have 1L1 which is a prostate tumor marker—a cholesterol binding protein; fatty acid synthase here is a primary source of long-chain fatty acids and so this is a large network of lipid metabolism that is likely affected in castrate-resistant prostate cancer. Of course, we think that other than growth that potentially tumor cell motility might also be affected by these metabolic transformations. Martin Hager who is an instructor in my group is giving a talk this afternoon in the metastasis section on a novel metastasis suppressor that we think is related to lipogenic metabolism. I encourage you to check that out.
We think that circulating cholesterol levels affect prostate cancer progression. Mike Lisanti’s group has recently published evidence that this is also true in breast cancer. We think that statin drugs are chemopreventive against aggressive prostate cancer, largely as the result of cholesterol lowering and not inhibition of isoprenoid synthesis and this really has to do with the fact that most statin drugs are poorly tissue-penetrant out at the periphery. Then ezetimibe is a novel compound, FDA-approved compound which can target cholesterol. Specifically we have seen potent effects in preclinical models and so this is a promising compound to look at the effects of cholesterol specifically in isolation from other metabolic effects.
In general, we think that the study of cholesterol metabolism and protein complexes that are associated with lipid-raft micro-domains that are cholesterol rich will provide novel insights into therapy.
I don’t have an acknowledgement slide but I just wanted to acknowledge my long-term collaborator in this area is Keith Solomon at Harvard and I showed work from Liyan Zhang and Jayoung Kim in my group. Thank you very much.