Fight Aging!

Here Matthew O'Connor of the SENS Research Foundation talks about the research that led to founding of Underdog Pharmaceuticals, a biotech startup incubated by the foundation to commercialize a means of targeting 7-ketocholesterol in atherosclerosis and other conditions. Oxidized cholesterols, and largely 7-ketocholesterol, are the primary cause of dysfunction in the macrophage cells normally responsible for preventing the build up of fatty plaques in blood vessel walls. That dysfunction is the cause of atherosclerosis, and the fact that the presence of oxidized cholesterols increases with age is one of the reasons why atherosclerosis is an age-related disease, and why young people don't exhibit the plaques that narrow and weaken blood vessels.

A sufficiently effective way of selectively clearing 7-ketocholesterol from the body should go a long way towards preventing and reversing atherosclerosis - and possibly other conditions as well. As noted here, there is evidence for 7-ketocholesterol to accumulate in other tissues and contribute to age-related conditions in other ways. The particular approach taken by the SENS Research Foundation scientists is to find a non-toxic molecule that selectively binds to the toxic molecule that one would like to remove, 7-ketocholesterol in this case, and deliver it to the body in volume. The bound molecules are then processed and excreted in the normal way. It will be interesting to see how the Underdog project progresses in the years ahead.

Matthew O´Connor presenting at Undoing Aging 2019

Thank you everybody. It is a real honor and a pleasure to be here today, speaking in a last minute slot. Don't volunteer to take over to speak for someone who didn't show up, because then you have to stay up all night, working on your slides. This slide is the SENS team from this last summer, a combination of the MitoSENS team and the team working on 7-ketocholesterol that Aubrey de Grey was just referring to, that we're calling the Underdog team. You can see me as the chief Underdog there.

I want to just spend a minute, Aubrey had mentioned that this is the tenth anniversary of SENS Research Foundation, and it is my 9th year with SENS Research Foundation. It has been a privilege to have started so early, to help build the lab, and I wish that I could show a time lapse photography of where it started and where it has gone, and also to be involved with some of these amazing interns. So many of these projects, including this one, are very intern driven. Regarding the MitoSENS team, people may not realize that all the progress Aubrey was talking about, publishing papers all of a sudden, didn't happen until after we hired Amutha, so you can see why that project got taken away from me and given to her. She is just a phenomenal mitochondrial biologist.

So this project that I'm about to tell you about involves a story about 7-ketocholesterol. I hope that a lot of you have heard about it. It is kind of one of the classic bad guy molecules that Aubrey has been talking about for a long time. My history with it goes back even further than SENS Research Foundation, as the foundation and Methuselah Foundation before it have been funding work on trying to get rid of this really nasty molecule 7-ketocholesterol for at least ten years, with work that we've funded in various places. I actually heard Aubrey give a talk about this, one of the best talks I've ever seen, at Rice University something like 15 years ago when I was in graduate school. So I've been thinking about it ever since then.

This project kind of started out as some ideas about how to think about this problem and how to get rid of this nasty molecule a different way. It started out as something that we were sort of dabbling in, and graduated from being my 20% project to something that is all-consuming now. On this slide, you can see causes of death worldwide, which probably looks familiar to you many of you. As Aubrey was just saying, atherosclerosis is the world's biggest killer. If you risk-adjust all of these diseases for what their underlying cause is, atherosclerosis is believed by world health organizations to kill about 44% of everybody. This molecule, 7-ketocholesterol, an oxidized form of cholesterol, is thought to be one of the earliest stages leading to atherosclerosis. But not everyone simplistically believes in that model; the classic model is that you eat too many hamburgers and you get too much cholesterol in your bloodstream, and that just sort of stochastically builds up, and eventually you get plaques, and eventually they rupture and you have heart attacks and strokes, and you die.

However 7-ketocholesterol is many, many, many times more toxic than LDL cholesterol is - so I call it the really bad cholesterol. It is by far the most common product of the reaction between a free radical and cholesterol, and so more often than not you get this stable 7-ketocholesterol which is extremely toxic. It has no useful purpose in your body. It can accumulate in the lysosomes of macrophages and can be an early step in the progression towards becoming a foam cell. Foam cells build up as a layer in atherosclerotic plaques, and 7-ketocholesterol is found inside them, as well as in the necrotic core of the plaque.

Let me change gears now and reveal the class of drug, the class of molecules, that we have been playing around with for the last few years. They are called cyclodextrins. You may or may not have heard of them, but they are a huge industry, and they have a huge variety of applications. The medical applications, despite the fact that cyclodextrins have been studied and applied for many decades, are only just starting to be realized. They come in three basic flavors, alpha, beta, and gamma, which are three different sizes, and they have six, seven, or eight different sugar rings that they are made out of. There are different forms of them, probably thousands of different ones that have been invented. They are extremely customizable and modifiable. Any one of these hydroxyl groups you can stick just about anything you want on it, doing synthetic chemistry.

The slide here shows just a few examples of common cyclodextrins that are used for various industrial purposes. Medically, they are mostly now only used as excipients, meaning as carrier molecules for small hydrophobic drugs. They are used in food: alpha-cyclodextrin is approved in the European Union as a bulk fiber supplement, so maybe you ate some of it this morning. There are versions of these that are extremely low toxicity household items, like Febreze. People can use cyclodextrins, use different formulations of them mixed with other guest molecules that stick into the cyclodextrin cavity, to engineer all kinds of materials such as self-healing gels. They look like jello, and you cut it in half and put it back together and there is no seam any more. Somebody in Japan built a car out of cyclodextrins, and there is a German scientist who is making self-healing paint for cars.

Now let me summarize a bit the history of using cyclodextrins themselves as the active component of drugs. That is what we're trying to do here, to engineer them to be drugs to target 7-ketocholesterol directly. The history here goes back to the 1990s, where this Australian group had a lot of foresight; they already knew that 7-ketocholesterol was a really toxic atherogenic molecule back then, this isn't a new discovery. There has been tons of evidence for that for decades. Cyclodextrins were just starting to be explored in the 80s and 90s as cholesterol binding drugs; different versions of them bind cholesterol well. So they went looking for a modified cyclodextrin that can specifically bind 7-ketocholesterol with a hypothesis, and they found hydroxpropyl-beta-cyclodextrin, which I will tell you about. After they published a bunch of papers, however, they abandoned it and never went into any animal studies.

However, there was a group in Texas that seems to have picked up on this idea, and there was an orphan disease called Niemann-Pick type C, which is a lysosomal storage disease, that hyperaccumulates 7-ketocholesterol. These patients, almost always children, are very sick and die very young. There are good mouse and cat models for this disease, and if you dose them with very high doses of hydroxpropyl-beta-cyclodextrin, you can rescue the small animals. People have started to look at this for atherosclerosis because hydroxpropyl-beta-cyclodextrin is so safe, but there is some debate as to whether you are targeting 7-ketocholesterol, and I'll present some evidence as to why I'm strongly in favor of one hypothesis over the other. For the Niemann-Pick studies, it is going into clinical trials, it is entering phase II, there is a lot of funding going into this now. Presumably because they've got their eyes on atherosclerosis rather than just Niemann-Pick, which is an extraordinarily rare disease.

This slide shows some older data on cyclodextrin, and this came from the Australians. They are soluablizing 7-ketocholesterol, and here is the effect that it has on cholesterol, which is to say nothing, while it does a great job soluablizing 7-ketocholesterol. I'm going to talk a lot about this assay, so if you don't understand it, hold tight and I'll explain it in more detail. 7-ketocholesterol is very toxic, so with increasing doses you kill cells in culture and this was work we funded at Rice University, kind of as a scientific control for the experiments that they were doing. They were playing with cyclodextrin; as I said, it has a lot of different functions. They were using it as a carrier molecule, and found that it was working better than anything else that they were using, while we were working on the earlier LysoSENS enzymatic process to rescue the cells from 7-ketocholesterol toxicity. Then more recently, work in that same lab led to a paper and a patent on the idea of using that same cyclodextrin to prevent and reverse lipofuscin from forming in cells.

So I'll leave that there as history and show you this happy and sad video of cats with Neimann-Pick disease. All three of these cats have this disease and their phenotype is pretty similar to that of the humans. It is a devastating disease. All of these cats have it, one of these cats has been treated with hydroxpropyl-beta-cyclodextrin at a very high dose. Apparently this video is famous in the FDA for how you get something fast tracked into humans by showing a heart-tugging video like this. Because it clearly is working very dramatically to help these cats.

This slide shows the assay that we use. It is a really simplistic assay to screen through many different compounds, the many different cyclodextrins, many different modifications to cyclodextrins that we've made. It is a simple turbidity assay that we've automated. If you dump most sterols that are not water soluable into an aqueous solution, they get cloudy. Then if you managed to add something to the solution that can soluablize the sterols, then the solution turns clear. That is an indication that they are binding to your target.

So we've automated it, you run many of these plates, read them in a plate reader, and you get data that looks like this next slide. I'm just reporting the percent turbidity, so start at 100% and then go down, or in some cases it becomes more cloudy. So these are some cyclodextrins that have been studied for Neimann-Pick disease, as I keep talking about, hydroxpropyl-beta-cyclodextrin is one of our favorites. It doesn't bind cholesterol, even at ridiculously high concentrations. With 7-ketocholesterol it has nice specificity. There's another one, sulphobutyl, that has also been tested somewhat in animal models for Neimann-Pick disease but hasn't really gone further than that, and doesn't seem to be as effective as hydroxpropyl-beta-cyclodextrin.

The safety profile for some cyclodextrins such as hydroxpropyl-beta-cyclodextrin is just phenomenal. It is much less toxic than aspirin, or something like that. So you can dose humans or animals with grams of it and they are fine. As you can see, hydroxpropyl-gamma-cyclodextrin, a bigger one, doesn't do anything in the assays.

We've checked many different cyclodextrins and run them through our screening process. This slide shows a bunch more from the catalogs. Some of them are better and some of them are worse. The point of showing this to you is to show how we gathered a ton of data about which cyclodextrins were interacting with which targets in which ways, and I deleted a whole bunch of data because I was practicing this this morning, and I had too much in the presentation. But we also screened a bunch of other targets, other sterols that you can order from the Sigma catalog. That helps us look at off-target effects.

We can also do it computationally. On this slide you see the former intern that Aubrey was bragging about before. She had the idea when she was an intern - we were already working on this project, and making some progress, and she said well why don't you just model them, you can learn so much about them. She said I'll figure it out. So now she's one of the world's experts in modeling cyclodextrins. Which is a very niche field. Only a few people in the world know how to do this.

What you see here is a molecular dynamic simulation of just a basic beta-cyclodextrin. We've done hundreds of different simulations on different tweaks and different cyclodextrins, mostly with just cholesterol and 7-ketocholesterol, trying to optimize selectivity binding. Also to optimize the affinity. So beta-cyclodextrin, just the core molecule, is known as a good cholesterol binding molecule, but we can learn additional things here. If you stick it in this way, both cyclodextrin and cholesterol are asymmetrical molecules and if you start with it the wrong way, it will go in the way that it prefers. It seems kind of nitty-gritty to talk about, but that kind of information is what helped us figure out all the different subtle ways in which cyclodextrins are binding their targets.

Now on this slide is a model of hydroxpropyl-beta-cyclodextrin binding 7-ketocholesterol extremely tightly and with specificity. To really get the lowdown on this, and this is all unpublished data, we'll have a publication coming out soon, that we're furiously writing now. Go see the posters at this conference, and grill me on this, because the author is brilliant, she's great at explaining it. So we started optimizing cyclodextrins, making little tweaks to them, and came up with some tricks on how to improve the affinity for them, starting out with hundreds of these models. This slide shows the kind of data you'll get from it, the geometry of how they are interacting, how closely in space they are interacting. Then the information, that I think is the most dramatic, about the affinity.

So when you find one that is actually going to grab on super-tightly, then you massively increase the free energy of interaction, or decrease it. You can see that this also corresponds with super-close, tight, and stable interaction between the two molecules. Those kinds of simulations are extremely high resolution. We do them for a whole microsecond, which is an eternity in terms of nanosecond resolution simulations of molecules. The chief way to do lots of simulations is to do molecular docking, and so using this method - this slide is, I think, 28 different simulations that are done quickly, doing just small little tweaks to a particular family of cyclodextrins. We look for areas such as here where you seem to get separation between cholesterol and 7-ketocholesterol.

That is what led us to the step shown on this slide, which is that we synthesized some new cyclodextrins and checked to see if they could increase their affinity and maintain specificity for 7-ketocholesterol. Up at the top here you have hydroxpropyl-beta-cyclodextrin, which doesn't bind cholesterol, it binds 7-ketocholesterol a little bit. However, note the change in scale down here, we're using lower dosing ranges. The modifications that we're making to these that I'm showing you here today, they all work phenomenally well to massively increase the affinity for our target 7-ketocholesterol. However, all of them also increase their affinity for cholesterol. You probably can't see all of them in this slide, but some of them are maintaining better specificity than others, and those are the ones we're most excited about.

So there are a couple of different families of new cyclodextrins that we've invented that have extremely high affinity for 7-ketocholesterol. One in particular that caught our eye that we're really excited about - again, hydroxpropyl-beta-cyclodextrin up here at the top of the slide, some nice specificity for 7-ketocholesterol. Down here, one of the new cyclodextrins that we made has an extremely high affinity for 7-ketocholesterol, and good specificity.

Like I was saying before, these are exciting molecules to work on because they are really safe, the different versions are edible, breathable, or injectable, as in happening in the Neimann-Pick trials. A classic safety test before going in vivo is to collect some blood from whatever hapless CEO or intern happens to be wandering down the wrong time, and then treat their blood with something, and if it lyses it, then it is killing their blood cells, and if it stays clear then it is not. All of the new cyclodextrins that we've made seem to be non-toxic in the pharmacological range, and as you go above that you get some variation. For the one we're most excited about, that had the highest specificity for 7-ketocholesterol, we couldn't find a dose at which we would kill any blood cells.

I'm going to try to make the case that animal models for atherosclerosis are useless, and that we should skip straight to humans. My argument is that mice, rodents, don't get atherosclerosis naturally, and when you do give it to them by knocking out their ability to metabolize cholesterol, they are just not able to take up cholesterol, to clear it, so that it just builds up into artificial plaques. We don't think that is a good model. So once again we're working in humans, we're stealing their blood, and then treating blood with the drug in concentrations that we think are realistic and for time periods that we think are realistic. We know a lot - so much work has been done on safety for cyclodextrins over the years. We know a lot about how quickly it is cleared from circulation in different animal systems and even humans.

Then what we're doing is instead of trying to measure serum 7-ketocholesterol levels, which basically don't exist, as 7-ketocholesterol isn't transported in HDL or LDL, and that is part of what makes 7-ketocholesterol so toxic, as it can't be transported out of cells. We're measuring our ability to release 7-ketocholesterol into the serum, and measuring it by mass spectrometry. The early data on our new cyclodextrins is looking good. In this slide here is hydroxpropyl-beta-cyclodextrin that can remove some 7-ketocholesterol from human blood cells, and the first new cyclodextrin that we made can do a lot more, a lot better, and with blood from multiple donors.

So to recap, there are some big players in this area that are excited about being able to treat an orphan disease like Neimann-Pick disease. They probably have their eyes on going after bigger indications like heart disease, but they think that they are treating cholesterol, which I think is crazy. I think that they are actually treating 7-ketocholesterol, and that they are having the success that they are having because of this. However our drug blows it away in terms of affinity for the target by something like tenfold.

Beyond atherosclerosis, 7-ketocholesterol is implicated in a number of diseases, as shown on this slide. It is one of those bad guys that accumulates in different tissues with age, and we don't even know yet everything that it is implicated in. Just as for senescent cells now that people are finding ways to kill them, all of a sudden you are finding aspects of aging that are being reversed when you kill senescent cells, and I predict that someday we'll see the same thing with 7-ketocholesterol. However, atherosclerosis and heart failure I think are great indications. Of course familial hypercholesterolemia is just an exaggerated way to get atherosclerosis. There is Neimann Pick disease, and we think are drug would work better than the ones that are out there now.

I don't want to try to claim that 7-ketocholesterol causes everything that has ever made everyone sick, but it does accumulate in macular degeneration and some cells in Alzheimer's disease. Our tools that we've developed to make new cyclodextrins, to test them, to model them, I think that this is a great platform. There are other toxic oxysterols that we could be going after that we haven't starting working on yet, but are good targets for this kind of technology. Basically any small hydrophobic molecule that is bioaccumulating is a potential target. So the space in this area looks crowded because there are tens of thousands of patents on cyclodextrins, however when you get down into using cyclodextrins as drugs themselves, then you are down into single digits, like four or five patents that are out there. So we think that we have pretty strong data and protections that will get there.

Because cyclodextrins have such a good regulatory profile and are so well established in how you can use them safely, industrially, in food, in medicine, we think that the pathway to turning this into an actually effective drug - that we can write a plan, and have it be realistic is good. We're talking to experts on formulation and manufacturing, and what you see on this slide are realistic timelines. We're developing some more assays that we think will be more effacious; assays that will demonstrate that we can reverse a foam cell phenotype for example, or remove 7-ketocholesterol from plaque samples that we're working on getting from human patients or cadavers. So we have a detailed month by month plan on how we think that we could get a drug to clinical trials in three years.

As Aubrey said, we're working and making great strides on turning this into a SENS Research Foundation spinout, wholly incubated and owned at the foundation. We're calling it Underdog Pharmaceuticals.