For Prospective Alzheimer's Drugs, It's All About Location, Location, Location

ResearchBlogging.orgOne of the more common questions I get is why they haven't found any drugs to treat Alzheimer's disease. (But they have, haven't they? What about cholinesterase inhibitors like Aricept? Ed. Those drugs mask the symptoms of Alzheimer's disease, but they do not change the clinical course.) Drug companies are particularly gifted at finding molecules to inhibit all the enzymes in our bodies. We know the enzymes involved in the etiology of Alzheimer's (as I will explain in a second). Why can't we inhibit them?

The problem with inhibiting the enzymes involved in Alzheimer's and hence treating the underlying illness is in how the enzymes work and where they are in the cell. These problems aren't necessarily fatal, though. Rajendran et al. show in the journal Science that by using some clever chemistry small molecules can still inhibit these enzymes. The authors demonstrate that by localizing your drug better -- making it go where the enzyme is -- you can get better results.

Background

This paper is not going to make a lot of sense unless you know the enzymes involved in the Alzheimer's. As I have talked about before (here and here), Alzheimer's is a disease of molecular crud. Specifically, the crud is a small peptide called ABeta. ABeta is produced from a protein in your brain called APP, the natural function of which is not entirely clear. What happens is that APP is cleaved in two steps by two successive enzymes (called the beta and gamma secretases) to form the peptide ABeta. ABeta then floats off into the synapse and forms accretions of crud called amyloid plaques. These plaques are a characteristic finding in Alzheimer's disease. Whether it is actually the plaques or smaller ABeta oligomers that kill neurons is still debated. What is clear is that if we could get rid of ABeta it would prevent the death of neurons.

Obviously scientists would really like to find a drug that prevents ABeta production.

ABeta creation from APP is depicted below.

i-7ceeb5a522c2d18dfc4fcb2ced78ac07-app.JPG

One of the complexities here is that APP is a transmembrane protein. The protein passes through the membrane and part is on the outside and part is on the inside of the cell. There are two pathways for APP processing in the cell. The benign pathway -- called the non-amyloidogenic pathway because it does not result in plaque formation -- starts when APP is cleavage by the alpha-secretase on the membrane right outside the cell. The resulting fragment is then cleaved in other ways, but it doesn't matter in this case because none of them result in the formation of ABeta. In contrast, the disease pathway -- called the amyloidogenic pathway because it can result in plaque formation -- starts by the cleavage of APP by the beta-secretase on the outside of the membrane. The fragment is then cleaved again this time inside the membrane by the gamma-secretase. The resulting fragment is ABeta which floats off to do all its bad bidness.

(Just to clarify some things. The alpha-, beta-, and gamma-secretases all have names; I am just not using them here. There are traditionally referred to by their activity as enzymes, and more than one protein can perform that function in some cases. Also, in the case of the gamma-secretase the cleavage is performed by a complex of proteins. Another thing: these processes are regulated eight-ways from Sunday, and I don't have time to talk about that. The resulting fragments from these cleavages that aren't ABeta often do important things too. It is a complex subject.)

If I was an enterprising drug scientist, I might conclude from the knowledge above: hey, why don't we inhibit either the beta or the gamma secretase? Then we could limit the formation of ABeta and hence the progress of Alzheimer's disease.

It turns out to be much harder than that. Gamma-secretase performs a variety of very important functions besides its role in making ABeta. When you inhibit these functions, bad things happens to the cell. Also, the gamma-secretase is performing a cleavage inside the membrane bilayer. We could manufacture a molecule that would be hydrophobic enough to get in there, but then you have problem making it water soluble enough to get it into the body in the first place. (You also have a problem with toxicity.) Beta-secretase at first seems simpler, but my diagram above is misleading. It turns out that the actual location of beta-secretase cleavage is inside the cell on the surface of what are called endosomes. Endosomes are in-pouchings of the cell membrane. The APP molecule is taken inside the cell in little membrane bubbles; that is where the actual cleavage takes place. You can't just douse a cell in drug and hope for the best. You actually have to get a drug that gets to where the Beta-secretase is active inside the cell.

Rajendran et al.

Rajendran et al. come up with a clever solution to this problem of getting a beta-secretase inhibitor to where the beta-secretase is active. They have an inhibitor of the beta-secretase enzyme, and they link it to a sterol (a lipid like cholesterol) to target it to where they want.

The authors show that this combination of sterol and drug is much more effective at limiting ABeta production than the drug alone. This is depicted in the figure below. (Panel A from Figure 1 of the paper.)

i-a3d66918d84984c6c8a6e7279fd22118-appproduction.jpg

What you are looking at is the amount of ABeta (soluble APP Beta = sAPPBeta = ABeta) produced in cultured cells (in this case HeLa cells) when the cells are subjected to the DMSO vehicle, the unconjugated drug, and the drug conjugated to the sterol. See how ABeta production declines with sterol conjugation. The last column is a drug that has been conjugated to the sterol in a different way. It suggests that in order for the drug to be effective the conformation with respect to the drug is important.

Critical to their interpretation of this data is the idea that the sterol-linked drug gets into endosomes. They show this using staining, depicted below. (Panel D of Figure 2 in the paper.)

i-bd3b5dfd0081f003c71f907772b1d4b9-applocalization.jpg

The green shows staining for the APP substrate. The white dots show the alpha-secretase BACE1. The red shows the sterol-linked drug when it has been labelled with a fluorescent molecule called rhodamine. The bottom right shows the overlay. See how their location overlaps. The picture looks very punctate because all of the molecules are located in bubble shaped endosomes inside the cell. The authors confirm here that their drug is localized into the desired location.

These results are all fine and good, but they were all performed in tissue culture. Does their sterol-linked drug actually work in an animal? The authors show that it does in a transgenic mouse model for Alzheimer's disease. (Mice do not naturally get Alzheimer's disease. Transgenic models using mice are discussed in this post.) This is shown in the figure below. (Panel B from Figure 4 in the paper.)

i-70dbea0b06f3b2fbf4db27f74030320d-appmouse.jpg

The authors injected their sterol-linked drug into the hippocampi of transgenic mice who would normally get something like Alzheimer's disease -- plaques, diminished memory capacity, etc. The hippocampus is a region particularly affected with the disease. You can see that the drug in the animal reduces the levels of ABeta, but only when it is conjugated to the sterol.

This is interesting and exciting work. It focuses on a concept that people who design drugs have known about for a long time, but it is a very good demonstration of it. That concept is called dimensionality. Dimensionality works something like this. Say I throw a drug into a culture full of cells. That drug can float throughout the culture (assuming it can get into the cells) and effect the target enzyme. However, because the drug can move in 3 dimensions, the concentration that I need in order to effectively inhibit my target enzyme is high. A large amount of drug is needed to guarantee that the enzyme doesn't work because the drug can just float away. Now consider a drug that is tethered to the membrane like in these experiments. Instead of being able to move in 3 dimensions, this drug can only move laterally in the membrane. It is more of a 2D drug. Because it is limited in its mobility, it is much more effectively targeted to inhibit that enzyme. This means that a much lower dose is required to be effective. By using dimensionality, the authors increase the effectiveness of their drug.

There is another aspect to Alzheimer's biology that the authors exploit that I didn't talk about. The enzymes involved in cleaving APP are localized in what are called lipid rafts. Lipid rafts are areas of the membrane that are more hypophobic -- more fatty -- than the surrounding membrane due the concentration of specific lipids like cholesterol. The actual cleavage takes place in these rafts. Knowing this if I wanted to increase the activity of a drug to inhibit these enzymes, I would want to target it to lipid rafts. The authors show that by selecting a sterol -- as opposed to some other tether molecule -- their drug localizes to rafts and is more effective. This is another way that they selectively target their drug.

The authors summarize the significance of their work:

By anchoring the inhibitor to the membrane, we achieved two goals: (i) The inhibitor became endocytosis-competent and gained access to endosomal beta-secretase; and (ii) we reduced the dimensionality of the otherwise soluble inhibitor, thereby enhancing the interaction between the inhibitor and the enzyme. Reaction rates between solutes and membrane receptors can be enhanced by reducing the dimensionality of the solute via nonspecific adsorption, and here this model has been realized in designing drug candidates. This model also explains why such a membrane-anchored version of the inhibitor would be superior to a soluble but membrane-permeable inhibitor. If tethered molecules (e.g., the enzyme and the inhibitor) were partitioned within microdomains, both the concentration and the interaction times of the components would be increased, as we have observed. By choosing sterol as a membrane anchor, the inhibitor is enriched in the vicinity of raft-associated beta-secretase, thus enhancing their interaction. The increased potency of the sterol-linked inhibitor confirms that the lipid environment and the subcellular localization of β-secretase regulate its activity. (Emphasis mine. Citations removed.)

This is some brilliant drug chemistry, but I do want to add some brief caveats. The first is that many, many drugs have been successfully employed in rodents but have failed to function in humans. I do not expect this drug or any other to be immediately applied to the treatment of Alzheimer's. We are still working on the problem, but we don't have solutions yet. Second, though the authors solve the problem of cellular access for the drug, they do not solve the problem of systemic access. In their animal model, they inject the drug directly into the brain. This is not how you would want to administer a drug to a patient. Ideally you would like to inject it into the bloodstream. It is not clear to me whether these sterol-tethered drugs would be effective if administered systemically.

Nonetheless, some great experiments and very illustrative of some key concepts in drug chemistry!

Hat-tip: Faculty of 1000

Rajendran, L., Schneider, A., Schlechtingen, G., Weidlich, S., Ries, J., Braxmeier, T., Schwille, P., Schulz, J.B., Schroeder, C., Simons, M., Jennings, G., Knolker, H., Simons, K. (2008). Efficient Inhibition of the Alzheimer's Disease -Secretase by Membrane Targeting. Science, 320(5875), 520-523. DOI: 10.1126/science.1156609

More like this