New Ways to Treat Alzheimer's and the Peripheral Sink Hypothesis

Blogging on Peer-Reviewed Research

You look away from a field for two seconds, and they get all crazy on you... I used to do only molecular biology focusing primarily on neurodegenerative diseases like Alzheimer's. Since I moved to a lab that studies behavior, I haven't been keeping up with the literature of Alzheimer's as much as I should so I was struck with how clever this paper is.

Sagare et al. look at a new way to lower the A-Beta levels in Alzheimer's disease by putting a recombinant protein into the blood that binds it. A-Beta is the protein in Alzheimer's disease that forms plaques and causes the pathology by killing neurons. By getting it out of the brain and into the blood, the authors show that you can improve performance measures in mice -- suggesting that this may be an effective treatment in humans.


There are a couple of important things to know about Alzheimer's research. One, mice do not get Alzheimer's disease naturally, so in order to study Alzheimer's disease a variety of transgenic -- genetically-modified -- mouse strains have been created. Though not universally true, the most common of these models over-express a protein called A-Beta. A-Beta is a peptide that is a portion of the protein APP (Alzheimer's Precursor Protein). What APP does normally in the brain is still a little fuzzy, but in the development of the the disease APP gets cleaved into A-Beta. A-Beta forms a sort of molecular crud called a plaque. We used to think that it was the plaques themselves that killed neurons, but now there is more evidence that it is actually soluble oligomers (dimers and trimers) that are toxic. (For more information on that read this review.)

In any case, A-Beta accumulation is clearly a problem in patients with Alzheimer's disease, so we would really like to develop ways to lower the levels of A-Beta -- whether it be in plaques or in oligomers.

Now initial attempts to do this focused on the production end of A-Beta. We looked at molecules that could inhibit the enzymes that cleave APP into A-Beta. It turns out that this is insanely difficult because APP is cleaved into A-Beta in a multiple step process that is regulated eight ways from Sunday. Further, the final step is performed by a large complex called the PS1-complex that actually cleaves the protein inside the membrane. It is a lot harder to develop drugs that A) can get in the brain, B) can inhibit an enzyme inside the membrane, and C) do not kill the patient. You have all sorts of solubility and toxicity issues.

Because of the difficulty in making inhibitors to A-Beta production, there are some other ways to get rid of A-Beta. One of those ways is to affect its degradation and/or trafficking, and this has spawned a whole field looking into what happens to A-Beta when it is created in the brain.

A-Beta Trafficking

It turns out that A-Beta floating around the brain is removed across the blood-brain-barrier (BBB, the layer of endothelial cells that separates the brain from the rest of the body) by a protein that is also involved in the trafficking of cholesterol -- Low-density lipoprotein receptor-related protein-1 (LRP). What happens is that LRP binds A-Beta in the brain and then transports it across the BBB into the blood in a process called transcytosis. Then the LRP extracellular domain is cleaved allowing the LRP bound to A-Beta to float away so that it can be degraded. (I think it is degraded in the liver...)

Anyway, that process is illustrated in this figure from Zlokovic, 2004 (Figure 1 in the paper):


There are some important things to note in this figure. First is that there is a protein that brings unbound A-Beta back into the CNS. It is called RAGE. So we are dealing with a steady-state process that mediates the levels of A-Beta in the CNS. Further, a variety of proteins have been tried to mop-up the A-Beta in the blood such as gelsolin and antibodies to A-Beta. These are depicted on the left.

This is the origin of the peripheral sink hypothesis. The idea is that if we could add some protein to the blood that grabbed all the A-Beta that got there and did not let it go, then we could prevent it from getting back into the brain. We could shift the equilibrium such that there was less A-Beta in the brain, and this might improve the symptoms of Alzheimer's disease.

Sagare et al.

This brings us to Sagare et al. Sagare et al. wanted to see whether you could add a recombinant version of LRP to the blood to bind up A-Beta and act as a peripheral sink. The recombinant version of the protein that they use is called sLRP, and it is the extracellular portion that is released into the blood with the normal receptor is cleaved.

They infused sLRP into mice that have been genetically engineered to get Alzheimer's disease (or the mouse equivalent...they get plaques and cognitive deficits). The authors found that the sLRP significantly lowered plaque burden in these mice.

This shown below (Figure 2A in the paper):


The left shows a brain stained for A-Beta (green) in an untreated mouse of the genetically engineered strain. The right shows a brain stained for A-Beta in a mouse treated with sLRP. See how the plaque burden is much lower.

That is good, but where did the A-Beta go? It went out into the blood. Here is a measurement of A-Beta (there are two types actually A-Beta 40 and A-Beta 42...the difference is not important) of the blood in treated and untreated mice (Figure 2C in the paper):


You can see the that total plasma concentration of A-Beta 40 and A-Beta 42 in the treated mice is higher, showing that the A-Beta is now in the blood rather than in the brain.

All of this is fine and good, but many Alzheimer's treatments have floundered on the rocks of improving cognitive function. Many attempted treatments are effective at removing A-Beta, but the animals or people don't actually get better. Fortunately, sLRP does cause a measureable improvement in the behavior of the mice. The authors tested a variety of behavioral measures, and they found that the treatment improved the mice nearly to the level of controls (not genetically modified):

Chronic treatment of mice heterozygous for the sw mutation of APP, the gene encoding Abeta precursor protein (APP+/sw, also called APPsw+/-, mice), with low-dose LRP-IV beginning at 6 months of age increased cerebral blood flow responses to whisker stimulation by 65% (Fig. 1c) and improved operant learning (Fig. 1d) and spatial (Fig. 1e) and recognition (Fig. 1f) memory almost to their levels in nontransgenic controls.

This is excellent stuff, but there is a good question if you have been paying attention. Other proteins have been tried as peripheral sinks before. Why does this one work, and those work not as well?

One reason is affinity. The authors test the affinity of sLRP for A-Beta, and they find that it is much higher than for other proteins.

Another reason is side-effects. I don't know if people remember this but they tried to use a vaccine A-Beta in people. The idea was that the antibodies to A-Beta would clear it from the brain and form a peripheral sink. It worked fine; however, the problem was that it also gave some people fatal encephalitis. Not good.

The idea with this protein is that you could get a high affinity molecule that would not cause encephalitis. The encephalitis issue looks promising in this case because the authors show that the recombinant protein itself does not cross the BBB.

In any case, this is some great research that shows a new avenue to treat Alzheimer's disease. This peripheral sink hypothesis and A-Beta trafficking research is definitely the thing to watch in the next couple years as we begin to understand more about how A-Beta moves around the body.

Full citation:

Clearance of amyloid-bold beta by circulating lipoprotein receptors
Abhay Sagare, Rashid Deane, Robert D Bell, Bradley Johnson, Katie Hamm, Ronan Pendu, Andrew Marky, Peter J Lenting, Zhenhua Wu, Troy Zarcone, Alison Goate, Kevin Mayo, David Perlmutter, Mireia Coma, Zhihui Zhong & Berislav V Zlokovic
Nature Medicine 13, 1029 - 1031 (2007)
Published online: 12 August 2007 | doi:10.1038/nm1635

Hat-tip: Faculty of 1000


More like this