To understand anything in the cell you need to understand nearly everything in the cell

Understanding how variants in one protein can either increase or decrease the risk of Parkinson’s disease requires understanding of the following: the lysosome, TMEM175, Protein kinase B, protein moonlighting, ion channel lysoK_GF, dopamine neurons among other things. So get ready for a deep dive into molecular and cellular biology.

It is now 50 years and 6 months since L-DOPA was released in the USA for Parkinson’s disease, and I was tasked as a resident by the chief with running the first L-DOPA clinic at the University of Colorado.  We are still learning about the disease as the following paper Nature vol. 591 pp. 431 – 437 ’21 will show. 

The paper describes an potassium conducting ion channel in the lysosomal membrane called LysoK_GF.  The channel is made from two proteins TMEM175 and protein kinase B (also known as AKT).

TMEM175 is an ion channel conducting potassium.  It is unlike any of the 80 or so known potassium channels.  It  contains two repeats of 6 transmembrane helices (rather than 4) and no pore loop containing the GYG potassium channel signature sequence. Lysosomes lacking it aren’t as acidic as they should be (enzymes inside the lysosome work best at acid pH).  Why loss of a potassium channel show affect lysosomal pH is a mystery (to me at least).

Genome Wide Association Studies (GWAS) have pointed to the genomic region containing TMEM175 as having risk factors for Parkinsonism.  Some variants in TMEM175 are associated with increased risk of the disease and others are associated with decreased risk — something fascinating as knowledge here should certainly tell us something about Parkinsonism.  

The other protein making up LysoK_GF is protein kinase B (also known as AKT). It is found inside the cell, sometimes associated with membranes, sometimes free in the cytoplasm. It is big containing 481 amino acids. Control of its activity is important, and Cell vol. 169 pp. 381 – 405 ’17 lists 21 separate amino acids which can be modified by such things as acetylation, phosphorylation, sumoylation, Nacetyl glucosamine, proline hydroxylation.  Well 2^21 is 2,097,152, so this should keep cell biologists busy for some time. Not only that some 100 different proteins AKT phosphorylates were known as 2017.  

TMEM175 is opened by conformational changes in AKT.  Normally the enzyme is inactive because the pleckstrin homology domain binds to the catalytic domain inhibiting enzyme activity as the substrate can’t get in.

Remarkably you can make a catalytically dead AKT, and it still works as a controller of TMEM175 activity — this is an example of a moonlighting molecule — for more please see — https://luysii.wordpress.com/2021/01/11/moonlighting-molecules/.

Normally the activity and conformation of AKT is controlled by the metabolic state of the cell (with 21 different molecular knob sites on the protein this shouldn’t be hard).  So the fact that AKT conformation controls TMEM175 conductivity which controls lysosome activity gives the metabolic state of the cell a way to control lysosomal function.  

Notice how to understand anything in the cell you must ask ‘what’s it for’, thinking that is inherently teleological. 

Now on to the two risk factors for Parkinsonism in TMEM175.  The methionine –> threonine mutation at amino acid #393 reduces the lysoK_GF current and is associated with an increased risk of parkinsonism, while the glutamine –> proline mutation at amino acid position #65 gives a channel which remains functional under conditions of nutrient starvation. 

The authors cultured dopamine neurons and found out that the full blooded channel LysoK_GF (TMEM175 + AKT) protected neurons against a variety of insults (MPTP — a known dopamine neuron toxin, hydrogen peroxide, nutrient starvation). 

TMEM175 knockout neurons accumulate more alpha-synuclein — the main constituent of the Lewy body of Parkinsonism.

So it’s all one glorious tangle, but it isn’t just molecular biological navel gazing, because it is getting close to one cause (and hopefully a treatment) of Parkinson’s disease.  

A research idea yours for the taking

Why would the gene for a protein contain a part which could form amyloid (the major component of the senile plaque of Alzheimer’s disease) and another part to prevent its formation. Therein lies a research idea, requiring no grant money, and free for you to pursue since I’ll be 80 this month and have no academic affiliation.

Bri2 (aka Integral TransMembrane protein 2B — ITM2B) is such a protein.  It is described in [ Proc. Natl. Acad. Sci. vol. 115 pp. E2752 – E2761 ’18 ] http://www.pnas.org/content/pnas/115/12/E2752.full.pdf.

As a former neurologist I was interested in the paper because two different mutations in the stop codon for Bri2 cause 2 familial forms of Alzheimer’s disease  Familial British Dementia (FBD) and Familial Danish Dementia (FDD).   So the mutated protein is longer at the carboxy terminal end.  And it is the extra amino acids which form the amyloid.

Lots of our proteins form amyloid when mutated, mutations in transthyretin cause familial amyloidotic polyneuropathy.  Amylin (Islet Amyloid Polypeptide — IAPP) is one of the most proficient amyloid formers.  Yet amylin is a protein found in the beta cell of the pancreas which releases insulin (actually in the same secretory granule containing insulin).

This is where Bri2 is thought to come in. It is also found in the pancreas.   Bri2 contains a 100 amino acid motif called BRICHOS  in its 266 amino acids which acts as a chaperone to prevent IAPP from forming amyloid (as it does in the pancreas of 90% of type II diabetics).

Even more interesting is the fact that the BRICHOS domain is found in 300 human genes, grouped into 12 distinct protein families.

Do these proteins also have segments which can form amyloid?  Are they like the amyloid in Bri2, in segments of the gene which can only be expressed if a stop codon is read through.  Nothing in the cell is perfect and how often readthrough occurs at stop codons isn’t known completely, but work is being done — Nucleic Acids Res. 2014 Aug 18; 42(14): 8928–8938.

I find it remarkable that the cause and the cure of a disease is found in the same protein.

Here’s the research proposal for you.  Look at the other 300 human genes containing the BRICHOS motif (itself just a beta sheet with alpha helices on either side) and see how many have sequences which can form amyloid.  There should be programs which predict the likelihood of an amino acid sequence forming amyloid.

It’s very hard to avoid teleology when thinking about cellular biochemistry and physiology.  It’s back to Aristotle where everything has a purpose and a design.  Clearly BRICHOS is being used for something or evolution/nature/natural selection/the creator would have long ago gotten rid of it.  Things that aren’t used tend to disappear in evolutionary time — witness the blind fish living in caves in Mexico that have essentially lost their eyes. The BRICHOS domain clearly hasn’t disappeared being present in over 1% of our proteins.

Suppose that many of the BRICHOS containing proteins have potential amyloid segments.  That would imply (to me at least) that the amyloid isn’t just junk that causes disease, but something with a cellular function. Finding out just what the function is would occupy several research groups for a long time.   This is also where you come in.  It may not pan out, but pathbreaking research is always a gamble when it isn’t stamp collecting.