F 1 D 0 2001 12 28 Insulin Receptor Talk
I heard this presentation about a year
ago, and it was suggested by Bill Livingstone
that we should post that presentation's
summary here. I concur.
Peter Ottensmeyer is a physicist, but has taken a
liking to modeling molecules. fpo@oci.utoronto.ca
He introduced the history of Electron Microscopy,
and then dug in to the meat of his talk.
>To: (Bill Livingstone) vege@f1d0.com
>Subject: insulin receptor
>Date: Thu, 07 Dec 2000 20:58:48 -0500
>Here is my summary of the lecture today.
I keep getting people asking me to take
pictures of their molecules, but this one
got my attention. Diabetes affects between
5-30% the population, depending on how you look.
(These are) images of bacteria, a virus, and
some pretty molecules. Studies have confirmed
that the pictures taken by Electron Microscopy
and the cartoons drawn are homogolous. So this
has been valuable, making such models.
Insulin receptor is is a P-collector protein. They
pump drugs out of the cell.
How do we get the 3D pictures? Also, how
do we deal with a 2D image of a 3D molecule?
Note that Electron Microscopy causes the
molecule itself to degrade!
To help us out nowadays, not only do we measure
the shadows, but we measure the high angle scatter
and the low angle scatter. All of this combines to
200 times more information than normal Electron
Microscopy. We count every possible electron!
I introduce "Ducks". You see, we
want to get all of the ducks in a row.
To measure ducks (this Insulin Receptor Protein)
we lay them down on a thin film (20 Angstrom) carbon film.
That holds them still, but they're still lying
down in a helter skelter fashion. As investigators,
we need to find out if we can learn the orientation
of the ducks!
We project in 3D and convert each projection to a
plane, and that plane is compressed to a line. Then
a computer can make inferences about the shape.
Using computer numerical means, we can rotate along
the line. All of this is an effort to determine the
orientation, that is, which end is the TOP. Which
side is (this) side, which is (that) side. It is
very hard to do.
Noise is a problem. We get as much as 20% noise
in a signal, and that makes it hard to decide
if we are looking at noise or images when we
rotate what we have.
We work toward a rotation matrix using physics
and math (Oops. Sorry about that. Uh, forget I said
that.) ...and gives us precise orientation. (This is)
a picture of SRP-54.
At the lab we get so many requests
to take pictures of molecules, but now
we just show people how to do it for themselves.
(You see, I won't do it anymore).
In spite of that, I decided to help out
the Insulin Receptor professors. "Insulin
Receptor is the most important molecule in
the world", says this one guy. (I dunno, says
me).
It doesn't just metabolise Glucose, but helps
synthesize proteins, lipids, glycogen... lots
of good stuff.
Insulin is not very big - only 6000 Daltons.
But we can connect it to a nugget of gold, and
then we can see it in pictures.
Here is a picture of the 3D structure, as a solid.
Which way is UP? Who knows. We rotate it around.
We are no wiser.
Now we go to a wire mesh, and try the same thing.
We know from that where the insulin connects, so
we found the top. Now we can start to untangle it
and it looks so beautiful!
The inside part of the receptor is Tyrosine Kinase.
Now, suddenly we can rotate and learn something.
Imagine a ship with 3 stacks, and a keel of Tyrosine
Kinase.
We need to ask ourselves, is our interpretation of
this in any way correct?
Well, we have lots of research done on
Tyrosine Kinase, via Xrays and more, so yes.
Also they've figured out how L1 and L2 work in
Australia (L1 and L2 are two of the lobes of the
insulin receptor).
The monamers have a large hydrophillic group. As
we animate we find positive and negative
amino acids do interractions that are a
nice surprise. Many hooks hold the insulin
when it comes.
Does it tell us anything? It lets the inside of the
cell know there is insulin waiting.
Tyrosine Kinase receptors all work in
monamer form. The ligand is what holds
it together, while the inside flops around.
To figure this out we tried
- a molecular model (but that didn't work)
- an opaque model (but that didn't work either)
- a bones model (that helped us see where things connect)
- a cylinder model (that simplified the whole interaction once done)
Insulin is a doublesided velcro which holds things
together. I tested my theory with lego! It was
faster than computer modelling. Occasionally there
is a transfer of a phosphate, even without insulin,
but in general, insulin is required for all of this
to happen.
We can explain 30+ papers with this new structure.
So.
Now we know how transmembrane
insulin receptor signalling
occurs.
Q: Prions? have you done any research into this?
A: not really, but diabetes affects more of the
population than prions.
(besides, I don't want to have to get that close to
BSE, and I like my steak, thank you)
That's all I know.
