Fixation of Endothelial Cells

Richard Aaron Warnock erawtech at leland.Stanford.EDU
Wed Mar 20 15:54:33 EST 1996


In article <4ipc9r$bel at mordred.gatech.edu>,
Richard Montes  <gt0441b at prism.gatech.edu> wrote:
>
>Does anybody have experience or know of literature about fixation of 
>endothelial cells and its effect on the membrane receptors on the surface 
>of the endothelial cells? I am trying to investigate the what is the 
>effect of the lateral diffusion of the receptors across the membrane to 
>cell adhesion. If I crosslink the membrane domains of these diffusing 
>proteins thereby anchoring them to eliminate diffusion, are they still 
>viable for adhesion?

Most likely. That is, if you *only* x-link adhesion molecules at the
membrane. In fact, clustering seems to be a recurrent theme of
adhesive interactions. You've got your work cut out for ya, there. A
lot (dare I say most?) endothelial adhesion molecules have either
stable or transient association with cytoskeletal components, either
directly, or linked through other molecules... On top of that, these
associations change with the state (activation, intercellular
associations, differentiation, etc.) of the cell. Obviously, if you
crosslink them to something else (other molecules, each other,
whatever...), you'll severely compromise these associations. For
example, read the literature concerning integrins and focal adhesion.

Also consider that cells will "patch" x-linked molecules (in the
absence of total cell surface fixation), showing that x-linking does
*not* "anchor" anything, just associates them.

>Second, are there differences in the fixative agents used, e.g., 
>glutaraldehyde, formaldehyde, metOH, etc?

Whoa! You've got some work there, too. No fixative will be appropriate
for all molecules (esp. if you want to perserve function). However,
these you've listed are particularly harsh, and alcohols will
percipitate the membrane... I personally know that you will pretty
much destroy (functionally) the bulk of leukocyte binding molecules.
I've had minor luck with .5% paraformaldehyde, but any left over
function (in my opinion) is probably due to unfixed molecules...

>Any input will be much appreciated. Please cross-email me privately 
>together with the response to the newsgroup. Thanks in advance.

One suggestion I have is to use Fab'2s agaist membrane proximal
domains from mAbs that you know won't block adhesion. For greater
x-linking, you could use whole mAb and 2nd stage... These have the
advantage of specificity, and won't screw up the rest of the surface
molecules (maybe...).

There are tons of bidirectional (chemical) x-linking schemes as
well. It's difficult to offer suggestions without knowing what type of
molecule(s)/function(s) your aiming at.

If you want to look at function of unpeturbed 3D surface
constellations, (ie static surface relationships) on function, you
might try inhibitors (eg energy specific, cytoskeletal, etc.).

Be aware that adhesion goes hand in hand with transmembrane signalling
(in and out), and any conclusions must take this in consideration.

Perhaps the your best bet will be to use molecular genetic techniques
to force your target to behave as you wish... 

Might be able to help more knowing more specifics.

Best wishes,

Aaron
-- 
"Nothing more is needed to destroy a man, than the conviction that his
life's work is useless."  -Antonin Artaud

erawtech at leland.stanford.edu (R. Aaron Warnock)



More information about the Cellbiol mailing list