Morning Anthony,
	If you are sectioning a material of small dimensions, then the
embedment MUST have sufficient hardness to hold the specimen rigid through
the cutting stroke.  At a minimum you will probably have to change to a
harder embedment.  With good infiltration and blocking, the 'compliance'
mismatch should disappear.  Finally, if you absolutely must embed the
compression device, then you will probably have to resort to sectioning with
a tungsten-carbide knife, a sapphire knife or a diamond knife, in each case
reducing the size of the specimen [depending on the material of which the
compressor is made - wood will 'make room' for the compressed material -
less for oak than pine].
	Let's say you want to do this with a hard material like bone.  Once
you have applied the compression (measured in p.s.i. or um of separation),
what do you select as a specimen?  Continuously compressed thru processing?
If that, then your problem is/will be diffusion of post-compression
solutions (isn't the compressor in the way?).  If you remove the compressor,
wait five minutes, and then preserve 'residual' compression, you should have
no problem with any piece of tissue that is of reasonable size.  In normal
tissue, during fixation (hours) and cross linking (days) with HCHO, much
more material will be moving around in the compressed [injured] tissue than
in un-compressed ['normal'].  The 'injury' caused by compression in soft
tissue will be to the cells first and then to other, extra cellular,
constituents [my assumption].  If that is true, then the compressed specimen
presents a much more uniform barrier to diffusion than non-compressed tissue
[again my assumption].  If you are testing dense connective tissue, then
diffusion of solutions will be an even greater problem if you hold the
compression thru processing, and you will be discounting a 'normal'
expansion and shrinkage of the tissue during the process.  
	With most tissues holding 70+% HOH, it is interesting to consider
that if external P increases equally along all vectors that converge on the
center of gravity of a cell, and HOH is incompressible, what damage might be
inflicted by such compressive forces.  Alternatively, by increasing pressure
along two opposing vectors, the alterations in morphology would be via
deformation along vectors at right angles to those applied forces, freedom
of deformation being the only limiting factor.  So, if deformation damage is
what you wish to study, then both the orientation of the tissue and the
freedom to deform are considerations.  For example, if a spherical cell is
so deformed, its orientation would not be a consideration, but if, on the
other hand, the cell is spindle-shaped, then its orientation is most
significant.  Similar thinking must be applied to filaments/fibers that are
randomly and non-randomly distributed in the tissue compartment.  If one is
dealing with tendon, then its orientation and freedom to deform at right
angles to the forces applied are significant matters to be considered.  The
entire experiment reduces to a problem of engineering the conditions of the
test and the means by which one 'fixes' the result of the test [the tissue]
such that the consequence of any test can be observed morphologically.
	With these considerations in mind, I would suggest flash-freezing
and removal of the compressive device followed by freeze-substitution at
-20oC in 100% EtOH or acetone, then direct embedment in glycol methacrylate
(GMA) (graded series of EtOH/GMA).  Flash freezing because you could then
compress with steel or lightly-Teflon-coated BRASS and get rapid cooling in
LN2, and then not worry about maintaining the compression.  The process of
substitution could take a few days or a couple months depending on the size
of the specimen.  My suggestion is not from experience but rather from a
desire to suggest some means by which treatments remain similar.  There
would be no problem with a waiting period following compression to study
'residual' compression effects.  I believe that the power of any experiment
lies in the ability of the designer to minimize variables (loose ends),
thus, all treatments [(including the force applied?) or (except compressive
force)] should be as much alike as possible.  
	Twelve turns of the screw might nicely serve the purpose of
eliciting a moan from a graduate student, and twenty might elicit a scream,
but if you want to know something about subject variance at the end, you
must use something like p.s.i. or, where freedom to flow/deform is not
restricted, distance between the faces of the compressive device.
	

Regards and good luck,

Fred Monson

Frederick C. Monson, PhD
Center for Advanced Scientific Imaging
Schmucker II Science Center, Room SS024
West Chester University of Pennsylvania
South Church Street & Rosedale Ave.
West Chester, PA, 19383
Phone & FAX:  610-738-0437
eMail:  fmonson@wcupa.edu
http://darwin.wcupa.edu/casi/
 

-----Original Message-----
From: Anthony Choo [mailto:amchoo@interchange.ubc.ca]
Sent: Monday, August 12, 2002 12:35 PM
To: HistoNet Server
Subject: embedding non-tissue material


Hello,

I am trying to embed some soft tissue under residual compression. (i.e. I
want to be able to look at compressed structures under the microscope.) I've
used nylon screws to apply the compression and embedded in JB4 plus.  The
specimen does not cut well on a glass microtome (I presume because the
plastic cannot be inflitrated giving me a compliance mismatch).  Any
suggestions for alterate materials?

(Note: I've considered using a device made of, say wood, however,
manufacturing the compression device would be more precise