Help: cell adhesion assay

Zbigniew Rudzki mprudzki at
Fri Aug 21 09:05:02 EST 1998

Sebastien PARIS wrote in message
<2.2f.32.19980820172858.0067f3b0 at>...
>I would like to try a cell adhesion assay.  Any informations and protocols
>are welcome

I have spent (or rather wasted) ~2 years trying to standardize the adhesion
assay of several cell lines to hyaluronan, chondrotin sulfates and
osteopontin in 96-well, 24-well and 6-well plates.  As an
obsessive-compulsive individual I have tested several hundreds different
protocols.  In my opinion there are several aspects that may this
preoccupation close to futile:

1.  Covering of the of the plate with the stuff your are testing the
adhesion to.
The protocols are numerous, depend on the particular substrate and you have
to pick up and test your own.  You may dry it or not dry it, do everything
at 37 C, 4C or ambient temperature, add BSE prior to the substrate or on the
top of it or not at all.  Even the type of plastic matters (those for tissue
culture have their own strong adhesive properties, so there is the issue of
blocking it).  Also the purity of the substrate is important -  the
components of extracellular matrix are usually sold as 99.9% pure or
something like this.  This remaining mysterious 0.1% may be actually
strongly adhesive.

2.  Removal of the "adhering" cells.  In most papers it is stated that "the
nonadherent cells were removed by washing with PBS"  or something like that.
In fact you will quickly realize that the amount of the cells that are
removed dramatically depends on the almost uncontrollable parameters like:
- the force you are using injecting/ aspirating anything from the wells,
- the angle of the pipette tip and its position in relation to the well,
- your current mood, day of the week, moon phase etc..

There are some protocols that propose different tricks, like centrifugation
(but you have to turn out the plate that per se removes some cells and seal
it firmly!;  also the effective centrifugation radius will differ between
the wells in one plate), use of the shear force generated by the stirring
bar with the plate floating freely above with its wells down etc..  All of
them involve mechanical manipulations with the plate and/or
pouring-aspirating fluids from the wells that detach the different amount of
cells in different wells and experiments.

3. counting of the remaining cells.  This is the relatively easiest element
of the procedure.  Most people use the substrates like MTT that is
metabolized by the adherent cells to colorimetrically measurable compound.
Of course adding/removing anything from the wells will profoundly change the
# of cells still sticking to the wells (MTT has to be solubilized in DMSO,
so you remove the medium and this results in loss of some additional cells).
"Manual" counting sounds easily on paper only.

4. finally the metabolic and receptor status of your cells is important.
Prior to the assay you have to remove them somehow.  Using trypsin or
collagenase will likely result in chewing out some cell surface receptors --
may be exactly those your adhesion depends on. So the cells should be
removed mechanically, that is difficult to standardize in terms of the % of
surviving cells and their dispersion in repeated experiments.  Clumps of the
cells that were not dispersed will behave differently than single cells in
any adhesion assay.

I really do not like to sound fatalistic, and my M.Sc. project in Montreal
was a success due to other research not involving the misfortunate adhesion
assays [and representing the University in epee    :)  ].  In my opinion the
acceptable adhesion assay would require complete automatization of all the
procedures.  However, even the good journals accept papers with the
protocols in which the adhesion assay description sounds really lousy for
someone who knows the cooking details -- apparently such a "stupid" essay is
usually performed by the success-oriented students and people who sign and
review the articles swallow the hook, not knowing the reality.
You will find plenty of protocols in which people did the assay in three
wells/condition (well, I wish they try three more), repeated the experiment
only twice, reported the standard error between the wells in their graphs
instead of standard deviation or did not mention their negative controls and
background adhesion.

Anyway, all the best

The best wishes for Professor Jean Tayot if he still works at the Depth.
Pathol. in Rouen, my favorite French city  :)

Zbigniew Rudzki, M.D., M.Sc., D.Sc
Department of Pathology
Jagiellonian University, Krakow, Poland

More information about the Methods mailing list