Antibody level determination

Dom Spinella dspinella at
Mon Oct 6 10:24:29 EST 1997

James Good writes: 
> Dear Reader
> I'm confused.
> There are several ways to determine the level of specific serum antibodies
> to a particular antigen : end point titre, ug antibody per ml serum,
> percentage of a reference serum, OD at a given dilution, etc. What I need
> to know is this : what are the pros and cons of each method? Is one more
> suitable than another for a given application? Which is better when it
> comes to calculating antibody isotype ratios (ie which gives a nearest
> approximation to the truth)?
> Thanks for any help
> James

Almost all these methods provide relatve (not absolute) measure of
antibody concentration.  The reason is that a "titer" (or an OD at a
given dilution when measuring by ELISA) is dependent upon several
factors besides the actual concentration of antibody in the serum.  One
of these factors is of course the assay which is used for measurement. 
Obviously if you perform an endpoint titation of a given antiserum using
a sensitive technique like RIA and an insensitive technique (say,
precipitation), the endpoint titers will be different. This is true even
using the same assay format -- a sensitive, optimized ELISA will provide
a higher apparent titer than a poorly optimized ELISA.  A second factor
is affinity of the antibody being measured.  A high-affinity antibody
will be detectable at higher dilutions than a lower affinity antibody at
the same concentration (and hence will appear to have a higher titer).

The most accurate method for assaying total quantities of each isotype
in a serun (assuming antigen specificity is irrelevant) is to actually
determine ug/ml.  However, this must be well-controlled with an
isotype-matched reference.  For example, if you wish to know the
concentration of each isotype in a serum, you could measure each
independently with a isotype-specific labelled antibody using a standard
curve generated from a purified monoclonal antibody of each isotype at
known concentrations.  However, if you wish to determine the relative
isotype levels of antigen-specific antibodies, the problem is more
difficult.  To be truly accurate, it is not enough to perform a series
of parallel ELISAs in which the second antibody is isotype specific. 
First, you don't know that the average affinity of each isotype for the
target antigen is the same (in fact, you can bet that its not -- IgM
antibodies are almost always of lower affinity than IgG for example). 
Also, you don't know that the affinity of the anti-isotype reagents is
the same (the anti IgG antibody may be of higher affinity than the
anti-IgA antibody for example and therefore provide a higher signal for
the same amount of target isotype).  Also, the isotypes differ with
respect to the number of antibody combining sites (molecule for
molecule, pentavalent IgM will bind more anti-IgM than IgG will bind
anti-IgG for example).

So depending upon how accurate you need to be, again the best (albeit
the most tedious) is to perform a standard curve with an affinity
purified reference antibody of each isotype.  Most peole don't go
through all that trouble and simply measure endpoint titers for each
isotype and accept the caveats in interpretation.  

I hope this is of some use to you.  Good luck.  -- Dom Spinella

More information about the Methods mailing list