Antibody repertiore (longish pedantic reply)

Tom Frey tomfrey at
Wed Apr 30 00:33:47 EST 1997

Malarchuk wrote:
>  I am currentlyenrolled in an Immunology course.  I understand that the
> human body has an antibody repertiore of 10e9 antibodies.  Our teacher
> indicated that the
> human body can make antibodies to common everyday plastic, how is that
> possible if plastic is only a relatively "recent" invention?

Antibody binding sites are generated by randomly combining v and j
minigenes (plus a d segment in heavy chains).  There is also some
junctional slop (N region diversity) around the recombination joints. 
This leads to a very high variability in two of the six loops that make
up the antibody binding site (one loop of three provided by each
chain).  Further mutation can also lead to changes in the antibody. 
10e9 seems an underestimate.

Now, to your real question ... antibodies are the receptors on B cells,
and if an antigen binds the B cell divides and produces more identical
or perhaps slightly mutant progeny.  Since the antibody binding sites
are relatively "infinitely variable" some indviduals will have some
antibodies that recognize almost anything.  The selection process will
mean that if you have any cells that recognize an antigen you will
develop more if you see that antigen, and the mutation process means
that if the antigen stays around you stand a good chance of developing
even higher affinity antibodies.  The system is set up to recognize
stuff you have never seen, since the short lifecycle of pathogens gives
them a mutation rate that can generate new antigens during the course of
an outbreak.

There are clearly further nuances.  For example four of the six loops
are inheritable, and are thus under some evolutionary pressure.  They
actually seem to skew the repertoire toward being able to recognize
common pathogens.  Also, T cells have some controlling roll and would
need to recognizes the antigen also.  It is usually at this level that
self-recognition is eliminated.  As a final complication (for this
overly long note, anyhow) an antigen like plastic (long chains of
similar stuff, ie polyvalent) may not need T-cell help.

More information about the Immuno mailing list