Re: Experiments which led to the elucidation of the structure of the antibody

Dear John:

Thank you for your interesting question.  It was actually something that I 
was asked about on my PhD program qualifying exam.  The first descriptions 
of antibodies (or gamma-globulins) were from Tiselius and Kabat, in 1939 (J. 
Exp. Med 69:119).  They immunized rabbits with ovalbumin (chicken egg white 
protein) and ran the immunized serum on a gel.  They saw four protein bands 
(globulins): albumin, alpha-globulin, beta-globulin and gamma-globulin.  
They then did an insightful experiment, where they adsorbed the serum 
against ovalbumin first, then ran the resultant "depleted" serum on a gel.  
They found that the gamma-globulin peak was dramatically reduced.  This 
experiment pointed out that there is some globulin in serum that binds 
directly to antigen.

The next step forward in our understanding of the immunoglobulin molecule 
came from ultracentrifugation and digestion experiments in the 1950s and 
1960s.  Rodney Porter and Gerald Edelman used ultracentrifugation to 
separate the immunoglobulin protein fraction into a 19S (high molecular 
weight) and a 7S (low molecular weight) fraction. Porter then used papain 
digestion of the 7S IgG molecule to yield the now well-known 2Fab and 1Fc 
per whole IgG molecule (these were resolved by gel electrophoresis).  The 
Fc, incidentally, was called Fc because it was a fragment that readily 
crystallized if chilled.  Alfred Nisonoff also digested the antibody 
molecule, but with pepsin, which yields F(ab')2 (which was about 100kD, a 
little over twice the size of the 45kD Fab fragment) and no Fc, because the 
Fc is digested away.  These researchers also noted that the Fab and F(ab')2 
portions of the molecules could precipitate antigen, but the Fc could not, 
so the Fab and F(ab')2 contained the antigen-combining region of the 
molecule.  After separating the immunoglobulin molecule into the antigen-
combining portion and the crystallizable portions, Edelman and Porter then 
went on to separate the molecule into heavy and light chains by 
mercaptoethanol reduction and alkylation, which cleaves intrachain disulfide 
bonds.  These experiments demonstrated that IgG was composed of two 50kD 
heavy chains and two 25kD light chains.  For their contributions to our 
understanding of the immunoglobulin molecule, Porter and Edelman shared the 
Nobel Prize in 1972.

So Porter, Edelman and Nisonoff elucidated the components of the 
immunoglobulin molecule in the experiments above, but the function of each 
portion remained unclear.  Porter addressed this question, too, in a series 
of seminal experiments.  He immunized goats with rabbit Fab fragments and 
rabbit Fc fragments, then asked whether the goat antisera would react with 
the rabbit heavy or light chains.  He found that the Fab antisera reacted 
with both the heavy and light chains, while the Fc antisera reacted with 
only the heavy chains.  Based on this obeservation, Porter and Edelman 
proposed that the IgG molecule consists of two identical disulfide-bonded 
heavy chains which are further disulfide-bonded to two identical light 
chains.  This insight has proven to be accurate up to today.

The next step in determining protein structure, beyond chain resolution, is 
to sequence the protein.  Initial attempts at sequencing were hindered by a 
lack of sufficient quantities of protein.  A human cancer, multiple myeloma, 
ultimately solved this problem.  Multiple myeloma is a cancer of plasma 
cells, which are antibody-secreting B cells, and the myeloma proteins can 
account for up to 95% of the serum immunoglobulin.  Many myeloma patients 
produce an excess of light chains, which can be filtered out of the blood by 
the kidney and end up in the urine.  These light chains in the urine, 
referred to as Bence-Jones proteins, provided enough protein for the first 
light chains to be sequenced.  When these light chains were compared, it was 
noted that the amino terminal half was variable, while the carboxy terminal 
half was constant and came in one of two flavors lambda or kappa.  Heavy 
chains were also sequenced from myeloma patient serum and compared.  Again, 
there was a variable amino terminal portion, and one of five constant 
regions: alpha, gamma, mu, delta or epsilon.

Finally, the immunoglobulin molecule was solved by X-ray crystallography.  I 
believe this was first done by LJ Harris in 1992 (Nature 360:369), but I may 
be wrong.  The polymeric structure of IgM was contributed by electron 
micrography by Davis, Roux, and Shulman (1988, European J. Immunology 18: 
1001). (These are cool pictures - you should check them out sometime).

These are the fundamental experiments that I know of that elucidated 
structure of the antibody molecule.  I hope it is helpful!

Sincerely

Ingrid Dodge, Mad Scientist

My favorite textbooks on immunology are:

Paul, WE Ed. Fundamental Immunology, 4th Ed.  New York: Lippincott-Raven 
Pub., 1999.

Kuby, J. Immunology, 3rd Ed.  New York: WH Freeman and Co., 1997 (from which 
I drew much of this discussion, as well as Paul above, and my history of 
immunology class notes).

Janeway and Abbas also have good introductory textbooks.