Some Questions About Immunity That Deserve More Attention:

a) Reading frame shifts inevitably result from most V(D)J recombinations of DNA sequence. Two out of three random recombinations between base sequences will produce a frame shift, and not only do frame shifts result in a radically different amino acid sequence from that point onward, but most of these produce chain terminating base sequences.

The joining of any Variable sequence to a Diversity sequence is not controlled so as to minimize frame shifts. Therefore two thirds of these recombinations produce radically different amino acid sequences; and so will two thirds of the recombination events between Diversity and Joining sequences, and two thirds of the Joining to Constant sequence recombinations.

That means only 1/3rd of 1/3rd of 1/3rd = 1/27 (= 4%) of heavy chain coding DNA won't have at least one frame shift in the resulting DNA that codes for their antibody heavy chains.

The light chains coding sequence will have only 1/9th (~10%) chance of not being frame-shifted. a tenth of a twenty seventh is less than half a percent.

Theoretical Immunologists have long recognized this paradox. The answer is supposed to be that ~99% of developing lymphocytes self-destruct.

Another part of the answer is that there are two heavy chain genes in a diploid cell and four light chain genes (two kappas and two lamdas). That's supposed to give you 6 (or is it fewer) more chances of getting at least one "Productive Recombination".

b) The key idea, seldom expressed explicitly, is that developing lymphocytes can somehow detect frame-shifted recombinations, and either suppress transcription of those antibody genes which have undergone a non-productive recombination, or cause that cell to undergo apoptosis.

One problem with such a hypothesis is how to detect frame shifts except based on synthesizing the actual protein. Nobody has a molecular mechanism that can look at a DNA sequence and detect frame shifts. But please tell me if you have ideas how to do this, or if there is evidence or good theories.

c) Another tricky situation arises when recombination produces a heavy chain DNA sequence and a light chain sequence that do NOT have a frame shift, and does it on the first try. Presumably that lymphocyte should make antibodies with no problems? Right? What's to worry about?

This cell, being diploid, has one more "Do-It-Yourself-Kit" for recombining VD and J sequence of DNA. For the light chains, diploid cells start with four such "kits"; in other words there can be three left over, etc,. The worry is that such a lymphocyte would get past the tolerance mechanism, and then later in life maybe loses the gene coding for the heavy chain or for one of the light chains. If the homologous chromosome then gets de-repressed, in the sense of becoming able to undergo V(D)J recombination, then that cell will become able to make antibodies that might be anti-self, because they will have escaped the chance of getting weeded out by the tolerance mechanism.

Cell. 2004 Sep 3;118(5):539-44.
The lingering enigma of the allelic exclusion mechanism.
Mostoslavsky R1, Alt FW, Rajewsky K.