We review our studies on the mechanism of somatic hypermutation of imm
unoglobulin genes. Most experiments were carried out using Ig transgen
es. We showed in these experiments that all required cis-acting elemen
ts are present within the 10-16 kb of a transgene. Only the Ig variabl
e region and its proximate flanks are mutated, not the constant region
. Several Ig gene enhancers are permissive for somatic mutation. Assoc
iation of the enhancer with its natural Ig promoter is not necessary.
Ho Never, the mutation process seems specific for Ig genes. No mutatio
ns were found in housekeeping genes from cells with high levels of som
atic hypermutation of their Ig genes. The Ig enhancers may provide the
Ig gene specificity. An exception may be the BCL6 gene, which was mut
ated in human but not in mouse B cells. Transcription of a region is r
equired for its mutability. When the transcriptional promoter located
upstream of the variable region is duplicated upstream of the constant
region, this region also becomes mutable. This suggests a model in wh
ich a mutator factor associates with the RNA polymerase at the promote
r, travels with the polymerase during elongation, and causes mutations
during polymerase pausing. The DNA repair systems, nucleotide excisio
n repair and DNA mismatch repair, are nor required. Our recent data wi
th an artificial substrate of somatic mutation suggest that pausing ma
y be due to secondary structure of the DNA or nascent RNA, and the spe
cific mutations to preferences of the mutator Factor.