Multiple immunoglobulin switch region homologies outside the heavy chain constant region locus

Abstract

Somatic rearrangement of gene segments enables higher organisms to generate additional genetic diversity from an essentially limited and otherwise linear genome. The only known example of this powerful mechanism in vertebrates is the immune system, in which active genes are somatically assembled from distant pieces of DNA^1–3. This process operates both in V–J (light chain) or V–D–J (heavy chain) joining, to form novel variable region genes from widely separated coding segments^4–9, and in heavy chain class switching, whereby assembled variable region genes are switched from one heavy chain gene to another^10–14. The latter process correlates with the sequential appearance of the same V region in association with different heavy chains during B-cell development^15–22. Conserved sequences adjacent to these genes may be required for the two processes^{4–6,23–28}. In the case of the heavy chain class switch from the μ to the α gene, such sequences involve 2–3-kilobase (kb) segments characterized by tandem pentanucleotide and higher-order repeats^24–26,28. We have previously shown that this region has been selectively conserved in both mouse and man and therefore changed little during the 70 Myr since these species have diverged²⁸. Using the switch segment as a hybridization probe, we now show that approximately 15 switch-like regions are present in germ-line DNA and that only one is deleted during the μ–α switch that has occurred in the IgA-producing cell line, S107. As the heavy chain constant region locus from μ through ε is deleted in this cell line (including all the other known heavy chain genes), we conclude that the remaining switch-like signals lie outside this region.