1096

Genes encoding intact immunoglobulins do not exist already assembled in the genome, ready for expression. Instead, the required gene segments are brought together and assembled in the course of B-cell development. The organization of the region of the genome containing the immunoglobulin genes is shown in Figure 23-15. In B cells, the DNA in this region is rearranged as described below to generate assembled and fully functional immunoglobulin-encoding genes in each B cell and its descendants. Although the rearrangement of heavy-chain genes occurs before the rearrangement of light-chain genes, we discuss light-chain genes first because of their less complex organization.

The immunoglobulin light-chain genes consist of clusters of V gene segments, followed downstream by a single C segment. Each V gene segment carries its own promoter sequence and encodes the bulk of the light-chain variable region, although a small piece of the nucleotide sequence encoding the light-chain variable region is missing from the V gene segment. This missing portion is provided by one of the multiple J segments located between the V segments and the single C segment in the unrearranged κ light-chain locus (see Figure 23-15a). (This J segment is a genetic element, not to be confused with the J chain, a polypeptide subunit of the pentameric IgM molecule and found also in association with IgA; see Figure 23-10.) In the course of B-cell development, commitment of a B-cell precursor to use a particular V gene segment—a random process—results in its physical juxtaposition with one of the J segments, again a random choice, to form an exon encoding the entire light-chain variable region (V_L). This DNA rearrangement not only generates an intact and functional light-chain gene, but also places the promoter sequence of the rearranged gene within controlling distance of enhancer elements, located downstream of the light-chain constant-region exon, that are required for its transcription. Only a fully rearranged light-chain gene is transcribed and subsequently translated into protein.

1097

1098

Recombination Signal Sequences Detailed DNA sequence analysis of the light-chain and heavy-chain regions revealed a conserved sequence element at the 3′ end of each V gene segment. This conserved element, called a recombination signal sequence (RSS), is composed of heptamer and nonamer sequences separated by a 23-bp spacer. At the 5′ end of each J segment, there is a similarly conserved RSS that contains a 12-bp spacer (Figure 23-16a). The 12- and 23-bp spacers separate the conserved heptamer and nonamer sequences by one and two turns of the DNA helix, respectively.

Somatic recombination is catalyzed by two enzymes, the RAG1 and RAG2 recombinases, which are expressed only in lymphocytes (Figure 23-17). Thus these rearrangements do not occur in any other cells of the body. Juxtaposition of the two gene segments to be joined is stabilized by the RAG1/RAG2 complex (Figure 23-16b). The recombinases then make a single-stranded cut at the exact boundary of each coding sequence and its adjacent RSS. Only gene segments that possess heptamer-nonamer RSSs with spacers of different lengths can engage in this type of rearrangement (the so-called 12/23-bp spacer rule). Each newly created –OH group at the site of cleavage then executes a nucleophilic attack on the complementary strand, creating a covalently closed hairpin for each of the two coding ends and double-strand breaks at the ends of the RSSs. Protein complexes that include the Ku70 and Ku80 proteins hold this complex together so that the ends about to be joined remain in close proximity: double-strand breaks in chromosomes need to be repaired, and thus the ends need to be held together for resolution and repair of these breaks to proceed. The RSS ends are then covalently joined without loss or addition of nucleotides, creating a circular reaction product (deletion circle) containing the intervening DNA, which is lost altogether. The hairpin ends of the coding segments undergoing recombination are then opened and finally joined as depicted in Figure 23-16c, completing the recombination process.

The recombination mechanism just described, called deletional joining, occurs when the V gene segment involved has the same transcriptional orientation as the other gene segments at the light-chain locus. Some V gene segments, however, have the opposite transcriptional orientation. These segments are joined to J segments by a mechanism, termed inversional joining, in which the V segment is inverted and the intervening DNA and RSSs are not lost from the locus.

Defects in the synthesis of RAG proteins obliterate the possibility of somatic gene rearrangements. As described below, the rearrangement process is essential for B-cell development; consequently, RAG deficiency leads to the complete absence of B cells. People with defects in RAG gene function suffer from severe immunodeficiency. Targeted deletion of RAG genes in mice likewise leads to a complete defect in immunoglobulin (and T-cell receptor) gene rearrangement, resulting in a developmental block in the generation of B and T lymphocytes.

Junctional Imprecision In addition to the random selection of V and J gene segments, processing of the intermediates created in the course of somatic recombination provides an additional means for expanding the variability of immunoglobulin sequences. This additional variability is created at the junction of the segments to be joined. The opening of the hairpins at the coding ends is a key step in this process: this opening may occur symmetrically or asymmetrically (see Figure 23-16c, steps 4 and 5). The protein Artemis, whose function requires the catalytic subunit of DNA-dependent protein kinase, carries out the opening of the hairpins.

1099

If the opening of a hairpin is asymmetric, a short, single-stranded palindromic sequence is generated. Filling in of this overhang by DNA polymerase results in the addition of several nucleotides, called P-nucleotides, that were not part of the original coding region of the gene segment in question. Alternatively, the overhang may be removed by an exonuclease, resulting in the removal of nucleotides from the original coding region. These possibilities apply equally to the V and the J coding regions. Symmetric opening of a hairpin retains all the original coding information. However, even if the hairpin is opened symmetrically, the ends of the DNA molecule tend to breathe, creating short single-stranded sequences, which may also be attacked by nucleases. Once the hairpins have been opened and the coding ends processed, the ends are ligated by two proteins, DNA ligase IV and XRCC4, generating a functional light-chain gene.

Inherent in this rearrangement process is junctional imprecision resulting in part from the addition and loss of nucleotides at the coding-region joints. When a V and a J segment recombine, the sequence and reading frame of the VJ product cannot be predicted. Only one in three recombination reactions results in a reading frame that is compatible with light-chain synthesis. The others produce frameshifts that do not encode functional proteins.

Light-chain diversity therefore arises not only from the combinatorial use of V and J gene segments, but also from junctional imprecision. Inspection of the three-dimensional structure of the light chain shows that the highly diverse joint generated as a consequence of junctional imprecision forms part of a loop—hypervariable region 3 (HV3)—that projects into the antigen-binding site and makes contact with antigen (see Figure 23-13b).