Search

Prototypical class switching in mouse and human immunoglobulin heavy chains occurs through recombination of tandem blocks of short repeats located 5' to each heavy chain constant region (CH) except C delta. Deletion of C mu in immunoglobulin D (IgD)-secreting murine plasmacytomas occurs illegitimately. We demonstrate here that in human IgD-secreting myeloma cells freshly isolated from patient bone marrow and in normal peripheral blood B lymphocytes, an IgD switch can occur through homologous recombination of a direct repeat consisting of a 442-bp sequence 1.5 kbp 3' of the JH complex and a 443-bp sequence that is duplicated almost perfectly (96% similarity) 1.7 kbp 5' of the C delta gene (442/443-base-pair [bp] repeat). This homologous recombination mechanism is not exclusive for IgD switching, since C mu deletion endpoints in two established IgD-secreting myeloma cell lines fall outside the 442/443-bp repeat. The 442/443-bp mediated recombination shows cell type specificity, and we propose that it represents a unique mode for increased levels of IgD secretion in humans.

The DNA coding for the human immunoglobulin D(IgD) heavy chain (delta, delta) has been sequenced including the membrane and secreted termini. Human delta, like that of the mouse, has a separate exon for the carboxyl terminus of the secreted form. This feature of human and mouse IgD distinguishes it from all other immunoglobulins regardless of species or class. The human gene is different from that of the mouse; it has three, rather than two, constant region domains; and its lengthy hinge is encoded by two exons rather than one. Except for the third constant region, the human and mouse genes are only distantly related.

Whether the immunoglobulin (Ig) heavy chain genes C mu and C delta are expressed singly or in combination, their transcripts undergo differentiation-specific alterations in membrane (M) versus secreted (S) forms as well as in abundance. To better understand this regulation, we have cloned cDNAs for human delta m and delta s to establish the 3' end of the C mu-C delta transcription unit. Steady state mRNA levels and transcription rates were then analyzed in normal and transformed human B cells representing different maturation and activation states. The ratio of micron/microsecond RNA and of delta m/delta s RNA correlated with developmental stage, with a higher ratio at earlier stages. Steady state ratios of total mu/delta RNA paralleled ratios of C mu/C delta nascent transcription, suggesting no major posttranscriptional control for differential expression. However, at all developmental stages, transcription termination occurred downstream of the micron exons, suggesting a strong posttranscriptional regulatory component for production of secreted versus membrane forms of mu RNA. The relative abundance of mature delta S RNA was considerably higher in the human than in the mouse, correlating with the increased levels of circulating IgD in the former species. Stimulation of human splenocytes with mitogens did not increase delta RNA; in fact, splenocytes activated with pokeweed mitogen were nearly devoid of delta RNA, and Staphylococcus aureus Cowan I caused only a minor change.

The entire nucleotide sequence (approximately 20 kbp) spanning the human immunoglobulin IgM (mu) and IgD (delta) heavy chain constant region genes has been determined from DNA of mu-delta producing chronic lymphocytic leukemic B cells. As in the murine IgM + IgD double-producing B cells, no rearrangement has occurred in the C mu-C delta region in the leukemic cells. The C mu locus is highly conserved between mouse and human with the exception of the nucleotide sequence between the C mu 4 and mu M1 exons, which has diverged dramatically. The intergenic sequence between human C mu and C delta is three times larger than the analogous region in the mouse and contains notable features absent from the mouse, including a 443 bp segment that is 96% identical to a 442 bp sequence that occurs just 3' to the heavy chain enhancer, a 366 bp sequence that is directly repeated with 76% homology, and 12 tandem copies of a 35 bp sequence. The human C delta gene contains two additional exons relative to mouse C delta, but shares with the mouse the unique distal location of both secreted and membrane coding segments. Several polymorphisms in the human population have been identified in the intergenic region and in C delta but not in C mu.

The data presented here suggest a model for isotype-specific regulation of IgA synthesis by Fc alpha R+ T cells (Figure 1). Immature mIgM+ +/- mIgD+ B cells are induced by T switch cells to express cell surface IgA (a phenotypic switch). If the T switch cell induces mIgA expression via a long primary RNA transcript from an unrearranged C alpha allele, the hypothetical intermediate switch B cell results (step 1); this may be the mechanism of heavy chain expression in memory B cells that express low levels of Ig. Alternatively, T switch cells may induce a DNA rearrangement in the CH locus of the B cell (a genotypic switch), which results in a deletion of all CH loci except C alpha (step 2). TH inducer cells promote maturation of mIgA+ B cells to IgA-secreting plasma cells. This may involve a DNA switch rearrangement (step 3) or the maturation of previously switched cells (step 4), and appears to be mediated via an IgABF with enhancing activity. Not shown in this figure, but inherent in this model, is a suppressive regulatory arm that may be mediated via IgABF with suppressive activity released from Fc alpha R+ suppressor T cells. Due to the presence of Fc alpha R on a variety of cell types, IgABF may suppress synthesis of IgA by acting not only on mIgA+ B cells but also on regulatory cells (T cells, B cells, and macrophages) bearing IgA bound to Fc alpha R. If the IgA system is analogous with the IgE system, mIgA-bearing B cells may be the direct target of IgABF. Binding of Ig to FcR has been shown to (a) increase the number of Fc receptors per cell, (b) enhance the number of cells expressing Fc receptors, (c) induce the release of IgBF that either suppress or enhance Ig secretion, and (d) effectively convert surface Ig- cells into surface Ig+ cells that are therefore receptive to IgBF. Thus, FcR+ cells may interact with IgBF and Ig via a regulatory network to stimulate or inhibit the immune response in an isotype-specific manner. Cell surface molecules (mIg, FcR) may serve as sensors that allow the cell to detect and respond to fluctuations in the levels of immune mediators that serve to modulate Ig synthesis and secretion. The relationship between IgBF and FcR is not known, nor is it known whether Fc receptors expressed by different cell types are encoded by the same gene and are controlled similarly.(ABSTRACT TRUNCATED AT 400 WORDS)