Lectures List

B and T cell development

The development of B and T cells differ in a number of important ways which will be discussed below. There are several common features though one of which, allelic exclusion, applies to both lineages and results in the ability of mature T and B cells to express a single antigen receptor. This process will be discussed first and will be exemplified with B cells but applies equally to T cell expressing a/b or g/d TCRs.

In diploid cells there exist both a maternal and a paternal allele of each gene and typically both alleles are expressed in a co-dominant manner. Antibody and T cell receptor genes are autosomal and therefore, in principle, individual cells could express two different antibodies or T cell receptors resulting from separate rearrangements at the maternal and paternal loci. As a result, however, each B or T cell would express several antigen specificities due to promiscuous assembly on the cell membrane of different H and L chain pairs (in the B cell) or diffferent a and b chains (in T cells). This does not occur because evolution has devised a strategy that allows the second allele to be inactivated when the first one (either maternal or paternal) has completed a successful rearrangement. This process is known as allelic exclusion.

The process of allelic exclusion

The fine mechanism that dictates the arrest or the start of rearrangement at the IgG or T cell receptor loci remains to be clarified but two important facts have been established: the first is the order in which rearrangement occurs, the second is that the arrest of further rearrangement requires functional polypeptide chains to be expressed.

In the Ig loci the order of events is as follows: both H chain alleles begin to rearrange and if a successfull DJ and VDJ rearrangement takes place on one of them this suppresses further rearrangement at the other allele and the cell begins to rearrange the k locus. If however the rearrangement of the first H chain allele is non productive, rearrangement continues on the second one and, if this is successful, the cell moves to k chain rearrangment. If the rearrangement of the second H chain allele is also non-successful the cell dies.

k chain rearrangement proceeds in a similar manner and rearrangement of the maternal and paternal alleles at the l locus only occurs if rearrangement of both alleles at the k locus has been non-productive. Thus the order is H, k and l (see Fig Allelic exclusion).

The statistics of this process are meaningful. With any rearrangement there are 2 chances out of 3 that the resulting sequence will be out of frame. However, the fact that both paternal and maternal alleles are utilized means that the cell has two attempts at each locus and, for a B cell that has successfully completed a H chain rearrangements, there are two L chain loci to be attempted. In spite of this the price to be paid to imprecisional joining is high and results in the majority of B cell progenitors that enter the B cell developmental programme dying as a result of their inability to assemble a functional antigen receptor. This is balanced, of course, by the extensive sequence diversity which imprecisional joining produces. Thus the number of cells that exit the system is considerably less that the number of cells that enters it but nearly all cells that assemble a functional receptor express specificities that are not merely encoded by their corresponding germ-line segments.

The development of B cells

The events leading to the formation of a functional BCR via gene recombination occur during B cell development and mark specific stages of differentiation of this lineage. Several other markers (such as the antigen recognized by the monoclonal antibody B220) accompany the development of these cells and have been useful for reconstructing B cell ontogeny. The B cell lineage starts with a cell (the pro-B cell) in which VDJ recombinase becomes active and a DJ rearrangement first occurs (Fig 2) .

In a subsequent stage, a complete VDJ rearrangement occurs and the cell expresses the H chain on the cell surface in association with an invariant (so-called surrogate) L chain. This stage corresponds to the pre-B cell and is followed by recombination of the L chain loci yielding a mature B cell in which the product of L chain rearrangement replaces the invariant L chain. This is the stage of the mature, vergin B cell. Upon antigenic stimulation and clonal expansion, a fraction of the B cell progeny develops in memory B cells and another fraction develops in plasmacells which secreted the antibody in a soluble form (see Fig Development and differentiation of B cells).

Expression of antibody genes

Each V gene segment contains a minimal promoter but these promoters bind RNA polymerase II very weakly and are fucntionally silent. Following VDJ rearrangement, however, these promoters become transcriptionally active because the rearrangement brings two enanchers close to the promoter of the V segments. The first of these enhancers is located in the J intron (ie the intron that separates the last J segment from the first Cm exon). The second enhancer is located in the intron that follows the last Ca exon. Similar enhancers are found in the k and l loci. In germ line DNA the V segment promoters are hundreds of thousands (or millions) nucleotides away from the enhancers.

After rearrangement, the V segment promoter comes within ~ 2 kb from the J enhancer and this results in transcription. There are strong similarities between the promoter structure of VH and Vk segments: both contain a TATA box and, upstream, a octanucleotide motif that interacts with B cell-specific DNA-binding proteins. The Ig enhancer sequences contain some of the B cell-specific sequences present in the promoter of the V segments (such as the octamer motif) as well as additional motifs. These include 'general' as well as B cell-specific sequences, one of which binds NF-kB, an important transcription factor that regulates transcription of the IL-2 and IFN-b genes as well.

The switch from membrane-bound to soluble antibody

A fraction of antigen-selected B cells differentiate in plasma cells, ie the antibody-producing cells. Plasma cells express a low number of membrane-bound antibody and, unlike the progenitor B cell, express large quantities of soluble antibody. This switch between membrane-bound and soluble is developmentally regulated and occurs by RNA processing.

The primary transcript contains a short hydrophilic sequence at the 3' end of the Cm4 exon. This is followed by a polyadenylation site, an intron, the M1 exon, another intron and the M2 exon followed by a second polyadenylation sequence. In cells expressing membrane-bound Ab, the second polyadenylation site is used and this leads to a splicing pattern of the primary transcript that produces the removal of the hydrophilic sequence at the 3' of the Cm4 exon. In cells that express soluble Ab, the first polyadenylation site is used and this leads to loss of the M1 and M2 exons that encode the transmembrane sequence.

A mechanism similar to the one discussed above (alternative use of polyadenylation sites) is responsible for simulteanous expression of m and d chains in mature B cells. However, in antigen-selected cells a different event takes place that leads to isotype switching (most often m-d to g). This is an entirely different process that involves further DNA recombination and the loss of the DNA separating the VDJ sequence and the constant segment resulting from the switch.

The recombinase substrate sites for this process (so called switch sites) consist of rather large DNA sequences composed of numerous and short repeat sequences and are located ~ 2 kb upstream from each C gene. As discussed in the lecture on somatic hypermutation of antibody genes, it is now known that class-switch recombination (isotype switch) is due to the activity of AID. It is also known that various cytokines secreted by TH cells (such as IL4) affect isotype switching.

T cell development

Progenitor T cells enter the thymus in the outer cortex. They express no TCR at this stage or RAG1 and RAG2 and they lack T cell-specific co-receptors such as CD4 ad CD8. Thus, throughout the early stages of development the T cell precursor is referred as a 'double negative' cell (CD4-/CD8-). These cells though express other markers that allow their identification (for example Thy-1, c-kit (the receptor for stem cell growth factor), CD44 (an adhesion receptor) and CD25 (the a chain of the IL-2 receptor).

Loss of expression of c-kit and CD44 precedes TCR rearrangement. g/d rearrangement is frequent in the early stages of T cell development and leads to the TCR g/d (CD3+) cells which predominates early in the early stages but accounts for only 5% (or less) of T cells in adults.

Other T cell progenitors rearrange the b/a loci (in this order) and cells with a successfull b chain rearrangement express the b chain on the cell surface in association with an invariant (pre T a chain), switch on expression of CD4 and CD8 and undergo rearrangement of the a chain. (double positive). Cells that fail to produce a functional rearrangement of the b or a chain die of apoptosis. Cells that succeed undergo selection for MHC restriction and against self reactivity (see below).

Evidence for MHC restriction

This was produced by Rolf Zinkernagel and Peter Doherty with a now classic experiments in which heterozygous (H-2 a/b) (A x B) F1 mice were thymectomized and lethally irradiated.

These mice were reconstituted with (A x B) F1 (H-2a/b) bone marrow and a thymus graft from B strain (H-2 b) mice. The mice were subsequently infected with the lymphocytic choriomengitis (LCM) virus.

Finally they were used as source of cytotoxic T cells which were tested on cells in culture for their ability to kill A strain (H-2 a) and B strain (H-2b) cells infected by the virus.

The result of the experiment demonstrated that infected B cells but not infected A cells were lysed. This indicated that selection had occurred in the reconstituted mouse for T cells that were able to interact with B strain (H-2 b) MHC and that these cells could kill virus-infected cells presenting antigens in association with H-2 b MHC but not with H-2 a MHC.

Effect of MHC on T cell development


Initial evidence for a direct role of self MHC proteins in positive selection of T cells came independently from antibody experiments (in which anti-class I and anti-class II antibodies in organ cultures of fetal thymus blocked maturation of CD8+ or CD4+ cells).

Subsequent and definitive evidence for the essential role of MHC in T cell development came from gene targeting experiments in mouse in which class I (-/-) mice were shown to lack CD8+ cells and class II (-/-) mice were shown to lack CD4+ cells (see Table Effect of MHC on T cell development). Hence the role of self-MHC in positive selection of T-cell during their development is now established beyond reasonable doubt.

Thymus histology

Positive selection for self-MHC restricted T cells occurs in the cortical region of the thymus, involves T cells at the stage of 'double-positives' (CD4+/CD8+) interacting with cortical epithelial cells.

Binding to MHC through their TCR confers to T cells a survival signal and, as a result, cells that fail to bind self-MHC die by apoptosis.

Cells that are positively selected for MHC binding migrate to the medulla of the thymus where they undergo a process of negative selection for self -reacting clones (see below).

Negative selection of self reacting T celln


Strong evidence for negative selection of self-reacting, MHC restricted cells was provided by an elegant experiment carried out by H von Boehmer and his colleagues.

The experiment involved analysis of T cells bearing a TCR a/b transgene specific for an antigen (the H-Y antigen) which is encoded by a gene on the Y chromosome (the H-Y antigen therefore is a self antigen in male but not in female transgenics).

The results of the experiment was that cells bearing the H-Y specific TCR were present in female but not in male animals. Negative selection for self-reacting T cells occurs in the medulla of the thymus, involves T cells at the stage of 'double-positives' (CD4+/CD8+) interacting with macrophage/dendritic cells. Cells selected against die by apoptosis.

  • Allelic exclusion
  • B cell development (stages and markers)
  • Ig enahncers
  • Isotype switch
  • Switch from membrane to soluble antibodies
  • T cell development (stages and markers)
  • Positive selection for MHC-restricted T cells
  • Negative selection for self-reactive T cells

[1] Zinkernagel RM, Doherty PC. H-2 compatibility requirement for T-cell-mediated lysis of target cells infected with lymphocytic choriomeningitis virus. Different cytotoxic T-cell specificities are associated with structures coded for in H-2K or H-2D. J Exp Med 141:1427 (1975)

[2] Zinkernagel RM. Cellular immune recognition and the biological role of major transplantation antigens. Nobel Lecture, 8th December 1996

[3] Doherty PC. Cell mediated immunity in virus infections. Nobel Lecture (1996)

[4] von Boehmer H, Kisielow P. Self-nonself discrimination by T cells. Science. 248: 1369 (1990)