Higher predicted replacement mutability of T cell receptor CDRs than T cell receptor FRs

Zheng et al. (1994) isolated individual germinal centers T cells and sequenced their immune receptor genes. Surprisingly, they found a number of mutations in these genes, raising the intriguing possibility that T cell receptors also undergo somatic hypermutation. The mutations that were found in these T cell receptors seemed to bear the mark of somatic hypermutation. That is, only germinal center T cells, already recruited in the immune response were affected, mutations were mostly found in the CDRs and they were concentrated in the hot spots described for B cell receptor hypermutation. This finding challenges one of the basic paradigms of self-nonself discrimination. It is generally believed that during their development in the thymus, T cells that bind any antigen with sufficiently high affinity are destined to die. On the contrary, when T cells bind antigens with high affinity outside of the thymus, they start replicating and performing their effector functions. T cell receptor mutation in the germinal centers may then turn a benign T cell into an auto-reactive one. Evidently, this finding raised a lot of controversy.

Kepler and Bartl (1998), looking for the presence of mutable motifs in T cell receptors CDRs concluded that some of the T-cell receptor chains resemble immunoglobulins in their mutability pattern, but a consistent trend could not be identified. Human TCR $V_\alpha$ did not show the immunoglobulin mutability pattern, whereas murine $V_\alpha$ did, although only when serine codons were excluded. TCR $V_\beta$ on the other hand, both human and murine, resembled the immunoglobulins.

 

Given that my results on immunoglobulin sequences are unambiguous, I applied the same analysis to human TCR receptor chains. If the FR/CDR mutability bias of immunoglobulin sequences were shared by T cells receptor sequences, we would have indirect evidence that T cell receptor genes also undergo somatic hypermutation, as proposed by Zheng et al. (1994). I used mainly human TCR $\alpha$ (n = 41) and $\beta$ (n = 54) chains. The number of $\gamma$ and $\delta$ sequences is considerably smaller (8 and 3, respectively). I used the IMGT alignments and CDR assignments.

Fig. shows the scatter plot of the predicted FR-CDR nucleotide mutability for the set of $TCR_\alpha$ (in blue) and $TCR_\beta$ (in red) sequences. For most $TCR_\beta$ sequences, the predicted CDR replacement mutability is higher than the FR mutability. For $TCR_\alpha$ this difference is not so clear, many sequences having, in fact lower FR than CDR mutability. Testing for evidence of mutability optimization with respect to somatic hypermutation, I find such evidence for $TCR_\beta$, but not for $TCR_\alpha$. The rank of the set average of FR mutability among 10⁴ variant sets with identical translation, but random codon usage, is 0.0103 in the case of $TCR_\beta$, and 0.3524 in the case of $TCR_\alpha$. CDR mutability ranks 0.9581 in the case of $TCR_\beta$, and 0.3941 in the case of $TCR_\alpha$.

These results allow a number of interpretations. They could indicate that somatic hypermutation currently affects T cell receptors, as suggested by Zheng et al. (1994). They could also indicate that the hypermutation mechanism does not operate in these genes currently, but it affected the ancestral receptor from which immunoglobulins and T cell receptors diverged. This feature would then have been preserved in immunoglobulin chains, but lost in T cells, whose sequence drifted away from the one that supported somatic mutation. The mutability pattern of immunoglobulins in other species also argues for somatic mutation being discovered early in phylogeny. This has also been suggested by M.Flajnik (personal communication, 1998). Finally, this result could indicate that the somatic hypermutation mechanism adapted to the codon bias of immunoglobulin genes. This codon bias could be shared by some of the T cell receptor genes, by virtue of the genealogical relationship between T cell and B cell receptors.