, , , ,

Question details: In Basic Immunology, 5th Edition, it was mentioned that “each TCR recognizes as few as one to three residues of the MHC-associated peptide”. How is that possible while retaining specificity to the antigen? It seems to me that any single amino acid could be common to more than 1 peptide.

Short answer: Structural constraints on how many amino acid residues TCRs can bind on short peptide sequences include

  • Several of the peptide’s amino acid residues are already engaged in binding firmly to an MHC (Major histocompatibility complex) molecule creating the pMHC (peptide bound to MHC molecule).
  • At the same time the TCR also needs to bind some amino acid residues of the MHC molecule itself.

Different parts of the TCR bind peptide and MHC. The unique process of somatic nucleotide insertions and deletions makes the TCR’s Complementarity-determining region (CDR)3 region hypervariable and thus quite flexible and effective in binding a few conserved residues on a MHC-bound peptide. This feature also makes TCRs cross-reactive, i.e., capable of binding >1 unique pMHC.

Longer answer on theoretical basis for why it should be so and some data on how TCRs actually bind peptides

The adaptive immune system is predicated on the notion of anticipatory defense. CD4 and CD8 T cells express the alphabeta TCR, composed of two protein chains, alpha and beta. Each undergoes somatic gene rearrangement, specifically V(D)J recombination, i.e., nucleotide insertions and deletions at the V(D) J junctions in the Complementarity-determining region (CDR)3 regions of each chain of the TCR. This means each human body iterates from scratch its adaptive defense armamentarium, something that a 2015 immune parameter analysis of 210 healthy monozygotic twin pairs only confirms (1).

At first glance, this seems reasonable enough. After all, TCRs on CD4 and CD8 T cells have evolved to recognize and bind not whole protein molecules, ‘antigens’, but rather presented by MHC molecules, tiny pieces thereof, ‘peptides’, 12- to 20-mer (12-20 amino acids in length) in the case of CD4s, and 8- to 14-mer in the case of CD8s.

On second glance, this imposes a heavy burden on such a recognition system. Zooming in from larger structures to much smaller ones means the number of peptides presented to T cells by MHC class I and II molecules should be very large. Why should? Zooming in to such magnification simply exponentially increases the number of targets, i.e., peptides. If MHC molecules didn’t present as diverse of a peptide pool as possible from within a cell, missed pathogen-derived peptides become missed opportunities weakening the anticipatory potential of the adaptive immune system.

Given that TCRs bind short peptide sequences presented by MHC molecules, how could recognition be based on other than a handful of amino acid residues? Much of the rest of the peptide binds the MHC molecule. Peptide binding to MHC is itself a critical filtering event of great consequence in adaptive immunity. Antigen processing generates many peptides of varying lengths during protein digestion but only a handful succeed in making it past several bottlenecks to successfully and tightly bind MHC in such a way that they get presented on the cell surface, a feature called Immunodominance.

Why Most T cell Receptors (TCRs) Need To Be Inherently Cross-Reactive, Capable Of Binding >1 pMHC Complex

T cell receptors (TCRs) bind bits of both MHC and peptide. TCRs are hypothesized to have been evolutionarily selected to recognize and bind MHC. Thus, considering just the TCR engagement with the peptide cuts off too small a slice of the pie since outcome, TCR-mediated biochemical activation of T cells, depends on other critical factors,

  • How the peptide is bound within the MHC, which in turn determines which of its amino acid residues are available to make contact with the TCR and how optimally that could happen.
  • How the peptide is bound within the MHC also determines how optimally key anchor residues of the MHC molecule itself are available to bind the TCR.
  • How optimally can T cell co-receptors CD3 (immunology), CD4 (for CD4+ T cells)/CD8 (for CD8+ T cells) bind MHC at the same time.

Corollary of the fact that TCRs recognize and bind peptides presented by MHC molecules is the need for correspondingly large repertoire of specific T cells, one specific enough to recognize and bind each peptide processed and presented by each MHC molecule. However, if each TCR bound only one specific peptide, a logical extension of the Clonal selection theory (2, 3), each individual would theoretically need >10^15 T cells. Why? Because the 20 amino acid alphabet predicates the possible number of peptides that could bind to MHC molecules to be in the range of 10^15 (4). But 10^15 monospecific T cells necessary for optimal anticipatory defense entails a body weight of >500kgs (4), clearly and simply a physical impossibility, something that Don Mason already demonstrated in 1998 with his absurd mouse cartoon (see below from 5).

Given the constraint that there can only be far fewer T cells with unique TCRs compared to the potential diversity of possible unique pMHC complexes, no surprise TCRs tend to be cross-reactive, i.e., capable of binding >1 pMHC complex. An elegant 2014 experimental study found each of 5 different mouse and human TCRs even capable of binding >100 different peptides presented by one MHC molecule (6). Thus far crystal structures of ~120+ TCR- pMHC complexes have been published. One of the most striking observations of such crystal structures is the tremendous flexibility in how TCRs bind to pMHC (7). Rather than binding unique peptides, a given TCR can bind many but TCR binding’s supposed to be very specific. Cross-reactivity is antithetical to specificity. Can the two be reconciled? Yes, by the fact that the common thread linking cross-reactive peptides that bind a given TCR is the presence of conserved motifs, i.e., one, two, three or more conserved residues at specific TCR-binding positions (see quote, emphasis mine, and figure below from 6).

‘TCR cross-reactivity is not achieved by each receptor recognizing a large number of unrelated peptide epitopes but rather through greater tolerance for substitutions to peptide residues outside of the TCR interface, differences in residues that contact the MHC, and relatively conservative changes to the residues that contact the TCR CDR loops. The segregation of TCR recognition and MHC binding allows for TCRs to simultaneously accommodate needs for specificity and cross-reactivity.’

Thus, this example shows that even though peptides like 2A and MCC are very different in their sequences, both successfully bind the same TCR 2B4 because the process of somatic nucleotide insertions and deletions makes the region that primarily makes contact with a peptide, the TCR’s CDR3 region, hypervariable. This endows TCRs with the capacity to be quite flexible in accommodating different peptide sequences and also be able to bind firmly by contacting only a handful of conserved residues on any given MHC-bound peptide.

Human epidemiological studies reveal the implications of T cell recognition being the way it is. A relatively obscure set of data epitomize not only the extent of T cell cross-reactivity but also suggest that such functionality enables a vast, connected immunoprotective landscape against disparate entities ranging from bacteria to virus to cancer.

~85% of malignant melanocytes express an antigen called HERV-K-MEL (8, 9, 10), product of a pseudo-gene incorporated in the HERV-K env gene. HERV (Human Endogenous retrovirus) in turn are endogenous retroviruses incorporated into the human genome over millions of years. Acquired between 3 and 6 million years back, HERV-K are the latest family (11), making them the only HERVs still capable of replicating in the human population within the last few million years. HERV-K appears to be involved in several stages of melanoma formation (12, 13, 14).

Spontaneous melanoma regressions have occasionally been reported in the literature, suggesting effective anti-melanoma immune responses occur in nature. But what are the coordinates of such immunity? Taking a leaf out of William Coley and his Coley’s toxins, the European Organization for Research and Treatment of Cancer (EORTC) established the Febrile Infections and Melanoma (FEBIM) working group, tasked to explore how prior infectious diseases and vaccines influenced melanoma risk.

Their studies thus far suggest the Tuberculosis (TB) BCG vaccine, the Vaccinia vaccine against small pox and the 17D Yellow fever vaccine provide some degree of protection against melanoma (15, 16, 17).

What could possibly link such disparate characters as BCG, Vaccinia, Yellow fever and Melanoma and what connects this story to TCRs and peptides? Turns out each of these really disparate agents, a bacterium and two unrelated viruses, express peptides with high sequence homology to the melanoma HERV-K-MEL peptide (see below from 18).

BCG, vaccinia and yellow fever vaccines are of course expected to induce specific immune responses against themselves. However, given they express proteins with high sequence homology, vaccine-specific cytotoxic CD8+ and helper CD4+ T cells would also include those cross-reactive to melanoma HERV-K-MEL peptide. This could prevent melanoma development in those vaccinees who retain robust memory immune responses against this cross-reactive peptide.

Such phenomena may underlie the observation that certain live vaccines like BCG and measles can protect against unrelated pathogens and even reduce rates of all-cause mortality (19, 20). And as more microbiota-immunity interactions get mined, such examples that at present seem unanticipated will become more commonplace.


1. Brodin, Petter, et al. “Variation in the human immune system is largely driven by non-heritable influences.” Cell 160.1 (2015): 37-47.

2. Jerne, Niels K. “The natural-selection theory of antibody formation.” Proceedings of the National Academy of Sciences 41.11 (1955): 849-857. http://www.pnas.org/content/41/1…

3. Jerne, Niels Kaj. “The somatic generation of immune recognition.” European journal of immunology 1.1 (1971): 1-9.

4. Sewell, Andrew K. “Why must T cells be cross-reactive?.” Nature Reviews Immunology 12.9 (2012): 669-677. http://www.tcells.org/scientific…

5. Mason, Don. “A very high level of crossreactivity is an essential feature of the T-cell receptor.” Immunology today 19.9 (1998): 395-404.

6. Birnbaum, Michael E., et al. “Deconstructing the peptide-MHC specificity of T cell recognition.” Cell 157.5 (2014): 1073-1087. http://www.cell.com/cell/pdf/S00…

7. Rudolph, Markus G., Robyn L. Stanfield, and Ian A. Wilson. “How TCRs bind MHCs, peptides, and coreceptors.” Annu. Rev. Immunol. 24 (2006): 419-466. http://nfs.unipv.it/nfs/minf/dis…

8. Kölmel, K. F., O. Gefeller, and B. Haferkamp. “Febrile infections and malignant melanoma: results of a case-control study.” Melanoma research 2.3 (1992): 207-212

9. Schiavetti, Francesca, et al. “A human endogenous retroviral sequence encoding an antigen recognized on melanoma by cytolytic T lymphocytes.” Cancer research 62.19 (2002): 5510-5516. https://www.researchgate.net/pro…

10. Grange, John M., et al. “Can prior vaccinations against certain infections confer protection against developing melanoma?.” Medical Journal of Australia 191.9 (2009): 478. http://citeseerx.ist.psu.edu/vie…

11. Sverdlov, Eugene D. “Retroviruses and primate evolution.” Bioessays 22.2 (2000): 161-171.

12. Muster, Thomas, et al. “An endogenous retrovirus derived from human melanoma cells.” Cancer research 63.24 (2003): 8735-8741. https://www.researchgate.net/pro…

13. Serafino, A., et al. “The activation of human endogenous retrovirus K (HERV-K) is implicated in melanoma cell malignant transformation.” Experimental cell research 315.5 (2009): 849-862. https://www.researchgate.net/pro…

14. Singh, Sarita, et al. “The role of human endogenous retroviruses in melanoma.” British Journal of Dermatology 161.6 (2009): 1225-1231.

15. Grange, John M., Bernd Krone, and John L. Stanford. “Immunotherapy for malignant melanoma–tracing Ariadne’s thread through the labyrinth.” European Journal of Cancer 45.13 (2009): 2266-2273.

16. Krone, Bernd, et al. “Protection against melanoma by vaccination with Bacille Calmette-Guerin (BCG) and/or vaccinia: an epidemiology-based hypothesis on the nature of a melanoma risk factor and its immunological control.” European Journal of Cancer 41.1 (2005): 104-117.

17. Mastrangelo, G., et al. “Does yellow fever 17D vaccine protect against melanoma?.” Vaccine 27.4 (2009): 588-591.

18. Cegolon, Luca, et al. “Human endogenous retroviruses and cancer prevention: evidence and prospects.” BMC cancer 13.1 (2013): 1. http://bmccancer.biomedcentral.c…

19. Goodridge, Helen S., et al. “Harnessing the beneficial heterologous effects of vaccination.” Nature Reviews Immunology (2016).

20. Muraille, Eric. “The Unspecific Side of Acquired Immunity Against Infectious Disease: Causes and Consequences.” Frontiers in microbiology 6 (2015). https://www.ncbi.nlm.nih.gov/pmc…