Cancer Immunotherapy: Golden Age for Computational Biologists
It’s a long belief that cancer cells can be killed by the immune system of patients themselves. As we know, in a flu, T cells in patients’ bodies can detect and kill infected cells. Cancer cells, although more malignant, are still cells. Why T cells in patients’ bodies don’t automatically find cancer cells and kill them? The truth is, at most times, T cells can’t recognize cancer cells. However, it doesn’t mean that we can’t teach T cells to find out cancer cells. Recently, scientists successfully modified T cells and made them more likely to bind with proteins on cancer cells. This method, called “immunotherapy”, has demonstrated dazzling therapeutic efficacy.
August 2017, FDA approved Kymriah, the first chimeric antigen receptor T cell (CAR-T) therapy, for the treatment of young patients with acute lymphoblastic leukemia (ALL).
March 2018, FDA approved Blincyto, the first-and-only approved bispecific CD19-directed CD3 T cell engager immunotherapy for Minimal Residual Disease (MRD).
April 2018, there are 216 clinical trials actively recruiting patients for T cell immunotherapy worldwide (data from http://clinicaltrials.gov).
Meanwhile, cancer immunotherapy steps into “big data” era thanks to the advances in high-throughput immune sequencing. Now we can analyze millions of T cell receptors in a single experiment, which provide a system-like perspective for scientists to understand antigen-specific interactions.
Click the picture to watch the video
This is a presentation given by Sisi Sarkizova and Michael Rooney. We suggest you watch the first 30min as a primer, for it covers most basic immunology knowledge you need to know.
Lymphocyte: The three major types of lymphocyte are T cells, B cells and natural killer(NK) cells
TCR (T cell receptor): a molecule found on the surface of T cells that is responsible for recognizing fragments of antigen as peptides bound to major histocompatibility complex (MHC) molecules. The binding between TCR and antigen peptides is of relatively low affinity and is degenerate: that is, many TCRs recognize the same antigen peptide and many antigen peptides are recognized by the same TCR
pMHC (peptide Major histocompatibility complex): a set of cell surface proteins in vertebrate, they bind to antigens and display them on the cell surface for recognition by T-cells.
Class I MHC molecules have β2 subunits: recognized by CD8 co-receptors.
Class II MHC molecules have β1 and β2 subunits: recognized by CD4 co-receptors.
HLA (human leukocyte antigen): pMHC in humans
Epitope: Each MHC molecule on the cell surface displays a molecular fraction of a protein, called an epitope.
Paratope: an antibody that binds to the epitope.
Complementarity-determining regions (CDRs): are part of the variable chains in immunoglobulins (antibodies) and T cell receptors, generated by B-cells and T-cells respectively, where these molecules bind to their specific antigen. A set of CDRs constitutes a paratope. As the most variable parts of the molecules, CDRs are crucial to the diversity of antigen specificities generated by lymphocytes.
Hypermutation: Somatic hypermutation is a cellular mechanism by which the immune system adapts to the new foreign elements that confront it
Neoantigens: protein sequences alteration from point mutations and gene fusions
V(D)J recombination: the unique mechanism of genetic recombination that occurs only in developing lymphocytes during the early stages of T and B cell maturation.
variable (V), joining (J), and in some cases, diversity (D) gene segments
With emerging data in immune sequencing, researchers have done various explorations.
3.1 organize groups of TCR sequences according to antigen specificities
e.g. Glanville et al. developed an algorithm called GLIPH (grouping of lymphocyte interactions by paratope hotspots);
3.2 explore determinants of epitope specificity
e.g. Dash et al. proposed a distance measure on the space of TCRs and built a distance-based classifier that can assign previously unobserved TCRs to characterized repertoires
3.3 predict affinity between antibody and antigen
e.g. Asti et al. used Maximum Entropy to predict antibody affinity for antigens from the space of Antibodies sequence.
3.4 determine properties cause differences in immunogenicity
e.g. Calis et al. developed a simple model to predict immunogenicity based on amino acids enrichment and positional importance.
[1] J. Glanville et al. Nature 547, 94–98 (2017)
[2] P. Dash et al. Nature 547, 89–93 (2017)
[3] L. Asti et al. PLoS Comput Biol 12(4): e1004870 (2016)
[4] J. J. A Calis et al. PLoS Comput Biol 9(10): e1003266 (2013)