Open access Review J Immunother Cancer: first published as 10.1136/jitc-2020-001111 on 27 August 2020. Downloaded from Antigen processing and presentation in cancer immunotherapy Maxwell Y Lee, Jun W Jeon, Cem Sievers, Clint T Allen To cite: Lee MY, Jeon JW, ABSTRACT is critical for antitumor immunity and that Sievers C, et al. Antigen Background Knowledge about and identification of T cancers can evade such immunity by immu- processing and presentation in cell tumor antigens may inform the development of T cell nodominance, display of immune check- cancer immunotherapy. Journal receptor-engineered adoptive cell transfer or personalized for ImmunoTherapy of Cancer points, or immunoediting for loss of specific cancer vaccine immunotherapy. Here, we review antigen 4 2020;8:e001111. doi:10.1136/ tumor antigens. Some cancers are composed processing and presentation and discuss limitations in jitc-2020-001111 of subclonal tumor cell populations that tumor antigen prediction approaches. harbor defects in antigen processing and Methods Original articles covering antigen processing Accepted 19 July 2020 and presentation, epitope discovery, and in silico T cell presentation, suggesting that these cancers epitope prediction were reviewed. may not be curable despite maximal activa- 5 Results Natural processing and presentation of antigens tion of T cell immunity. For those cancers is a complex process that involves proteasomal proteolysis without defects in antigen processing and of parental proteins, transportation of digested peptides presentation, knowledge of immunogenic T into the endoplasmic reticulum, loading of peptides cell antigens may assist in the engineering of onto major histocompatibility complex (MHC) class I immunotherapeutics designed to control and molecules, and shuttling of peptide:MHC complexes to eradicate cancer.6 the cell surface. A number of T cell tumor antigens have For a peptide to serve as a natural T cell been experimentally validated in patients with cancer. antigen, the parental protein must first Assessment of predicted MHC class I binding and total score for these validated T cell antigens demonstrated a be processed so that its peptides can be wide range of values, with nearly one-third of validated presented on an MHC I molecule. Natural processing and presentation of an antigen is antigens carrying an IC50 of greater than 500 nM. Conclusions Antigen processing and presentation is a a complex and multifactorial process, which complex, multistep process. In silico epitope prediction is still subject to active research. The avail- techniques can be a useful tool, but comprehensive able peptide repertoire is constrained by how experimental testing and validation on a patient- by- patient proteins are cleaved, trimmed, loaded onto http://jitc.bmj.com/ basis may be required to reliably identify T cell tumor MHC, and translocated to the cell surface. antigens. Specificity within this process is likely an explanation for why the vast majority of peptides predicted to be high- affinity MHC BACKGROUND binders are unable to elicit T cell responses.7 Immunotherapy has revolutionized cancer Validated T cell antigens are often tumor- treatment. At present, various therapeutics specific as they are derived from expressed on September 28, 2021 by guest. Protected copyright. for the enhancement or replacement of tumor- specific mutated genes. An alternative T cell antitumor immunity have Food and theory for why many peptide candidates do Drug Administration approval for a variety of not elicit T cell responses and, thus, are not cancers.1 2 The core principles driving current antigens, is that these mutated peptides are immunotherapies began to take shape in the very similar to their non-mutated counter- late 20th century with the elucidation of the parts, leading to central tolerance and elim- mechanism by which T cells recognize anti- ination of potentially reactive T cells. © Author(s) (or their gens presented by the major histocompati- Despite challenges, there has been success employer(s)) 2020. Re- use bility complexes (MHC).3 T cell antigens are in using T cells to target specific anti- permitted under CC BY-NC. No commercial re- use. See rights specific peptide sequences that are recog- gen–MHC combinations. The cancer testis and permissions. Published by nized by CD8 or CD4 T cells when presented antigen NY- ESO-1 is one of the best exam- BMJ. on MHC I or MHC II molecules, respec- ples.8 As a validated, highly immunogenic NIDCD, National Institutes of tively. Neoantigens are peptides derived from antigen, the use of the NY- ESO-1 T cell exper- Health, Bethesda, Maryland, USA tumor- specific mutations which have not been imental system has set the stage for an abun- Correspondence to previously recognized by the body’s immune dance of T cell and cancer biology research 9 Dr Clint T Allen; system. Experimental evidence supports over the past two decades. Furthermore, its clint. allen@ nih. gov that tumor antigen recognition by T cells use in clinical trials has led to advances in Lee MY, et al. J Immunother Cancer 2020;8:e001111. doi:10.1136/jitc-2020-001111 1 Open access J Immunother Cancer: first published as 10.1136/jitc-2020-001111 on 27 August 2020. Downloaded from adoptive cell transfer immunotherapy, including the first lumen by deubiquitinating and unfolding them. The successful treatment of a non- melanoma tumor using T immunoproteasomes are capped by an alternative 11S cell receptor (TCR)-engineered T cells.9 The successful complex that is also IFN-γ inducible.14 As a result of the application of NY- ESO-1- specific T cells suggests that different proteolytic subunits and capping structures, the epitope- specific, T cell- based immunotherapy may allow cleavage site specificity differs between constitutive and personalized immunotherapy. In this review, we focus on immunoproteasomes.15 the mechanisms behind natural processing and presenta- The characterization of proteasomal activity includes tion of T cell antigens and summarize current methods two basic approaches. In vitro study involves co- incuba- and limitations of antigen prediction and validation. tion of parental proteins with proteasomes. The resulting peptides are analyzed using mass spectrometry.16 Several proteolysis prediction techniques are trained on in BIOLOGY OF ANTIGEN PRESENTATION vitro datasets.17 In vivo proteolysis study involves the MHC molecule measurement of peptides that have been eluted from Class I MHC molecules are polypeptides composed of a an peptide:MHC complexes and subjected to mass spec- 18 polymorphic heavy chain that associates with a constant trometry. A limitation of this strategy is that only a small β2- microglobulin (β2m) subunit, whereas class II MHC proportion of cleavage sites can be identified in this molecules are composed of an α and β polypeptide manner as most cleaved peptides do not end up being chain. Class I MHC molecules are generally recognized presented on MHC complexes. Furthermore, the influ- by CD8 cytotoxic T cells, whereas class II MHC molecules ence of other cellular proteases cannot be deconvoluted are recognized by CD4 T cells. For the purposes of this from pure proteasome function. Several prediction tech- review, we will only discuss class I MHC molecules and niques are trained on in vivo datasets and are generally their interactions with CD8 T cells. The mechanisms superior at predicting in vivo proteolysis as compared 19 underlying epitope prediction and discovery for class II with prediction techniques trained on in vitro datasets. MHC molecules, while important, are beyond the scope of this article. The peptide binding groove of an MHC Peptide loading onto MHC I I molecule preferentially binds 8–11mer peptides. The The peptide loading complex (PLC) is a multi- subunit, top surface of the groove, where both the MHC protein endoplasmic reticulum (ER) membrane complex which and bound peptide are exposed, is the portion of the coordinates peptide translocation into the ER, editing, peptide:MHC complex detected by a TCR.10 and loading onto MHC I molecules. One of the core Humans possess three MHC I genes: human leucocyte components is the transporter associated with antigen antigen A (HLA- A), HLA- B, and HLA- C. These genes processing (TAP), which is made up of the TAP1 and encode highly polymorphic proteins, particularly in the TAP2 subunits. Each subunit contains a transmembrane peptide interacting region, allowing different HLA mole- domain and a nucleotide binding domain which interact cules to bind different sets of peptides.11 This presents to mediate a channel for peptide movement between the 12 http://jitc.bmj.com/ a challenge to the prediction of MHC binding peptides cytosol and ER lumen. TAP preferentially translocates as prediction algorithms must take into account HLA- peptides with 9–16 residues, although longer peptides up specific binding preferences. to 25–30 amino acids in length can also be translocated with lower efficiency.20 From proteins to peptides The specificity of TAP has been well studied. On the For a peptide to be presented on an MHC I molecule, C- terminal end of the peptide, TAP selects for the pres- it must be processed by the cellular antigen processing ence of an aromatic, hydrophobic, or positively charged on September 28, 2021 by guest. Protected copyright. machinery. This begins with proteolysis of parental terminal amino acid such as phenylalanine, tyrosine, argi- protein precursors. Exogenous proteins that are internal- nine, or leucine.21 On the N- terminal end, the first three ized by the cell through processes such as phagocytosis residues have significant effects on peptide binding. and endocytosis are processed in the late endosome and Aromatic and hydrophobic side chains are favored and primarily presented on MHC II. Endogenous proteins are an N- terminal arginine is optimal.22 Proteasomal speci- processed by the proteasome and are presented mainly ficity at the C-terminal residue is relatively non-specific on MHC I.12 and aligns with the restraints of both TAP and MHC Proteasomes are proteases that break down misfolded, I molecules.23 However, the N- terminal ends of TAP- damaged, aberrant, or ubiquitinylated proteins.