Ideally, CD8+ T-cell responses against virally infected or malignant cells are defined at the level of the specific peptide and restricting MHC class I element, a determination not yet made in the dog. To advance the discovery of canine CTL epitopes, we sought to determine whether a putative classical MHC class Ia gene, Dog Leukocyte Antigen (DLA)-88, presents peptides from a viral pathogen, canine distemper virus (CDV). To investigate this possibility, DLA-88*508:01, an allele prevalent in Golden Retrievers, was expressed as a FLAG-tagged construct in canine histiocytic cells to allow affinity purification of peptide-DLA-88 complexes and subsequent elution of bound peptides. Pattern analysis of self peptide sequences, which were determined by liquid chromatography-tandem mass spectrometry (LC-MS/MS), permitted binding preferences to be inferred. DLA-88*508:01 binds peptides that are 9-to-12 amino acids in length, with a modest preference for 9- and 11-mers. Hydrophobic residues are favored at positions 2 and 3, as are K, R or F residues at the C-terminus. Testing motif-matched and -unmatched synthetic peptides via peptide-MHC surface stabilization assay using a DLA-88*508:01-transfected, TAP-deficient RMA-S line supported these conclusions. With CDV infection, 22 viral peptides ranging from 9-to-12 residues in length were identified in DLA-88*508:01 eluates by LC-MS/MS. Combined motif analysis and surface stabilization assay data suggested that 11 of these 22 peptides, derived from CDV hemagglutinin, large polymerase, matrix, nucleocapsid, and V proteins, were processed and presented, and thus, potential targets of anti-viral CTL in DLA-88*508:01-bearing dogs. The presentation of diverse self and viral peptides indicates that DLA-88 is a classical MHC class Ia gene.

No abstract

Development of effective immunotherapy for chemoresistant malignancies can be advanced by studies in spontaneous cancer models, such as the dog. A crucial first step, T‐cell epitope discovery, can be assisted by determination of binding motifs of common dog leukocyte antigen (DLA) class Ia allotypes. Boxers are popular, inbred dogs with increased risks of relevant target cancers and restricted MHC diversity. We sought to identify the motif of DLA‐88*034:01, a breed‐dominant allotype, to assist peptide prediction from tumor antigens. Mass spectrometry of eluted peptides showed a preference for nonamers with conserved amino acid preferences: basic at position (P)1; hydrophobic at P2; acidic at P4; histidine at P6; and phenylalanine at P9. This data should expedite finding epitopes restricted by this DLA‐88 allotype.

Cancer‐testis antigens (CTAs) are a category of self proteins aberrantly expressed in diverse malignancies, mostly solid tumours, due to epigenetic de‐repression. Normally expressed only in fetal or gametogenic tissues, CTAs are tantalizing immunotherapy targets, since autoimmunity risks appear minimal. Few prevalent CTAs have been identified in human hematologic cancers, and just two in their veterinary counterparts. We sought to discover new CTAs in canine hematologic cancers such as histiocytic sarcoma (HS) and lymphoma to foster immunotherapy development. To accomplish this, the ligandome binding the dog leukocyte antigen (DLA)‐88*508:01 class I allele overexpressed in an HS line was searched by mass spectrometry to identify possible CTA‐derived peptides, which could serve as CD8+ T‐cell epitopes. Twenty‐two peptides mapped to 5 human CTAs and 12 additional proteins with CTA characteristics. Expression of five promising candidates was then evaluated in tumour and normal tissue by quantitative and end‐point RT‐PCR. The ortholog of an established CTA, IGF2BP3, had unexpectedly high expression in peripheral blood mononuclear cells (PBMCs). Four other testis‐enhanced proteins were also assessed. AKR1E2, SPECC1 and TPX2 were expressed variably in HS and T‐cell lymphoma biopsies, but also at high levels in critical tissues, including kidney, brain and marrow, diminishing their utility. A more tissue‐restricted candidate, NT5C1B, was detected in T‐cell lymphomas, but also at low levels in some normal dog tissues. These results illustrate the feasibility of discovering canine CTAs by a reverse approach, proceeding from identification of MHC class I‐presented peptides to a comparative RNA expression survey of tumours and normal tissues.

Canine spontaneous cancers may offer greater fidelity than rodent models in advancing clinical immunotherapies. Boxers in particular are distinguished as study subjects by their popularity, and high incidence of human‐relevant cancers. Further, the MHC class I allele DLA‐88*034:01, with a known motif, dominates the breed, facilitating discovery of shared CTL responses against mutation‐origin neoepitopes by standard prediction methods. We experimentally confirmed the allomorph's binding motif by developing an MHC surface stabilization assay. The assay validated four DLA‐88*034:01‐presented peptides from canine distemper virus, ubiquitously administered in routine vaccines, for positive controls in future CTL studies. In turn, these viral peptides substantiated motif‐based prediction for DLA‐88*034:01. The study adds new tools for studying neoepitope‐specific CTL in Boxers to foster canine comparative oncology.