Table 3.
Unique tumor antigens identified through breast cancer genome sequencing and predicted binding to HLA-A2.
| Gene Symbol | Mutation Location 1 | Mutation Type | Amino Acid Sequence 2 | Predicted Affinity 3 | |
|---|---|---|---|---|---|
|
| |||||
| mutated | wild type | ||||
| DDX11 | 12; 31122692 | SNV (T > G) | QEDFMAELYRGLEAGKIGIFE | 15 | 25 |
| PTCH2 | 1; 45068225 | SNV (C > G) | CHGFSHKFMHSQEELLLGGMA | 26 | 21 |
| PARVA | 11; 12496610 | insertion | SFAFELMQDGMEGLEKPKPRPE | 32 | 18192 |
| JAK2 | 9; 5040714 | SNV (T > C) | QWRHDFVHGWTKVPVTHETQE | 55 | 31 |
| DYNC2H1 | 11; 102687902 | SNV (G > A) | EQISKKDNTHQAHALFSLAWF | 60 | 19004 |
| CMV pp65 | N/A | N/A | NLVPMVATV 4 | N/A | 60 |
| PPPDE1 | 1; 242935580 | SNV (A > T) | LQSCLPKEWLSPAALQSSVSQ | 65 | 102 |
| SHE | 1; 152723308 | deletion | AVFDSIPEVVHYYSLSKGQNT• 5,6 | 66 | 18757 |
| SLC44A1 | 9; 107137789 | SNV (G > A) | LKTLSDVQKFTEINGSALCSY | 94 | 82 |
| JAK2 | 9; 5040714 | SNV (T > C) | QWRHDFVHGWTKVPVTHETQE | 111 | 291 |
| NALCN | 13; 100688137 | SNV (A > T) | VIGTTLHVYPELYHSQFTYFQ | 127 | 1293 |
| GP100280 | N/A | N/A | YLEPGPVTA 4 | N/A | 135 |
| GUK1 | 1; 226395989 | SNV (C > A) | MAGSQKEEIMQPQQGVPFQES | 162 | 331 |
| DHDDS | 1; 26646672 | SNV (G > A) | FLNVCFAYTSHHEISNAVREM | 186 | 378 |
| MAP3K8 | 10; 30789749 | SNV (C > T) | DLGALAGYFNLVRGLPTLEYG | 216 | 961 |
| DHDDS | 1; 26646672 | SNV (G > A) | FLNVCFAYTSHHEISNAVREM | 414 | 889 |
| FAM107B | 10; 14603968 | SNV (C > T) | ELQKVMEKRKQDQVIKQKEEE | 429 | 949 |
Mutation location is from Ensembl build 54. The first number is the chromosome; the second indicates the first mutated nucleotide;
The indicated peptide sequences (21-mers) were screened for candidate epitopes of 8–11 amino acids in length. The minimal epitope with the highest predicted affinity is underlined. Amino acids that differ from the wildtype sequence are indicated in red;
Predicted affinity (nM) based on the NetMHC 3.2 prediction algorithm [59]. The NetMHC 3.2 server predicts binding of peptides to a number of different HLA alleles using artificial neural networks (ANNs) and weight matrices [60-62]. Affinity scores of <50 nM indicate strong binding, whereas scores between 50 and 500 nM indicate weak binding. Similar data were obtained using a second prediction algorithm [63];
Commonly used immunodominant peptides derived from cytomegalovirus (CMV pp65) [64] and melanoma (gp100280) [65] are highlighted in yellow;
Please note that one limitation of next-generation sequencing technologies is that it can be very difficult to identify and validate small structural variants such as insertions or deletions. Robust computer algorithms have been established for the identification of these structural variants or indels [66]. Because indels are frequently frame shift mutations, they significantly alter the amino acid sequence, and may be more likely to be recognized by the immune system;
• indicates stop codon.