Pan-cancer expression profile of NK1R
We observed that NK1R is significantly up-regulated in 7 tumors: GBM, GBMLGG, LGG, LIHC, PAAD, ALL and LAML, while down-regulated in 23 tumors: UCEC, BRCA, CESC, LUAD, ESCA, STES, KIRP, KIPAN, COAD, COADREAD, PRAD, STAD, HNSC, KIRC, LUSC, WT, SKCM, BLCA, THCA, READ, UCS, ACC and KICH, which signifies the universal differential expression of NK1R in human tumors (Fig. 1).
Prognostic signature of NK1R among human tumors
In the field of NK1R associated with prognosis, 5 types of tumors containing TARGET-LAML, TCGA-STES, STAD, BLCA and LAML were significantly in poor prognosis when NK1R was highly expressed. Low expression of NK1R suggested preferentially a shorter OS in 11 tumor types: GBMLGG, LGG, BRCA, CESC, SARC, KIPAN, HNSC, KIRC, PAAD, ACC and KICH (Fig. 2A). Furthermore, our Kaplan-Meier curve revelaed that NK1R expression level is able to be an index to evaluate the prognosis (Fig. 2B). To evaluate NK1R prognostic model, we found that the best prediction performance for NK1R was in GBMLGG and LGG by receiver operating characteristic (ROC) curve.
The AUC value of 1 year, 3 years and 5 years are all greater than 0.7 (Fig. 2C). We subsequently analyzed the correlation between NK1R and clinical signatures, including the TNM stage, grade, gender and age, and we found that the expression of NK1R was significantly related to different clinical signatures in terms of different cancer types (Fig. 3A-3G).
Correlation between NK1R and tumor immune infiltration
For stromal cell scores, immune infiltration in 34 cancer types was significantly correlated with NK1R, in which positive correlation contained TCGA-BRCA, CESC, LUAD, ESCA, STES, KIRP, KIPAN, COAD, COADREAD, PRAD, STAD, HNSC, KIRC, LUSC, LIHC, SKCM/SKCM-P/SKCM-M, BLCA, THCA, MESO, READ, UVM, PAAD, TGCT, LAML, PCPG, ACC, DLBC, CHOL, TARGET-WT and NB, and negative correlation contained TCGA-GBMLGG and LGG (Fig. 6).
For immune scores, there were 31 types of human tumors significantly correlated with NK1R, containing 27 positively correlated types which are TCGA-BRCA, LUAD, ESCA, STES, KIRP, KIPAN, COAD, COADREAD, PRAD, STAD, HNSC, LIHC, SKCM/SKCM-P/SCKM-M, BLCA, THCA, MESO, READ, OV, UVM, PAAD, PCPG, ACC, DLBC, CHOL and TARGET-NB, as well as 4 positively correlated types which are TCGA-GBMLGG, LGG, TGCT and TARGET-LAML (Fig. 7).
Then we calculated ESTIMATE score, which was based on the evaluation of both stromal and immune cells. We observed that NK1R expression was significantly correlated with immune infiltration in 34 cancer types. 31 types were positively correlated, such as TCGA-BRCA, CESC, LUAD, ESCA, STES, KIRP, KIPAN, COAD, COADREAD, PRAD, STAD, HNSC, KIRC, LUSC, LIHC, SKCM/SKCM-P/SKCM-M, BLCA, THCA, MESO, READ, OV, UVM, PAAD, LAML, PCPG, ACC, DLBC, CHOL, TARGET-NB, and 3 types were negatively correlated, such as TCGA-GBMLGG, LGG and TARGET-LAML (Fig. 8).
Through TIMER analysis, we found that all of the immune infiltration cells we studied were associated with elevated NK1R expression in PRAD, SKCM-M, THCA, SKCM, BRCA, COADREAD, LUAD, COAD, STES, LIHC, STAD, PAAD and OV, while parts of the immune infiltration cells were related to raised NK1R expression in HNSC, LUSC, CESC, ESCA, KIRP, BLCA, MESO, CHOL, PCPG, READ, KICH, KIPAN, UCS, KIRC and ACC. The macrophage infiltration in DLBC was negatively correlated with NK1R, but CD8 + T cell, neutrophil and DC infiltration was positively correlated. In LGG, TGCT, GBMLGG and UCEC, all of the significant immune infiltration was negatively correlated with NK1R (Fig. 9A-9B).