PD-L1 Activity Is Associated with Partial EMT and Metabolic Reprogramming in Carcinomas

Abstract

Immune evasion and metabolic reprogramming are hallmarks of cancer progression often associated with a poor prognosis and frequently present significant challenges for cancer therapies. Recent studies have highlighted the dynamic interaction between immunosuppression and the dysregulation of energy metabolism in modulating the tumor microenvironment to promote cancer aggressiveness. However, a pan-cancer association among these two hallmarks, and a potent common driver for them—epithelial-mesenchymal transition (EMT)—remains to be done. This meta-analysis across 184 publicly available transcriptomic datasets as well as The Cancer Genome Atlas (TCGA) data reveals that an enhanced PD-L1 activity signature along with other immune checkpoint markers correlate positively with a partial EMT and an elevated glycolysis signature but a reduced OXPHOS signature in many carcinomas. These trends were also recapitulated in single-cell, RNA-seq, time-course EMT induction data across cell lines. Furthermore, across multiple cancer types, concurrent enrichment of glycolysis and PD-L1 results in worse outcomes in terms of overall survival as compared to enrichment for only PD-L1 activity or expression. These results highlight potential functional synergy among these interconnected axes of cellular plasticity in enabling metastasis and multi-drug resistance in cancer.

Mutually exclusive teams-like patterns of gene regulation characterize phenotypic heterogeneity along the noradrenergic-mesenchymal axis in neuroblastoma

“…Further, activity of the NOR geneset correlated negatively with the Hallmark EMT signature, while MES scores correlated positively, suggesting similarity between NMT and EMT at a transcriptomic level (Figure S2a). PD-L1 levels were found to be positively associated with both the hallmark EMT and partial EMT signatures, as well as glycolysis and FAO levels across datasets, reminiscent of observations made in carcinomas 22 , 50 (Figure S2).…”

Mutually exclusive teams-like patterns of gene regulation characterize phenotypic heterogeneity along the noradrenergic-mesenchymal axis in neuroblastoma

“…Further, activity of the NOR geneset correlated negatively with the Hallmark EMT signature while MES scores correlated positively, suggesting similarity between NMT and EMT at a transcriptomic level (Figure S2A). PD-L1 levels were found to be positively associated with both hallmark EMT and partial EMT signature, as well as glycolysis and FAO levels across datasets, reminiscent of observations made in carcinomas (Muralidharan et al, 2022;Sahoo et al, 2021) (Figure S2).…”

“…The NOR/MES antagonism was also distinctly observed in our meta-analysis of 75 bulk RNA-seq datasets, and in 10 datasets where we evaluated different experimental perturbations capable of inducing a phenotypic switch along the noradrenergic/mesenchymal axis. Functionally speaking, we observed a positive association of MES phenotype with both glycolysis and PD-L1 signatures, reminiscent of mesenchymal state traits seen in multiple carcinomas (Muralidharan et al, 2022). Expression levels of B4GALNT1 that synthesizes GD2, a cell surface marker reported to be a therapeutic target for NB patients (Sait and Modak, 2017), correlated negatively with MES, but positively with a NOR phenotype.…”

A systems-level analysis of the mutually antagonistic roles of RKIP and BACH1 in dynamics of cancer cell plasticity

“…We noticed resonating trends in these datasets as noted in TCGA samples: RKIP inversely correlated with ssGSEA scores of RPMS and ferroptosis gene signature, while BACH1 correlated positively with ssGSEA scores of BPMS and ferroptosis signature (figure 2e(i)(ii)). Further, both RPMS and BPMS scores correlate positively with those of PD-L1 signature but negatively with OXPHOS (figure 2e(ii)(iii)), reminiscent of observations of PD-L1 activity being anti-correlated with OXPHOS across carcinomas [39]. Across these datasets, RPMS correlates positively with glycolysis but negatively with FAO, while BPMS shows relatively weaker trends (figure 2e(iii)).…”