The Opinion article from Gabriel Ichim and Stephen W. G. Tait, 'A fate worse than death: apoptosis as an oncogenic process' (Nat. Rev. Cancer 16, 539–548; 2016)1, highlights intrinsic mechanisms of failure of commonly used cancer therapies that are heavily dependent on apoptosis. Homeostatic mechanisms are triggered by the presence of dying cells within tissues, favouring the survival of residual tumoural cells in the tumour microenvironment2. Understanding these homeostatic mechanisms will provide actionable targets to enhance the effectiveness of therapeutic strategies used in the clinic.

The presence of apoptotic cells with subtumorigenic numbers of mouse melanoma cells promoted tumour engraftment and growth, in a mechanism dependent on the recruitment of inflammatory cells3. In a separate study, it was shown that clearance of dying cells by tumour macrophages shifts them towards the pro-tumorigenic M2 phenotype by mechanisms involving the receptor for the lipid mediator platelet-activating factor (PAF), a member of the 1-alkyl,2-acylglycerophosphocholine subclass of lipids4.

Lipid mediators are generated in the tumour microenvironment and can modulate tumour growth. Prostaglandin E2 (PGE2) is among the most studied molecular effectors of tumour repopulation following chemotherapy5. PGE2 is derived from the metabolism of membrane phospholipids, through the action of phospholipase A2 (PLA2). Both Ca2+-dependent PLA2 (cPLA2) and Ca2+-independent PLA2 (iPLA2; also known as PLA2G6)6 cleave phosphatidylcholine into arachidonic acid and lysophosphatidylcholine (Fig. 1). Although arachidonic acid can be enzymatically converted into prostanoids, including PGE2, lysophosphatidylcholine is converted into PAF through the action of lysophosphatidylcholine acyltransferases (LPCAT1, LPCAT2, LPCAT3 and LPCAT4)7. Therefore, in situations in which PGE2 is produced, depending on the levels of LPCAT expression, PAF is also produced. In a positive feedback loop, through the activation of its receptor (PAFR), PAF induces either activation of PLA2 (Ref. 8) or induction of cyclooxygenase 2 (COX2) expression9,10, leading to the further release of PGE2. The contribution of PAFR-dependent signalling in tumour repopulation was also demonstrated in its response to chemotherapy and radiotherapy. PAF protects tumour cells from drug-induced cell death, and the administration of PAFR antagonists blocks its protective effect11,12,13

Figure 1: Pro-oncogenic effects of platelet-activating factor (PAF) and PAF-like molecules after a cell death stimulus.
figure 1

In response to a death stimulus, activated caspase-3 (CASP3) promotes activation of phospholipase A2 (PLA2), which cleaves membrane phospholipids such as phosphatidylcholine (PC), releasing arachidonic acid, which can be converted into prostaglandin E2 (PGE2) and lysophosphatidylcholine (LPC or lyso-PAF (LPAF)), which can then be converted into PAF through the activity of lysophosphatidylcholine acyltransferase (LPCAT). PAF acts in a positive autocrine loop to further activate PLA2. PAF also interacts with intracellular PAF receptor (PAFR), activating nuclear factor-κB (NF-κB), which in turn induces cyclooxygenase 2 (COX2) transcription, thus potentiating PGE2 production. Both PAF and PGE2 act as pro-survival stimuli in tumour cells, promoting proliferation of the surviving cells and leading to tumour repopulation. Activation of PAFR in macrophages promotes reprogramming towards an anti-inflammatory and pro-tumorigenic profile (M2 genes). Moreover, irradiation and chemotherapy generate a wide range of PAFR ligands (PAF-like) in the tumour microenvironment that act as PAFR agonists (dashed arrows), amplifying the effect of PAFR activation in tumour cells and macrophages.

PowerPoint slide

Radiation therapy is among the most potent inducers of PAFR ligands through the oxidation of phospholipids14, whose formation is independent of LPCAT. Regardless of their origin, however, PAFR ligands favour PGE2 production. Experimental radiation therapy was more effective in tumour-engrafted PAFR-knockout mice, as compared with wild type, owing to decreased tumour repopulation in animals devoid of PAFR signalling. Moreover, PAFR ligands induce tissue remodelling and angiogenic cytokine production, through macrophage polarization towards the M2 phenotype. The accumulating evidence for the involvement of PAFR-dependent signalling in the survival response of tumours to chemo- and radiotherapy supports the concept that PAFR antagonists could be used for adjuvant cancer therapy. Several antagonists have been developed in the past, but none has yet been tested for clinical use in cancer. PAFR antagonists are likely to complement strategies that target tumour-associated macrophages (colony-stimulating factor 1 receptor (CSF1R) antagonists) and PGE2 (COX2 inhibitors), which, together, will improve the efficacy of apoptogenic therapies.

There is a reply to this Correspondence by Ichim, G. & Tait, S. W. G. Nat. Rev. Cancer (2017)