PGE2 as a signal link between tumor “seed” and microenvironment “soil”

While metastatic tumor cells and their microenvironment are considered as “seed” and “soil”, the functional link between them is critical. Only when the intrinsic programs of tumor cells are matched with the extrinsic signaling and supplies from tumor microenvironment, metastasis could be succeed.
PGE2 as a signal link between tumor “seed” and microenvironment “soil”

Lung tumorigenesis and metastasis are associated with chronic inflammation. From ecological view, metastatic tumor cells are evolved under selection pressure from microenvironment. Thus, the oncogenic program of cancer cells ought to be compatible to the signaling network of microenvironment. If not, cancer cells may either become apoptotic, senescent or quiescent. Conversely, quiescent or dormant tumor cells can be awaked when the extrinsic signaling from microenvironment become compatible, such as the conditions following chemotherapy, radiotherapy, aging, immunosuppression or inflammation.

      The pathological process of tumorigenesis and metastasis in vivo is complicated. For mechanistic analysis, it is important that an animal model used can recapitulate the pathological process and features occurred in human. In general, human lung cancer develops at the age of 50-70, indicating that there is a “smoldering” period of tumor cells in “aged” or inflammatory microenvironment of lung tissue. However, in most transgenic mouse models with oncogenes, lung cancer develops only in 2-5 months, corresponding to 20-30s of age in human. There is lack of a “smoldering” period in those models. Consequently, the epigenetic process and oncogenic program can be very different.  

       GPRC5A is a newly identified lung tumor suppressor. Gprc5a-knockout (ko) mice develop spontaneous lung tumor in 1.5 to 2 years (1), corresponding to 50-70s in human. Even treatment with carcinogen still takes one year to induce lung tumor in Gprc5a-ko mice (2). Gprc5a-ko mouse lungs are susceptible to multiple inflammatory stimuli (3-4). Gprc5a-ko leads lung epithelial cells exhibits activated NF-kB signaling, with increased levels of chemokines and pro-inflammatory cytokines (3). Thus, tumorigenesis and metastasis in Gprc5a-ko mouse lung experiences a “smoldering” period via reciprocal signaling between tumor cells and inflammatory or “aged” microenvironment, a process similar in human.

In Gprc5a-ko mouse model, we find that Ptges/PGE2 is a major upregulated signaling during lung tumorigenesis and metastasis. PGE2 is known as a major mediator of inflammation, acting through autocrine or paracrine fashion. Dysregulated COX2/PGE2 has been reported in many types of cancers. However, there are controversies regarding the roles and mechanisms of PGE2 since most of the results were obtained from analyses in vitro, in immune-deficient nude mice, or in animal models that do not resemble human diseases. Thanks to Gprc5a-ko mouse model, we were able to perform comprehensive in vivo study in a pathological microenvironment similar to human.

       Interestingly, block of PGE2 signaling in Ptges-ko mouse tumor cells, although inhibits the EMT-like and stem-like features in vitro (5), does not affect tumor growth and lung metastasis in immune-deficient mice, but does so in immune-competent mice (6). This suggests that Ptges/PGE2-mediated oncogenic effects in tumor cells are mainly through immune-evasion. Indeed, Ptges-ko tumor cells are susceptive to T cell-mediated cytotoxicity, a character that could not be seen in vitro and in nude mice. More interestingly, Ptges-ko tumor cells are still able to form metastasis in Gprc5a-ko mouse lung, but do not form tumor in subcutaneous tissue. This suggests that the immunity in Gprc5a-ko mouse lung, an inflammatory tissue microenvironment, is suppressed, whereas the immunity in subcutaneous tissue is still functional. Mechanistically, PGE2 induces polarization of M2 type macrophage directly; and PGE2 induces MDSC recruitment indirectly via various chemokines and cytokines, which is crucial for T cell inhibition. Thus, PGE2 acts as a key signal to orchestrate the intrinsic program of tumor cells and extrinsic signaling of immunosuppression in microenvironment for tumorigenesis or metastasis. Importantly, targeting PTGES/PGE2 signaling via Cay10526, an inhibitor of PTGES, restores immunity, and inhibits lung metastasis in Gprc5a-ko mice. Of notion, Cay10526 used in therapy, was 10 times lower than that used in nude mice as previously reported, suggesting the effect was not due to direct cytotoxicity (7). Thus, effective therapy on lung metastasis can be achieved by targeting the link between “seed” and “soil”, which restores host immunity and endows tumor cells susceptible to T cell-mediated cytotoxicity, rather than killing cancer cells by drug directly.  Taken together, PGE2 is a signal link between tumor “seed” and its microenvironment “soil”. PGE2 promotes lung metastasis and tumorigenesis mainly through immunosuppression.


  • Tao, Q, et al. Identification of the retinoic acid-inducible Gprc5a as a new lung tumor suppressor gene. J Natl Cancer Inst. 99(22):1668-82, 2007.
  • Zhong S, et al. Lung tumor suppressor GPRC5A binds EGFR and restrains its effector signaling. Cancer Research. 75:1801-1814, 2015.
  • Deng, J. et al. Gprc5a-knockout leads to NF-kB activation in airway epithelium and promotes lung inflammation and tumorigenesis. Cancer Prev Res (Phila Pa). 3(4):424-37, 2010.
  • Liao Y, et al. Gprc5a-deficiency confers susceptibility to endotoxin-induced acute lung injury via NF-kB pathway. Cell Cycle, 14(9):1403-1412, 2015.
  • Tong Wang, et al. Stabilization of PTGES by deubiquitinase USP9X promotes metastatic features of lung cancer via PGE2 signaling. American Journal of Cancer Research. 9(6):1145-1160, 2019.
  • Tong Wang, et al. PTGES/PGE2 signaling links immunosuppression and lung metastasis in Gprc5a-knockout mouse model. Oncogene, 2020. doi: 10.1038/s41388-020-1207-6.
  • Jing B, et al. Interleukin-6/STAT3 signaling orchestrates pre-metastatic niche formation and immunosuppressive traits in lung. Cancer Research, 80(4):784-797, 2020.


Please sign in or register for FREE

If you are a registered user on Nature Portfolio Cancer Community, please sign in