Cancer exosomes induce tumor innervation

Abstract

Patients with densely innervated tumors suffer with increased metastasis and decreased survival as compared to those with less innervated tumors. We hypothesize that in some tumors, nerves are acquired by a tumor-induced process, called axonogenesis. Here, we use PC12 cells as an in vitro neuronal model, human tumor samples and murine in vivo models to test this hypothesis. When appropriately stimulated, PC12 cells extend processes, called neurites. We show that patient tumors release vesicles, called exosomes, which induce PC12 neurite outgrowth. Using a cancer mouse model, we show that tumors compromised in exosome release are less innervated than controls. Moreover, in vivo pharmacological blockade of exosome release similarly attenuates tumor innervation. We characterize these nerves as sensory in nature and demonstrate that axonogenesis is potentiated by the exosome-packaged axonal guidance molecule, EphrinB1. These findings indicate that tumor released exosomes induce tumor innervation and exosomes containing EphrinB1 potentiate this activity.

Tumor-infiltrating nerves create an electro-physiologically active microenvironment and contribute to treatment resistance

“…Moreover, since some published work support a more stringent isolation of sEVs, we further purified some preparations by Optiprep density gradient centrifugation [33, 34]. Consistent with previous findings, CD9 and CD81 positive sEVs were present in fraction 8 (Supplemental Figure 3A) as well as in sEVs purified by differential ultracentrifugation alone (“crude”) [19]. Satisfied that HGSOC cell lines release sEVs and our methodology successfully purifies them from conditioned media, we tested their axonogenic potential utilizing PC12 cells, a rat pheochromocytoma cell line, as a surrogate for neurite outgrowth activity.…”

“…Therefore, we tested whether sEVs from HGSOC cell lines possess neurite outgrowth activity. sEVs from conditioned media were purified by differential ultracentrifugation as previously described [19]. To validate our methodology, we analyzed purified sEVs by atomic force microscopy and found their size (84-130nm) was consistent with sEVs (Figure 3A).…”

“…To test this, we turned to a previously generated set of isogenic cell lines in which the parental cell line (A2780) is platinum sensitive, while two independently derived lines (C30, CP-70) are platinum resistant [55, 56]. Purified sEVs from these cell lines express EV markers (Figure 8F); equal amounts were tested on PC12 cells as previously described (Figures 3 & Supplemental 3) [18, 19]. While sEVs from the CP70 treatment resistant line induced significantly more neurite outgrowth from PC12 cells than sEVs from the treatment sensitive parental A2780 line, sEVs from the second treatment resistant line, C30, did not (Figure 8G).…”

“…Likewise, pain in recurrent or late stage cervical cancer remains a challenge [68-70]. As opposed to the sensory innervation we identified in cancers [18, 19], several groups have identified sympathetic and parasympathetic innervation in other tumors (prostate, liver and breast) [3, 5, 17, 71]. Recent work demonstrates varying roles for sympathetic/parasympathetic tumor-infiltrating nerves including modulation of inflammation [72], immune cell functions [17] and mediating stress effects on disease progression [73-75].…”

“…Dishes (150mm 2 ) were seeded with 500,000 cells and cultured in medium containing 10% fetal calf serum which was depleted of sEVs by overnight ultracentrifugation at 110,000×g. Conditioned medium from cultured cells was harvested 48 hours later and sEVs were purified by differential ultracentifugation as described by Madeo et al [19]. Briefly, conditioned medium was centrifuged at 300×g for 10min at 4°C to pellet cells.…”