Generating therapy-induced bone-growing NEPC cell line.
Increasing evidence indicates NED plays a critical role in PC cells escape of ADT, including resistant to a new generation of anti-androgen therapies(9–13). To test if GRP/GRP-R targeted therapy is sufficient to inhibit growth of castration-resistant prostate adenocarcinoma and NED PC cells, first, we generated an ADT-resistant PC cell by treating androgen dependent LNCaP PC cells with MDV3100 for more than 3 months (named LNCaP-MDV). Our studies show that long-term ADT with anti-androgens induces PC cell transdifferentiation to NEPC, as demonstrated by morphology, gene expression profiles (Fig. 1). Morphologically, LNCaP-MDV cells developed many long cytoplasmic processes, with secondary and tertiary neurotic-like branching, well beyond 3 to 5 times the length of the cell body (Fig. 1A). Relative to the parental cells, LNCaP-MDV had increased NE markers (chromogranin A, synaptophysin and NSE) expression (Fig. 1B). Further, although full-length AR (AR-FL) expression (at protein level; Fig. 1E) is slightly decreased and AR variants (AR-V7) expression is significantly increased (Fig. 1D, E). These results indicate long-term treatment with anti-androgens induces NED in PC cells.
In order to test if therapy-induced NED allows cancer cells to grow in the bone microenvironment, LNCaP-MDV, the therapy-induced NEPC (tNEPC) cells were inoculated into the bone of nude mice by intratibial injection. Small animal X-ray radiograph imaging was performed to monitor bone lesion development. Mice were sacrificed for histological analysis at 8 weeks post-inoculation. It is well known that LNCaP cells are difficulted to grow in murine bone following intratibial or intrafemural injection(23). As expected, no growth (0/5) was detected by either X-ray radiographic imaging or histological analysis for control LNCaP cells grafted into the bone after 8 weeks (Fig. 1F left panel and G a, b). Most surprisingly, long-term ADT enabled LNCaP-MDV cells to colonize and grow in the bone by intratibial injection (Fig. 1F right panel, G c, d, e and f). These results indicate that long-term ADT can changes the characteristics of the LNCaP to LNCaP-MDV cells that enables a non-bone-growing PC cell to progress to a bone-growing PC cell, respectively. Additionally, the LNCaP-MDV cell line is the first model of tNEPC that grows in the bone.
Blocking of GRP/GRP-R signaling efficiently inhibits NF-κB activity and ARVs (AR-V7) expression in PC cells.
Previously, we demonstrated that axis of ADT ◊ GRP/GRP-R ◊ NF-κB ◊ ARVs is an important mechanism that contributes PC progression to CRPC(19). In addition, we have demonstrated that GRP-R expression is high in 99% of the PC patients, including NEPC(19). Most importantly, as a cell surface protein, GRP-R is easily targeted by drugs. As a selective GRP-R antagonist, RC-3095 has been shown to have anti-inflammatory properties in many types of murine models(24–26). In order to determine if blocking of GRP/GRP-R signaling efficiently inhibits NF-κB activity and ARVs expression thereby control CRPC progression, 22RV1, an androgen-independent AR-FL and AR-V7 positive prostate adenocarcinoma cell, and LNCaP-MDV, a therapy-induced tNEPC cells were treated with RC-3095. NF-κB activity in PC cells was measured using the NGL reporter that is a NF-κB responsive vector which has Luciferase and Green Fluorescent Protein (GFP) reporter genes(27). The results show that blocking GRP/GRP-R signaling using RC3095 is sufficient to inhibit NF-κB activity in both of prostate adenocarcinoma (22RV1) and tNEPC (LNCaP-MDV) cells (Fig. 2A). In addition, blocking of GRP/GRP-R signaling efficiently inhibits ARVs (AR-V7) expression in these PC cells (Fig. 2B, C and D).
Blocking of GRP/GRP-R signaling increases anti-androgen sensitivity in CRPC cells.
It is known that both wild-type AR-FL and ARVs regulate AR target genes and contributes to PC cells survival and progression. Blocking GRP/GRP-R signaling alone may inhibit ARVs expression (Fig. 2B and C) but may not be sufficient to block both wild-type AR-FL and ARVs activity. However, anti-androgens (such as Bicalutamide and MDV3100) will block AR-FL activity efficiently in PC cells. Therefore, it is possible that blocking ARVs expression by blocking GRP/GRP-R signaling may reverse anti-androgen insensitive CRPC cells to become anti-androgen sensitive PC cells. To investigate if blocking of GRP/GRP-R signaling increases anti-androgen sensitivity in CRPC cells, first, we investigate if blocking of GRP/GRPR signaling is sufficient to block AR activity in CRPC cells. 22RV1 cells were treated with RC3095, MDV3100 or RC3095 + MDV3100. AR activity was measured using ARR2PB-Luc vector, AR responsive reporter vector(28). The results show that although RC3095 efficiently inhibits AR-V7 expression (Fig. 2B and D), RC3095 alone treatment failed to inhibit AR activity and PSA expression significantly (Fig. 3A and B). Also, as expected, MDV3100 alone treatment failed to inhibit PSA expression significantly in 22RV1 cells (Fig. 3B). However, when RC3095 (10− 6M) was present, MDV3100 efficiently inhibited PSA expression and AR activity in 22RV1 cells (Fig. 3B). These results indicate that although RC3095 is sufficient to inhibit AR-V7 expression efficiently (Fig. 2B and D), it cannot inhibit AR-FL activity efficiently; and, MDV3100 may inhibit AR-FL activity but it cannot block AR-V7 activity. To block AR activity efficiently in 22RV1 cells, both of RC3095 (blocks AR-V7 expression/activity) and MDV3100 (blocks AR-FL activity) are needed (Fig. 3B). To further confirm this observation, 22RV1 and LNCaP-MDV tNEPC cells were treated with RC3095 alone or in combination of RC3095 with MDV3100. Our studies show that RC3095 or MDV3100 alone had no significant effect on the growth rate of 22RV1 and LNCaP-MDV cells (Fig. 3C and D). However, when the cells were treated with anti-androgen (MDV3100) plus GRP-R antagonist (RC3095), the growth rate of the cells was significantly inhibited (Fig. 3C and D). These results indicate that blocking of GRP/GRP-R signaling is sufficient to decrease AR-V7 expression thereby increasing responsiveness of CRPC and tNEPC to the anti-androgen treatment.
Blocking of GRP/GRP-R signaling in combination with ADT is sufficient to control CRPC tumor growth in vivo.
In order to test if blocking of GRP/GRP-R signaling alone or in combination with anti-androgens is sufficient to control CRPC tumor growth in vivo, 22RV1 cells were placed into the right flank of 6–7 weeks old male athymic nude mice by subcutaneous (s.c.) injection. After the primary tumor size reaches 3 to 4 mm in diameter (about 2–3 weeks), the mice were treated with RC3095 (20ug/day, sc) alone or in combination with ADT for two weeks. Castration or MDV3100 (10mg/kg/day, gavage) was used to mimic ADT. The results show that after treatment with either RC-3095, MDV3100 or castration alone, tumors continued to grow and there was no significant difference in the tumor size compared to the control group (treated with vehicle). However, when the mice were treated with RC3095 combined with castration or MDV3100, the tumor growth was significantly inhibited (Fig. 4A and B). Immunohistochemical (IHC) staining shows that blocking of GRP/GRP-R signaling (RC3095 treatment) efficiently inhibits AR-V7 expression (Fig. 5A) and the number of the cancer cells stained by Ki67 (proliferation marker) in the tumor from mice treated with RC3095 plus ADT was significantly lower than that the mice treated with RC3095, MDV3100 or castration alone (Fig. 5A and B). In addition, when the mice were treated with RC3095 combined with castration or MDV3100, the serum PSA levels were significantly inhibited (Fig. 5C). These results strongly support that blocking of GRP/GRP-R signaling in combination with ADT is sufficient to control CRPC tumor growth.
Blocking of GRP/GRP-R signaling in combination with ADT efficiently inhibits tNEPC tumor growth in the bone.
Unlike other types of cancer, patients with advanced PC develop osseous metastasis and the initial metastasis of PC is almost strictly limited to bone which is often the only site of spread even in late disease(29, 30). A clinical study shows that NED is positive in up to 52% of patients with bone metastasis(14) and up to 15% of the CRPC bone metastasis are tNEPC(15). To test if blocking of GRP/GRP-R signaling or in combination with anti-androgens is sufficient to control tNEPC tumor growth in the bone, LNCaP-MDV cells were placed into the right tibia of 6–7 week old male athymic nude mice by direct intratibial injection. Tumor formation and growth were monitored by a small animal x-ray radiograph once a week. After bone tumor formation (6 weeks after grafting), the mice were treated with RC3095 (20ug/day, sc) alone or in combination with ADT for two weeks. Although conventional x-ray radiograph is an easy and convincing method to monitor tumor growth and detect bone lesion, it was difficult to quantitate the tumor size or bone lesions for assessment of treatment response (Fig. 6A). However, a significant increase in necrosis was observed in the combination treatment (RC3095 + Castration or RC3095 + MDV3100) groups by histological analysis (Fig. 6A). In addition, the number of the cancer cells stained by Ki67 in the tumor from mice treated with RC3095 plus ADT was significantly lower than that the mice treated with RC3095, MDV3100 or castration alone (Fig. 6B).
Taken together, these results strongly indicate that blocking of GRP/GRP-R signaling in combination with ADT is a potential new approach to control CRPC tumor growth, including ADT induced tNEPC.