Orally active lysophosphatidic acid receptor antagonist attenuates pancreatic cancer invasion and metastasis in vivo

Abstract

Pancreatic cancer is highly metastatic and has a poor prognosis. However, there is no established treatment for pancreatic cancer. Lysophosphatidic acid (LPA) has been shown to be present in effluents of cancers and involved in migration and proliferation in a variety of cancer cells, including pancreatic cancer cells, in vitro. In the current study, we examined whether an orally active LPA antagonist is effective for pancreatic cancer tumorigenesis and metastasis in vivo. Oral administration of Ki16198, which is effective for LPA(1) and LPA(3), into YAPC-PD pancreatic cancer cell-inoculated nude mice significantly inhibited tumor weight and remarkably attenuated invasion and metastasis to lung, liver, and brain, in association with inhibition of matrix metalloproteinase (MMP) accumulation in ascites in vivo. Ki16198 inhibited LPA-induced migration and invasion in several pancreatic cancer cells in vitro, which was associated with the inhibition of LPA-induced MMP production. In conclusion, Ki16198 is a promising orally active LPA antagonist for inhibiting the invasion and metastasis of pancreatic cancer cells. The inhibitory effects of the antagonist on invasion and metastasis in vivo may be partially explained by the inhibition of motility activity and MMP production in cancer cells.

Figure 1

Effects of Ki16425 and Ki16198 on motility responses to lysophosphatidic acid (LPA) and epidermal growth factor in pancreatic cancer cell lines. (a,b) YAPC‐PD cells were treated with the indicated concentrations of Ki16425 (circle) or Ki16198 (triangle) and then assayed to measure migration response to 100 nM LPA (a) and invasion response to 1 μΜ LPA (B). (c to e) Effects of Ki16425 (10 μΜ) and/or Ki16198 (10 μΜ) on invasion responses to 1 μM LPA and 10 ng/mL EGF in Panc‐1 (c), CFPAC‐1 (d), or BxPC‐3 pancreatic cancer cells (e). Data are the means ± SD of the representative result of three separate experiments.

Figure 2

Effect of lysophosphatidic acid (LPA) administration on metastasis or invasion to liver in vivo. Nude mice were inoculated with YAPC‐PD cells. LPA or phosphate‐buffered saline (PBS) was intraperitoneally injected every day from day 0 to day 7 after the inoculation of the cells. On day 7, mice were killed and individual invasive or metastasis activity to liver (left panel) was assessed as the ratio of mRNA expression of human glyceraldehyde 3‐phosphate dehydrogenase (GAPDH) to mouse GAPDH (n = 9). Right panel shows the means ± SE of their activities. *Effect of LPA administration was significant (P < 0.05).

Figure 3

Effect of Ki16198 on tumor volume, ascites formation, and metastasis or invasive activity. Twenty nude mice were inoculated with YAPC‐PD cells. Ki16198 or vehicle was orally administered to each 10 mice every day from day 0 to day 28, when mice were killed. Three control mice and one Ki16198‐treated mouse died before day 28. Tumor size (a) and ascites volume (b) were measured. **The effect of Ki16198 was significant (P < 0.01). The invasive or metastasis activities to liver (c), lung (d), and brain (e) were also measured. Results are shown as an individual activity (left panel) and means ± SE of their activities (right panel). The effect of Ki16198 was significant (*P < 0.05; **P < 0.01).

Figure 4

Effect of Ki16198 on matrix metalloproteinase (MMP) activity in ascites. The ascites collected in Figure 3 was used. The representative gelatin zymography (a) and the means ± SE of the respective MMP (proMMP‐9, active MMP‐9, proMMP‐2, and active MMP‐2) activity evaluated by densitometry (b) are shown. **The effect of Ki16198 was significant (P < 0.01).

Figure 5

Effects of Ki16198 and Ki16425 on matrix metalloproteinase (MMP) activity and mRNA expression in YAPC‐PD cells. YAPC‐PD cells were incubated for 48 h (a,b) or 12 h (c) with or without of epidermal growth factor (10 ng/mL) or lysophosphatidic acid (LPA) (1 μΜ) in the presence or absence of 10 μM Ki16198 or 10 μM Ki16425 in RPMI1640 medium containing 0.1% bovine serum albumin (BSA) to measure MMP activities (a,b) and MMP mRNA expression (c). The representative gelatin zymography (a) and the means ± SE of proMMP‐9 and proMMP‐2 activities evaluated by densitometry (b) are shown. Note that unidentified band was sometimes but not always detected in the upper position of proMMP‐2 without any constant effect of agonists. In (c), the mRNA expression was expressed as percentages of the control value without test agents: the control value (a relative ratio to glyceraldehyde 3‐phosphate dehydrogenase [GAPDG] × 10³) for MMP‐9 mRNA and MMP‐2 mRNA was 0.78 ± 0.14 and 0.54 ± 0.11, respectively. *The effect of EGF or LPA was significant (P < 0.05).