Depletion of serotonin and selective inhibition of 2B receptor suppressed tumor angiogenesis by inhibiting endothelial nitric oxide synthase and extracellular signal-regulated kinase 1/2 phosphorylation

Abstract

The effects of serotonin (5-HT) on tumor growth are inconsistent. We investigated whether a decreased level of 5-HT affected tumor growth using 5-HT transporter knockout (5-HTT(-/-)) mice, which showed 5-HT depletion. When cancer cells were injected subcutaneously into both 5-HTT(-/-) and 5-HTT(+/+) mice, the tumor growth was markedly attenuated in 5-HTT(-/-) mice. Serotonin levels in the blood, forebrain, and tumors of 5-HTT(-/-) mice bearing tumors were significantly smaller than those of their 5-HTT(+/+) littermates. However, 5-HT did not increase cancer cells' proliferation in vitro. When we applied 5-HTT inhibitors to the wild mice bearing tumors, they did not inhibit tumor growth. The endothelial nitric oxide synthase (eNOS) expressions in tumors were reduced in 5-HTT(-/-) mice compared with 5-HTT(+/+) mice. Stimulations with 5-HT (1-50 microM) induced eNOS expressions in human umbilical vein endothelial cell (HUVEC) in a concentration-dependent manner. When we measured activations of multiple signaling pathways by using a high-throughput phosphospecific antibodies platform, 5-HT stimulated the extracellular signal-regulated kinase 1/2 (ERK1/2) in HUVEC. Moreover, we found that the physiological level of 5-HT induced phosphorylation of both ERK1/2 and eNOS in HUVEC. Human umbilical vein endothelial cell expressed both 5-HT(2B) and 5-HT(2C) receptors. SB204741, a specific 5-HT(2B) receptor inhibitor, blocked 5-HT-induced ERK1/2 and eNOS phosphorylations, whereas RS102221, a specific 5-HT(2C) receptor inhibitor, did not in HUVEC. SB204741 reduced microvessel density in tumors and inhibited the proliferation of HUVEC in vitro. These results suggest that regulation of 5-HT and 5-HT receptors, especially the 5-HT(2B) receptor, may serve as a therapeutic strategy in cancer therapy.

Figure 1

Effects of 5-HT on tumor growth in mice. (A) A total of 1 x 10⁶ LLC cells were implanted into 5-HTT^-/- and 5-HTT^+/+ mice. Tumor volumes were calculated from tumor measurements scored on the indicated day. Results are presented as the mean tumor volume ± SD (n = 6, 7 per group). Insert shows Western blot analysis of 5-HTT protein in tumors of 5-HTT^-/- and 5-HTT^+/+ mice. (B) A total of 1 x 10⁶ B16F0 cells were implanted into 5-HTT^-/- and 5-HTT^+/+ mice. Tumor volumes were calculated from tumor measurements scored on the indicated day. Results are presented as the mean tumor volume ± SD (n = 6 per group). (C) Serotonin in whole blood, forebrain, and tumors of 5-HTT^-/- and 5-HTT^+/+ mice on day 23 (LLC) after implantation. Serotonin in tumors of 5-HTT^-/- and 5-HTT^+/+ mice on day 19 (B16F0) after implantation. Values represent the mean ± SD (n = 6, 7 per group). Statistically significant (*P < .05, **P < .01) compared with 5-HTT^+/+ mice.

Figure 2

Paroxetine and fluvoxamine did not decrease tumor growth. (A) Mice were injected s.c. with LLC on day 0 and were treated with saline (closed squares), paroxetine (circles), or fluvoxamine (open squares) from day 6, every 2fs days. (B) Mice were injected s.c. with KLN205 on day 0 and were treated with saline (closed squares), paroxetine (circles), or fluvoxamine (open squares) from day 6, every 2 days. Tumor volumes were calculated from tumor measurement scored on the indicated day. Results are presented as the mean tumor volume ± SD. (C, D) On day 23 (LLC) or 33 (KLN205) after implantation 5-HT in whole blood and tumors of mice with paroxetine, fluvoxamine, or saline. *Statistically significant (P < .05) compared with saline-treated mice. The values represent the mean ± SD (n = 8 per group).

Figure 3

Association between 5-HT and eNOS. (A) LLC tumors from 5-HTT^-/- mice decreased eNOS. (B) Effects of 5-HT on 5-HT-induced eNOS in cultured HUVEC. Serotonin or saline was added to cultured HUVEC for 24 hours. Cells were collected for extraction and immunoblot analysis with antibodies to eNOS and β-actin. (C) Effects of serotonin on 5-HT-induced signaling pathways in cultured HUVEC. Saline or 5-HT, 1 µM, was added to cultured HUVEC for 10 minutes. Phosphorylation of intracellular signaling molecules was assessed by using the human Phospho-MAPK Array kit. The columns present the results of the densitometric analysis of the dot images corresponding to the phosphorylation status of individual protein. *Statistical significance was determined by 2-tailed-Student's t test.

Figure 4

Stimulation of 5-HT_2B receptor induced the phosphorylation of eNOS through ERK1/2 in HUVEC. (A) Effect of doses of 5-HT incubated for 10 minutes on phosphorylation of ERK1/2 and eNOS in HUVEC. (B) Serotonin induced activation of ERK1/2 and eNOS. Human umbilical vein endothelial cells were treated with 1-µM 5-HT for the indicated time. (C) Inhibition of ERK1/2 activity by U0126 and PD98059. Human umbilical vein endothelial cells were preincubated with 10-µM U0126 or 10-µM PD98059 for 30 minutes, then treated with 1-µM 5-HT for 5 minutes. (D) Reverse transcription-polymerase chain reaction analysis of selected 5-HT receptors in HUVEC. The bands representing the 5-HT_2A receptor (301 bp), the 5-HT_2B receptor (363 bp), the 5-HT_2C receptor (354 bp), and GAPDH (268 bp) are indicated. (E) Western blot analysis to demonstrate the expression of 5-HT_2B and 5-HT_2C receptors in LLC, KLN205, and HUVEC. The proteins were detected with polyclonal 5-HT_2B receptor and polyclonal 5-HT_2C receptor antibodies. Mice forebrain expressed 5-HT_2B and 5-HT_2C receptors as a positive control. (F) Inhibition of ERK1/2 activity by SB204741. Human umbilical vein endothelial cells were preincubated with 10-µM SB204741 or 10-µM RS102221 for 30 minutes, then treated with 1-µM 5-HT for 5 minutes. Phosphorylation of ERK1/2 and eNOS was determined by Western blot analysis using phosphospecific antibodies.

Figure 5

SB204741 reduced tumor growth. (A) Mice were injected s.c. with KLN205 on day 0 and were treated with saline (closed squares), SB204741 (circles), or RS102221 (open squares) from day 6, every 2 days. (B) Mice were injected s.c. with LLC on day 0 and were treated with saline (closed squares), SB204741 (circles), or RS102221 (open squares) from day 6, every 2 days. Tumor volumes were calculated from tumor measurement scores on the indicated day. Results are presented as the mean tumor volume ± SD. (n = 8 per group). *Statistically significant (P < .05) compared with saline-treated mice. (C) When the diameter of the tumors reached 1 cm, mice were killed. Bars, 100 µm. Hematoxylin and eosin-stained sections of KLN205 or LLC tumors (left). Representative sections of tumors stained for factor VIII as a vascular endothelial marker (right; original magnification, x200).

Figure 6

SB204741 treatment induced a decrease in vessel density in tumor tissues. (A) Microvessel densities were calculated. Results are indicated as the mean ± SD of eight mice in each group. The difference in MVD between control and SB204741 treatment mice was statistically significant (*P < .05). N.S.: not significant by one-way ANOVA with Fisher's least significant difference test. (B) Location of phospho-eNOS in the endothelial cells. The LLC tumor tissues were stained with anti-factor VIII-related Ag (green) and anti-phospho-eNOS (red). The images were merged. Bars, 50 µm.