5-HT, NGF, Ondansetron, SB271046, SB258585, Clozapine and PD98059 were obtained from Sigma-Aldrich (St. Louis, MO, USA). MDL7222 was obtained from Santa Cruz Biotechnology, Inc. (Santa Cruz, CA, USA).

PC12 cells (RIKEN Tsukuba Institute, Tsukuba, Japan) were routinely cultured in Dulbecco’s modified Eagle’s medium (DMEM; Sigma-Aldrich) containing 10% fetal bovine serum (Life Technologies, Gaithersburg, MD, USA), 10% horse serum (Life Technologies), 50 U/ml penicillin and 50 μg/ml streptomycin (Life Technologies). The cells were incubated at 37ºC in a 5% CO₂ atmosphere. For each experiment, 3×10⁵ cells were spread onto a 60 mm dish (Becton Dickinson, Franklin Lakes, NJ, USA). Following incubation for two days, the cells were used for RNA extraction and transwell migration assay. To obtain differentiated cells, the medium was changed to NGF-supplemented medium (DMEM containing 100 ng/ml NGF, 1% horse serum, 50 U/ml penicillin and 50 μg/ml streptomycin) and incubated for one day.

Total RNAs were extracted using TRIzol reagent (Invitrogen, Carlsbad, CA, USA), then cDNAs were synthesized with ReverTra Ace (Toyobo, Osaka, Japan) according to the manufacturer’s instructions. PCRs were performed with a denaturation step at 95ºC for 5 min, followed by 35 (Fig. 2A), 29 (Fig. 2B) or 35 cycles (Fig. 2C) of denaturation at 95ºC for 30 sec, primer annealing at 55ºC for 30 sec, and primer extension at 72ºC for 30 sec. TATA binding protein (TBP) was used as a control. The primers used were: 5-HT_1A (forward) 5′-CTCTGTTGCTGGGTACTCTCATT/(reverse) 5′-AGTCTATAGGGTCGGTGATAGCC-3′, 5-HT_1B 5′-3′-, 5-HT_2A 5′-TGTACGTGAACCAAGTCAAAGTG-3′/ 5′-GTAGATGATGGGGTTGATGAGAG-3′, 5-HT_2A 5′-ATG CTGAAAACAGAACCAACCT-3′/5′-ACATCCAGGTAAAT CCAGATCG-3′, 5-HT_2B 5′-TCGTCAAGATTACGG TGGTATG-3′/5′-CACCATCTTTTCTGGTGATGAA-3′, 5-HT_2C 5′-ATAGGGGGCAACATTCTTGTTAT-3′/5′-ACAGGGATAGGAACTGAAACTCC-3′, 5-H_T3, 5′-GGAA GTCTCCAAGCATTCCTTAT-3′/5′-ACGTAGAACTTC ATTTCCGCATA-3′, 5-HT₄ 5′-CCAATATTGTGGAC CCTTTCATA-3′/5′-GACTGGCTTCTTTTCAAGCTACA, 5-HT_5A 5′-AAGATTTACAAGGCTGCGAAGTT-3′/5′-ACT GATGAGCTCCGTAACAAAGA-3′, 5-HT_5B 5′-CTGG ATCGCTACTGGACTATCAC-3′/5′-GTGA ATACCGTCTCA GACTCCTG-3′, 5-HT₆ 5′-CTGGGAATGTTCTTTGT CACCT-3′/5′-GAAGCGGAGTCTGAATCTGA GTT-3′, 5-HT₇ 5′-ACTTCTTCTGCAACGTCTTCATC-3′/5′-GCG GCCTTGTAAATCTGATAGTA-3′, TBP 5′-TGCTGGCGG TTTGGCTAGGTTTCTGC-3′/5′-GGTCAGAGTTTGAGAA TGGAAGAGTT-3′.

PC12 cell suspension containing 2×10⁵ cells in DMEM was applied to each upper well of the transwell chamber (Becton Dickinson), which was previously coated with type I collagen (50 μg/ml, Becton Dickinson) on both sides. In the bottom well, DMEM with or without motogen, NGF (100 ng/ml) and/or 5-HT (0.1–10 μM) were applied to allow the cells to migrate across filters (8 μm pore size). Ondansetron (1 μM, 5-HT₃ antagonist), MDL7222 (1 μM, 5-HT₃), SB271046 (1 μM, 5-HT₆), SB258585 (1 μM, 5-HT₆) and PD98059 (20 μM, ERK inhibitor) were also applied in some of the experiments. The transwell migration assay was performed at 37ºC for 5 h. After removal of the remaining cells on the top side of filters using cotton swabs, the filters were fixed with 4% PFA/PBS for 15 min. After incubation with Hoechst 33258 (Nacalai Tesque, Kyoto, Japan) at room temperature for 5 min, the cells on the bottom side of transwell inserts were washed three times with PBS and examined by fluorescence microscopy (IX83; Olympus, Tokyo, Japan). Ten images were captured randomly for one experiment and the number of nuclei in a 600×600 μm in each image was counted. Image J was used for counting as previously described (14). Relative percentages of the cell number to the average of the cell number in the control experiments (DMEM only in bottom well) were plotted on the graphs.

P-values were calculated by one- or two-way ANOVA followed by Tukey’s HSD. Data are presented as mean ± standard error.

To determine the effect of 5-HT on PC12 cell migration, the transwell migration assay was used (15,16). Cells were spread onto the upper side of the transwell insert and the number of cells migrating across the filter was counted. Addition of 5-HT to the bottom well together with DMEM significantly increased the migrating cells (Fig. 1A). 5-HT-induced migration was identified in a dose-dependent manner (Fig. 1B). An amount of 1 μM 5-HT significantly increased cell migration, while 0.1, 0.5, 5.0 and 10 μM 5-HT did not show significant changes compared with the control. NGF is known to induce PC12 cell migration (15,16), thus we added 5-HT together with NGF. NGF and 5-HT induce migration in an additive manner.

We examined which serotonin receptor mediates PC12 cell migration induced by 5-HT. Previously, it was reported that 5-HT₃ enhanced neurite outgrowth induced by NGF in PC12 cells (17). However, little is known regarding the expression and molecular function of other 5-HT receptors in PC12 cells.

mRNA expression of 5-HT receptors was examined by RT-PCR and 5-HT₆ and 5-HT₃ were found to be expressed (Fig. 2A) in undifferentiated and differentiated (1 day after application of 100 ng/ml NGF) PC12 cells (Fig. 2B and C).

Using antagonists against 5-HT₃ and 5-HT₆ receptors, we investigated which 5-HT receptor mediates the migration induced by 5-HT. Addition of Ondansetron (5-HT₃ antagonist, 1 μM) or MDL7222 (5-HT₃, 1 μM) with 5-HT did not reveal any significant changes (Fig. 3A) while antagonists of 5-HT₆, SB258585 (1 μM) or SB271046 (1 μM) inhibited the migration induced by 5-HT (Fig. 3A). The data indicate that 5-HT₆ but not 5-HT₃ mediates serotonin-induced PC12 cell migration.

5-HT₆ receptor is known to be coupled with Gs protein which mediates the accumulation of cAMP (18). To examine whether cAMP pathway is involved in 5-HT₆ signaling in order to induce migration, Clozapine, which was reported to inhibit cAMP accumulation mediated by 5-HT₆, was used (19). Clozapine (1 μM) significantly reduced the effect of 5-HT on migration (Fig. 4A). Increasing of cAMP activates ERK through protein kinase A-Rap1 (20). Thus, we examined whether the inhibition of ERK activation affects 5-HT-induced PC12 cell migration. Application of ERK inhibitor, PD98059 (20 μM), significantly suppressed the cell migration induced by 5-HT (Fig. 4B). These data indicate that cAMP and ERK are involved in the 5-HT₆-mediated signaling pathways in order to induce PC12 cell migration.

In the present study, we have demonstrated that 5-HT induced PC12 cell migration in a transwell migration assay in a dose-dependent manner (Fig. 1A). A number of studies have indicated that 5-HT affects proliferation of neural cells (7,21–23) as well as neurite outgrowth (17,24–26), however, its effect on neural migration remains to be clarified. Findings of previous studies suggest 5-HT affects cortical neuron migration during prenatal development (7–9,13). Depletion of 5-HT by injection of DL-P-chlorophenylalanine (PCPA, an inhibitor of 5-HT synthesis) during the E12-E17 stage arrested migration and disorganized the positioning of cortical neurons (7), suggesting a positive effect of 5-HT on migration. By contrast, cortical slices exposed to high doses of 5-HT (100–400 μM) arrested the migration of GABergic neurons (8,9). Although findings of those studies reported negative effects of 5-HT, the concentration of 5-HT was considerably higher than that of another study focusing on the prenatal cortex (100–200 fmol/mg) (7). The effect of 5-HT on neural migration therefore remains to be elucidated. A positive effect of 5-HT at the concentration of 1 μM was observed, which is similar to the in vivo results obtained in that study.

Blocking of 5-HT stimulation by 5-HT₆ antagonists, SB271046 and SB258585 (Fig. 3B) indicate that 5-HT requires 5-HT₆ receptor but not 5-HT₃ to induce PC12 cell migration (Fig. 3A). 5-HT₆ coupled with Gs-protein is known to elevate the cAMP level that stimulates ERK via protein kinase A (19,26). Inhibition of cAMP accumulation by Clozapine (19) and ERK by PD98059 (27) reduced the 5-HT effect on the migration (Fig. 4).

Previous studies suggest that NGF showed a positive effect on PC12 cell migration by activating ERK signaling (15,16). EGF and cAMP pathways activate ERK independently in order to promote PC12 cell differentiation (5,28,29). Similarly, it is possible that the 5-HT₆-cAMP activates an independent pathway from the NGF-ERK signal to yield an additive effect on PC12 cell migration (Fig. 1A).

Antagonists of 5-HT₃ did not show any effects on PC12 cell migration in our experiment (Fig. 3A) although 5-HT was expressed (Fig. 2A), suggesting independence of 5-HT₃ signaling pathway from cAMP or ERK. Another possibility is that 5-HT₃ is not sensitive to a 5-HT induction in the experiments of this study. Homma et al(17) suggest a difference of 5-HT sensitivity between differentiated (pretreated with NGF for 3 days) and undifferentiated PC12 cells to enhance neurite outgrowth mediated by 5-HT₃. In their experiment, 50 μM 5-HT were required to enhance neurite outgrowth of undifferentiated PC12 cells whereas 5 μM of 5-HT were sufficient for differentiated cells. In our experiment, undifferentiated PC12 cells were treated with 1 μM 5-HT together with 5-HT₃ antagonists (Fig. 3A). Higher concentrations of 5-HT may be required to activate the 5-HT₃-dependent pathway in order to modulate PC12 cell migration.

5-HT₃ and 5-HT₆ are proposed as candidates to modulate neural migration (30). Although we observed an effect of 5-HT₆ but not 5-HT₃ on PC12 migration, it is possible 5-HT₃ also modulated the PC12 cell migration in different conditions, such as in the presence of NGF. Transwell migration assay with PC12 cells is useful in the study of molecular mechanisms of the neural migration induced by 5-HT or by NGF in vitro.