5-HT7 receptor-dependent intestinal neurite outgrowth contributes to visceral hypersensitivity in irritable bowel syndrome

Abstract

Irritable bowel syndrome (IBS) is characterized by visceral hypersensitivity (VH) associated with abnormal serotonin/5-hydroxytryptamine (5-HT) metabolism and neurotrophin-dependent mucosal neurite outgrowth. The underlying mechanisms of VH remain poorly understood. We investigated the role of 5-HT₇ receptor in mucosal innervation and intestinal hyperalgesia. A high density of mucosal nerve fibres stained for 5-HT₇ was observed in colonoscopic biopsy specimens from IBS patients compared with those from healthy controls. Staining of 5-HT₃ and 5-HT₄ receptors was observed mainly in colonic epithelia with comparable levels between IBS and controls. Visceromotor responses to colorectal distension were evaluated in two mouse models, one postinfectious with Giardia and subjected to water avoidance stress (GW) and the other postinflammatory with trinitrobenzene sulfonic acid-induced colitis (PT). Increased VH was associated with higher mucosal density of 5-HT₇-expressing nerve fibres and elevated neurotrophin and neurotrophin receptor levels in the GW and PT mice. The increased VH was inhibited by intraperitoneal injection of SB-269970 (a selective 5-HT₇ antagonist). Peroral multiple doses of CYY1005 (a novel 5-HT₇ ligand) decreased VH and reduced mucosal density of 5-HT₇-expressing nerve fibres in mouse colon. Human neuroblastoma SH-SY5Y cells incubated with bacteria-free mouse colonic supernatant, 5-HT, nerve growth factor, or brain-derived neurotrophic factor exhibited nerve fibre elongation, which was inhibited by 5-HT₇ antagonists. Gene silencing of HTR7 also reduced the nerve fibre length. Activation of 5-HT₇ upregulated NGF and BDNF gene expression, while stimulation with neurotrophins increased the levels of tryptophan hydroxylase 2 and 5-HT₇ in neurons. A positive-feedback loop was observed between serotonin and neurotrophin pathways via 5-HT₇ activation to aggravate fibre elongation, whereby 5-HT₃ and 5-HT₄ had no roles. In conclusion, 5-HT₇-dependent mucosal neurite outgrowth contributed to VH. A novel 5-HT₇ antagonist could be used as peroral analgesics for IBS-related pain.

Fig. 1. Immunostaining of nerve fibres in colonoscopic biopsy specimens from healthy subjects and IBS patients for PGP9.5 and 5-HT receptor subtypes.

A Representative immunofluorescence images of staining for PGP9.5 (a pan neuronal marker, green) and nuclear staining with Hoechst dye (blue) in colonic mucosal biopsies from healthy control (HC) subjects (N = 12) and IBS patients (N = 13). Bar: 50 μm. B Quantification of PGP9.5 immunofluorescence intensity in mucosal area. Two to three images were taken for each individual, and an average of 30 images per group was scored in a blinded manner. C Staining for 5-HT₇ (green) and with Hoechst dye (blue) in colonic biopsies. Bar: 50 μm. D Merged images of PGP9.5 (green), 5-HT₇ (red) and Hoechst (blue) staining in colonic mucosal tissues from IBS patients. Arrowheads indicate the colocalization of PGP9.5 and 5-HT₇. Bar: 10 μm. E Staining for 5-HT₃ (green) and Hoechst (blue) in colonic biopsies. Bar: 50 μm. F Staining for 5-HT₄ (green) and Hoechst (blue) in colonic biopsies. Bar: 50 μm.

Fig. 2. Activation of 5HT₇ played a critical role in intestinal hyperalgesia in two IBS-like mouse models.

Two mouse models with visceral hypersensitivity were investigated. A Timeline of the first model. a Mice were post-infected with Giardia and subjected to water avoidance stress (designated the GW model). The uninfected unstressed control (Ctrl) groups were pair-fed saline and left in cages unhandled. Electrode planting was performed at least 14 days prior to measurement of visceromotor responses (VMRs) to colorectal distension. b Pilot studies indicated the presence of Giardia trophozoites in the intestine on postinfection day 4 and clearance around 14 days. The absence of trophozoites in intestine were noted on postinfection day 42, confirming the self-limiting status of infection. c The plasma corticosterone levels as an indicator of psychological stress. *P < 0.05 vs.  Ctrl. N = 9–10/group. B Timeline of the second model. a Mice were intracolonically injected trinitrobenzene sulfonic acid (TNBS) and after resolution of colitis, the VMRs to colorectal distension was measured (designated the Post-TNBS (PT) model). The sham groups (Sham) were injected the same volume of saline. b and c Pilot studies showing myeloperoxidase (MPO) activity and histopathological score in mouse colonic tissues on various days after TNBS administration. d Pilot studies showing VMRs expressed as the area under curve (AUC) on various days after TNBS administration. Mice post-resolution of colitis on day 24 was chosen as the time point for the PT model. *P < 0.05 vs. Sham. N = 6–10/group. C, D Representative images of colon histology in Ctrl and GW mice, as well as those in Sham and PT mice after TNBS administration for 2, 7, 14, 24, and 51 days. Bar: 50 μm. E, F Western blots showing increased 5-HT₇ protein levels in the colonic mucosal tissues of GW and PT mice compared to those of their respective controls. G, H Quantitative PCR results of transcript levels of tryptophan hydrolase 1/2 (Tph1/2), 5-HT receptor subtypes (Htr3, Htr4, and Htr7), nerve growth factor (Ngf), brain-derived neurotrophic factor (Bdnf), and receptor subunits of p75^NTR (Ngfr), TrkA (Ntrk1), and TrkB (Ntrk2) in the colonic tissues of GW and PT mice. *P < 0.05 vs. Ctrl or Sham. N = 6/group. I, J Intestinal pain levels in GW and PT mice after intraperitoneal (i.p.) treatment with a selective 5HT₇R antagonist SB269970 (SB7) or vehicle (veh) 90 min prior to pain measurement. *P < 0.05 vs. Ctrl+veh or Sham+veh. ^#P < 0.05 vs. GW+vehicle or PT+vehicle. N = 6–9/group.

Fig. 3. Oral administration of a novel 5-HT₇ receptor ligand exerted potent analgesic effects in two IBS-like animal models.

Mice were perorally (p.o.) administered CYY1005 (CYY) at various doses for the measurement of intestinal pain. A Analgesic effects of p.o. administered single dose of CYY (1.5 and 5 mg/kg) in the GW model. *P < 0.05 vs. Ctrl+vehicle, ^#P < 0.05 vs. GW+vehicle. B Analgesic effects of p.o. administered multiple dose (m.d.) of CYY (1 and 3 mg/kg) for ten consecutive days in the GW model. *P < 0.05 vs. Ctrl+vehicle (m.d.), ^#P < 0.05 vs. GW+vehicle (m.d.). C Analgesic effect of p.o. administered single dose of CYY (5 mg/kg) in the PT model. *P < 0.05 vs. Sham+vehicle, ^#P < 0.05 vs. PT+vehicle. D Analgesic effects of p.o. administered multiple dose (m.d.) of CYY (3 mg/kg) for ten consecutive days in the PT model. No effects on the baseline values were seen in Sham control mice administered m.d. of CYY. *P < 0.05 vs. Sham+vehicle (m.d.), ^#P < 0.05 vs. PT+vehicle (m.d.). E Administration of CYY at concentrations of 5, 10, or 20 mg/kg for single dose p.o. had no effect on the baseline values in Sham control mice. F Administration of CYY higher than 5 mg/kg for single dose p.o. did not result in more robust analgesic effects in PT mice. ^#P < 0.05 vs. PT+vehicle. N = 6–8/group.

Fig. 4. Increased mucosal nerve fibre densities were correlated with higher 5-HT₇ expression in the colonic tissues of GW and PT mice.

A,B Immunostaining for PGP9.5 in the colonic tissues of GW and PT mice and mice administered multiple doses (m.d.) of CYY. C, D Immunostaining for 5-HT₇ in the colonic tissues of GW and PT mice and mice administered CYY (m.d.). The insets showed magnified images of the staining. Bar: 50 μm. E, F Quantification of immunoreactivity in the colonic mucosal tissues of GW and PT mice. Fluorescence intensity of (a) PGP9.5 and (b) 5-HT₇ per area was quantified using an imaging software in a total of 25 images from each mouse group. *P < 0.05 vs. Ctrl or Sham; ^#P < 0.05 vs. GW or PT. N = 6–8/group.

Fig. 5. Mouse colonic supernatant induced neurite outgrowth in human SH-SY5Y cells in a 5-HT₇-dependent manner.

Differentiated SH-SY5Y cells were incubated with bacteria-free mouse colonic supernatant, and nerve fibre elongation was assessed. A Nerve fibre length of SH-SY5Y cells exposed to mouse colonic supernatant obtained from Ctrl, GW, or GW+CYY (m.d.) mice. a Average neurite length, and (b) Percentage of neurons with neurite longer than 50 μm. B Nerve fibre length of SH-SY5Y cells exposed to mouse colonic supernatant obtained from sham, PT, or PT+CYY (m.d.) mice. a Average neurite length, and b Percentage of neurons with neurite longer than 50 μm. C, D Representative figures of neurons after exposure to mouse colonic supernatant. Bar: 50 μm. E, F Pretreatment with a selective 5-HT₇ antagonist SB7 inhibited neurite outgrowth caused by incubation with colonic supernatant from GW and PT mice in a dose-dependent manner. A total of 300–400 neurons from each group were used for quantification of nerve fibre length. The bar graph represented the mean ± SEM of each group. *P < 0.05 vs. Ctrl or Sham. ^#P < 0.05 vs. GW or PT.

Fig. 6. Stimulation with exogenous serotonin induced neurite outgrowth and neurotrophin upregulation in a 5-HT₇-dependent manner.

Human neuroblastoma SH-SY5Y cells differentiated by retinoic acid were stimulated with 5-HT (1 μM) for 48 h. A, B Average nerve fibre length and percentage of neurons with neurite longer than 50 μm in SH-SY5Y cells stimulated with 5-HT, in the absence or presence of CYY (10 μM, a 5-HT₇ receptor ligand), SB7 (10 μM, a 5-HT₇ antagonist), ALN (10 μM, a 5-HT₃ antagonist), and GR (10 μM, a 5-HT₄ antagonist). A total of 250–300 neurons from each group were used for quantification of nerve fibre length. The bar graph represented the mean ± SEM of each group. *P < 0.05 vs. w/o; ^#P < 0.05 vs. 5-HT + vehicle. C Nerve fibre length after stimulation with a 5-HT₇ agonist LP211 (0, 0.1, or 1 μM). *P < 0.05 vs. 0. D Gel images showing gene silencing of HTR7 in SH-SY5Y cells by infection with lentiviral shRNA oligonucleotides (sh7R) compared to those with mock plasmid controls (pLKO). The bands are products of semi-quantitative PCR. E Quantitation PCR results indicated that only HTR7, but not HTR3 or HTR4, gene expression was knocked down by sh7R. *P < 0.05 vs. pLKO. N = 6/group. F Photoimages showing reduced 5-HT₇ protein expression and nerve fibre length in the HTR7-knockdown cells. Bright field views (upper panel) and Immunostaining of 5-HT₇ (lower panel). Bar: 50 μm. G Quantitative results of 5-HT₇ intensity per area in the HTR7-knockdown cells. *P < 0.05 vs. pLKO. H Nerve fibre length in the HTR7-knockdown cells. a Average fibre length. b Percentage of neurons with neurite longer than 50 μm. A total of 250–300 neurons from each group were used for analysis of the gene knockdown experiments. The bar graph represented the mean ± SEM of each group. *P < 0.05 vs. pLKO.

Fig. 7. Enhanced neurotrophin expression by activation of 5-HT₇.

A, B Transcript levels of NGF and BDNF, and receptor subunits TrkA (NTRK1), TrkB (NTRK2), and p75^NTR (NGFR) in SH-SY5Y cells after exposure to 5-HT and LP211 for 24 and 48 h. The neuroblastoma cells which are untreated (UT) or those differentiated with retinoic acid (RA) served as controls. *P < 0.05 vs. UT; ^#P < 0.05 vs. RA. N = 6/group. C The 5-HT-induced upregulation of NGF and BDNF gene expression was inhibited by 5-HT₇ antagonists (CYY or SB7) but not by antagonists to 5-HT₃ (ALN) or 5-HT₄ (GR). *P < 0.05 vs. w/o; ^#P < 0.05 vs. 5-HT + vehicle. N = 6/group.

Fig. 8. Reciprocal aggravation of serotonin and neurotrophin pathways via 5-HT₇ activation in neurons.

Human SH-SY5Y cells were stimulated with NGF or BDNF (100 ng/ml) for 48 h and the neurite length and protein levels were measured. A Representative immunostaining images of PGP9.5, TPH2, and 5-HT₇ in cells with or without neurotrophin stimulation. Bar: 50 μm. B Longer nerve fibre length was observed after neurotrophin stimulation. a Average nerve fibre length and b percentage of neurons with neurite longer than 50 μm. *P < 0.05 vs. w/o. C Increased immunofluorescent intensity of (a) TPH2 and (b) 5-HT₇ in neurons after neurotrophin stimulation. *P < 0.05 vs. w/o. A total of 250–300 neurons from each group were used for quantification. The bar graph represented the mean ± SEM of each group. D Transcript levels of tryptophan hydroxylases (TPH1 and TPH2) and 5-HT receptor subtypes (HTR3, HTR4, and HTR7) in neurons after neurotrophin stimulation. The neuroblastoma cells which are untreated (UT) or those differentiated with retinoic acid (RA) served as controls. *P < 0.05 vs. UT; ^#P < 0.05 vs. RA. N = 6/group. E Knockdown of NTRK1 gene in SH-SY5Y cells by lentiviral shRNA targeting TrkA (shTrkA) suppressed the NGF-induced TPH2 and HTR7 overexpression. The mock plasmid control is labeled as pLKO. a Gel images of semi-quantitative PCR band products showing gene silencing of NTRK1. b, c Quantitative PCR results showing reduction of TPH2 and HTR7 mRNA levels in the NTRK1-knockdown cells upon NGF stimulation. F Knockdown of NTRK2 gene by lentiviral shRNA targeting TrkB (shTrkB) decreased the BDNF-induced TPH2 and HTR7 overexpression. a Gel images of semi-quantitative PCR band products showing gene silencing of NTRK2. b, c Quantitative PCR results showing reduction of TPH2 and HTR7 mRNA levels in the NTRK2-knockdown cells upon BDNF stimulation. *P < 0.05 vs. pLKO. N = 6/group. G Neurotrophin-mediated nerve fibre elongation was inhibited by pretreatment with 5-HT₇ antagonists CYY and SB7. *P < 0.05 vs. UT; ^#P < 0.05 vs. vehicle. A total of 250–300 neurons from each group were used for quantification. The bar graph represented the mean ± SEM of each group. H Proposed schema of a positive-feedback loop between serotonin and neurotrophin pathways via 5-HT₇ activation for induction of neurite outgrowth and visceral hypersensitivity.