Hemokinin-1 (HK-1) is a newly identified tachykinin, originating from the immune system rather than neurons, and may participate in the immune and inflammatory response. In colonic mucosa of patients with inflammatory bowel disease (IBD), up-regulation of the TAC4 gene encoding HK-1 and increased production of prostaglandin E2 (PGE2) occur. Our aim was to examine the mechanistic link between human HK-1 and PGE2 production in normal human colon. Exogenous HK-1 (0.1 μM) for 4 h evoked an increased PGE2 release from colonic mucosal and muscle explants by 10- and 3.5-fold, respectively, compared with unstimulated time controls. The HK-1-stimulated PGE2 release was inhibited by the tachykinin receptor antagonists (S)1–2-[3-(3,4-dichlorophenyl)-1-(3-isopropoxyphenylacetyl)piperidin-3-yl]ethyl-4-phenyl-l azonia-bicyclo[2.2.2]octane (SR140333) [neurokinin-1 (NK1)] and N-[(2S)-4-(4-acetamido-4-phenylpiperidin-1-yl)-2-(3,4-dichlorophenyl)butyl]-N-methylbenzamide (SR48968) [neurokinin-2 (NK2)] and was also inhibited by the cyclooxygenase (COX)-2 inhibitor N-(2-cyclohexyloxy-4-nitrophenyl)methanesulfonamide) (NS-398) but not by the COX-1 inhibitor 5-(4-chlorophenyl)-1-(4-methoxyphenyl)-3-trifluoromethylpyrazole (SC-560). A parallel study with substance P showed similar results. Molecular studies with HK-1-treated explants demonstrated a stimulatory effect on COX-2 expression at both transcription and protein levels. It is noteworthy that this was coupled with HK-1-induced down-regulation of 15-hydroxyprostaglandin dehydrogenase (15-PGDH) mRNA and protein expression. Immunoreactivity for 15-PGDH occurred on inflammatory cells, epithelial cells, platelets, and ganglia. This finding provides an additional mechanism for HK-1-evoked PGE2 increase, in which HK-1 may interfere with the downstream metabolism of PGE2 by suppressing 15-PGDH expression. In conclusion, our results uncover a novel inflammatory role for HK-1, which signals via NK1 and NK2 receptors to regulate PGE2 release from human colonic tissue, and may further explain a pathological role for HK-1 in IBD when abnormal levels of PGE2 occur.
Tachykinins are a family of small biologically active peptides, originally identified in neurons. The best known mammalian tachykinins are substance P (SP), neurokinin (NK) A, and NKB, which act mainly via NK1, NK2, and NK3 receptors, respectively (Pennefather et al., 2004). Another tachykinin, hemokinin-1 (HK-1; sequence TGKASQFFGLM-NH2), has now been identified in humans (Kurtz et al., 2002; Page, 2004). HK-1 was found in immune cells, rather than in neurons (Zhang et al., 2000), and it shares a similar amino acid sequence and receptor selectivity with SP (Kurtz et al., 2002; Pennefather et al., 2004). In the intestine, SP and NKA act as neurotransmitters to modulate gut functions such as motility, secretion, blood flow, and inflammation (Castagliuolo et al., 1997; Holzer and Holzer-Petsche, 1997).
SP can also be released from non-neuronal cells (Castagliuolo et al., 1997; Moriarty et al., 2001) and is considered to act as an autocrine, paracrine, or endocrine regulator of secretory, immune, and inflammatory responses. HK-1 has been implicated in lymphoid (Zhang et al., 2000; Zhang and Paige, 2003; Berger et al., 2010) and myeloid (Janelsins et al., 2009) cell proliferation and development. This leads to the speculation that the intestinal immune and inflammatory actions of SP could be shared by HK-1. However, the actions of HK-1 in the gastrointestinal system remain largely unknown.
It has been hypothesized that tachykinins might be implicated in inflammatory bowel disease (IBD), by interrupting the cross-talk between the neural and immune systems and promoting acute and chronic intestinal inflammation, thus contributing to the motor, secretory, and immunological disturbances that characterize IBD (Holzer, 1998). This hypothesis was supported by evidence that overexpression of SP and its preferred NK1 receptor occurred in the intestinal mucosa of patients with IBD (Mazumdar and Das, 1992; Goode et al., 2000). Furthermore, animal studies also suggested a critical role for SP/NK1 receptor in amplifying intestinal inflammation (Kataeva et al., 1994; Castagliuolo et al., 1997; Di Sebastiano et al., 1999) and even initiating colitis (Lin et al., 2009). Recently, our molecular studies demonstrated that the TAC4 gene encoding HK-1, together with the TAC1 gene encoding SP and the TACR1 gene encoding NK1 receptor, were dramatically up-regulated in the colonic mucosa of patients with ulcerative colitis (Liu et al., 2011).
Prostaglandin E2 (PGE2) is one of a number of eicanosoids that are synthesized via cyclooxygenase (COX) and metabolized by 15-hydroxyprostaglandin dehydrogenase (15-PGDH) (Otani et al., 2006). PGE2 is the major prostaglandin of the intestine, and its role during intestinal inflammation is controversial. PGE2 and prostaglandin EP4 receptor are required for maintaining the mucosal barrier integrity in experimental models of IBD, presumably through the enhancement of epithelial regeneration (Jiang et al., 2007), but at higher concentrations PGE2 can reduce colonic mucosal barrier integrity (Lejeune et al., 2010). Studies on EP4 receptor knockout mice also suggested a protective role for PGE2 against inflammation (Kabashima et al., 2002). On the other hand, PGE2 may lead to impaired mucosal healing or excessive fibrosis by inhibiting the migration of the colonic lamia propria fibroblasts from patients with Crohn's disease (Rieder et al., 2010). PGE2 production is increased in IBD (Baumeister et al., 1996), and it has been suggested as a mediator and/or prognostic marker in IBD (Wiercińska-Drapalo et al., 1999; Sheibanie et al., 2007).
PGE2 and tachykinins have two-way interactions, in that tachykinins can increase levels of PGE2 in rat intrapulmonary bronchi (Szarek et al., 1998), whereas PGE2 also stimulates SP release in rat sensory neurons (White, 1996). Although it is clear that abnormal levels of SP, HK-1, and PGE2 all occur in IBD, and functional interaction between SP and PGE2 has been reported in human colonic epithelial cells (Koon et al., 2006), a mechanistic link between HK-1 and PGE2 production remains unknown. Given the potential pathological role of HK-1 in IBD, our aim was to explore the inflammatory role for HK-1 in normal human colonic tissue.
Materials and Methods
Patients and Specimens.
Colon ring segments approximately 5 cm in length were obtained from 20 male and 15 female patients undergoing colectomy for carcinoma (age range 41–72 years). Most specimens were collected from the sigmoid colon (n = 27) with some also from the ascending colon (n = 7), and descending colon (n = 1). Segments were taken 10 to 20 cm away from the site of carcinoma; any specimen appearing macroscopically inflamed or showing abnormal histological features was discarded. Patients who had obstruction or had undergone radiation or chemotherapy were excluded from our study. This project was approved by the Human Ethics Committees of the University of New South Wales (HREC08310) and South Eastern Sydney and Illawarra Area Health Service (SESIAHS 06/69).
Tissue Explant Preparation.
Surgically resected specimens were obtained within 5 min of removal from patients, immediately placed in ice-cold Krebs-Henseleit solution pregassed with carbogen (95% O2 and 5% CO2), and kept on ice until dissection. Specimens were dissected into smooth muscle and mucosa layers and kept in Krebs-Henseleit solution at 4°C overnight. For muscle explants, the mucosa, submucosa, serosa, and taenia coli were removed, leaving the circular muscle with a thin layer of longitudinal muscle. For mucosal explants, the submucosa was removed. Adjacent explants (3 × 5 mm; approximate weight 60–120 mg) of colonic mucosa or muscle were placed in 2-ml tissue baths containing Krebs-Henseleit solution maintained at 37°C and aerated with carbogen. Explants were not placed under tension and were allowed to equilibrate for 1 h before being washed and exposed to different treatments before aliquots of bath fluids were collected and snap-frozen for later PGE2 measurement. At the end of treatments, explants were snap-frozen for RNA or protein extraction as described below.
Tachykinin-Stimulated PGE2 Release in Colonic Tissue Explants.
Time-course experiments were initially performed where mucosal and muscle explants were incubated in the presence or absence of human HK-1 (0.1 μM) for 0, 1, 2, 4, and 8 h. In concentration-response experiments, mucosal and muscle explants were challenged with HK-1 (10 pM to 10 μM) for 4 h (found to be the optimum time period).
The effects of selective tachykinin receptor antagonists were examined. The NK1 receptor antagonist (S)1–2-[3-(3,4-dichlorophenyl)-1-(3-isopropoxyphenylacetyl)piperidin-3-yl]ethyl-4-phenyl-l azonia-bicyclo[2.2.2]octane (SR140333) (0.1 μM) or the NK2 receptor antagonist N-[(2S)-4-(4-acetamido-4-phenylpiperidin-1-yl)-2-(3,4-dichlorophenyl)butyl]-N-methylbenzamide (SR48968) (0.1 μM) was added 1 h before the initiation of HK-1 or SP. For experiments investigating COX inhibitors, the nonselective COX inhibitor indomethacin (1 μM), the selective COX-1 inhibitor 5-(4-chlorophenyl)-1-(4-methoxyphenyl)-3-trifluoromethylpyrazole (SC-560) (0.1 μM), or the selective COX-2 inhibitor N-(2-cyclohexyloxy-4-nitrophenyl)methanesulfonamide) (NS-398) (60 μM) was added 30 min before the addition of HK-1 or SP. These concentrations of SC-560 and NS-398 were based on their COX isoform selectivity (Barnett et al., 1994; Johnson et al., 1995). Some tissue explants were also preincubated with the nerve blocking agent tetrodotoxin (TTX; 0.1 μM) for 30 min. After the pretreatment with inhibitors, explants were exposed to HK-1 or SP for 4 h.
PGE2 levels released into bath fluid were measured by using a PGE2 ELISA kit (R&D System, Minneapolis, MN) according to the manufacturer's instructions. Results (in picogram/milligram of tissue) were expressed as S.E.M.
Quantitative Real-Time PCR.
After incubation, total RNA was extracted from mucosal and muscle explants by using the TRIzol method followed by a DNase treatment to remove contaminating DNA. A SuperScript III First Strand Synthesis kit (Invitrogen, Carlsbad, CA) was used to generate cDNA according to the manufacturer's recommendations and was then subjected to real-time quantitative PCR (qPCR) using Maxima SYBR Green qPCR Master Mix (Fermentas, Burlington, Canada) and a thermal cycler (Eppendorf AG, Hamburg, Germany). The forward primer (fp) and reverse primer (rp) for each gene used in PCR were: COX-1 (official gene name PTGS1), fp 5′-ctc cca gga gta cag cta cga-3′ and rp 5′-cca gca atc tgg cga gag a-3′; COX-2 (PTGS2), fp 5′-ggc ttc cat tga cca gag cag-3′ and rp 5′-gcc gag gct ttt cta cca ga-3′; 15-PGDH (HPGD), fp 5′-ttg gaa gac tgg aca ttt tgg-3′ and rp 5′-cct tca cct cca ttt tgc tt-3′; and GAPDH, fp 5′-cat gag aag tat gac aac agc ct-3′ and rp 5′-agt cct tcc acg ata cca aag t-3′. The PCR cycle consisted of 95°C for 2 min, followed by 40 cycles of 95°C for 15 s, 60°C for 15 s, and 68°C for 20 s. In the final step, the melting curve analysis was carried out during gradual temperature elevation from 60 to 95°C. Experiments were conducted in duplicate, using GAPDH as a housekeeping gene (HKG) and a control smooth muscle sample as a calibrator in each PCR run (Liu et al., 2011). After amplification, the mRNA level for each gene was expressed as fold change, in which each target gene was normalized to GAPDH and expressed relative to the calibrator using the formula: fold change = 2−ΔΔCt, where ΔΔCt = [Ct(target) − Ct(HKG)]sample −[Ct(target) − Ct(HKG)]calibrator (Liu et al., 2011).
Protein was extracted from human colonic mucosal strips after incubation in lysis buffer (10 mM Tris-HCl, 1 mM EGTA, and 10 mM EDTA) in the presence of a protease inhibitor cocktail (Roche, Castle Hill, Australia). Proteins (20 μg) in the tissue lysates were subjected to 10% SDS-polyacrylamide gel electrophoresis and transferred to polyvinylidene difluoride membrane. The membrane was loaded onto blot holders of the SNAP i.d. Protein Detection System (Millipore Corporation, Billerica, MA) and subjected to immunodetection according to the manufacturer's protocol. Primary antibodies used were mouse monoclonal anti-COX-1 (1:1000), rabbit polyclonal anti-COX-2 (1:200), rabbit polyclonal anti-15-PGDH (1:200), and rabbit polyclonal anti-GAPDH (1:8000). Secondary antibodies used were horseradish peroxide-conjugated goat anti-rabbit and mouse IgG (1:2000). The proteins were detected by using the Enhanced Chemiluminescence reaction kit (Millipore Corporation). Bands of COX-1, COX-2, 15-PGDH, and GAPDH were scanned and subjected to densitometry analysis, and the results were expressed relative to GAPDH.
Full-thickness pieces of normal human colon (n = 6) were fixed in Zamboni's solution. Paraffin-embedded sections were cut and mounted on poly-l-lysine-coated slides. Reduction of peroxidases was accomplished by incubation in H2O2 (3%) for 5 to 10 min at room temperature followed by nonspecific protein blocking (2% horse or goat serum) for 30 min. Sections were incubated with primary antibodies overnight at room temperature. Primary antibodies used were mouse monoclonal COX-1 antibody (1:500), mouse monoclonal COX-2 antibody (1:100), and rabbit polyclonal 15-PGDH antibody (1:50). Sections were then washed thoroughly and incubated with biotinylated horse anti-mouse or goat anti-rabbit secondary antibodies (1:2000) for 2 h, followed by avidin-biotinylated enzyme complex for 1 h. Immunoreactivity (IR) was visualized by reaction with the substrate diaminobenzidine. Each experiment included a negative control incubated with secondary antibody alone; these negative controls consistently exhibited no immunoreactivity.
Immunohistochemistry was also performed in tissue explants (n = 7) incubated in the presence or absence of HK-1 (0.1 μM) for 4 h. The experimental procedures were the same as mentioned above.
Data for PGE2 ELISA and Western blot densitometry were normally distributed and expressed as mean ± S.E.M. The means were compared using unpaired or paired Student's t test. In addition, one-way analysis of variance (ANOVA) was used with a post (Newman-Keuls) test to determine the difference between groups. Real-time PCR data were nonparametric and expressed as median. Thus, the Wilcoxon test was used to compare pair-matched medians. All data analyses were performed by using Prism (Version 5; GraphPad Software Inc., San Diego, CA). P values < 0.05 were considered to indicate a significant difference.
HK-1 (American Peptides, Sunnyvale, CA) and SP (Auspep, Melbourne, Australia) were reconstituted in 0.01 M acetic acid containing 1% β-mercaptoethanol and stored as aliquots at −20°C. The PGE2 ELISA kit was purchased from R&D System, the DNase treatment kit was from Promega (Madison, WI), and the TRIzol kit and SuperScriptIII cDNA Synthesis Kit were from Invitrogen. Maxima SYBR Green qPCR Master Mix was purchased from Fermentas. Indomethacin and TTX were purchased from Sigma-Aldrich (Sydney, Australia). NS-398, and SC-560 (Cayman Chemical, Ann Arbor, MI) were dissolved in dimethyl sulfoxide. Antibodies for COX-1 (monoclonal), COX-2 (monoclonal; polyclonal), and 15-PGDH (polyclonal) were obtained from Cayman Chemical. Anti-GAPDH antibody (polyclonal) and secondary antibodies (anti-mouse IgG; anti-rabbit IgG) for Western blot were purchased from Sigma-Aldrich, and biotinylated secondary antibodies (anti-mouse IgG; anti-rabbit IgG) for immunohistochemistry were from Vector Laboratories (Burlingame, CA). SR140333 and SR48968 were gifts from Sanofi-Synthélabo Recherche (Montpellier, France). All drug solutions were stored as aliquots at −20°C, except NS-398, which was stored at room temperature. All other reagents were of analytical grade.
Time Course of HK-1-Induced PGE2 Production in Mucosa and Muscle.
To investigate whether HK-1 causes PGE2 production in colonic muscle and mucosa, paired tissue explants were incubated with or without HK-1 (0.1 μM) for 0 h to 8 h. HK-1 induced PGE2 release into bath fluid in a time-dependent manner (Fig. 1). There was a trend for an increase in PGE2 release with the time in unstimulated tissues, but it was not significant. After incubation with HK-1 for 1 and 2 h, the PGE2 levels were no different from unstimulated paired time controls. In the mucosa, PGE2 levels showed a 10-fold increase (p < 0.01) after 4 h and a 2-fold increase (p < 0.05) after 8 h compared with paired time controls (Fig. 1A). In muscle, there was a 3.5-fold increase in PGE2 release (p < 0.01) after 4 h, whereas there was no significant change after 8 h (Fig. 1B). Thus, 4 h was considered as the optimal time point and was used for all subsequent experiments for mucosa and muscle.
Concentration-Response Relationship of HK-1-Induced PGE2 Production in Mucosa and Muscle.
After 4-h incubation with HK-1, PGE2 release from both mucosa and muscle showed a bell-shape curve in response to HK-1 in the range of 0.1 pM to 10 μM, with the maximum response at 0.1 μM HK-1 (Fig. 2). In mucosal explants, the effects of HK-1 at 0.01, 0.1, and 1 μM (Fig. 2A) were significantly different from the nonstimulated control, whereas the effect of HK-1 in muscle was statistically different from control only at 0.1 μM (Fig. 2B). In all subsequent experiments, 0.1 μM was used as the optimal concentration of HK-1.
Effect of Inhibitors on Responses to HK-1 and SP in Mucosal Explants.
First, we investigated whether PGE2 production depended on the activation of tachykinin receptors. HK-1-induced PGE2 production was inhibited to a similar extent by both the NK1 receptor antagonist SR140333 (44% of inhibition) and the NK2 receptor antagonist SR48968 (51% of inhibition) (Fig. 3A). In parallel, SP (0.1 μM) also stimulated PGE2 production, and this was inhibited by both SR140333 and SR48968 in a virtually identical manner compared with HK-1 (Fig. 3B). The NK1 and NK2 antagonists alone did not affect the basal level of PGE2 release at 4 h (n = 7; data not shown).
To determine which COX isoform was involved in HK-1- and SP-induced PGE2 production, mucosal strips were pretreated with the nonselective COX inhibitor indomethacin, the COX-1-selective inhibitor SC-560, or the COX-2-selective inhibitor NS-398. Both HK-1- and SP-induced PGE2 production was unaffected by SC-560 (Fig. 4). However, it was notable that indomethacin (n = 8; p < 0.05) and NS-398 (n = 8; p < 0.05) not only blocked, but significantly reduced, HK-1-stimulated PGE2 release to below unstimulated baseline levels (Fig. 4).
To determine the role of neurons in PGE2 release, mucosal explants were incubated in the presence and absence of TTX (1 μM). TTX alone significantly increased PGE2 release by 217 ± 78% over the basal level (n = 8; p < 0.05), but TTX did not affect HK-1-induced PGE2 production (205 ± 138%; n = 8).
HK-1-Regulated COX and 15-PGDH Gene Expression in Mucosal and Muscle Explants.
Paired explants were incubated in the presence and absence of HK-1 (0.1 μM) for 4 h, and expression of COX-1, COX-2, and 15-PGDH mRNA in the explants was then determined by using real-time reverse transcription-PCR. Although HK-1 had no effect on the levels of COX-1 mRNA (Fig. 5, A and B), it enhanced the expression of COX-2 mRNA by 2.4- and 1.5-fold in mucosal and muscle explants, respectively (Fig. 5, C and D) compared with nonstimulated paired controls. No significant up-regulation of COX-2 expression induced by HK-1 was seen at earlier incubation time points (data not shown).
Exposure of colonic mucosa to HK-1 (0.1 μM) significantly reduced 15-PGDH mRNA expression by 1.7-fold compared with nonstimulated control samples (Fig. 5E). We also attempted to demonstrate the expression of 15-PGDH mRNA in muscle explants, but expression was too low (data not shown).
HK-1-Regulated COX and 15-PGDH Protein Expression in Mucosal Explants.
Western blot was used to determine the protein expression of COX-1, COX-2, and 15-PGDH in paired mucosal explants after incubation with or without HK-1 (0.1 μM) for 4 h (Fig. 6A). COX-1 protein expression was unaltered by HK-1, but COX-2 expression was significantly increased with a 2.7-fold change (Fig. 6). It is noteworthy that HK-1 caused a 1.6-fold reduction (p < 0.05) of 15-PGDH protein expression (Fig. 6).
Localization of COX-1, COX-2, and 15-PGDH Immunoreactivity in Unstimulated Normal Human Colon.
At the antibody concentrations used here, the IR of COX-1 in epithelial, stromal, intestinal smooth muscle, and vascular smooth muscle cells was more pronounced than that of COX-2 and 15-PGDH (Table 1; Figs. 7 and 8). COX-1 IR was widespread in human colonic mucosa, muscle, and ganglia, as well as in vascular smooth muscle (Fig. 7). COX-2 IR was generally weaker than COX-1, except in submucosal ganglia where intensity was similar. In the epithelium, COX-2 IR was localized primarily at the cell membrane, whereas COX-1 IR showed cytoplasmic and nuclear expression (Figs. 7 and 8).
Immunostaining for 15-PGDH was mainly cytoplasmic, of moderate intensity similar to that of COX-2. 15-PGDH staining was highly localized, to regions of the mucosa and submucosal and myenteric ganglia (Figs. 7 and 8). It is noteworthy that nuclear staining of 15-PGDH was also observed in epithelial cells (Figs. 7C and 8C) and myenteric ganglia (Fig. 7L). Furthermore, weak 15-PGDH staining was seen on goblet cells (Fig. 8C).
In blood vessels, prominent immunostaining of granulocyte- and monocyte-like cells was observed with COX-1, whereas inmmunostaining with COX-2 and 15-PGDH was moderate. Platelets showed immunoreactivity for all three enzymes (Fig. 8, G–I).
HK-1-Induced Alteration of COX and 15-PGDH Immunoreactivity in Human Colonic Mucosa.
The immunoreactive staining for COX-1, COX-2, and 15-PGDH seemed less dense in the surface epithelial cells of HK-1-treated tissue explants compared with nontreated control tissues (Fig. 9). COX-1 IR in other mucosal structures seemed unaffected by HK-1 (Fig. 9, A and B). HK-1 evoked a clear increase of COX-2 IR and reduction of 15-PGDH IR in crypt epithelial cells and immune cells in lamina propria (Fig. 9, C–F), whereas changes were less noticeable in smooth muscle and submucosal regions.
SP and the NK1 receptor are up-regulated in IBD human intestine and in experimental animal models of IBD (Goode et al., 2000; Liu et al., 2011). Studies also demonstrate that PGE2 plays an important role in the pathogenesis of IBD (Sheibanie et al., 2007). HK-1 has been implicated in immune/inflammation, but little has been reported about its effects in the human colon. In the current work, we investigated the effect of human HK-1 on PGE2 production in the normal human colon and precisely defined its actions on COX and 15-PGDH. To our knowledge, this is the first study simultaneously evaluating the proinflammatory effects of HK-1 and SP in normal human gut. This extends previous observations showing that exogenous SP stimulates PGE2 production in human colonic epithelial cells (Koon et al., 2006) and provides an original finding that, similar to SP, HK-1 has the ability to induce PGE2 release in human colon.
The effect of HK-1 to enhance PGE2 release in both colonic mucosa and muscle was time- and concentration-dependent. Unusually, a bell-shaped concentration curve was observed, with the maximum response at 0.1 μM HK-1. A similar phenomenon has been seen in studies with SP in human polymorphonuclear leukocytes (Gallicchio et al., 2009). This bell-shape response indicates that at higher concentrations desensitization may occur, or tachykinins may bind to other tachykinin receptor subtypes or activate different mechanisms that have inhibitory effects on PGE2 production. Although significant PGE2 production was elicited in both muscle and mucosa, the greater effects in mucosa (10-fold change) suggest a prominent role for HK-1 in that location. In colonic inflammation, immune cells in the lamina propria could be the primary source of HK-1, because endogenous HK-1 is preferentially expressed in lymphocytes and macrophages in mice (Zhang et al., 2000; Zhang and Paige, 2003) and humans (Klassert et al., 2008). This is in line with increased expression of the TAC4 gene encoding HK-1 in colonic mucosa in ulcerative colitis (Liu et al., 2011).
HK-1 shows strong preference for the NK1 compared with NK2 or NK3 receptors (Bellucci et al., 2002). Our present findings show that HK-1 and SP signal via both NK1 and NK2 receptors to induce PGE2 release in colonic mucosa. It has been previously reported that SP/NK1 receptors causes PGE2 release from rat intrapulmonary bronchi (Szarek et al., 1998) and porcine jejunum (Thorbøll et al., 1998). Another study has demonstrated that SP stimulates PGE2 release in human colonic epithelial cells stably transfected with NK1 receptors (Koon et al., 2006), but such interactions have not been extensively studied in native human intestine. The pathological implication of NK1 receptor in inflammation has been demonstrated by the beneficial effects of NK1 antagonists in experimental models of colitis (Cutrufo et al., 1999; Di Sebastiano et al., 1999) and by our previous work showing an up-regulation of TACR1 gene in colonic mucosa of patients with IBD (Liu et al., 2011).
In addition to showing the activation of the NK1 receptor pathway, we provide new findings that NK2 receptors also influence SP- and HK-1-evoked PGE2 production in the human colon. A relationship between NK2 receptors and PGE2 release has been previously shown in alveolar macrophages (Brunelleschi et al., 1992), urinary bladder (Tramontana et al., 2000), and normal human colon (Burcher et al., 2008). A relationship between NK2 receptors and inflammation was shown in human IBD by Renzi et al. (2000) who demonstrated marked increases in the expression of NK1 receptors in epithelia and vascular endothelia, whereas NK2 receptors were increased in inflammatory cells of lamina propria and activated eosinophils around mucosal crypts. The simultaneous involvement of both NK1 and NK2 receptors has also been reported in animal models of inflammation (Cutrufo et al., 1999) and hypersecretion (Turvill et al., 2000). Our study showing the active role of NK1 and NK2 receptors in PGE2 production may help to explain the significance of up-regulated NK1 and NK2 receptors in the pathophysiology of IBD. Although NK1 and NK2 receptors have significant roles in gut pathophysiology, few drugs targeting these individual receptors have entered the clinic for human use. A promising NK2 receptor antagonist is being tested for irritable bowel syndrome and may have an effect in inflammation (Quartara et al., 2009). Our study implies that a combined use of both receptor antagonists may be more therapeutically beneficial.
The steady-state cellular levels of PGE2 depend on the relative rate of COX-dependent biosynthesis and 15-PGDH-mediated catabolism. In this study, we demonstrated that HK-1-induced PGE2 release requires the involvement of COX-2 rather than COX-1. Furthermore, HK-1 induces PGE2 release at the same concentrations and incubation times found to induce COX-2 gene and protein expression, which were both up-regulated to a similar extent. Strikingly, we have also provided new insights on the effect of HK-1 on expression of 15-PGDH, the key enzyme responsible for PGE2 inactivation. Coupled with the marked up-regulation of COX-2, HK-1 significantly inhibited 15-PGDH expression at both the mRNA and protein levels. This finding suggests that HK-1 may have dual effects on PGE2 production, not only by increased PGE2 synthesis through up-regulation of COX-2, but also by decreased 15-PGDH expression and, hence, reduced degradation of PGE2. This, we believe, is the most novel finding of our study. This is in line with a previous study showing a marked reduction of 15-PGDH in colonic mucosa from patients with IBD (Otani et al., 2006).
Our immunohistochemical data in colonic mucosa showed COX-2 as well as COX-1 expression primarily in epithelial cells (upper crypts) and immune cells in lamina propria, with some immunoreactivity also localized to submucosal ganglia. They also revealed that 15-PGDH has a similar pattern of immunoreactivity as COX-1 and COX-2 and represent another novel finding that 15-PGDH is also expressed in neuronal regions. In line with our findings that HK-1 increased PGE2 in muscle explants, we also found expression of COX-1, COX-2, and 15-PGDH in myenteric ganglia, as well as weaker expression in longitudinal and circular muscle. Our previous studies have shown that indomethacin enhances contractile responses to NKA in human colonic muscle strips from control and diverticular disease patients, suggesting NK2 receptor-stimulated release of prostanoids that relax smooth muscle (Burcher et al., 2008). COX-1 and COX-2 have been shown to be present in the neuromuscular compartment of human colon where they modulate cholinergic excitatory control of colonic motility (Fornai et al., 2005). Further studies using confocal microscopy are necessary to confirm which neuronal and other markers are colocalized with COX and 15-PGDH. We were surprised to find that our data showed that TTX enhanced basal PGE2 release, but had no effect on HK-1-stimulated PGE2 production. This suggests that there is an inhibitory neuronal component that can reduce PGE2 release. However, the lack of effect of TTX on HK-1-induced response is suggestive of epithelial and immune cells as the major cellular source of HK-1-COX-2-derived PGE2.
In summary, our study has demonstrated a novel role for HK-1 as an important stimulant for PGE2 release in human colon and explored the mechanism involved. This is the first demonstration that HK-1, acting via both NK1 and NK2 receptors, up-regulates COX-2 expression, an effect shared by SP. We further demonstrate that HK-1-induces 15-PGDH down-regulation, and this represents an additional mechanism by which HK-1 increases PGE2 production. This study reveals a mechanistic link between HK-1 and PGE2 in humans and provides new insights into a potential functional role of HK-1 in the pathophysiology of colonic inflammation.
Participated in research design: Dai and Liu.
Conducted experiments: Dai.
Contributed new reagents or analytic tools: Perera and King.
Performed data analysis: Dai, Burcher, and Liu.
Wrote or contributed to the writing of the manuscript: Dai, Southwell, Burcher, and Liu.
We thank Changfa Qu and Jason Le for specimen collection and Fei Shang and the Histology and Microscopy Unit at the University of New South Wales for technical support.
This study was supported by the National Health and Medical Research Council of Australia [Grant ID 568861]. B.R.S. was supported by the Victorian Government's Operational Infrastructure Support Program.
This article represents partial fulfillment for L.D.'s Ph.D thesis in pharmacology.
Article, publication date, and citation information can be found at http://jpet.aspetjournals.org.
- substance P
- prostaglandin E2
- 15-hydroxyprostaglandin dehydrogenase
- inflammatory bowel disease
- glyceraldehyde-3-phosphate dehydrogenase
- housekeeping gene
- analysis of variance
- enzyme-linked immunosorbent assay
- polymerase chain reaction
- quantitative PCR
- forward primer
- reverse primer
- (S)1–2-[3-(3,4-dichlorophenyl)-1-(3-isopropoxyphenylacetyl)piperidin-3-yl]ethyl-4-phenyl-l azonia-bicyclo[2.2.2]octane
- Received July 25, 2011.
- Accepted September 27, 2011.
- Copyright © 2012 by The American Society for Pharmacology and Experimental Therapeutics