SEARCH

SEARCH BY CITATION

Keywords:

  • colitis;
  • ion transport;
  • PKR1;
  • PROK2;
  • prokineticin;
  • visceral pain

Abstract

  1. Top of page
  2. Abstract
  3. Materials and methods
  4. Results
  5. Discussion
  6. Acknowledgments and Disclosures
  7. Author Contribution
  8. References
  9. Supporting Information

Background  Prokineticin 2 (PROK2) is an inflammatory cytokine-like molecule expressed predominantly by macrophages and neutrophils infiltrating sites of tissue damage. Given the established role of prokineticin signaling on gastrointestinal function, we have explored Prok2 gene expression in inflammatory conditions of the gastrointestinal tract and assessed the possible consequences on gut physiology.

Methods  Prokineticin expression was examined in normal and colitic tissues using qPCR and immunohistochemistry. Functional responses to PROK2 were studied using calcium imaging and a novel antagonist, Compound 3, used to determine the role of PROK2 and prokineticin receptors in inflammatory visceral pain and ion transport.

Key Results Prok2 gene expression was up-regulated in biopsy samples from ulcerative colitis patients, and similar elevations were observed in rodent models of inflammatory colitis. Prokineticin receptor 1 (PKR1) was localized to the enteric neurons and extrinsic sensory neurons, whereas Pkr2 expression was restricted to sensory ganglia. In rats, PROK2-increased intracellular calcium levels in cultured enteric and dorsal root ganglia neurons, which was blocked by Compound 3. Moreover, PROK2 acting at prokineticin receptors stimulated intrinsic neuronally mediated ion transport in rat ileal mucosa. In vivo, Compound 3 reversed intracolonic mustard oil-induced referred allodynia and TNBS-induced visceral hypersensitivity, but not non-inflammatory, stress-induced visceral pain.

Conclusions & Inferences  Elevated Prok2 levels, as a consequence of gastrointestinal tract inflammation, induce visceral pain via prokineticin receptors. This observation, together with the finding that PROK2 can modulate intestinal ion transport, raises the possibility that inhibitors of PROK2 signaling may have clinical utility in gastrointestinal disorders, such as irritable bowel syndrome and inflammatory bowel disease.

Abbreviations:
UC

ulcerative colitis

MO

mustard oil

CRD

colorectal distension

DRG

dorsal root ganglia

GI

gastrointestinal

WAS

water avoidance stress

VMR

visceromotor response

TTX

tetrodotoxin

IBD

inflammatory bowel disease

IBS

irritable bowel syndrome

PROK1 (EG-VEGF) and PROK2 (Bv8) were first isolated during studies aimed at identifying biologically active peptides from black mamba venom and frog skin.1–3 Orthologous genes were identified in a range of species, including man, where they encode small peptides of around 80–90 amino acids containing a highly conserved N-terminal AVITGA motif that is critical for their biologic activity.4–6 The early observation that these peptides caused contraction of isolated GI tract led to them being termed prokineticins,4 and considerable progress has since been made in expanding our understanding of prokineticin signaling, both at the molecular and physiologic level (reviewed in Refs.7,8).

The activity of the prokineticins is now known to be mediated by two G-protein coupled receptors, PKR1 and PKR2, which are similarly conserved across species.9–11In vitro, binding and functional assays have shown that PROK1 and PROK2 are capable of binding to and signaling through both receptors via Gq, Gi, and Gs pathways.12 As a consequence of this redundancy, prokineticin signaling in vivo is thought to depend on tissue-specific expression of the ligands, receptors, and associated G-proteins. Indeed, while prokineticins have been shown to be expressed in a wide range of overlapping adult and embryonic tissues, numerous examples of tissue-specific differences in prokineticin expression have emerged in recent years.13 For example, PROK2, but not PROK1, expression appears to predominate in activated immune cells.14

Although the role of prokineticin signaling has been elaborated in a number of diverse physiologic processes, including inflammation, haematopoeisis, angiogenesis, circadian rhythms, and reproduction (reviewed in Refs.7,15,16), questions still remain about the importance of these peptides in regulating the function of the gastrointestinal tract, the organ in which they were first shown to have an effect. Research on the potential role of prokineticin signaling in the GI tract has largely focused on their regulation of GI motility through empirical observations on the effect of exogenously applied peptide ligands. Thus, ex vivo, PROK1 and PROK2, and their non-mammalian homologs, have been shown to modulate contraction of smooth muscle derived from various regions of the GI tract,1–4,17 although no effect on gastric or colonic contractility was observed by Bassil et al.,18 and in vivo, PROK1 has recently been shown to stimulate upper GI transit.19 While such findings have led to the suggestion that prokineticins, or modulators of prokineticin signaling, may represent novel therapeutic agents for gastrointestinal motility disorders,1 only limited work has directly explored alterations in prokineticin signaling in GI pathophysiology.20,21

In this study, we have observed a marked increase in Prok2 gene expression in the GI tract during inflammation, both in biopsy samples taken from patients diagnosed with ulcerative colitis (UC) and in tissue samples taken from various preclinical models of colitis. We demonstrate that elevated levels of PROK2 can elicit effects on visceral sensory perception and ileal ion transport, which together suggest that inhibitors of prokineticin signaling may have clinical utility in inflammatory GI disorders.

Materials and methods

  1. Top of page
  2. Abstract
  3. Materials and methods
  4. Results
  5. Discussion
  6. Acknowledgments and Disclosures
  7. Author Contribution
  8. References
  9. Supporting Information

Animals

Male C57BL/6 mice and female C3H/HeN mice (20–25 g) (Harlan, UK) were used in the mustard oil and Citrobacter rodentium models, respectively. Male Sprague–Dawley rats (200–250 g) were used for the TNBS and water-avoidance stress models. All animals were housed in groups of six (rats) or eight (mice) in an animal care unit with normal day-night lighting conditions. All the studies reported here were approved by a local ethical review process and conformed to the UK Animals (Scientific Procedures) Act 1986. Animals had free access to normal chow and water throughout the experiments, unless stated otherwise.

Real time PCRs

RNA and cDNA were prepared according to standard protocols, and qRT-PCR was performed essentially as described in.22 Detailed methodologies can be found in the Supplementary Materials and Methods section.

Immunohistochemistry

Formalin-fixed, wax-embedded tissues with no overt signs of pathology were purchased from Asterand (Detroit, MI, USA); informed consent had been sought and received from all donors. All tissues were used in accordance with the Human Tissue Act 2004 (UK). The samples were processed and stained according to standard protocols either manually or using a Ventana Discovery XT (Ventana Medical Systems Inc., Tucson, AZ, USA). In the case of manual fluorescent IHC, following de-waxing and rehydration, tissue sections were subject to antigen retrieval in citrate buffer (DAKO, Ely, UK) according to the manufacturer’s recommended protocol. The sections were subsequently washed three times in PBS prior to being blocked in PBS containing 5% donkey serum, 1% BSA, and 0.3% triton X100. The sections were then incubated overnight at 4 °C in PBS containing 1% BSA plus 1 : 100 dilution of rabbit anti-PKR1 (ab13084 – Abcam, Cambridge, UK) and, if appropriate, 1 : 1000 dilution of chicken anti-NF200 (ab4680 – Abcam). Sections were then rinsed three times in PBS before being incubated in FITC and Cy3-conjugated secondary antibodies (Jackson ImmunoResearch) for 3 h at 4 °C. Finally, the sections were rinsed for a further three times in PBS before being mounted on Vectashield containing DAPI. Two-color chromogenic IHC was performed on tissue sections using the Ventana Discovery XT, essentially according to the manufacturer’s recommended protocol using a guinea pig anti-TRPV1 antibody (GP14100 – Neuromics, Edina, MN, USA) in addition to the PKR1 antibody described above. Sections were subject to antigen retrieval in (citrate buffer, DAKO) and peroxidase block (S2023, DAKO) prior to staining. The DAB component of the staining was performed manually using a DAB substrate kit (SK4100, Vector Labs, Burlingame, CA, USA) to allow greater control of color development.l

Intracellular calcium imaging of neurons

DRG neurons were prepared from adult rats using methods described elsewhere,23 and cultured in minimum essential medium (MEM, 1x) with Earles salts, l-glutamine, supplemented with 10% fetal bovine serum, 1% Penicillin/Streptomycin (Invitrogen, Paisley, UK), and 50 ng mL−1 Nerve Growth Factor (Promega, Madison, WI, USA). Myenteric neuronal cultures were prepared from rat tissue as described previously.22 Neurons grown on 28 mm glass coverslips were loaded with Fluo-4AM (Invitrogen) at 37 °C for 1 h in HEPES buffer saline solution (HBSS) containing 140 mmol L−1 NaCl, 5 mmol L−1 KCl, 10 mmol L−1 HEPES, 10 mmol L−1 Glucose, 1 mmol L−1 MgCl2, 2 mmol L−1 CaCl2, pH7.4. Images were acquired every second through an Olympus IX81 microscope, using CellR software and a CCD Hamamatsu ORCA-ER camera. Regions of interest were selected off-line, once the neuronal status of the cell had been confirmed by depolarizing the cells with 75 mmol L−1 K+ at the end of each experiment. Fluorescence intensity was expressed as the fluorescence at each time point relative to basal levels (F/F0).

Ussing’s chambers

Short circuit current (Isc) in isolated ileal mucosa in response to stimulation by PROK2 was measured in Ussing’s chambers according to standard techniques. Terminal ileum was isolated from adult C57BL/6 male mice and rinsed in Krebs Henseliet Solution (KHS; 117 mmol L−1 NaCl, 25 mmol L−1 NaHCO3, 4.7 mmol L−1 KCl, 2.5 mmol L−1 CaCl2, 1.2 mmol L−1 MgSO4, 1.2 mmol L−1 KH2PO4, and 11.1 mmol L−1 glucose). The longitudinal and circular muscle layers were removed by dissection, and the mucosa was mounted on an Ussing’s chamber with a 3 mm window diameter (Harvard Apparatus, Edenbridge, UK). The apical side of the mucosa was bathed in modified KHS with equimolar amounts of mannitol replacing glucose, and the basolateral side was bathed in KHS. Both sides were continually circulated through the half chambers, maintained at 37 °C, and continuously bubbled with 95% O2/5% CO2. Short circuit currents (Isc) were measured by a standard procedure using a dual voltage clamp (World Precision Instruments, Sarasota, FL, USA), with current compensation for series resistance between the voltage-sensing electrodes. The trans-epithelial voltage was temporarily set at 2 mV above zero every few minutes. The Isc were continuously recorded using a PowerLab 8/30 (AD Instruments, Oxford, UK). Half maximal effects were determined using GraphPad Prism 4 software (GraphPad Software, CA, USA).

Animal models

MO and TNBS models  Models of mustard oil (MO)-induced visceral allodynia and TNBS-induced visceral hypersensitivity were performed essentially as described in Refs. (24) and (25). Briefly, in the case of MO-induced visceral allodynia, MO was administered intracolonically to mice that had been fasted overnight. Withdrawal thresholds to application of von Frey filaments to the abdomen were measured as a test for referred allodynia prior to (baseline) and 48 h following MO administration. The lowest amount of force required to elicit a response was recorded as withdrawal threshold (in grams). For TNBS-induced visceral hypersensitivity, TNBS or vehicle was administered intracolonically to habituated rats. After 7 days, and following an overnight fast, the animals were placed in a Bollman cage, and colorectal distension (CRD) was performed by graded increases in intensity of phasic isobaric distensions (baseline = 0, 10, 20, 30, 40, 50, and 60 mmHg, 2 min each) using a computer-controlled barostat system. The EMG activity of the abdominal musculature was recorded, and the visceromotor response (VMR) to distension was calculated as the area under the curve and expressed as a percentage of the maximal response at 60 mmHg. More detailed information on both models can be found in the Supplementary Materials and Methods section.

Water-avoidance stress model  Stress-induced visceral hypersensitivity was induced using a water-avoidance stress (WAS) paradigm. Animals were either placed in a high walled cage with a centrally placed 6 cm3 platform (sham stress) for 1 h or placed in an identical cage with water (at room temperature) filled up to 1 cm from the top of the platform for 1 h. Twenty-four hours following either sham stress or WAS, visceral hypersensitivity was assessed using CRD as described briefly above, and in more detail in the Supplementary Materials and Methods.

Citrobacter rodentium infection model  Mice were infected with Citrobacter rodentium as described elsewhere.26 Briefly, C. rodentium ICC 169 were grown overnight in Luria-Bertani broth supplemented with 50 μg mL−1 nalidixic acid. Forty milliliters of bacteria were pelleted by centrifugation and resuspended in 2 mL of sterile PBS. Mice were then gavaged with 200 μL of the bacterial suspension or with PBS alone, culled at 9 days post infection, and colons were removed for further processing.

Drugs

ω-conotoxin GVIA and tetrodotoxin citrate were obtained from Tocris. Capsaicin, TNBS, carbachol, mustard oil, histamine, and diclofenac were obtained from Sigma (Dorset, UK). PROK2, DSS, and morphine were obtained from Fitzgerald Industries International (Acton, MA, USA), MP Biomedicals (Illkirch, France), and Macfarlan Smith (Edinburgh, UK), respectively. Compound 3 and alosetron were synthesized in-house. ω-conotoxin GVIA, tetrodotoxin citrate, and DSS were prepared in water. TNBS and mustard oil were prepared in 50% and 70% ethanol, respectively. Diclofenac and capsaicin were prepared in 0.5% methylcellulose and DMSO, respectively. Morphine and alosetron were prepared in saline, and Compound 3 was prepared in 25% cremophor/saline.

Statistical analysis

All statistical analyses were conducted using Systat (Cranes Software International (Bangalore, India) and Origin 7.5 (Origin Lab, Northampton, MA, USA) software. Data are expressed as the mean ± SEM, and were subjected to Student’s t-tests, the Mann–Whitney U-test, or a 1-way anova followed by a Dunnett post test, as described in the figure legends. A P value of ≤0.05 was considered statistically significant with *indicating < 0.05, indicating < 0.01, and indicating < 0.001.

Results

  1. Top of page
  2. Abstract
  3. Materials and methods
  4. Results
  5. Discussion
  6. Acknowledgments and Disclosures
  7. Author Contribution
  8. References
  9. Supporting Information

Prokineticin 2 expression in visceral inflammation

A marked up-regulation of Prok2 gene expression was seen in colonic mucosal biopsy samples derived from patients with active UC compared with control samples taken from comparable regions of the gut, but which showed no overt signs of pathology (Fig. 1A). Levels of Prok2 positively correlated with IL-1β expression, (Fig. 1B) suggesting that increased Prok2 levels were a direct consequence of inflammation. To explore this further, expression of the Prok2 gene in hapten-, chemical-, and infectious-inflammatory preclinical models was monitored (Fig. 2). In untreated control animals, Prok2 transcripts were virtually undetectable in full thickness colon samples in all the models. TNBS-induced colitis in rats resulted in a marked increase (>7 fold) in expression of Prok2 after 7 days, (Fig. 2A) which was in contrast to relatively modest changes in Prok1 and Pkr1 levels (Supplementary Fig. S1). During acute DSS or MO-induced inflammatory colitis, Prok2 levels also increased markedly in line with the onset of colitis as denoted by an increase in IL-1β expression (Fig. 2B and C). Prok1 and Pkr1 levels remained unaltered in both cases (Supplementary Fig. S1). Following C. rodentium infection of C3H/HeN mice, colitis slowly developed with peak levels of bacterial shedding and weight loss observed by day 11 post infection (data not shown). Mice culled on day 9, when symptoms of infection first manifest, showed marked increases in Prok2 transcript levels, with expression levels again correlating with IL-1β levels (Fig. 2D).

image

Figure 1. Prok2 and IL-1β mRNA expression in colonic biopsies were quantified using TissueScan qRT-PCR panels (Origene). (A) Comparison of Prok2 expression in biopsies from ulcerative colitis patients and non-inflammatory (normal) controls. Values are mean ± SEM from n ≥ 8 donors per group. < 0.01 compared with controls using Mann–Whitney U-test. (B) Correlation of Prok2 transcript levels with levels of IL-1β in matched samples. CT (cycle threshold) values for each sample were normalized against β-actin levels.

Download figure to PowerPoint

image

Figure 2.  Expression of prokineticin 2 and IL-1β mRNA in animal models of colitis. Intra-colonic instillation of TNBS (A) or mustard oil (C), supplementation of the drinking water with DSS (B), or oral gavage of Citrobacter rodentium (D), was used to induce colitis. Full thickness samples of colon near the site of chemical instillation (TNBS and MO models) or tissue from distal colon (DSS or C. rodentium models) were analyzed for Prok2 and IL-1β mRNA expression using qRT-PCR analysis. Black columns represent Prok2 expression, and red points/lines denote IL-1β expression. Values are mean ± SEM from n = 6–8 donors per group. < 0.01 compared with controls using Student’s t-test. CT, cycle threshold; Veh, vehicle; D2, day 2; MO, mustard oil; C. rodentium, Citrobacter rodentium.

Download figure to PowerPoint

PKR1 expression in intrinsic enteric neurons and extrinsic sensory neurons

To better understand the potential physiologic significance of the elevated Prok2 expression seen in colitis, studies were undertaken to explore the expression pattern of its receptors, PKR1 and PKR2. PKR1 expression was readily detectable in normal human GI tract tissue, with staining apparent in the myenteric and submucosal ganglia throughout the stomach, small intestine, and colon (Fig. 3A). No overt muscle/mucosal staining was observed in any of the samples. In these studies, no staining was observed if equivalent amounts of a non-specific IgG were used in place of the primary antibodies, or if the primary antibodies were simply omitted from the experiment (data not shown). Analysis of Pkr1 mRNA expression using real time PCR not only confirmed expression of the gene in human GI tissues but also revealed expression in a variety of other tissues, including nodose and dorsal root ganglia (Fig. 3B). The latter observation is consistent with previous reports of Pkr1 mRNA expression in murine DRG27,28 and with our own observations of PKR1 immunostaining in rat DRG, where it appears to be co-expressed with TRPV1 in small/medium diameter cell bodies (Fig. 3D and E). In contrast, Pkr2 was found to be highly expressed in the brain, DRG, bone marrow, spleen, and testis only. Expression levels in other tissues, including the GI tract were much lower (Fig. 3C). In rat DRG, Pkr2 expression was observed in cell bodies (Fig. 3F). Taken together, these data suggest that elevated PROK2 levels in inflammation could activate PKR1 expressed in enteric neurons and extrinsic sensory neurons, and possibly PKR2 in extrinsic sensory neurons, to affect multiple aspects of GI physiology including ion transport and pain sensation.

image

Figure 3.  Expression of prokineticin receptors in human and rat tissues. (A) Immunohistochemical analysis of PKR1 expression in the myenteric and submucosal ganglia of human stomach, ileum, and colon. Pkr1 (B) and Pkr2 (C) expression in a range of human tissues determined using qRT-PCR. (D, E) Immunohistochemical of PKR1 distribution in rat DRG showing an absence of expression in large diameter, NF200 positive neurons (D), but co-expression with TRPV1 in presumptive nociceptive sensory neurons (E). (F) Distribution of Pkr2 mRNA by in situ hybridization in rat DRG. The IHC/ISH images shown are representative of the data obtained from n = 3 donors/animals.

Download figure to PowerPoint

PROK2 signaling promotes enteric-mediated ion transport in the ileum

In light of the above observations, the effect of prokineticin signaling on murine ileal ion transport was determined using a novel prokineticin receptor antagonist, 6-[(2-Amino-pyridin-3-ylmethyl)-amino]-1,3-bis-(4-methoxy-benzyl)-1H-[1,3,5] triazine-2,4-dione (hereafter referred to as Compound 3).29In vitro, PROK2 induced increases in intracellular Ca2+([Ca2+]i) in both hPKR1 and hPKR2 transfected CHO cells (EC50 = 9.1 ± 3.3 nmol L−1 and 18.4 ± 6.7 nmol L−1 respectively) (Supplementary Fig. S2A). Compound 3 blocked activation of hPKR1 by PROK2 with an IC50 = 0.5 ± 0.1 μmol L−1, and was similarly active (IC50 = 0.7 ± 0.1 μmol L−1) at hPKR2 (Supplementary Fig. S2B). No activity of Compound 3 was detected in a Novartis safety panel of 60 different receptors, proteases, and kinases at concentrations >10 μmol L−1, confirming its selectivity for prokineticin receptors (data not shown). PROK2 also stimulated increases in [Ca2+]i in cultured ileal myenteric neurons through a mechanism that was inhibited by Compound 3 (Fig. 4A). Addition of PROK2 to preparations of ileal mucosa also induced a concentration-dependent increase in Isc with an EC50 of 0.6 ± 0.1 nmol L−1, and a maximal response of 65.3 ± 10 μAmp cm−2 (Fig. 4B and C). The reduced responses observed at higher PROK2 concentrations probably result from desensitization during the cumulative addition protocol used, because, in separate experiments, the highest concentration (22.7 μmol L−1) induced a more marked ΔIsc when added alone (123.6 ± 12.5 μAmp cm−2, mean ± SEM, n = 5). Addition of Compound 3 (0.5 μmol L−1) 30 min prior to the addition of PROK2 increased the EC50 to 3.7 ± 1.1 nmol L−1 in contiguous preparations from the same animal (= 0.005) (Fig. 4D), but had no effect on the response to carbachol or histamine (Fig. 4E). Changes in Isc were induced via a neuronally driven pathway as pretreatment of the tissue with 1 μmol L−1ω-conotoxin GVIA led to a significant decrease in the Isc response to addition of 2.5 nmol L−1 PROK2 (Fig. 4F). Similarly, pretreatment with 100 nmol L−1 of TTX for 30 min reduced the PROK2-mediated Isc response by 64% (= 0.01, n = 5, data not shown). However, the Isc response was not affected by pretreatment with capsaicin (1 μmol L−1) for 30 min (Fig. 4F), which depletes extrinsic afferent terminals of neurotransmitter, suggesting the involvement of PKR1 expressed on enteric neurons.

image

Figure 4.  PROK2-mediated ion transport (Isc) in mouse ileal mucosa. (A) Effect of 5 μmol L−1 Compound 3 on PROK2-induced increases in [Ca2+]i in cultured myenteric neurons (n = 198–237). Data are expressed as the fluorescence at each time point relative to basal levels (F/F0). < 0.001 compared with control using Student t-test. (B) Representative trace of Isc in response to increasing PROK2 concentration and dose response curve (C) from n = 4 separate experiments. (D) Inhibition of PROK2-induced responses following pretreatment with 0.5 μmol L−1 of Compound 3. (E) Isc responses to carbachol and histamine were unaffected by pretreatment of the tissue with Compound 3. EC50 values are mean ± SEM from n = 4 experiments. (F) Effect of pretreatment with 1 μmol L−1ω-conotoxin GVIA or 1 μmol L−1 capsaicin on PROK2-induced Isc responses. Mean ± SEM data shown from n = 6 tissue samples per group. *< 0.05 compared with vehicle using Student’s paired t-test.

Download figure to PowerPoint

Prokineticin inhibition reverses inflammatory visceral pain

In primary cultures of rat DRG neurons, PROK2-induced a marked increase in [Ca2+]i in approximately 30% of the cells (Fig. 5A and C), and responses were inhibited by preincubation with the inhibitor Compound 3 (5 μmol L−1) (Fig. 5B and C). Compound 3 had no effect on the response of cultured DRG neurons to agonists of TRPV1 or TRPA1 (Fig. 5D and E). Together, these findings support the notion that functional prokineticin receptors are present in extrinsic sensory neurons, and that their activity can be blocked by Compound 3. To determine whether elevated Prok2 levels observed in colitis might play a role in visceral pain, Compound 3 was administered intravenously, 30 min prior to assessment of MO-induced referred allodynia to the abdomen. The PKR antagonist dose-dependently reversed the reduction in mechanical response thresholds to von Frey hairs, measured 48 h after MO instillation into the colon (Fig. 5F). Compound 3 (10 mg kg−1, i.v.) significantly increased the mean threshold response when compared with vehicle-treated animals to levels comparable with the positive control Alosetron (3 mg kg−1, p.o.). Diclofenac (30 mg kg−1, p.o.) had no effect on pain sensation in this model (Supplementary Fig. S3A). Compound 3 was similarly effective in a second model of inflammatory visceral pain, TNBS-induced colitis, where a significant and prolonged increase in the sensory response to mechanical distention of the distal colon is observed following TNBS instillation into the colon.30 In this model, Compound 3 (10 mg kg−1, i.v.), 30 min prior to initiation of CRD resulted in a statistically significant reversal of the TNBS-induced hypersensitivity in VMR to balloon distension at pressures of 50 and 60 mmHg (Fig. 5G). Morphine (1 mg kg−1, p.o.), included as a positive control, also produced significant reversal of TNBS-induced visceral hypersensitivity in this study, whereas Diclofenac (30 mg kg−1, p.o.) did not (Supplementary Fig. S3B).

image

Figure 5.  Antinociceptive effects of a prokineticin receptor antagonist in models of inflammation-induced visceral hypersensitivity. (A–C) Effect of PROK2 on [Ca2+]i in cultured DRG neurons (A, C) and the effect of preincubation with Compound 3 (B, C). Data are expressed as the fluorescence at each time point relative to basal levels (F/F0). Inserts show representative pseudocolor Ca2+ images before and after stimulation of cells with PROK2. Effect of 5 μmol L−1 Compound 3 on increases in [Ca2+]i induced by (D) the TRPV1 agonist capsaicin (1 μmol L−1, n > 75 from n = 4 separate experiments) or (E) the TRPA1 agonist MO (30 μmol L−1, n > 29 from three separate experiments). (F) Effect of Compound 3 (3–10 mg kg−1) on visceral allodynia induced by intracolonic administration of 0.25% mustard oil. Effects were monitored as changes in the withdrawal threshold in response to the application of von Frey filaments to the abdomen, 30 min after intravenous administration of Compound 3. Alosetron (3 mg kg−1, p.o.) was included as a positive control. (G) Effect of Compound 3 (10 mg kg−1, i.v.) on visceral hypersensitivity as assessed by the visceromotor response to colonic balloon distension in male rats, 7 days following intracolonic administration of 40 mg kg−1 TNBS. Morphine (1 mg kg−1, s.c.) was included as a positive control. Data are mean ± SEM from n = 6–8 animals per group. *< 0.05, < 0.01 and < 0.001 compared with vehicle using 1-way anova followed by Dunnett’s post test.

Download figure to PowerPoint

To investigate whether the PKR antagonist might also be efficacious in a non-inflammatory model of visceral pain, Compound 3 was tested in a model of acute WAS-induced visceral hypersensitivity.31 In this model, no increases in Prok2 transcript levels were observed (Fig. 6A), suggesting that a different mechanism underlies the visceral hypersensitivity compared with the acute inflammatory models. Visceral hypersensitivity was observed 24 h after exposure to WAS for 1 h and Compound 3 (10 mg kg−1, i.v.) 30 min prior to CRD failed to induce a statistically significant reversal of visceral pain in this model (Fig. 6B). In contrast, morphine (1 mg kg−1, s.c.) induced a complete reversal of the acute stress-induced elevated pain sensation.

image

Figure 6.  Effect of a prokineticin receptor antagonist in a model of stress-induced visceral hypersensitivity. (A) Expression of Prok2 and IL-1β mRNAs in the distal colon following water avoidance stress. Black columns represent Prok2 expression, whereas red points/lines denote IL-1β expression. (B) Effect of intravenous administration of Compound 3 (1–10 mg kg−1) on stress-induced visceral hypersensitivity to colonic balloon distension. Morphine (1 mg kg−1, s.c.) was included as a positive control. Data are mean ± SEM from n = 8 animals per group, and were analyzed using Student’s t-test (qRT-PCR) or 1-way anova followed by a Dunnett post test (WAS-induced visceral hypersensitivity), *< 0.05 and < 0.01. CT, cycle threshold; WAS, water avoidance stress; VMR, visceromotor response; Veh, vehicle.

Download figure to PowerPoint

Discussion

  1. Top of page
  2. Abstract
  3. Materials and methods
  4. Results
  5. Discussion
  6. Acknowledgments and Disclosures
  7. Author Contribution
  8. References
  9. Supporting Information

Although structurally and phylogenically distinct from other cytokines, recent data suggests that the prokineticins are functionally related to chemokines with a strong pro-inflammatory activity.15 PROK2, in particular, has been shown to be preferentially expressed by cells of monocyte–granulocyte lineage and has been observed at high levels in infiltrating neutrophils associated with appendicitis and tonsillitis.14 Moreover, PROK2 has been reported to induce haematopoietic responses in vivo and in vitro to increase neutrophil and monoctye cell counts and stimulate monocyte/macrophage migration, survival, and induce IL-1 and IL-12 secretion.14,32

To explore the role of PROK2 in inflammatory conditions of the GI tract, we investigated levels of Prok2 mRNA in biopsy samples taken from patients with active UC and observed a marked elevation in Prok2 expression that correlated with the extent of inflammation as judged by increased IL-1β expression. Similar changes in Prok2 mRNA expression in rodent models of chemically induced and infectious-colitis were also observed, with the onset of colitis in each case being accompanied by an increase in Prok2 expression, suggesting that it may be a key mediator of the inflammatory response. The origin of the increased Prok2 in the present study remains undefined because of a lack of suitable tools to explore its expression, but it could be due to either de novo synthesis in the GI tract or the infiltration of activated immune cells. Based on published data showing the expression of Prok2 in infiltrating neutrophils14 and activated macrophages,32 it is tempting to speculate that these cells, which are known to play major roles in colitic diseases, are the source of the increased levels of Prok2. Recent findings from Kimball et al.20 suggested an increase in the expression levels of Pkr1 and Prok1 in a model of mustard oil-induced colitis. Interestingly, our own data from the same model suggests that Pkr1 and Prok1 levels were unchanged in samples collected after 48 h. This most likely reflects differences in the experimental design as the increases described by Kimball et al., were transient, being maximal at 2–6 h and resolved within 24 h.

To assess the potential physiologic significance of these findings, we first investigated the distribution of prokineticin receptors in a range of tissues. In human and rodent tissues, prokineticin receptors are expressed in extrinsic sensory neurons, raising the possibility that elevated PROK2 may modulate sensory perception during periods of gut inflammation. Consistent with this, we found that a PKR antagonist was able to reverse pain behavior in two models of inflammatory visceral pain in which Prok2 levels are elevated. In contrast, Compound 3 was less effective at reversing pain sensation in a model of non-inflammatory visceral pain in which Prok2 levels remained low. These data on visceral pain are in line with earlier observations that PROK2 plays an important role in somatic nociception. For example, topical delivery of Bv8, the frog homolog of PROK2, to the rat paw results in a 50% decrease in the nociceptive threshold to pressure applied to the injected paw.33,34 In contrast, mice lacking PROK2 display a marked reduction in nociception induced by thermal and chemical stimuli.33 Giannini and colleagues also established a link between Prok2 expression in granulocytes and the onset and development of pain behaviors in a model of Freund’s adjuvant-induced paw inflammation.35

The exact mechanism by which elevated PROK2 might lead to visceral nociception remains to be elucidated. Several reports support the idea that PROK2 released by inflammatory cells may drive inflammatory pain by acting both as a chemoattractant for monocytes and macrophages and as a stimulator of the release of a range of proinflammatory and proalgesic cytokines.14,15,32,35–37 Other studies suggest that prokineticins may act directly on sensory neurons in a process involving the sensitization of TRPV1, a key mediator of inflammatory pain that is expressed highly in nociceptive neurons.27 Expression of TRPV1 is more prevalent in afferent neurons innervating the GI tract than somatic tissues,38 suggesting that this pathway may be important in visceral pain sensation. Our observation that the non-steroidal anti-inflammatory drug, Diclofenac, was ineffective at alleviating pain in the TNBS and mustard visceral hypersensitivity models suggests that acute inhibition of inflammation is not sufficient to provide analgesia. This favors the hypothesis that at least part of the pro-nociceptive behavior of PROK2 comes from a direct activation of PKRs expressed by sensory neurons rather than an indirect modulation of the inflammatory process; however, further studies are needed to confirm this.

We also noted that prokineticin receptors, primarily PKR1, are expressed within the gut. The expression of PKR1 in enteric neurons raised the possibility that prokineticin signaling may also modulate intrinsic gut reflexes, such as those regulating secretion. A recent report by Ralbovsky and colleagues suggested that PROK1/PKR1 signaling may have a role to play in regulating secretion in the GI tract.39 In this study, we have extended this to show that PROK2-mediated activation of PKR1 on enteric neurons may be an important regulator of ion transport within the GI tract. The physiologic significance of PROK2-induced changes in Isc needs to be confirmed; however, as PROK1 induces ileal fluid secretion,19,39 the raised PROK2 levels in colitis might stimulate secretion. Although PROK2-mediated Isc involves the enteric nerve system because effects were blocked by inhibitors of neuronal activity (ω-conotoxin GVIA or TTX) and not affected by capsaicin pretreatment, the inhibition of ileal Isc was not complete, suggesting that there may be a non-neuronal component to prokineticin-mediated ion transport. This is supported by a recent study conducted by Wade et al.19 in which the authors demonstrate that PROK1-mediated secretion in the rat gut is insensitive to TTX, and thus likely to be mediated by receptors expressed in other cell types, potentially enterocytes. However, our own data suggests that there is no expression of PKRs in the epithelial layer, and so further work will be required to elucidate the cell type mediating this aspect of PROK2-mediated ion transport.

Further studies are also needed to establish the receptor responsible for mediating the responses observed in this study. Our analyses of PKR expression, which reveal only very low levels of Pkr2 in the gut, might suggest a more prominent role for PKR1 in local GI responses to elevated PROK2. However, the absence of selective antagonists for PKR1 and PKR2 means that an involvement of PKR2 cannot be ruled out at this stage using currently available pharmacological tools. It would be interesting in future studies to assess visceral nociceptive and secretory responses in Pkr1 and Pkr2 knockout mice/tissues. It may also be informative to explore whether the increases in Prok2 expression noted here are underpinned by increases in one or other of the known splice variants for this gene. In the study by Chen et al.,12 the authors noted that the truncated form of PROK2, PK2β, shows selectivity in vitro for PKR1 vs PKR2. If the increases in Prok2 seen here reflect increases only in PK2β, then this might support a role for PKR1 in visceral nociception and ion transport.

The expression of prokineticins in the GI tract and the effect of exogenous prokineticins on aspects of gut function were first reported around a decade ago; however, few studies since then have been aimed at understanding the regulation of prokineticin expression in a disease context. In this study, we have shown that Prok2 mRNA levels are highly increased during inflammation in human and rodent GI tract. Further studies are needed to demonstrate that the increases in Prok2 gene expression reported in this article translate into elevated protein levels and increased signaling. However, our findings raise the possibility that in the context of disease, increased PROK2 signaling through prokineticin receptors expressed by extrinsic sensory neurons innervating the GI tract and enteric neurons may play a role in mediating visceral pain perception and epithelial ion transport. This data, coupled with studies demonstrating that prokineticins induce GI muscle contraction and transit4,19 suggests that blocking PROK2 signaling by, for example, antagonism of prokineticin receptors might alleviate symptoms, in particular, pain and diarrhea, associated inflammatory bowel disease, or irritable bowel syndrome.

Acknowledgments and Disclosures

  1. Top of page
  2. Abstract
  3. Materials and methods
  4. Results
  5. Discussion
  6. Acknowledgments and Disclosures
  7. Author Contribution
  8. References
  9. Supporting Information

The authors thank Philip Smith for setting up the DSS animal model. The authors, all of whom are employees of Novartis, had complete access to all the data that supports this publication and declare that no financial or other conflict of interest exists in relation to the content of the article. The funders had no role in study design, data collection and analysis, decision to publish, or preparation of the manuscript.

Author Contribution

  1. Top of page
  2. Abstract
  3. Materials and methods
  4. Results
  5. Discussion
  6. Acknowledgments and Disclosures
  7. Author Contribution
  8. References
  9. Supporting Information

RPW, EL, AR, MP, SL, CM, ST, FZ, and GB performed the experiments and analyzed the data. RPW prepared the figures and wrote the first draft of the manuscript. MSN designed the research study, participated in the analysis of data, and prepared the final version of this paper.

References

  1. Top of page
  2. Abstract
  3. Materials and methods
  4. Results
  5. Discussion
  6. Acknowledgments and Disclosures
  7. Author Contribution
  8. References
  9. Supporting Information

Supporting Information

  1. Top of page
  2. Abstract
  3. Materials and methods
  4. Results
  5. Discussion
  6. Acknowledgments and Disclosures
  7. Author Contribution
  8. References
  9. Supporting Information

Figure S1. Expression of Pkr1 and Prok1 genes in TNBS, acute DSS, and MO-induced colitis. Colitis was induced by intra-colonic instillation of TNBS (A), by supplementation of the drinking water with DSS (B, C), or by intracolonic instillation of mustard oil (D). Modest increases in Prok1 and Pkr1 expression were observed following TNBS-treatment (A), but expression levels were unchanged following DSS- (B, C) or MO- (D) treatment. For the TNBS and MO models, full thickness samples of colon from around the site of chemical instillation were used for the qRT-PCR analyses. Comparable tissue from distal colon was taken from the animals treated with DSS. In each case, values are mean ± SE from n = 6–8 donors per group. Statistical significance in (A) was determined using Student’s t-test. *< 0.05; < 0.01 compared with controls. CT, cycle threshold; Veh, vehicle; D2, day 2; MO, mustard oil.

Figure S2. Inhibition of PROK2-mediated increases in [Ca2+]i by Compound 3. hPKR1 and hPKR2 were transiently transfected using FUGENE6 into the Aequorin parental CHO-A12 cells. Transfected cells were plated and incubated at 37 °C for 18–24 h. For agonist experiments, hPROK2 was added to the cells and luminescence read immediately on a Lumilux Cellular Screening Platform. For antagonist experiments, Compound 3 was added and incubated for 30 min. PROK2 was subsequently added at the EC80 concentration and luminescence read immediately. (A) PROK2 increased [Ca2+]i in hPKR1 and hPKR2 transfected CHO cells. (B) This response was blocked by preincubation of the cells with Compound 3. The data shown are representative traces from a single experiment. Studies were repeated three times to generate the EC50 and IC50 values referred to in the text.

Figure S3. Diclofenac (30 mg kg−1, p.o.) has no effect in models of MO- (A) and TNBS-induced (B) visceral pain. In each case, diclofenac was administered 1 h prior to assessment of visceral allodynia/hypersensitivity. Values shown are mean ± SE from n ≥ 6 donors per group. Data were analyzed using 1-way anova followed by a Dunnett post test with a P value of ≤0.05 being considered statistically significant.

Data S1. Supplementary materials and methods.

FilenameFormatSizeDescription
NMO_1804_sm_FigS1.tif279KSupporting info item
NMO_1804_sm_FigS2.tif255KSupporting info item
NMO_1804_sm_FigS3.tif91KSupporting info item

Please note: Wiley Blackwell is not responsible for the content or functionality of any supporting information supplied by the authors. Any queries (other than missing content) should be directed to the corresponding author for the article.