- Email: [email protected]
Altered distribution of small-conductance calcium-activated potassium channel SK3 in Hirschsprung’s disease
Journal of Pediatric Surgery 50 (2015) 1659–1664
Contents lists available at ScienceDirect
Journal of Pediatric Surgery journal homepage: www.elsevier.com/locate/jpedsurg
Altered distribution of small-conductance calcium-activated potassium channel SK3 in Hirschsprung’s disease David Coyle, Anne Marie O’Donnell, Prem Puri ⁎ National Children’s Research Centre, Our Lady’s Children’s Hospital, Crumlin, Dublin, Ireland
a r t i c l e
i n f o
Article history: Received 11 December 2014 Accepted 22 January 2015 Key words: SK3 Hirschsprung’s disease Purinergic Inhibitory junction potential P2Y PDGFRα
a b s t r a c t Purpose: SK3 channels are voltage-independent Ca2+-dependent K+ channels that play a key role in regulating smooth muscle membrane potential during purinergic inhibitory neurotransmission in the colon. Dysmotility problems are common after a properly performed pull-through operation for Hirschsprung’s disease (HSCR). We hypothesised that ganglionic bowel just proximal to the transition zone is abnormal and designed this study to investigate SK3 channel expression in HSCR. Methods: Entire resected bowel specimens were collected at the time of pull-through surgery for HSCR (n = 6). Control colonic specimens were obtained at the time of colostomy closure in patients following anorectoplasty (n = 6). SK3 protein expression was assessed qualitatively using immunofluorescence with confocal microscopy and quantitatively using western blot (WB) analysis. Results: Positive SK3 immunofluorescence was seen in the mucosa and in all three smooth muscle layers and the myenteric plexus in control specimens. SK3 immunofluorescence co-localised with PDGFRα. A band was detected at ~70 kDa on WB. SK3 protein expression was barely detectable in aganglionic tissue and markedly reduced in the ganglionic bowel of 3 patients with HSCR compared to controls. Conclusion: Decreased SK3 expression in ganglionic bowel may explain the basis of persistent bowel symptoms in some patients following a properly performed pull-through operation for HSCR. © 2015 Elsevier Inc. All rights reserved.
Hirschsprung’s disease (HSCR) is the most common congenital gut motility disorder, with 90% of cases presenting in the neonatal period. The functional intestinal obstruction in HSCR is due to the presence of a tonically contracted aganglionic segment of colon [1]. The gold standard treatment for HSCR is a pull-through operation to resect the aganglionic bowel to the level of the ganglionic bowel, which is then anastomosed just proximal to the dentate line. While the majority of patients have a satisfactory outcome after a properly performed pullthrough operation, many patients continue to have persistent bowel symptoms such as constipation, soiling and recurrent enterocolitis at long-term follow-up. Transition zone pull-through or residual aganglionosis may be the cause of persistent bowel symptoms in some of these patients, but many have no identifiable cause for their ongoing bowel dysfunction. There is little available information on the histopathological basis of persistent bowel symptoms following a properly performed pull-through operation [2,3]. Nitrergic and purinergic enteric motor neurons are among the key mediators of colonic inhibitory neurotransmission, and consequently smooth muscle relaxation. The post-junctional response to inhibitory
Abbreviations: SK3, small-conductance calcium-activated potassium channel type 3; PDGFRα+, platelet-derived growth factor alpha-positive; HSCR, Hirschsprung’s disease; IJP, inhibitory junction potential; SMC, smooth muscle cell; ICC, interstitial cells of Cajal. ⁎ Corresponding author at: National Children’s Research Centre, Our Lady’s Children’s Hospital, Crumlin, Dublin 12, Ireland. Tel.: +353 1 4096420. E-mail address: [email protected] (P. Puri). http://dx.doi.org/10.1016/j.jpedsurg.2015.01.013 0022-3468/© 2015 Elsevier Inc. All rights reserved.
neurotransmission consists of generation of two inhibitory junction potentials (IJP): The fast and slow IJPs [4]. The slow IJP is of low amplitude, generates smaller potentials than the fast IJP, and is dependent on nitrergic neurotransmission [5]. The fast IJP is a high-amplitude potential which causes rapid membrane repolarisation, closing voltagegated Ca2+ channels and thus eliciting smooth muscle relaxation. The fast IJP is dependent on activation of small-conductance Ca 2 +activated K + (SK3) channels by purinergic neurotransmission mediated through P2Y1 receptors (P2RY1) [4,6]. Expression of the P2RY1 has been demonstrated to be absent in the sub-mucosal and myenteric plexuses of aganglionic bowel from children with HSCR, while there was abundant immunoreactivity in ganglionic colon and in controls [7]. SK3 channels are voltage-independent Ca 2+-dependent K + channels that play a key role in regulating smooth muscle cell (SMC) membrane potential during purine-mediated inhibitory neurotransmission in the colon. SK3 gene expression has previously been demonstrated to be reduced in aganglionic bowel, suggesting that there is dysregulation of spontaneous electrical rhythmicity of SMC in HSCR [8]. Expression of SK3 in the colon has been mainly localised to platelet derived growth factor-α positive (PDGFRα +) cells, also known as fibroblastlike cells [9,10]. PDGFRα+ cells are closely associated with, but distinct from, interstitial cells of Cajal (ICCs); morphologically resembling ICCs, but not expressing c-kit. These cells are widely distributed within the myenteric plexus region and in the circular and longitudinal muscle layers throughout the colon [6]. Three types of cells – SMCs, ICCs and PDGFRα + cells – collectively form a syncytium known as the SIP
1660
D. Coyle et al. / Journal of Pediatric Surgery 50 (2015) 1659–1664
Table 1 Clinical details for patients with HSCR whose pull-through colonic specimens were utilised for this study. Case
Age at surgery (months)
Gender (M/F)
Neonatal diagnosis (Y/N)
Distance of proximal margin from most proximal extent of transition zone (mm)
A B C D E F
7 3 4 5 14 5
M M M M M M
Y Y Y Y Y Y
130 35 70 60 140 35
syncytium. Together, they elicit specific expression of receptors and ion channels in response to neurotransmitters and neuropeptides [11]. As our previous work indicates that purinergic neurotransmission is normally regulated in the ganglionic bowel in HSCR, we postulated that abnormalities further downstream in the fast IJP pathway may contribute to persistent bowel symptoms seen in some patients after a properly performed pull-through operation. We therefore hypothesised that SK3 protein expression is reduced in HSCR, not only in the aganglionic bowel, but also in ganglionic bowel. We tested this hypothesis by examining the distribution of SK3 expression qualitatively and quantitatively in colonic tissue from children with HSCR and in healthy controls. We also evaluated the expression profile of P2RY1 in the same specimens to provide further insights into aberrations in the IJP pathway in HSCR. 1. Materials and methods 1.1. Specimen collection Approval was received from the local ethics advisory committee to carry out the study. Informed parental consent for subject participation was obtained. We retrieved fresh entire resected colonic specimens from patients undergoing pull-through operations for Hirschsprung's disease at Our Lady’s Children’s Hospital and Children’s University Hospital (n = 6). The full circumference of the proximal resection margin was confirmed to be normally ganglionated at the time of surgery by immunohistochemistry. The most proximal and distal margins of the
specimen were utilised for experimental purposes. The distance of the proximal resection margin from the transition zone ranged from 4 to 12 cm. Clinical details of our patient cohort are included in Table 1. Colonic controls were retrieved at the time of sigmoid/descending colostomy closure from children who had previously undergone anorectoplasty for imperforate anus (n = 6). 1.2. Double-labelling immunofluorescence Fresh resected specimens were embedded in OCT compound (VWR, Ireland [361603E]) and snap frozen in liquid nitrogen. Ten micron sections were cut and were washed in PBS with 1% Triton X-100, and were incubated in 10% bovine serum albumin blocking solution (BSA, SigmaAldrich, Ireland [A2153-50G]) for 1 h at room temperature to prevent non-specific binding. Samples were incubated in primary antibody diluted in 5% BSA at 4 °C overnight (Mouse anti-KCNN3 antibody 1:200: Abnova, Taipei, Taiwan [H00003782-A01]; Rabbit anti-anoctamin 1 antibody 1:200: Santa Cruz Biotechnology [sc135235]; Rabbit anti-α-SMA antibody 1:100: Novus biologicals, UK [NB600-531]; Rabbit anti-platelet derived growth factor (PDGFRα) antibody 1:200: Santa Cruz Biotechnologies [sc-338]; Rabbit anti-protein gene product 9.5 antibody 1:200: Dako diagnostics, Ireland [Z5116]). Samples were rinsed in 1% PBS with 0.05% Tween® (Sigma-Aldrich, Ireland [P1379-500ML]) (PBST) and were incubated in species-specific secondary antibody for 60 min at room temperature (Donkey anti-rabbit antibody Alexa Fluor® 488: Abcam, UK [ab150073]; Donkey anti-mouse antibody Alexa Fluor® 594: Abcam, UK [ab150116]). Samples were rinsed for minimum 60 min prior to incubation with DAPI nuclear counterstain (Thermo Scientific, Ireland [EN62248]). Following additional rinsing with PBST sections were mounted with glass coverslips using Mowiol® 4-88 fluorescence mounting medium (Sigma Aldrich, Ireland [81381-50G]). Specimens were visualised using laser scanning confocal microscopy (LSM700 Confocal Microscope, Carl Zeiss MicroImaging GmbH, Jena, Germany). 1.3. Protein extraction and western blot Full-thickness sections of colon were homogenised using a tissue homogeniser in RIPA lysis buffer (Sigma-Aldrich, Ireland [R0278])
Fig. 1. (A) Confocal microscopy imaging demonstrating high levels of SK3 (red) immunofluorescence (IF) in the mucosa seen on the left of the image, with low levels of expression in the muscularis mucosa of this aganglionic specimen. Cell nuclei are stained with DAPI (blue); (B) confocal microscopy series showing the close spatial arrangement of ANO1 positive ICC fibres (green) and SK3 positive fibres of PDGFRα+ cells (red) in the circular muscle layer.
D. Coyle et al. / Journal of Pediatric Surgery 50 (2015) 1659–1664
containing 1% protease inhibitor cocktail (Sigma Aldrich, Ireland [P2714]). Soluble and insoluble fractions were then separated by centrifugation at 4 °C at 4000 rpm over 30 min. The protein concentration of the supernatant was determined by Bradford assay (Sigma Aldrich, Arklow, Ireland [B6916]) using a standard curve generated from known concentrations of BSA. Protein concentrations were standardised by addition of Laemmli lysis buffer (Sigma-Aldrich Ltd, Arklow, Ireland [38733]) and protein lysate was loaded onto a NuPAGE® Novex Bis-Tris gel (Biosciences, Dublin, Ireland [NP0321BOX]) in NuPAGE® MES SDS running buffer (Biosciences, Dublin, Ireland [NP0002]) for electrophoresis at 150 V for 2 h. Proteins were then transferred from the gel to a 0.45 nm PVDF membrane at 30 V for 90 min. Membranes were blocked in 3% dried skimmed milk dissolved in PBST for 1 h to prevent non-specific antibody binding and were then incubated in primary antibody overnight at 4 °C (Mouse antiKCNN3 antibody 1:1,000 dilution: Abnova, Taipei, Taiwan [H00003782A01]; Mouse anti-P2RY1 antibody 1:500, Abnova: Taipei, Taiwan [H00005028-M01]; Mouse anti-GAPDH antibody 1:5,000: Abcam, UK [ab9484]). Membranes were rinsed in PBST for 4 h and were then
1661
incubated in species-specific secondary antibody (Rabbit anti-mouse antibody 1:10,000: Abcam, UK [ab6728]) for 90 min at room temperature. Following further rinses with PBST for 1 h, membranes were incubated in chemiluminescent substrate (SuperSignal™ West Pico Chemiluminescent Substrate, Thermo-Fischer, Ireland [34079]) for 5 min at room temperature before transfer to a chemiluminescence cassette for blot visualisation. Semi-quantification of protein expression was carried out by performing densitometry, normalising expression of SK3 and P2RY1 against that of GAPDH, using ImageJ software (an open-access software available from http://imagej.nih.gov/ij/). 2. Results 2.1. Confocal microscopy Strong SK3 immunofluorescence (IF) was identified in the mucosal layer (Fig. 1A). This appeared to be localised to the membranous domain primarily. In the circular and longitudinal muscle layers SK3 expression
Fig. 2. (A) Confocal microscopy imaging demonstrating moderate co-localisation (yellow) of SK3 (red) with α-SMA (green) in the circular muscle of this ganglionic specimen, (B) SK3 colocalisation (yellow) with PDGFRα (green) in the myenteric plexus. A lack of co-localisation of SK3 IF (red) with both ANO1 (green) in (C) and PGP 9.5 (green) (D) in the myenteric plexus.
1662
D. Coyle et al. / Journal of Pediatric Surgery 50 (2015) 1659–1664
was seen in a filamentous pattern, forming a network of fibres adjacent to, but not co-localising with, ANO-1 positive fibres of ICCs (Fig. 1B). There was occasional co-localisation of SK3 IF with α-SMA in the smooth muscle layers (Fig. 2A). In the myenteric and submucosal plexuses SK3 expression was seen to co-localise with PDGFRα in cells consisting of a central cell body with multiple projections, consistent with PDGFRα + cells (Fig. 2B). Again, SK3-positive fibres were seen in close association with both ICCs and with PGP 9.5-positive nerve cells and fibres (Fig. 2C and D respectively). The density of SK3 positive fibres in the deep smooth muscle layers was considerably lower in aganglionic tissue compared to ganglionic tissue in HSCR, while SK3-positive fibres seemed to form a much denser network in the colonic controls when compared with the HSCR specimens (Fig. 3). The intensity of IF in the myenteric plexus was broadly similar between aganglionic and ganglionic samples as well as controls. 2.2. Western blot analysis A band at ~ 70 kDa was seen on western blot analysis, consistent with the predicted molecular weight for SK3 (Fig. 4A and B). Protein expression was normalised against expression of the loading control protein, GAPDH. SK3 expression was broadly reduced in aganglionic bowel
compared to ganglionic bowel in HSCR. In 3/6 patients SK3 expression in ganglionic bowel was reduced relative to expression levels in colonic controls (Fig. 4A and B). On western blot assay for P2RY1 a single band was detected between 28 and 38 kDa (predicted molecular weight 31.5 kDa). Expression of P2RY1 was significantly reduced in aganglionic specimens relative to both ganglionic and colonic control specimens, between which expression levels were similar (Fig. 4C and D). 2.3. Clinical follow-up Clinical follow-up was available on all patients for a median of 12–17 months post-operatively. Two patients presented with anastomotic strictures in the early follow-up period, but their symptoms resolved with dilatations. No patient developed post-operative enterocolitis. Two patients required oral laxative therapy to maintain a regular bowel habit but all patients were off laxative therapy by one year post-operatively. 3. Discussion The SK3 channel, also known as KCa2.3, is encoded by the KCNN3 gene and is one of a family of small-conductance potassium channels (SK1–
Fig. 3. This confocal micrograph demonstrates the differential expression of SK3 (red) seen in different tissue types, with little IF seen in the aganglionic smooth muscle layers, moderate IF seen in the ganglionic smooth muscle layers, and abundant expression in a filamentous pattern seen in the healthy control smooth muscle layers.
D. Coyle et al. / Journal of Pediatric Surgery 50 (2015) 1659–1664
1663
Fig. 4. Western blot analysis for SK3 seen in (A) and (B), showing reduced expression in all aganglionic samples and reduced SK3 protein expression in 3 of the ganglionic samples in HSCR, while expression in the colonic control specimens was relatively uniform. Western blot analysis for P2RY1 is shown in (C) and (D), showing reduced expression aganglionic bowel in HSCR only, while expression in ganglionic bowel in HSCR and controls is similar. Graphical representations of expression levels for SK3 and P2RY1 are shown in boxplots in (E) and (F) respectively.
SK3). It is found abundantly in the brain, but is also found commonly in smooth muscle-containing tissues around the body, such as the myometrium, detrusor, colon, rectum, and in arteries and arterioles [12–15]. Activation of SK3 channels is dependent on calmodulin-dependent binding of Ca 2+[16]. In GI smooth muscle cells, changes in intracellular Ca2+ concentration are thus transduced into alterations in membrane potential via outward currents produced by activated SK3 channels [17]. SK channel expression was first examined in the gastro-intestinal tract in 2001 by Fujita et al., who, using RT-PCR and immunohistochemistry, localised expression of SK3 to c-kit positive ICCs in the myenteric plexus and smooth muscle layers [18]. Later work by this group subsequently demonstrated that SK3 expression was actually present in ckit mutant mice in whom ICCs were absent. Using immune-electron microscopic studies they localised its expression to c-kit negative fibroblast-like cells with similar morphological features to ICCs but with smaller gap junction connections to smooth muscle cells [10]. More detailed descriptions of these fibroblast-like cells have since been performed. Specifically, they express PDGFR-α, and form part of a network that also includes ICCs and smooth muscle cells, known as the SIP syncytium [19,6]. Kurahashi et al. carried out a detailed functional and morphological study of PDGFRα+ cells, and showed that application of ATP, ADP and β-nicotinamide adenine dinucleotide (β-NAD) elicited large-amplitude K + currents which were apamin-sensitive. They showed that SK3 channels are highly expressed in PDGFRα + cells and activation of these channels contributes to the outward currents activated by purines in the intact muscles in response to enteric inhibitory neurotransmission [6]. In children with Hirschsprung’s disease (HSCR), functional intestinal obstruction results from the presence of a non-relaxing tonically contracted aganglionic segment. While the primary defining abnormality in HSCR is the absence of ganglion cells for varying extents of colon, abnormalities in expression profiles for various neurotransmitters, receptors and other key proteins involved in inhibitory neurotransmission
have been identified by many investigators. Nitric oxide synthase (NOS) and vasoactive intestinal peptide have both been found to be deficient in the aganglionic segment [20,21]. We have previously demonstrated that expression of the receptors P2RY1 and P2RY2, which have been shown to mediate purinergic neurotransmission in the human colon, is essentially absent in the submucosal and myenteric plexuses of aganglionic bowel in HSCR [7,22]. Taken together, our results suggest that down-regulation of the mediators of inhibitory neurotransmission in aganglionic colon leads to unopposed cholinergic activity and a tonic hyper-contractile state. Immunohistochemical studies carried out by Vanderwinden et al. have previously demonstrated SK3 expression to follow a similar pattern to that observed in our study, albeit they did not detect any difference in SK3 expression between ganglionic colon in HSCR (n = 3) and normal healthy control colon [9]. Our study has demonstrated that SK3 channels are expressed in the mucosa, smooth muscle layers and myenteric plexus, and are primarily expressed by PDGFRα+ cells in the human colon. The differential expression of SK3 between aganglionic, ganglionic and colonic control bowel in a proportion of patients in the study raises questions regarding its potential role in ganglionic bowel dysmotility in patients with persistent bowel symptoms after pull-through operation. We have demonstrated that P2RY1 expression is similar in ganglionic bowel in HSCR compared to controls, suggesting that the purinergic neurotransmission component of the fast IJP is normally regulated in ganglionic bowel. Reduced SK3 expression in the most proximal part of the resected ganglionic bowel of some patients suggests that despite normal purinergic inhibitory neurotransmission, the fast component of the IJP is attenuated in the ganglionic bowel of these patients and smooth muscle relaxation does not take place normally. Given that these changes are only present in half of our patient cohort it is uncertain whether the observed histopathological changes are dependent on the length or extent of ganglionic bowel removed during the pull-through operation in these patients. As all six patients are under the age of 15 months it is not possible to determine their
1664
D. Coyle et al. / Journal of Pediatric Surgery 50 (2015) 1659–1664
bowel function at the present time. No patient has presented with postoperative enterocolitis at follow-up and none are requiring laxatives to aid defecation beyond one year of follow-up. Long-term follow-up clinical studies are required to determine if the abnormalities in SK3 expression seen here in some patients are likely to translate into clinically significant symptoms later in life. In summary, expression of the small-conductance Ca 2 +-activated K + channel, SK3, predominantly co-localises with PDGFRα in the mucosa and in the smooth muscle layers (tunica muscularis and muscularis mucosa) where they form a dense network in close association with projections of ICCs. SK3 expression is markedly deficient in aganglionic bowel in HSCR, and moderately reduced in the ganglionic bowel of half of those specimens examined compared with healthy controls, while P2RY1 expression is only reduced in aganglionic bowel compared to ganglionic bowel and healthy controls. These findings implicate SK3 in the pathogenesis of HSCR and further functional studies are required to determine if the moderately reduced expression levels of SK3 observed in the ganglionic bowel of some patients are associated with clinically significant dysmotility.
Acknowledgements We wish to acknowledge the assistance and guidance of Dr. Luiz Alvarez and the Departments of Histopathology in Our Lady’s Children’s Hospital, Crumlin and Children’s University Hospital, Temple St, Dublin. This research was funded by grants from the National Children’s Research Centre/Children’s Medical Research Foundation, Our Lady’s Children’s Hospital, Crumlin and from Temple Street Children’s University Hospital, Dublin. The authors declare no competing interests. DC, AMOD, and PP were involved in study conception. Specimen collection was performed by DC. Experimental work was performed by DC and AMOD. The paper was drafted by DC and PP. All authors have reviewed, revised and approved the final manuscript. References [1] Holschneider AM, Puri P. Hirschsprung's disease and allied disorders. 3rd ed. Berlin Heidelberg: Springer Verlag; 2008. [2] Menezes M, Pini Prato A, Jasonni V, et al. Long-term clinical outcome in patients with total colonic aganglionosis: a 31-year review. J Pediatr Surg 2008;43(9):1696–9.
[3] Menezes M, Corbally M, Puri P. Long-term results of bowel function after treatment for Hirschsprung's disease: a 29-year review. Pediatr Surg Int 2006;22(12):987–90. [4] Hwang SJ, Blair PJ, Durnin L, et al. P2Y1 purinoreceptors are fundamental to inhibitory motor control of murine colonic excitability and transit. J Physiol 2012;590(Pt 8):1957–72. [5] Keef KD, Du C, Ward SM, et al. Enteric inhibitory neural regulation of human colonic circular muscle: role of nitric oxide. Gastroenterology 1993;105(4):1009–16. [6] Kurahashi M, Zheng H, Dwyer L, et al. A functional role for the 'fibroblast-like cells' in gastrointestinal smooth muscles. J Physiol 2011;589(Pt 3):697–710. [7] O' Donnell AM, Puri P. Deficiency of purinergic P2Y receptors in aganglionic intestine in Hirschsprung's disease. Pediatr Surg Int 2008;24(1):77–80. [8] Piotrowska AP, Solari V, Puri P. Distribution of Ca2 + -activated K channels, SK2 and SK3, in the normal and Hirschsprung's disease bowel. J Pediatr Surg 2003;38(6):978–83. [9] Vanderwinden JM, Rumessen JJ, de Kerchove d'Exaerde Jr A, et al. Kit-negative fibroblast-like cells expressing SK3, a Ca2+-activated K+ channel, in the gut musculature in health and disease. Cell Tissue Res 2002;310(3):349–58. [10] Fujita A, Takeuchi T, Jun H, et al. Localization of Ca2+-activated K+ channel, SK3, in fibroblast-like cells forming gap junctions with smooth muscle cells in the mouse small intestine. J Pharmacol Sci 2003;92(1):35–42. [11] Koh SD, Rhee PL. Ionic conductance(s) in response to post-junctional potentials. J Neurogastroenterol Motil 2013;19(4):426–32. [12] Herrera GM, Pozo MJ, Zvara P, et al. Urinary bladder instability induced by selective suppression of the murine small conductance calcium-activated potassium (SK3) channel. J Physiol 2003;551(Pt 3):893–903. [13] Chen MX, Gorman SA, Benson B, et al. Small and intermediate conductance Ca(2+)activated K+ channels confer distinctive patterns of distribution in human tissues and differential cellular localisation in the colon and corpus cavernosum. Naunyn Schmiedebergs Arch Pharmacol 2004;369(6):602–15. [14] Pierce SL, Kresowik JD, Lamping KG, et al. Overexpression of SK3 channels dampens uterine contractility to prevent preterm labor in mice. Biol Reprod 2008;78(6): 1058–63. [15] Taylor MS, Bonev AD, Gross TP, et al. Altered expression of small-conductance Ca2+-activated K+ (SK3) channels modulates arterial tone and blood pressure. Circ Res 2003;93(2):124–31. [16] Kong ID, Koh SD, Bayguinov O, et al. Small conductance Ca2+-activated K+ channels are regulated by Ca2+-calmodulin-dependent protein kinase II in murine colonic myocytes. J Physiol 2000;524(Pt 2):331–7. [17] Ro S, Hatton WJ, Koh SD, et al. Molecular properties of small-conductance Ca2+-activated K+ channels expressed in murine colonic smooth muscle. Am J Physiol Gastrointest Liver Physiol 2001;281(4):G964–73. [18] Fujita A, Takeuchi T, Saitoh N, et al. Expression of Ca(2+)-activated K(+) channels, SK3, in the interstitial cells of Cajal in the gastrointestinal tract. Am J Physiol Cell Physiol 2001;281(5):C1727–33. [19] Iino S, Horiguchi K, Horiguchi S, et al. c-Kit-negative fibroblast-like cells express platelet-derived growth factor receptor alpha in the murine gastrointestinal musculature. Histochem Cell Biol 2009;131(6):691–702. [20] O'Kelly TJ, Davies JR, Tam PK, et al. Abnormalities of nitric-oxide-producing neurons in Hirschsprung's disease: morphology and implications. J Pediatr Surg 1994;29(2): 294–9 [discussion 9–300]. [21] Kusafuka T, Puri P. Altered mRNA expression of the neuronal nitric oxide synthase gene in Hirschsprung's disease. J Pediatr Surg 1997;32(7):1054–8. [22] Gallego D, Hernandez P, Clave P, et al. P2Y1 receptors mediate inhibitory purinergic neuromuscular transmission in the human colon. Am J Physiol Gastrointest Liver Physiol 2006;291(4):G584–94.
Contents lists available at ScienceDirect
Journal of Pediatric Surgery journal homepage: www.elsevier.com/locate/jpedsurg
Altered distribution of small-conductance calcium-activated potassium channel SK3 in Hirschsprung’s disease David Coyle, Anne Marie O’Donnell, Prem Puri ⁎ National Children’s Research Centre, Our Lady’s Children’s Hospital, Crumlin, Dublin, Ireland
a r t i c l e
i n f o
Article history: Received 11 December 2014 Accepted 22 January 2015 Key words: SK3 Hirschsprung’s disease Purinergic Inhibitory junction potential P2Y PDGFRα
a b s t r a c t Purpose: SK3 channels are voltage-independent Ca2+-dependent K+ channels that play a key role in regulating smooth muscle membrane potential during purinergic inhibitory neurotransmission in the colon. Dysmotility problems are common after a properly performed pull-through operation for Hirschsprung’s disease (HSCR). We hypothesised that ganglionic bowel just proximal to the transition zone is abnormal and designed this study to investigate SK3 channel expression in HSCR. Methods: Entire resected bowel specimens were collected at the time of pull-through surgery for HSCR (n = 6). Control colonic specimens were obtained at the time of colostomy closure in patients following anorectoplasty (n = 6). SK3 protein expression was assessed qualitatively using immunofluorescence with confocal microscopy and quantitatively using western blot (WB) analysis. Results: Positive SK3 immunofluorescence was seen in the mucosa and in all three smooth muscle layers and the myenteric plexus in control specimens. SK3 immunofluorescence co-localised with PDGFRα. A band was detected at ~70 kDa on WB. SK3 protein expression was barely detectable in aganglionic tissue and markedly reduced in the ganglionic bowel of 3 patients with HSCR compared to controls. Conclusion: Decreased SK3 expression in ganglionic bowel may explain the basis of persistent bowel symptoms in some patients following a properly performed pull-through operation for HSCR. © 2015 Elsevier Inc. All rights reserved.
Hirschsprung’s disease (HSCR) is the most common congenital gut motility disorder, with 90% of cases presenting in the neonatal period. The functional intestinal obstruction in HSCR is due to the presence of a tonically contracted aganglionic segment of colon [1]. The gold standard treatment for HSCR is a pull-through operation to resect the aganglionic bowel to the level of the ganglionic bowel, which is then anastomosed just proximal to the dentate line. While the majority of patients have a satisfactory outcome after a properly performed pullthrough operation, many patients continue to have persistent bowel symptoms such as constipation, soiling and recurrent enterocolitis at long-term follow-up. Transition zone pull-through or residual aganglionosis may be the cause of persistent bowel symptoms in some of these patients, but many have no identifiable cause for their ongoing bowel dysfunction. There is little available information on the histopathological basis of persistent bowel symptoms following a properly performed pull-through operation [2,3]. Nitrergic and purinergic enteric motor neurons are among the key mediators of colonic inhibitory neurotransmission, and consequently smooth muscle relaxation. The post-junctional response to inhibitory
Abbreviations: SK3, small-conductance calcium-activated potassium channel type 3; PDGFRα+, platelet-derived growth factor alpha-positive; HSCR, Hirschsprung’s disease; IJP, inhibitory junction potential; SMC, smooth muscle cell; ICC, interstitial cells of Cajal. ⁎ Corresponding author at: National Children’s Research Centre, Our Lady’s Children’s Hospital, Crumlin, Dublin 12, Ireland. Tel.: +353 1 4096420. E-mail address: [email protected] (P. Puri). http://dx.doi.org/10.1016/j.jpedsurg.2015.01.013 0022-3468/© 2015 Elsevier Inc. All rights reserved.
neurotransmission consists of generation of two inhibitory junction potentials (IJP): The fast and slow IJPs [4]. The slow IJP is of low amplitude, generates smaller potentials than the fast IJP, and is dependent on nitrergic neurotransmission [5]. The fast IJP is a high-amplitude potential which causes rapid membrane repolarisation, closing voltagegated Ca2+ channels and thus eliciting smooth muscle relaxation. The fast IJP is dependent on activation of small-conductance Ca 2 +activated K + (SK3) channels by purinergic neurotransmission mediated through P2Y1 receptors (P2RY1) [4,6]. Expression of the P2RY1 has been demonstrated to be absent in the sub-mucosal and myenteric plexuses of aganglionic bowel from children with HSCR, while there was abundant immunoreactivity in ganglionic colon and in controls [7]. SK3 channels are voltage-independent Ca 2+-dependent K + channels that play a key role in regulating smooth muscle cell (SMC) membrane potential during purine-mediated inhibitory neurotransmission in the colon. SK3 gene expression has previously been demonstrated to be reduced in aganglionic bowel, suggesting that there is dysregulation of spontaneous electrical rhythmicity of SMC in HSCR [8]. Expression of SK3 in the colon has been mainly localised to platelet derived growth factor-α positive (PDGFRα +) cells, also known as fibroblastlike cells [9,10]. PDGFRα+ cells are closely associated with, but distinct from, interstitial cells of Cajal (ICCs); morphologically resembling ICCs, but not expressing c-kit. These cells are widely distributed within the myenteric plexus region and in the circular and longitudinal muscle layers throughout the colon [6]. Three types of cells – SMCs, ICCs and PDGFRα + cells – collectively form a syncytium known as the SIP
1660
D. Coyle et al. / Journal of Pediatric Surgery 50 (2015) 1659–1664
Table 1 Clinical details for patients with HSCR whose pull-through colonic specimens were utilised for this study. Case
Age at surgery (months)
Gender (M/F)
Neonatal diagnosis (Y/N)
Distance of proximal margin from most proximal extent of transition zone (mm)
A B C D E F
7 3 4 5 14 5
M M M M M M
Y Y Y Y Y Y
130 35 70 60 140 35
syncytium. Together, they elicit specific expression of receptors and ion channels in response to neurotransmitters and neuropeptides [11]. As our previous work indicates that purinergic neurotransmission is normally regulated in the ganglionic bowel in HSCR, we postulated that abnormalities further downstream in the fast IJP pathway may contribute to persistent bowel symptoms seen in some patients after a properly performed pull-through operation. We therefore hypothesised that SK3 protein expression is reduced in HSCR, not only in the aganglionic bowel, but also in ganglionic bowel. We tested this hypothesis by examining the distribution of SK3 expression qualitatively and quantitatively in colonic tissue from children with HSCR and in healthy controls. We also evaluated the expression profile of P2RY1 in the same specimens to provide further insights into aberrations in the IJP pathway in HSCR. 1. Materials and methods 1.1. Specimen collection Approval was received from the local ethics advisory committee to carry out the study. Informed parental consent for subject participation was obtained. We retrieved fresh entire resected colonic specimens from patients undergoing pull-through operations for Hirschsprung's disease at Our Lady’s Children’s Hospital and Children’s University Hospital (n = 6). The full circumference of the proximal resection margin was confirmed to be normally ganglionated at the time of surgery by immunohistochemistry. The most proximal and distal margins of the
specimen were utilised for experimental purposes. The distance of the proximal resection margin from the transition zone ranged from 4 to 12 cm. Clinical details of our patient cohort are included in Table 1. Colonic controls were retrieved at the time of sigmoid/descending colostomy closure from children who had previously undergone anorectoplasty for imperforate anus (n = 6). 1.2. Double-labelling immunofluorescence Fresh resected specimens were embedded in OCT compound (VWR, Ireland [361603E]) and snap frozen in liquid nitrogen. Ten micron sections were cut and were washed in PBS with 1% Triton X-100, and were incubated in 10% bovine serum albumin blocking solution (BSA, SigmaAldrich, Ireland [A2153-50G]) for 1 h at room temperature to prevent non-specific binding. Samples were incubated in primary antibody diluted in 5% BSA at 4 °C overnight (Mouse anti-KCNN3 antibody 1:200: Abnova, Taipei, Taiwan [H00003782-A01]; Rabbit anti-anoctamin 1 antibody 1:200: Santa Cruz Biotechnology [sc135235]; Rabbit anti-α-SMA antibody 1:100: Novus biologicals, UK [NB600-531]; Rabbit anti-platelet derived growth factor (PDGFRα) antibody 1:200: Santa Cruz Biotechnologies [sc-338]; Rabbit anti-protein gene product 9.5 antibody 1:200: Dako diagnostics, Ireland [Z5116]). Samples were rinsed in 1% PBS with 0.05% Tween® (Sigma-Aldrich, Ireland [P1379-500ML]) (PBST) and were incubated in species-specific secondary antibody for 60 min at room temperature (Donkey anti-rabbit antibody Alexa Fluor® 488: Abcam, UK [ab150073]; Donkey anti-mouse antibody Alexa Fluor® 594: Abcam, UK [ab150116]). Samples were rinsed for minimum 60 min prior to incubation with DAPI nuclear counterstain (Thermo Scientific, Ireland [EN62248]). Following additional rinsing with PBST sections were mounted with glass coverslips using Mowiol® 4-88 fluorescence mounting medium (Sigma Aldrich, Ireland [81381-50G]). Specimens were visualised using laser scanning confocal microscopy (LSM700 Confocal Microscope, Carl Zeiss MicroImaging GmbH, Jena, Germany). 1.3. Protein extraction and western blot Full-thickness sections of colon were homogenised using a tissue homogeniser in RIPA lysis buffer (Sigma-Aldrich, Ireland [R0278])
Fig. 1. (A) Confocal microscopy imaging demonstrating high levels of SK3 (red) immunofluorescence (IF) in the mucosa seen on the left of the image, with low levels of expression in the muscularis mucosa of this aganglionic specimen. Cell nuclei are stained with DAPI (blue); (B) confocal microscopy series showing the close spatial arrangement of ANO1 positive ICC fibres (green) and SK3 positive fibres of PDGFRα+ cells (red) in the circular muscle layer.
D. Coyle et al. / Journal of Pediatric Surgery 50 (2015) 1659–1664
containing 1% protease inhibitor cocktail (Sigma Aldrich, Ireland [P2714]). Soluble and insoluble fractions were then separated by centrifugation at 4 °C at 4000 rpm over 30 min. The protein concentration of the supernatant was determined by Bradford assay (Sigma Aldrich, Arklow, Ireland [B6916]) using a standard curve generated from known concentrations of BSA. Protein concentrations were standardised by addition of Laemmli lysis buffer (Sigma-Aldrich Ltd, Arklow, Ireland [38733]) and protein lysate was loaded onto a NuPAGE® Novex Bis-Tris gel (Biosciences, Dublin, Ireland [NP0321BOX]) in NuPAGE® MES SDS running buffer (Biosciences, Dublin, Ireland [NP0002]) for electrophoresis at 150 V for 2 h. Proteins were then transferred from the gel to a 0.45 nm PVDF membrane at 30 V for 90 min. Membranes were blocked in 3% dried skimmed milk dissolved in PBST for 1 h to prevent non-specific antibody binding and were then incubated in primary antibody overnight at 4 °C (Mouse antiKCNN3 antibody 1:1,000 dilution: Abnova, Taipei, Taiwan [H00003782A01]; Mouse anti-P2RY1 antibody 1:500, Abnova: Taipei, Taiwan [H00005028-M01]; Mouse anti-GAPDH antibody 1:5,000: Abcam, UK [ab9484]). Membranes were rinsed in PBST for 4 h and were then
1661
incubated in species-specific secondary antibody (Rabbit anti-mouse antibody 1:10,000: Abcam, UK [ab6728]) for 90 min at room temperature. Following further rinses with PBST for 1 h, membranes were incubated in chemiluminescent substrate (SuperSignal™ West Pico Chemiluminescent Substrate, Thermo-Fischer, Ireland [34079]) for 5 min at room temperature before transfer to a chemiluminescence cassette for blot visualisation. Semi-quantification of protein expression was carried out by performing densitometry, normalising expression of SK3 and P2RY1 against that of GAPDH, using ImageJ software (an open-access software available from http://imagej.nih.gov/ij/). 2. Results 2.1. Confocal microscopy Strong SK3 immunofluorescence (IF) was identified in the mucosal layer (Fig. 1A). This appeared to be localised to the membranous domain primarily. In the circular and longitudinal muscle layers SK3 expression
Fig. 2. (A) Confocal microscopy imaging demonstrating moderate co-localisation (yellow) of SK3 (red) with α-SMA (green) in the circular muscle of this ganglionic specimen, (B) SK3 colocalisation (yellow) with PDGFRα (green) in the myenteric plexus. A lack of co-localisation of SK3 IF (red) with both ANO1 (green) in (C) and PGP 9.5 (green) (D) in the myenteric plexus.
1662
D. Coyle et al. / Journal of Pediatric Surgery 50 (2015) 1659–1664
was seen in a filamentous pattern, forming a network of fibres adjacent to, but not co-localising with, ANO-1 positive fibres of ICCs (Fig. 1B). There was occasional co-localisation of SK3 IF with α-SMA in the smooth muscle layers (Fig. 2A). In the myenteric and submucosal plexuses SK3 expression was seen to co-localise with PDGFRα in cells consisting of a central cell body with multiple projections, consistent with PDGFRα + cells (Fig. 2B). Again, SK3-positive fibres were seen in close association with both ICCs and with PGP 9.5-positive nerve cells and fibres (Fig. 2C and D respectively). The density of SK3 positive fibres in the deep smooth muscle layers was considerably lower in aganglionic tissue compared to ganglionic tissue in HSCR, while SK3-positive fibres seemed to form a much denser network in the colonic controls when compared with the HSCR specimens (Fig. 3). The intensity of IF in the myenteric plexus was broadly similar between aganglionic and ganglionic samples as well as controls. 2.2. Western blot analysis A band at ~ 70 kDa was seen on western blot analysis, consistent with the predicted molecular weight for SK3 (Fig. 4A and B). Protein expression was normalised against expression of the loading control protein, GAPDH. SK3 expression was broadly reduced in aganglionic bowel
compared to ganglionic bowel in HSCR. In 3/6 patients SK3 expression in ganglionic bowel was reduced relative to expression levels in colonic controls (Fig. 4A and B). On western blot assay for P2RY1 a single band was detected between 28 and 38 kDa (predicted molecular weight 31.5 kDa). Expression of P2RY1 was significantly reduced in aganglionic specimens relative to both ganglionic and colonic control specimens, between which expression levels were similar (Fig. 4C and D). 2.3. Clinical follow-up Clinical follow-up was available on all patients for a median of 12–17 months post-operatively. Two patients presented with anastomotic strictures in the early follow-up period, but their symptoms resolved with dilatations. No patient developed post-operative enterocolitis. Two patients required oral laxative therapy to maintain a regular bowel habit but all patients were off laxative therapy by one year post-operatively. 3. Discussion The SK3 channel, also known as KCa2.3, is encoded by the KCNN3 gene and is one of a family of small-conductance potassium channels (SK1–
Fig. 3. This confocal micrograph demonstrates the differential expression of SK3 (red) seen in different tissue types, with little IF seen in the aganglionic smooth muscle layers, moderate IF seen in the ganglionic smooth muscle layers, and abundant expression in a filamentous pattern seen in the healthy control smooth muscle layers.
D. Coyle et al. / Journal of Pediatric Surgery 50 (2015) 1659–1664
1663
Fig. 4. Western blot analysis for SK3 seen in (A) and (B), showing reduced expression in all aganglionic samples and reduced SK3 protein expression in 3 of the ganglionic samples in HSCR, while expression in the colonic control specimens was relatively uniform. Western blot analysis for P2RY1 is shown in (C) and (D), showing reduced expression aganglionic bowel in HSCR only, while expression in ganglionic bowel in HSCR and controls is similar. Graphical representations of expression levels for SK3 and P2RY1 are shown in boxplots in (E) and (F) respectively.
SK3). It is found abundantly in the brain, but is also found commonly in smooth muscle-containing tissues around the body, such as the myometrium, detrusor, colon, rectum, and in arteries and arterioles [12–15]. Activation of SK3 channels is dependent on calmodulin-dependent binding of Ca 2+[16]. In GI smooth muscle cells, changes in intracellular Ca2+ concentration are thus transduced into alterations in membrane potential via outward currents produced by activated SK3 channels [17]. SK channel expression was first examined in the gastro-intestinal tract in 2001 by Fujita et al., who, using RT-PCR and immunohistochemistry, localised expression of SK3 to c-kit positive ICCs in the myenteric plexus and smooth muscle layers [18]. Later work by this group subsequently demonstrated that SK3 expression was actually present in ckit mutant mice in whom ICCs were absent. Using immune-electron microscopic studies they localised its expression to c-kit negative fibroblast-like cells with similar morphological features to ICCs but with smaller gap junction connections to smooth muscle cells [10]. More detailed descriptions of these fibroblast-like cells have since been performed. Specifically, they express PDGFR-α, and form part of a network that also includes ICCs and smooth muscle cells, known as the SIP syncytium [19,6]. Kurahashi et al. carried out a detailed functional and morphological study of PDGFRα+ cells, and showed that application of ATP, ADP and β-nicotinamide adenine dinucleotide (β-NAD) elicited large-amplitude K + currents which were apamin-sensitive. They showed that SK3 channels are highly expressed in PDGFRα + cells and activation of these channels contributes to the outward currents activated by purines in the intact muscles in response to enteric inhibitory neurotransmission [6]. In children with Hirschsprung’s disease (HSCR), functional intestinal obstruction results from the presence of a non-relaxing tonically contracted aganglionic segment. While the primary defining abnormality in HSCR is the absence of ganglion cells for varying extents of colon, abnormalities in expression profiles for various neurotransmitters, receptors and other key proteins involved in inhibitory neurotransmission
have been identified by many investigators. Nitric oxide synthase (NOS) and vasoactive intestinal peptide have both been found to be deficient in the aganglionic segment [20,21]. We have previously demonstrated that expression of the receptors P2RY1 and P2RY2, which have been shown to mediate purinergic neurotransmission in the human colon, is essentially absent in the submucosal and myenteric plexuses of aganglionic bowel in HSCR [7,22]. Taken together, our results suggest that down-regulation of the mediators of inhibitory neurotransmission in aganglionic colon leads to unopposed cholinergic activity and a tonic hyper-contractile state. Immunohistochemical studies carried out by Vanderwinden et al. have previously demonstrated SK3 expression to follow a similar pattern to that observed in our study, albeit they did not detect any difference in SK3 expression between ganglionic colon in HSCR (n = 3) and normal healthy control colon [9]. Our study has demonstrated that SK3 channels are expressed in the mucosa, smooth muscle layers and myenteric plexus, and are primarily expressed by PDGFRα+ cells in the human colon. The differential expression of SK3 between aganglionic, ganglionic and colonic control bowel in a proportion of patients in the study raises questions regarding its potential role in ganglionic bowel dysmotility in patients with persistent bowel symptoms after pull-through operation. We have demonstrated that P2RY1 expression is similar in ganglionic bowel in HSCR compared to controls, suggesting that the purinergic neurotransmission component of the fast IJP is normally regulated in ganglionic bowel. Reduced SK3 expression in the most proximal part of the resected ganglionic bowel of some patients suggests that despite normal purinergic inhibitory neurotransmission, the fast component of the IJP is attenuated in the ganglionic bowel of these patients and smooth muscle relaxation does not take place normally. Given that these changes are only present in half of our patient cohort it is uncertain whether the observed histopathological changes are dependent on the length or extent of ganglionic bowel removed during the pull-through operation in these patients. As all six patients are under the age of 15 months it is not possible to determine their
1664
D. Coyle et al. / Journal of Pediatric Surgery 50 (2015) 1659–1664
bowel function at the present time. No patient has presented with postoperative enterocolitis at follow-up and none are requiring laxatives to aid defecation beyond one year of follow-up. Long-term follow-up clinical studies are required to determine if the abnormalities in SK3 expression seen here in some patients are likely to translate into clinically significant symptoms later in life. In summary, expression of the small-conductance Ca 2 +-activated K + channel, SK3, predominantly co-localises with PDGFRα in the mucosa and in the smooth muscle layers (tunica muscularis and muscularis mucosa) where they form a dense network in close association with projections of ICCs. SK3 expression is markedly deficient in aganglionic bowel in HSCR, and moderately reduced in the ganglionic bowel of half of those specimens examined compared with healthy controls, while P2RY1 expression is only reduced in aganglionic bowel compared to ganglionic bowel and healthy controls. These findings implicate SK3 in the pathogenesis of HSCR and further functional studies are required to determine if the moderately reduced expression levels of SK3 observed in the ganglionic bowel of some patients are associated with clinically significant dysmotility.
Acknowledgements We wish to acknowledge the assistance and guidance of Dr. Luiz Alvarez and the Departments of Histopathology in Our Lady’s Children’s Hospital, Crumlin and Children’s University Hospital, Temple St, Dublin. This research was funded by grants from the National Children’s Research Centre/Children’s Medical Research Foundation, Our Lady’s Children’s Hospital, Crumlin and from Temple Street Children’s University Hospital, Dublin. The authors declare no competing interests. DC, AMOD, and PP were involved in study conception. Specimen collection was performed by DC. Experimental work was performed by DC and AMOD. The paper was drafted by DC and PP. All authors have reviewed, revised and approved the final manuscript. References [1] Holschneider AM, Puri P. Hirschsprung's disease and allied disorders. 3rd ed. Berlin Heidelberg: Springer Verlag; 2008. [2] Menezes M, Pini Prato A, Jasonni V, et al. Long-term clinical outcome in patients with total colonic aganglionosis: a 31-year review. J Pediatr Surg 2008;43(9):1696–9.
[3] Menezes M, Corbally M, Puri P. Long-term results of bowel function after treatment for Hirschsprung's disease: a 29-year review. Pediatr Surg Int 2006;22(12):987–90. [4] Hwang SJ, Blair PJ, Durnin L, et al. P2Y1 purinoreceptors are fundamental to inhibitory motor control of murine colonic excitability and transit. J Physiol 2012;590(Pt 8):1957–72. [5] Keef KD, Du C, Ward SM, et al. Enteric inhibitory neural regulation of human colonic circular muscle: role of nitric oxide. Gastroenterology 1993;105(4):1009–16. [6] Kurahashi M, Zheng H, Dwyer L, et al. A functional role for the 'fibroblast-like cells' in gastrointestinal smooth muscles. J Physiol 2011;589(Pt 3):697–710. [7] O' Donnell AM, Puri P. Deficiency of purinergic P2Y receptors in aganglionic intestine in Hirschsprung's disease. Pediatr Surg Int 2008;24(1):77–80. [8] Piotrowska AP, Solari V, Puri P. Distribution of Ca2 + -activated K channels, SK2 and SK3, in the normal and Hirschsprung's disease bowel. J Pediatr Surg 2003;38(6):978–83. [9] Vanderwinden JM, Rumessen JJ, de Kerchove d'Exaerde Jr A, et al. Kit-negative fibroblast-like cells expressing SK3, a Ca2+-activated K+ channel, in the gut musculature in health and disease. Cell Tissue Res 2002;310(3):349–58. [10] Fujita A, Takeuchi T, Jun H, et al. Localization of Ca2+-activated K+ channel, SK3, in fibroblast-like cells forming gap junctions with smooth muscle cells in the mouse small intestine. J Pharmacol Sci 2003;92(1):35–42. [11] Koh SD, Rhee PL. Ionic conductance(s) in response to post-junctional potentials. J Neurogastroenterol Motil 2013;19(4):426–32. [12] Herrera GM, Pozo MJ, Zvara P, et al. Urinary bladder instability induced by selective suppression of the murine small conductance calcium-activated potassium (SK3) channel. J Physiol 2003;551(Pt 3):893–903. [13] Chen MX, Gorman SA, Benson B, et al. Small and intermediate conductance Ca(2+)activated K+ channels confer distinctive patterns of distribution in human tissues and differential cellular localisation in the colon and corpus cavernosum. Naunyn Schmiedebergs Arch Pharmacol 2004;369(6):602–15. [14] Pierce SL, Kresowik JD, Lamping KG, et al. Overexpression of SK3 channels dampens uterine contractility to prevent preterm labor in mice. Biol Reprod 2008;78(6): 1058–63. [15] Taylor MS, Bonev AD, Gross TP, et al. Altered expression of small-conductance Ca2+-activated K+ (SK3) channels modulates arterial tone and blood pressure. Circ Res 2003;93(2):124–31. [16] Kong ID, Koh SD, Bayguinov O, et al. Small conductance Ca2+-activated K+ channels are regulated by Ca2+-calmodulin-dependent protein kinase II in murine colonic myocytes. J Physiol 2000;524(Pt 2):331–7. [17] Ro S, Hatton WJ, Koh SD, et al. Molecular properties of small-conductance Ca2+-activated K+ channels expressed in murine colonic smooth muscle. Am J Physiol Gastrointest Liver Physiol 2001;281(4):G964–73. [18] Fujita A, Takeuchi T, Saitoh N, et al. Expression of Ca(2+)-activated K(+) channels, SK3, in the interstitial cells of Cajal in the gastrointestinal tract. Am J Physiol Cell Physiol 2001;281(5):C1727–33. [19] Iino S, Horiguchi K, Horiguchi S, et al. c-Kit-negative fibroblast-like cells express platelet-derived growth factor receptor alpha in the murine gastrointestinal musculature. Histochem Cell Biol 2009;131(6):691–702. [20] O'Kelly TJ, Davies JR, Tam PK, et al. Abnormalities of nitric-oxide-producing neurons in Hirschsprung's disease: morphology and implications. J Pediatr Surg 1994;29(2): 294–9 [discussion 9–300]. [21] Kusafuka T, Puri P. Altered mRNA expression of the neuronal nitric oxide synthase gene in Hirschsprung's disease. J Pediatr Surg 1997;32(7):1054–8. [22] Gallego D, Hernandez P, Clave P, et al. P2Y1 receptors mediate inhibitory purinergic neuromuscular transmission in the human colon. Am J Physiol Gastrointest Liver Physiol 2006;291(4):G584–94.