The Pelvis–Kidney Junction Contains HCN3, a Hyperpolarization-Activated Cation Channel That Triggers Ureter Peristalsis

View metadata, citation and similar papers at core.ac.uk brought to you by CORE

provided by Elsevier - Publisher Connector original article http://www.kidney-international.org & 2010 International Society of Nephrology

see commentary on page 483 The pelvis–kidney junction contains HCN3, a hyperpolarization-activated cation channel that triggers ureter peristalsis Romulo Hurtado1, Gil Bub2 and Doris Herzlinger1

1Department of Physiology and Biophysics, Weill Medical College of Cornell University, New York, New York, USA and 2Department of Physiology, Anatomy and Genetics, University of Oxford, Oxford, UK

Peristaltic waves of the ureteric smooth muscles move urine The excretory and regulatory functions of the kidney are down from the kidney, a process that is commonly defective dependent on filtration of blood at the glomerulus; in congenital diseases. To study the mechanisms that control reabsorption of solutes and fluids required for homeostasis the initiation and direction of contractions, we used video by the renal tubules; and the movement of excess fluids, microscopy and optical mapping techniques and found that solutes, and metabolic wastes into the bladder for storage electrical and contractile waves began in a region where the prior to excretion.1,2 This latter process, the transport of renal pelvis joined the connective tissue core of the kidney. urine from the kidney to the bladder, is controlled, in large Separation of this pelvis–kidney junction from more distal part, by coordinated contractions of the smooth-muscle coat urinary tract segments prevented downstream peristalsis, investing the renal pelvis and ureter.3 Congenital defects in indicating that it housed the trigger for peristalsis. Moreover, this peristaltic process are common and can cause a wide cells in the pelvis–kidney junction were found to express spectrum of pathologies ranging from obstruction and isoform 3 of the hyperpolarization-activated cation on permanent, pressure-induced renal damage to more subtle channel family known to be required for initiating electrical conditions such as persistent infection due to reflux of urine activity in the brain and heart. Immunocytochemical from the bladder back into the kidney.4–7 Despite the high and real-time PCR analyses found that hyperpolarization- incidence of acquired and congenital defects that impair activated cation-3 is expressed at the pelvis–kidney junction upper urinary tract peristalsis, the molecular mechanisms where electrical excitation and contractile waves originate. that control the initiation site and unidirectional nature of Inhibition of this channel caused a loss of electrical activity this process remain poorly understood. at the pelvis–kidney junction and randomized the origin In animals with uni-papillary kidneys, such as the mouse, of electrical activity in the urinary tract, thus markedly the smooth-muscle coat investing the upper urinary tract perturbing contractions. Collectively, our study demonstrates begins where the renal pelvis joins the connective tissue core that hyperpolarization-activated cation-3 channels play of the kidney (Figure 1).8 It is extremely thin at this site, the a fundamental role in coordinating proximal-to-distal pelvis–kidney junction (PKJ), but gradually thickens toward peristalsis of the upper urinary tract. This provides insight the ureteral–pelvic junction. Although cells along the entire into the genetic causes of common inherited urinary tract length of the smooth-muscle coat surrounding the renal disorders such as reflux and obstruction. pelvis and ureter are excitable, under normal conditions Kidney International (2010) 77, 500–508; doi:10.1038/ki.2009.483; contractions always initiate at the PKJ and move distally, in a published online 23 December 2009 coordinated, wave-like manner, down the ureter. These KEYWORDS: obstructive nephropathy; obstructive uropathy coordinated contractile waves are myogenic, occurring in the absence of nervous input, and current hypotheses predict that they are triggered by a specialized cell population at the PKJ.9–11 Specifically, large numbers of atypical smooth- muscle cells are found at the PKJ adjacent to the smooth muscle and when isolated, they exhibit spontaneous depolarizations that are greater in frequency than the pelvic smooth- muscle cells. Moreover, atypical smooth-muscle-cell depolar- Correspondence: Doris Herzlinger, Department of Physiology and Biophy- izations precede contractile activity in tissue strips taken sics, Weill Cornell Medical College of Cornell University, 1300 York Ave, New from the wall of the renal pelvis. Thus, it is likely that York, NY 10021, USA. E-mail: [email protected] presence of spontaneous electrical activity at the PKJ localizes Received 25 August 2009; revised 6 October 2009; accepted 13 October the origin of upper urinary tract peristalsis and coordinates 2009; published online 23 December 2009 downstream contractile activity. However, the molecular

500 Kidney International (2010) 77, 500–508 R Hurtado et al.: Role of PKJ in triggering ureteral peristalsis original article

loss of electrical activity at the PKJ and randomized the site where spontaneous membrane depolarizations initiate along c the urinary tract. Thus, we have shown, for the ﬁrst time, that inhibition of HCN channel activity results in loss of electrical m activity at the PKJ, and perturbs both the origin and propagation of peristaltic waves along the upper urinary pa tract. In addition, these results provide new insight into pkj possible genetic causes of urine reﬂux and abnormal ureteral peristalsis.

RESULTS p upj Unidirectional waves of contractile and electrical activity initiate at the PKJ We devised an explant system to directly and continuously analyze waves of contractile and electrical activity from the u PKJ through to the distal ureter. Briefly, ureters with adjoining kidneys were isolated from adult male mice and the PKJ and renal pelvis were exposed by bisecting the kidneys along their sagittal plane (Figure 2b and c). Direct Figure 1 | The upper urinary tract is composed of the kidney video-microscopic examination of explants prepared in this and ureter. Blood is filtered in the kidney by the glomeruli and manner showed that contractile waves always initiated at the plasma ultrafiltrate is modified as it passes through the tubular nephron segments and collecting ducts located in the cortical PKJ and then moved distally through the renal pelvis and (c) and medullary (m) zones of the organ. The collecting ducts then down the ureter (n ¼ 25) (Figure 2d and e, and coalesce into several large ducts in the papilla (pa), and urine Supplementary Movie S1). Contractile waves occurred 4–9 flows from these ducts into the renal pelvis (p). The renal pelvis, times/min in the explanted urinary tracts, similar to a funnel shaped structure between the renal tubules and the 17 ureter (u), joins the connective tissue core of the kidney at the frequencies reported in vivo. To analyze electrical activity pelvis–kidney junction (pkj). Both the renal pelvis and the ureter in explants, changes in membrane potential were detected have a water impermeable epithelium and smooth muscle coat, using the voltage-sensitive dye RH237.18 Explants were which thickens at the ureter–pelvis junction (upj) and functions to carry urine, in relatively unmodified form, from the kidney loaded with the dye and changes in the intensity of to the bladder. fluorescence signals across the upper urinary tract were recorded over time. Results of these studies showed that mechanisms underlying spontaneous depolarizations at spontaneous electrical excitation initiates at the PKJ and the PKJ have yet to be established and a definitive relation- propagates distally in a coordinated wave-like manner ship between this electrical activity and peristalsis remains (n ¼ 15) (Figure 3). Collectively, these video-microscopic unclear. and optical mapping data demonstrate that both electrical We used a candidate gene approach to identify ion and contractile activity initiate at the PKJ. channels that play a role in initiating urinary tract peristalsis at the PKJ. Specifically, we hypothesized that spontaneous Proximal urinary tract tissues containing the PKJ are required membrane depolarizations observed at the PKJ may be for peristalsis controlled, in part, by members of the hyperpolarization- We next asked whether the PKJ was required to initiate activated cation (HCN) ion channel family. HCN channels contractile waves that propagate distally down the renal have been shown to be required for spontaneous membrane pelvis and the ureter. Urinary tract explants were prepared as depolarizations exhibited by pacemaker cells of the heart and described above and then further dissected to sequentially for spontaneous depolarizations that underlie autorhythmic remove portions of the urinary tract as shown in Figure 4 electrical activity in the brain.12–16 To test this hypothesis we (n ¼ 4). Initial dissections removed major portions of the devised a novel explant system to analyze contractile and kidney, including the papilla and the outer regions of the electrical activity in the murine urinary tract from the PKJ cortex and medulla (Figure 4a, illustration; Figure 4a, through to the distal ureter. Reverse transcription-PCR (RT- representative sample; Supplementary Movie S2). Real-time PCR) and immunohistochemical techniques were used to video microscopy of these explants showed that the upper identify HCN-family members expressed in the urinary tract urinary tract continued to contract in a proximal-to-distal and strikingly, a high level of HCN3 expression was observed wave, initiating at the PKJ (n ¼ 4) (Figure 4d–f). These at the PKJ. Most importantly, inhibition of HCN channel data showed that kidney tissues, including cortex, medulla, activity disrupted coordinated, proximal-to-distal contrac- and renal papillae, were not required for initiating tions of the upper urinary tract smooth-muscle coat, causing peristalsis. it to twitch along its entire length. Optical mapping To test whether the PKJ was required for triggering the experiments showed that HCN channel inhibition caused a observed proximal-to-distal contractions, we separated the

Kidney International (2010) 77, 500–508 501 original article R Hurtado et al.: Role of PKJ in triggering ureteral peristalsis

c m pa pa pkj

p p u u

pa pkj

0 s u 4 s 7 s 8 s Figure 2 | Upper urinary tract contractions begin at the pelvis–kidney junction. Kidneys were bisected as shown in (a) to expose the inside of the organ (b and c). Cortical (c), medullary (m) and papillary (pa) zones are visible in bisected kidneys as well as the renal pelvis (p) and ureter (u). At higher power (c), the pelvis-kidney junction (pkj) can be seen. Bisected kidneys were analyzed by real-time video microscopy and contractile activity was recorded (d-g). Peristaltic waves occurred 4–9 times/min and always initiated at the pkj (e, arrow). As can be seen in f, contractions move distally (arrow) toward the ureteral-pelvic junction and then down the ureter (arrow, g).

PKJ and the proximal renal pelvis from the more distal HCN2 expression has been localized to the inner regions of the urinary tract (Figure 4g, illustration; medullary collecting ducts,20 and to spatially localize Figure 4h, representative sample; Supplementary Movie S3). HCN3 protein expression we performed chromogenic Real-time video microscopy showed that the proximal immunohistochemistry using 80-mm-thick urinary tract renal pelvis still connected to the PKJ continued contracting sections. HCN3 was highly expressed in the upper urinary in a coordinated manner, initiating at the PKJ (Figure 4j) tract at the PKJ and in the renal papilla (Figure 5d and e). and propagating distally (Figure 4k). In contrast, when Paraffin sections of tissues processed for immunohistochem- separated from the PKJ, distal urinary tract segments lost ical detection of HCN3 (Figure 5f) showed that HCN3 coordinated contraction and remained motionless (Figure expression was localized to cells residing within the 4i–l). Thus, these data show that the proximal region of the connective tissue coat investing the renal pelvis. To further urinary tract containing the PKJ is required for driving define the cell population expressing HCN3 we performed peristalsis. double immunofluorescence for the detection of smooth- muscle actin and HCN3 (Figure 5g–i) using 12-mm-frozen HCN 3 is highly expressed at the PKJ sections. As observed by chromogenic techniques, HCN3 Collectively, the tissue separation experiments described expression was detected at the PKJ (Figure 5g) within the above combined with our video-microscopic and optical connective tissue layer of the urinary tract, and now could be mapping analyses suggest that spontaneous electrical activity localized to a cell population sub-adjacent to the smooth- observed at the PKJ initiates upper urinary tract peristalsis. muscle actin-positive (Figure 5h) smooth-muscle coat of the To further test this hypothesis we used a candidate gene urinary tract (Figure 5i, overlay). Importantly, detection of approach to characterize the molecular mechanisms under- smooth-muscle actin alone (Figure 5j and l) showed that lying spontaneous depolarizations observed at the PKJ. green autofluorescence (Figure 5j) was observed in the Specifically, we tested whether members of the HCN channel renal tubules, but not the connective surrounding the renal family, which are required for initiating cardiac contractions pelvis. Thus, a cell population at the PKJ that underlies the and rhythmic electrical activity in the brain, play a role in most proximal tip of the smooth-muscle coat expresses high initiating urinary tract peristalsis.12–16 level of HCN ion channels. Collectively, these morphological We first determined whether mRNAs encoding HCN- data combined with the known function of HCN channels in family members were expressed at the PKJ. The proximal- other systems, raise the possibility that HCN ion channel most region of the upper urinary tract (Figure 5a, schematic activity may trigger peristalsis at the PKJ. illustration) containing the PKJ was isolated from other kidney tissue and RT-PCR analysis was performed for HCN channel activity is required for coordinated upper HCN1–4 (Figure 5b). Abundant levels of HCN3 and low urinary tract peristalsis levels of HCN2 mRNAs were detected in the proximal To determine whether HCN channel activity is required for urinary tract. Control experiments assaying brain tissue upper urinary tract peristalsis, we used the well-characterized showed expression of all four HCN channel isoforms (Figure HCN ion channel inhibitor, ZD7288. Both video-microscopic 5c), as previously reported.19 and optical mapping analyses were performed using urinary

502 Kidney International (2010) 77, 500–508 R Hurtado et al.: Role of PKJ in triggering ureteral peristalsis original article

pkj

pkj pkj p

u 0 s 1.2 s

0 s 0 s

2.2 s 2.7 s

3 s 3 s

3.2 s 4 s 4 s Figure 3 | Electrical excitation of the upper urinary tract initiates at the pelvic–kidney junction. (a–e) Explants were loaded with the voltage sensitive dye RH237 and changes in membrane potential were detected by monitoring changes in fluorescence intensity over time, as noted. In these frames (a–e), fluorescent intensity represented by color (green to white) marked the electrical activities. Changes in membrane potential occurred in a wave, initiating at the pelvic–kidney junction (pkj) (a), and moving distally down the pelvis (b, c) to the ureter (d, e). These data were used to generate an isochronal map of electrical activity over time (f). In the isochronal map shown in f, time is 7 s 7 s represented by color and ranges from red, demarking the initial site of electrical activity, to blue, demarking the last electrically Figure 4 | Proximal urinary tract tissues containing the active site recorded. As can be seen in the captured still images pelvic–kidney junction are required to drive peristalsis. (a-e), the most proximal zone of the renal pelvis, including the PKJ, Portions of the upper urinary tract were dissected away from the was first to depolarize (red). Sequential electrical activity (yellow to ureter to determine the minimum tissue required for peristalsis. blue) was then observed moving distally in a coordinated wave-like The initial cuts (a, schematic illustration; b, representative sample) fashion down the renal pelvis and the ureter. removed the outer cortex, majority of the medulla, and the papilla. The remaining renal pelvis and ureter tissue continued to contract in a coordinated proximal-to-distal direction 4–9 times/ tract explants incubated with increasing concentrations of min (c-f). Contractions initiated at the pelvic–kidney junction (d) inhibitor. Control explants (n ¼ 8) incubated with buffer and moved distally down the renal pelvis and ureter (e and f). To further localize the tissues driving peristalsis, the pelvic–kidney alone exhibited proximal-to-distal, contractile (Figure 6a–d) junction was separated from the more distal regions of the pelvis and electrical excitation (Figure 6i and j) waves, as previously (g, schematic illustration; h, representative sample). The tissue seen, initiating at the PKJ and moving distally down the containing the pkj continued to contract (i-l), initiating at the pkj urinary tract. (j) and moving distally (k). In contrast, distal segments separated from the PKJ lost coordinated contractile activity and remained Beginning with 15 mM ZD7288, subtle changes in urinary motionless. tract peristalsis and electrical excitation patterns were

Kidney International (2010) 77, 500–508 503 original article R Hurtado et al.: Role of PKJ in triggering ureteral peristalsis

HCN1HCN2HCN3HCN4GAPDH HCN1HCN2HCN3HCN4GAPDH

Kidney Brain

p pkj pkj pa

ra ra

HCN3 SMA Overlay

pkj pkj pkj

ra ra ra

Auto –flourecence SMA Overlay

pkj pkj pkj

ra ra ra

Figure 5 | HCN3 is highly expressed in the pelvic–kidney junction. To determine if HCN channels play a role in urinary tract peristalsis HCN mRNA (a–c) and protein (d–l) expression levels were analyzed. For mRNA expression analyses, the proximal region of the upper urinary tract was isolated (a, schematic illustration) and RT-PCR from total RNA was performed (b). Abundant levels of HCN isoform 3 (HCN3) mRNA and faint levels of isoform 2 (HCN2) were detected in urinary tract tissue samples (b), whereas all four HCN isoforms (HCN1-4) were detected in control brain tissue (c). Chromagenic immunohistochemistry on 80 mM thick vibratome sections (d and e) show that HCN3-expressing cells (purple reaction product) were localized to the pelvis–kidney junction (pkj), which is adjacent to the renal artery (ra). HCN3 expression is also detected in the tubular epithelia of the renal papilla (pa). Examination of paraffin sections prepared from vibratome samples processed for immunohistochemical detection of HCN3 antibody binding (f) demonstrate that HCN3-expressing cells are underneath the surface of the renal pelvis (p) and exhibit a spindle-shaped morphology. The purple precipitate indicative of HCN antibody binding was not detected in control tissue sections incubated with secondary antibody reagents in the absence of primary antibody (data not shown). To further characterize the HCN3-positive cell population present at the PKJ, indirect immunofluorescence microscopy was used to detect HCN3-expressing cells and smooth muscle in the same tissue section (g-i). HCN3 antibody binding was visualized with alexa-488 conjugated secondary antibodies (g and i) and smooth muscle was labeled with cy3-anti-smooth muscle actin (SMA, h and i). As can be seen in g, HCN3 staining (green) was detected in a cell population at the pkj in close proximity to the renal artery (ra). Smooth muscle actin staining (h, red) and overlay (i) of both HCN3 (green) and SMA (red) antibody binding demonstrate that these HCN3-expressing cells at the PKJ are located underneath the SMA-positive smooth muscle layer that invests the renal pelvis. Examination of control sections incubated with Alexa-488 secondary antibody alone (j), cy3-anti-SMA alone (k), or overlay of both (l) demonstrate that the renal artery and renal tubules exhibit non-specific staining (j and l). Scale bar ¼ 100 mm. observed (data not shown). Strikingly, at 30 mM ZD7288 initiated at the PKJ and propagated distally, contractile (n ¼ 8) coordinated, proximal-to-distal contractions of the activity at the PKJ was absent in explants treated with 30 mM urinary tract musculature were lost. Instead, rapid, twitch- ZD7288 (Supplementary Movie S5) and replaced by random like contractile activity was observed, which was difﬁcult to twitches in more distal, focal domains of the urinary tract. capture in time-lapse still images (Figure 6e–h), but could Optical mapping analyses are consistent with this ZD7288- easily be observed in continuous video recordings included in induced twitch-like behavior. In the presence of 30 mM Supplementary Movies S4 and S5. Whereas control explants ZD7288, electrical activity at the PKJ was lost (Figure 6k (Supplementary Movie S4) exhibited contractions that and l) but was not completely abolished. The coordinated

504 Kidney International (2010) 77, 500–508 R Hurtado et al.: Role of PKJ in triggering ureteral peristalsis original article

pkj

0 s 3 s 4 s 7 s

pkj

0 s 3 s 4 s 7 s

pkj pkj

3.5 s 3.5 s Figure 6 | HCN3 channel activity is required for coordinated peristalsis and electrical propagation of the upper urinary tract. To determine if HCN channels are functionally important for urinary tract peristalsis we used the well characterized HCN channel inhibitor ZD7288. Explants were placed in a bathing tyrodes solution with and without 30 mM ZD7288. Explants exposed to tyrodes solution alone maintained coordinated peristalsis (a-d), initiating at the pkj (b) and propagating distally down the upper urinary tract (c, d). Explants treated with 30 mM ZD7288 (e-h) lost coordinated contractile activity. Explants treated with 30 mM ZD7288 exhibited twitch-like contractile activity difficult to capture in still images, but easily detectable in video recordings (supplemental data). Optical mapping of explants kept in tyrodes solution alone contained proximal-to-distal electrical excitation patterns (i, still image; j, isochronal map) initiating (red) at the pkj and moving distally down the urinary tract. Optical mapping of explants treated with 30 mM ZD7288 contained desynchronized electrical excitation patterns (k, still image; l, isochronal map) with loss of electrical excitation at the pkj, and initiation site (red) randomized within the urinary tract. waves of electrical activity that initiated at the PKJ in control electrical and contractile activity initiate at the junction explants were replaced by a random and asynchronized between the renal pelvis and the connective tissue core of the excitation pattern in 30 mM ZD7288-treated explants. Thus, at kidney, the PKJ. These results are in agreement with those of 30 mM concentration of inhibitor, both electrical and previous studies analyzing electrical activity and smooth- contractile activity at the PKJ is replaced by randomized muscle contractions of the upper urinary tract.21–25 activity along the urinary tract, resulting in abnormal Most importantly, our work begins to elucidate the peristalsis. In contrast to 30 mM ZD7288, we were unable to molecular mechanisms underlying spontaneous electrical detect either electrical or contractile activity at much higher activity at the PKJ, and establish a deﬁnitive relationship concentration (90 mM, data not shown). between this activity and coordinated peristalsis. We Collectively these functional analyses indicate that both hypothesized that spontaneous membrane depolarizations contractile and electrical activities of the upper urinary tract localized to the PKJ were mediated, in part, by members of musculature are sensitive to HCN channel inhibition, in a the HCN ion channel family. Members of this cation-selective dose-dependent manner. These data combined with those ion channel family are required for spontaneous electrical from HCN3 expression analyses of the urinary tract indicate activity in the brain and heart. All four HCN isoforms that spontaneous depolarizations mediated by HCN ion (HCN1–4) are expressed in the brain, and HCN channel channels at the PKJ play a fundamental role in driving activity has been shown to be required for spontaneous coordinated, proximal-to-distal contractions of the upper depolarization in neurons.26–28 In the heart, HCN2 and urinary tract. HCN4 have been shown to be required for spontaneous electrical activity of the cardiac pacemaker cells of the 14,15 DISCUSSION sinaotrial node. RT-PCR analyses described in this report In this study we devised methods to assay contractile and demonstrate that high levels of HCN3 and barely detectable electrical activities characterizing upper urinary tract peri- levels of HCN2 are expressed in the proximal region of the stalsis in mouse. Our data show that proximal-to-distal upper urinary tract. Immunohistochemical studies showed

Kidney International (2010) 77, 500–508 505 original article R Hurtado et al.: Role of PKJ in triggering ureteral peristalsis

that HCN2 expression is restricted to the renal tubules,20 smooth-muscle actin immunostaining as compared with that whereas we show that HCN3 expression is localized to the of the adjacent smooth-muscle-cell population, indicative of connective tissue layer underlying the smooth-muscle coat at few contractile filaments. Currently, the nature of sponta- the PKJ. Consistent with this HCN expression pattern, neous activity detected in atypical smooth-muscle cells is inhibition of HCN channel activity caused loss of sponta- unknown, and our data suggest that HCN channel activity neous electrical activity at the PKJ, desynchronized the maybe the underlying mechanism. Further characterization electrical propagation in the upper urinary tract, and caused of both atypical and HCN-positive cell types in the upper abnormal peristalsis. urinary tract will determine whether they are in fact the same Perturbation of electrical activity and peristalsis of the cell population.47 Alternatively, it is possible that HCN upper urinary tract with ZD7288 occurred in a dose- channel activity indirectly triggers smooth-muscle contrac- dependent manner. ZD7288 is the most widely used and tions at the PKJ. For example, HCN channel activity may be best characterized HCN channel blocker. Inhibition of HCN important for release of prostaglandins that support urinary channel activity by ZD7288 has been used in the nervous tract peristalsis. system to alleviate neuropathic pain via reversal of ectopic In conclusion, the data presented in this study raise the spontaneous depolarizations of neurons,29–32 and in the heart possibility that HCN channel mutations that disrupt as a bradycardiac agent to slow down abnormally fast heart spontaneous depolarizations at the PKJ may perturb upper rates via inhibition of spontaneous activity of the sinoatrial urinary tract peristalsis and cause hydronephrosis. Hydrone- node.33–36 Importantly, we observed inhibition of electrical phrosis, or pressure-induced enlargement of the renal pelvis, activity at the PKJ at a concentration of ZD7288 just below is detected in up to 1% of fetuses by routine ultrasound and the IC50 documented for HEK239 kidney cell line transfected is a leading cause of renal failure in the pediatric popula- with HCN channel constructs. At much higher concentration.49 Genetic studies using the mouse indicate that tions of ZD7288 all electrical activity was abolished, hydronephrosis can be caused by several different mechan- indicating requirement of ion channel inhibition for more isms. For example, uroplakin mutations can cause hydrone- fundamental aspects of smooth-muscle physiology.37–40 phrosis due to obliteration of the ureteral lumen by Collectively, these drug titration studies combined with overgrowth of the urothelium.50,51 Mutations in AQP2, the HCN expression analyses strongly support the hypothesis ADH-sensitive water channel, causes hydronephrosis due to that HCN channel activity plays a specialized role in the sheer volume of fluids that must be excreted in the coordinating upper urinary tract peristalsis. absence of water reabsorption by the collecting duct system.52 Our finding that HCN channel activity is necessary for Finally, mice with defects in the formation and/or organiza- proximal-to-distal upper urinary tract peristalsis agrees with tion of the ureteral smooth-muscle coat develop hydrone- published studies implicating HCN channels in coordinating phrosis due to absence of contractile forces that propel urine essential physiological processes. In the brain, sensory from the kidney to the bladder.53–56 Although these mouse thalamic neurons spontaneously depolarize after initial mutants prove that physical barriers to urine flow and afferent excitatory inputs, and deletion of their HCN channel defective ureteral smooth-muscle differentiation can cause activity (HCN2-KO) disrupts spontaneous depolarizations, hydronephrosis, this condition often occurs in the absence of altering the firing behavior of the neurons in a physiologically any physical barrier or obvious defects in the ureteral similar manner to epilepsy.41 In the heart, spontaneous smooth-muscle coat. Studies analyzing the physiology of electrical activity of sinoatrial node cells are responsible for mice harboring HCN3 mutations and/or defects in the HCN- maintaining coordinated heartbeat, and HCN channel expressing cell types at the PKJ may reveal a possible cause deletion in these cells (HCN2-KO and HCN4-KO) causes for such functional urinary tract obstructions and provide cardiac arrhythmia.41–43 In fact, comparisons between novel diagnostic tools for identifying fetuses at risk for renal spontaneous electrical activity of the heart and upper urinary damage due to impaired urine flow out of the kidney. tract have been drawn by others, based on similarities in electrical activity, including electromyograms recorded from MATERIALS AND METHODS pacemaker cells at the sinoatrial node and PKJ.22,23,44 Our Animals work shows for the first time that these similarities are likely All mice were housed at the Weill Medical College of Cornell to be due to the role of HCN ion channels in initiating University Animal Facility and treated according to the Research spontaneous depolarizations in both systems. Animal Resource Center Guidelines. Adult and postnatal mouse pups were purchased from Taconic Farms (Germantown, NY, USA). The spatial distribution and cellular characteristics of the HCN-expressing cells described in this work are similar to Kidney explants those described for atypical smooth-muscle cells thought to 45–48 Kidneys were isolated from adult, 5-week-old mice and were placed be pacemakers of the upper urinary tract. Spatially, both in Tyrodes Solution containing sodium bicarbonate (Sigma, cell populations are localized to the most proximal regions of St Louis, MO, USA) and cut using a no. 22 surgical blade (Bard-Parker, the urinary tract, the PKJ, and are found within the Franklin Lakes, NJ, USA) to expose the junction between the pelvis connective tissue underlying the smooth-muscle-cell layer and the renal parenchyma. Kidneys were bisected by cutting of the urinary tract. Furthermore, both cell types have low longitudinally alongside the renal papillae and urinary tract.

506 Kidney International (2010) 77, 500–508 R Hurtado et al.: Role of PKJ in triggering ureteral peristalsis original article

Explants containing approximately half the renal cortex and andembeddedinOCTcompound.Frozensectionsof12mmthickness medulla, but the entire papilla, pelvis, and ureter were transferred were cut, rehydrated with PBS, permeabilized with 0.2% Triton X-100 onto 24-mm Costar Transwell Permeable Supports (0.4 mm poly- in PBS for 20 min, and blocked with 1% normal donkey serum in PBS carbonate membrane) (Corning, Corning, NY, USA) and placed for 20 min. Rat anti-HCN3 monoclonal antibody (clone TLL6C5; into six-well tissue culture plates (Costar) containing 1.5 ml Tyrodes Millipore, Billerica, MA, USA) was used at a concentration of 1/700 Solution per well, and 800 ml of Tyrodes Solution were added directly and mouse anti-smooth-muscle actin conjugated to Cy3 (Sigma) was on top of the kidney. Plates were then placed on top of a rocking used at 1/2200, both diluted in 1% normal donkey serum in PBS, at nutrator housed inside a 5% CO2 incubator, allowing a bathing 37 1C for 3 h and 1 h, respectively. After five washes with PBS, the solution of Tyrodes to pass over the kidney explant without completely sections were incubated with Dyelight 488-conjugated goat anti-rat submerging it. A ZD7288 (Tocris Bioscience, Ellisville, MI, USA) stock (Jackson ImmunoResearch, West Grove, PA, USA) at a concentration solution of 100 mM in phosphate-buffered saline (PBS) was stored at of 1/1000, diluted in 1% normal donkey serum in PBS, for 1.5 h at 20 1C diluted one-half with Hybri-Max dimethylsulfoxide (Sigma) 37 1C. For vibratome sections kidneys were embedded in 7% low- and brought to final concentrations, as noted, in Tyrodes (dimethyl- melting-point agarose. Vibratome sections of 80 mmwerecut, sulfoxide was maintained below 0.2% at all times.) The inhibitor and permeabilized, and blocked with TSP (0.5% Triton X-100, 0.1% control solutions with the vehicle alone were added directly on top of saponin), 1% normal donkey serum in TBS overnight at room explants and changed every half hour for a total of 2 h. temperature, and then at 37 1C for 3 h. HCN3 was added at a concentration of 1/700 in blocking solution for 6 h at room Optical mapping temperature. After washing overnight in TSP, sections were incubated Propagation of cellular excitation and activation sequences of kidney with alkaline phosphatase-conjugated anti-rat (Jackson ImmunoR- explants were determined by optical mapping techniques in a esearch) at a concentration of 1/1200 in blocking solution for 3 h at manner similar as previously described.57,58 In short, kidney room temperature. After washing with TSP, chromagenic detection 59 explants were stained with the voltage-sensitive dye RH237 (50 mM was performed by NBT-BCIP staining as previously described. in Tyrodes) for 10 min in a 5% CO2 incubator. Spread of excitation throughout the urinary tract was monitored by changes in DISCLOSURE fluorescence signals of RH237 using a Cardioplex 80 80 pixel All the authors declared no competing interests. CCD mounted on a MacroScope-2a system (Redshirt Imaging, Boston, MA, USA) at an acquisition speed of 125 Hz for 9600 ms. ACKNOWLEDGMENTS The excitation light (528 nm) was provided by a filtered tungsten This work was supported by NIHDK45218 awarded to DH and a halogen light source and long-pass-filtered (4650 nm; Chroma BBSRC grant from the UK to GB. We thank Dr Takashi Mikawa and Dr Technology, Rockingham, VT, USA) before being imaged by the David Christini for providing optical mapping apparatuses and laboratory space. camera. Cytochalasin-D was used at 75 mM to inhibit muscular contractions since optical mapping cannot be performed using SUPPLEMENTARY MATERIAL moving tissues. Data were acquired using the Redshirt Cardioplex Movie S1. Figure 2 movie. software and analyzed by custom-written software. Movie S2. Figure 4b movie. Movie S3. Figure 4h movie. RT-PCR Movie S4. Figure 6 control movie. Total RNA isolated from adult brain and upper urinary tracts with Movie S5. Figure 6 30 mM ZD7288 movie. the adjoining kidney tissue was extracted according to the TRIzol Supplementary material is linked to the online version of the paper at (Invitrogen, Carlsbad, CA, USA) reagent protocol, treated with http://www.nature.com/ki DNase-I (Roche, Indianapolis, IN, USA), and converted to cDNA using the SuperScript cDNA Synthesis system (Invitrogen). cDNA REFERENCES was amplified by PCR using the following primers: HCN1, 1. Barbuti A, DiFrancesco D. Control of cardiac rate by ‘funny’ channels in 50CCCTCAGTCCTAAATACAGACC30 (forward) and 50GGAGCGT health and disease. Ann N Y Acad Sci 2008; 1123: 213–223. GTTCCTTCACCT30 (reverse); HCN2, 50CCGTCATCCACACCA-A 2. Biel M, Wahl-Schott C, Michalakis S et al. Hyperpolarization-activated 0 0 cation channels: from genes to function. Physiol Rev 2009; 89: 847–885. AGC3 (forward) and 5 GGCTGGTTATTGCGTGAGC (reverse); 3. Eaton DC, Pooler JP. Vander’s Renal Physiology. Physiology. 7th edn. 0 0 HCN3, 5 CCTAACATAC-TGCCCTTTATCACC3 (forward) and WCB McGraw: Boston, 1998. 50GTGGATTCTCACTTGGTGTGGAC30 (reverse); HCN4, 50CCAA 4. Chevalier RL, Peters CA. Congenital urinary tract obstruction: Proceedings GAACTTTCCCGAGTGCC (forward) and 50CCTAATCA-CAGAAA of the State-Of-The-Art Strategic Planning Workshop–National Institutes AACCTGAAGG (reverse); and GAPDH, 50ATGACATCAAGAAGGT of Health, Bethesda, Maryland, USA, 11–12 March 2002. Pediatr Nephrol 0 0 0 2003; 18: 576–606. GGTG3 (forward) and 5 CATACCAGGAAATGAGCTTG3 (reverse). 5. Chevalier RL. Pathogenesis of renal injury in obstructive uropathy. Curr To analyze DNA sequence fidelity, PCR products were gel-extracted Opin Pediatr 2006; 18: 153–160. using a gel extraction kit (Qiagen, Germantown, MD, USA) and 6. Rosen S, Peters CA, Chevalier RL et al. The kidney in congenital subcloned into PCRII plasmid using the TOPO-TA cloning system ureteropelvic junction obstruction: a spectrum from normal to (Invitrogen), and sent to the Cornell University Life Sciences Core nephrectomy. J Urol 2008; 179: 1257–1263. 7. Becker A, Baum M. Obstructive uropathy. Early Hum Dev 2006; 82: 15–22. Laboratories Center for sequencing of inserts. All respective PCR 8. Gosling JA, Dixon JS. Species variation in the location of upper urinary products showed 100% fidelity with the DNA sequences listed in the tract pacemaker cells. Invest Urol 1974; 11: 418–423. National Institute of Health NCBI database. 9. Santicioli P, Maggi CA. Myogenic and neurogenic factors in the control of pyeloureteral motility and ureteral peristalsis. Pharmacol Rev 1998; 50: 683–722. Immunohistochemistry 10. Klemm MF, Exintaris B, Lang RJ. Identification of the cells underlying P8 kidneys were fixed in Bouins fixative (Sigma) at 4 1C for 6 h. For pacemaker activity in the guinea-pig upper urinary tract. J Physiol 1999; frozen sections, kidneys were cryoprotected by 30% sucrose in PBS 519(Part 3): 867–884.

Kidney International (2010) 77, 500–508 507 original article R Hurtado et al.: Role of PKJ in triggering ureteral peristalsis

11. Lang RJ, Exintaris B, Teele ME et al. Electrical basis of peristalsis in the 36. Marshall PW, Rouse W, Briggs I et al. ICI D7288, a novel sinoatrial node mammalian upper urinary tract. Clin Exp Pharmacol Physiol 1998; 25: modulator. J Cardiovasc Pharmacol 1993; 21: 902–906. 310–321. 37. Felix R, Sandoval A, Sanchez D et al. ZD7288 inhibits low-threshold Ca(2+) 12. Santoro B, Grant SG, Bartsch D et al. Interactive cloning with the SH3 channel activity and regulates sperm function. Biochem Biophys Res domain of N-src identifies a new brain specific ion channel protein, with Commun 2003; 311: 187–192. homology to eag and cyclic nucleotide-gated channels. Proc Natl Acad Sci 38. Shin KS, Rothberg BS, Yellen G. Blocker state dependence and trapping in USA 1997; 94: 14815–14820. hyperpolarization-activated cation channels: evidence for an intracellular 13. Ludwig A, Zong X, Jeglitsch M et al. A family of hyperpolarization- activation gate. J Gen Physiol 2001; 117: 91–101. activated mammalian cation channels. Nature 1998; 393: 587–591. 39. Akaike T, Jin MH, Yokoyama U et al. T-type Ca2+ channels promote 14. Ludwig A, Zong X, Stieber J et al. Two pacemaker channels from human oxygenation-induced closure of the rat ductus arteriosus not only by heart with profoundly different activation kinetics. EMBO J 1999; 18: vasoconstriction but also by neointima formation. J Biol Chem 2009; 284: 2323–2329. 24025–24034. 15. Ishii TM, Takano M, Xie LH et al. Molecular characterization of the 40. Zeng X, Keyser B, Li M et al. T-type (alpha1G) low voltage-activated hyperpolarization-activated cation channel in rabbit heart sinoatrial node. calcium channel interactions with nitric oxide–cyclic guanosine J Biol Chem 1999; 274: 12835–12839. monophosphate pathway and regulation of calcium homeostasis in 16. Seifert R, Scholten A, Gauss R et al. Molecular characterization of a slowly human cavernosal cells. J Sex Med 2005; 2: 620–630; discussion 630–623. gating human hyperpolarization-activated channel predominantly 41. Ludwig A, Budde T, Stieber J et al. Absence epilepsy and sinus expressed in thalamus, heart, and testis. Proc Natl Acad Sci USA 1999; 96: dysrhythmia in mice lacking the pacemaker channel HCN2. EMBO J 2003; 9391–9396. 22: 216–224. 17. Dwyer TM, Schmidt-Nielsen B. The renal pelvis: machinery that 42. Stieber J, Herrmann S, Feil S et al. The hyperpolarization-activated concentrates urine in the papilla. News Physiol Sci 2003; 18:1–6. channel HCN4 is required for the generation of pacemaker action 18. Kim JH, Dunn MB, Hua Y et al. Imaging of cerebellar surface activation potentials in the embryonic heart. Proc Natl Acad Sci USA 2003; 100: in vivo using voltage sensitive dyes. Neuroscience 1989; 31: 613–623. 15235–15240. 19. Notomi T, Shigemoto R. Immunohistochemical localization of Ih 43. Herrmann S, Stieber J, Stockl G et al. HCN4 provides a ‘depolarization channel subunits, HCN1–4, in the rat brain. J Comp Neurol 2004; 471: reserve’ and is not required for heart rate acceleration in mice. EMBO 241–276. J 2007; 26: 4423–4432. 20. Bolivar JJ, Tapia D, Arenas G et al. A hyperpolarization-activated, cyclic 44. Orbelli L, Von Brucke IV E. Beitrage zur Physiologie der autonom nucleotide-gated, (Ih-like) cationic current and HCN gene expression in innervierten muskulatur. Pflugers Arch Gesamte Physiol Menschen Tiere renal inner medullary collecting duct cells. Am J Physiol Cell Physiol 2008; 1910; 133: 23. 294: C893–C906. 45. Dixon JS, Gosling JA. The fine structure of pacemaker cells in the pig renal 21. Weiss RM, Tamarkin FJ, Wheeler MA. Pacemaker activity in the upper calices. Anat Rec 1973; 175: 139–153. 46. Dixon JS, Gosling JA. The musculature of the human renal calices, pelvis urinary tract. J Smooth Muscle Res 2006; 42: 103–115. and upper ureter. J Anat 1982; 135: 129–137. 22. Weiss RM, Wagner ML, Hoffman BF. Wenckebach periods of the ureter. A 47. Lang RJ, Hashitani H, Tonta MA et al. Spontaneous electrical and Ca further note on the ubiquity of the Wenckebach phenomenon. Invest Urol signals in the mouse renal pelvis that drive pyeloureteric peristalsis. Clin 1968; 5: 462–467. Exp Pharmacol Physiol 2009, doi:10.1111/j.1440-1681.2009.05226.x. 23. Bozler E. The activity of the pacemaker previous to the discharge of a 48. Lang RJ, Hashitani H, Tonta MA et al. Spontaneous electrical and Ca2+ muscular impulse. Am J Physiol 1942; 136:9. signals in typical and atypical smooth muscle cells and interstitial cell of 24. Longrigg N. Minor calyces as primary pacemaker sites for ureteral activity Cajal-like cells of mouse renal pelvis. J Physiol 2007; 583: 1049–1068. in man. Lancet 1975; 1: 253–254. 49. Chevalier RL. Obstructive uropathy: state of the art. Pediatr Med Chir 2002; 25. Seki N, Suzuki H. Electrical properties of smooth muscle cell membrane in 24: 95–97. renal pelvis of rabbits. Am J Physiol 1990; 259: F888–F894. 50. Hu P, Deng FM, Liang FX et al. Ablation of uroplakin III gene results in 26. Steriade M, Timofeev I. Neuronal plasticity in thalamocortical networks small urothelial plaques, urothelial leakage, and vesicoureteral reflux. during sleep and waking oscillations. Neuron 2003; 37: 563–576. J Cell Biol 2000; 151: 961–972. 27. McCormick DA, Bal T. Sleep and arousal: thalamocortical mechanisms. 51. Wu XR, Kong XP, Pellicer A et al. Uroplakins in urothelial biology, function, Annu Rev Neurosci 1997; 20: 185–215. and disease. Kidney Int 2009; 75: 1153–1165. 28. Pape HC. Queer current and pacemaker: the hyperpolarization-activated 52. McDill BW, Li SZ, Kovach PA et al. Congenital progressive cation current in neurons. Annu Rev Physiol 1996; 58: 299–327. hydronephrosis (cph) is caused by an S256 L mutation in aquaporin-2 29. Harris NC, Constanti A. Mechanism of block by ZD 7288 of the that affects its phosphorylation and apical membrane accumulation. Proc hyperpolarization-activated inward rectifying current in guinea pig Natl Acad Sci USA 2006; 103: 6952–6957. substantia nigra neurons in vitro. J Neurophysiol 1995; 74: 2366–2378. 53. Mahoney ZX, Sammut B, Xavier RJ et al. Discs-large homolog 1 regulates 30. Chaplan SR, Guo HQ, Lee DH et al. Neuronal hyperpolarization-activated smooth muscle orientation in the mouse ureter. Proc Natl Acad Sci USA pacemaker channels drive neuropathic pain. J Neurosci 2003; 23: 2006; 103: 19872–19877. 1169–1178. 54. Chang CP, McDill BW, Neilson JR et al. Calcineurin is required in 31. Lee DH, Chang L, Sorkin LS et al. Hyperpolarization-activated, cation- urinary tract mesenchyme for the development of the nonselective, cyclic nucleotide-modulated channel blockade alleviates pyeloureteral peristaltic machinery. J Clin Invest 2004; 113: mechanical allodynia and suppresses ectopic discharge in spinal nerve 1051–1058. ligated rats. J Pain 2005; 6: 417–424. 55. Yu J, Carroll TJ, McMahon AP. Sonic hedgehog regulates proliferation and 32. Sun Q, Xing GG, Tu HY et al. Inhibition of hyperpolarization-activated differentiation of mesenchymal cells in the mouse metanephric kidney. current by ZD7288 suppresses ectopic discharges of injured dorsal root Development 2002; 129: 5301–5312. ganglion neurons in a rat model of neuropathic pain. Brain Res 2005; 56. Caubit X, Lye CM, Martin E et al. Teashirt 3 is necessary for ureteral 1032: 63–69. smooth muscle differentiation downstream of SHH and BMP4. 33. BoSmith RE, Briggs I, Sturgess NC. Inhibitory actions of ZENECA ZD7288 Development 2008; 135: 3301–3310. on whole-cell hyperpolarization activated inward current (If) in guinea- 57. Bub G, Glass L, Publicover NG et al. Bursting calcium rotors in cultured pig dissociated sinoatrial node cells. Br J Pharmacol 1993; 110: 343–349. cardiac myocyte monolayers. Proc Natl Acad Sci USA 1998; 95: 10283–10287. 34. Berger F, Borchard U, Gelhaar R et al. Effects of the bradycardic agent ZD 58. Hall CE, Hurtado R, Hewett KW et al. Hemodynamic-dependent 7288 on membrane voltage and pacemaker current in sheep cardiac patterning of endothelin converting enzyme 1 expression and Purkinje fibres. Naunyn Schmiedebergs Arch Pharmacol 1994; 350: differentiation of impulse-conducting Purkinje fibers in the embryonic 677–684. heart. Development 2004; 131: 581–592. 35. Berger F, Borchard U, Gelhaar R et al. Inhibition of pacemaker current by 59. Hurtado R, Mikawa T. Enhanced sensitivity and stability in two-color in the bradycardic agent ZD 7288 is lost use-dependently in sheep cardiac situ hybridization by means of a novel chromagenic substrate Purkinje fibres. Naunyn Schmiedebergs Arch Pharmacol 1995; 353: 64–72. combination. Dev Dyn 2006; 235: 2811–2816.

508 Kidney International (2010) 77, 500–508