I. Introduction

Top Next References

The main, and possibly only, function of the pyeloureteral complex is to ensure the unidirectional transport of urine from the kidney to the urinary bladder. An extremely efficient set of mechanisms operates to achieve this target: the basic process regulating ureteral peristalsis is myogenic, being initiated by spontaneously active pacemaker cells in the renal pelvis. In recent years, the mechanisms of excitability and electromechanical coupling in ureter cells have been extensively investigated; several pharmacological tools have been instrumental for a further understanding of the myogenic regulation of pyeloureteral motility.

A role of innervation in regulating pyeloureteral motility has been advocated from time to time (See Boyarsky and Labay, 1969; Weiss, 1992; Amann, 1993) but its definition has remained elusive until recently. In particular, a convincing proof for a local control of ureteral motility by cholinergic or noradrenergic nerves has not been obtained. The existence of a prominent sensory innervation of the renal pelvis and ureter is self-evident when considering that renal colic arising from the presence/passage of a stone along the ureter is one of the most intense and vivid forms of visceral pain. In the past few years, important advancements have been made through the recognition of the role of sensory innervation in determining local changes of pyeloureteral motility and in producing neurogenic inflammation.

In both areas of investigation (myogenic and neurogenic regulation of pyeloureteral function), the use of novel drugs and pharmacological tools has been instrumental to unravel principles of physiological or pathophysiological relevance. Our purpose is to bring together the new information obtained regarding the myogenic and neurogenic factors regulating pyeloureteral motility, ureteral pain, and inflammation and to discuss this recently obtained information on the background of the previous literature.

	II. The Myogenic Theory of Ureteral Peristalsis

Top Previous Next References

The roots of the myogenic theory of ureteral peristalsis can be traced back to the pioneering work of Engelmann (1869) who localized the origin of peristaltic pressure waves in the renal pelvis and proposed that the impulse for ureteral contraction passes from one ureter cell to another, the whole ureter working as a syncitium. The basic organization of the pyeloureteral complex has a loose resemblance to that regulating cardiac excitation and contractility (fig. 1). In both organs, the electrical and mechanical activities are initiated by spontaneously active cells (in the atrium and renal pelvis, respectively) and are then conducted to regions that are not normally active unless driven by a pacemaker (the ventricles and ureter, respectively); in addition, both the ureter and cardiac tissues have a plateau-type action potential (Bozler, 1942a,b; Burnstock and Prosser, 1960). However, as evidenced at several places in this review, the analogy holds true only at a very superficial level.

View larger version (23K): [in this window] [in a new window]   Fig. 1.   Schematic drawing illustrating the basic organization of the pyeloureteral complex (renal pelvis/ureter). Ureteral peristalsis originates from the renal pelvis where urine produced in the kidney collected: the renal pelvis is spontaneously active (panels A and B) to generate a pacemaker activity which propagates to the ureter thus determining the propulsion of boluses of urine toward the bladder. The ureter is normally quiescent (panel C) although latent pacemakers exist at all levels of the pyeloureteral tract which can be excited to fire action potentials (panel D) and contractions by depolarizing stimuli. Excitation of latent pacemakers by noxious/irritant stimuli can produce antiperistaltic waves of contraction producing urine reflux toward the kidney.

The main pieces of evidence indicating that ureteral peristalsis is essentially a myogenic process are as follows: (a) peristalsis occurs in the isolated pyeloureteral complex with characteristics similar as those observed in vivo, providing that the renal pelvis is attached to the ureter (Macht, 1916; Finberg and Peart, 1970; Malin et al., 1970); (b) an efficient peristalsis occurs, with characteristics similar to those observed before surgery, after denervation of the ureter (Wharton, 1932), in autotransplanted ureters (O'Conor and Dawson-Edwards, 1959), and even in a reversed ureter (Melick et al., 1961); (c) peristalsis of the isolated pyeloureteral tract is unaffected by application of tetrodotoxin or other neuron blocking agents (Golenhofen and Hannappel, 1973; Teele and Lang, 1998).

Some debate has developed in the literature about the exact localization of a putatively unique pacemaker responsible for initiation of peristaltic waves. Most authors have localized the pacemaker in supraureteral structures (calyces and proximal renal pelvis) (Benjamin, 1971; Constantinou, 1978; Kiil, 1957; Hannappel and Golenhofen, 1974a; Morita et al., 1981; Tsuchida et al., 1981). Kobayashi (1964, 1965) found the pacemaker at the border between the renal pelvis and calyces. Golenhofen and Hannappel (1973) presented evidence indicating that the utmost renal ends of the renal pelvis are involved in the pacemaker process. The other prevailing opinion indicated the pyeloureteral junction as the main pacemaker (Barr, 1971; Weiss et al., 1967; Weiss, 1971). The issue about the location of the pacemaker is complicated by the existence of remarkable species-related differences in the anatomical organization of the renal pelvis and calyces (Kiil and Setekleiv, 1973).

Because a frequency gradient exists in the renal pelvis in terms of autorhythmicity, the most proximal portions of the renal pelvis are thought to generate the physiological rhythm of peristalsis. The existence of specialized cells capable of generating a spontaneous rhythm in the renal pelvis had been suggested in early anatomical studies. Gosling and Dixon (1972, 1974) described special smooth muscle cells with pale-staining cytoplasm present in the renal calyx and pelvis but not in the ureter and discussed the possibility that these cells may have a special "pacemaker" quality. Lang et al. (1998) recently described the presence of fibroblasts-like cells, resembling the interstitial cells of Cajal that are responsible for pacemaking activity in the smooth muscle of the intestine, to be present in the proximal part of the guinea pig renal pelvis. Electrophysiological evidence also indicates an enrichment in the density of pacemaker cells at the level of the proximal renal pelvis, as reviewed in Section III.B.

According to the current data, it seems that the whole renal pelvis acts as a pacemaker to generate all-or-none propulsion of urine and that a proximal-to-distal gradient exists for setting the frequency of spontaneous activity. When the proximal, middle, or distal renal pelvis was dissected, spontaneous activity was detected from each subregion with similar frequencies as those observed before dissection (Constantinou et al., 1978; Constantinou and Yamaguchi, 1981), i.e., every subregion of the renal pelvis behaves as an oscillator.

A "multiple coupled oscillators" model has been developed to explain the ability of the whole renal pelvis to act as the pacemaker and to respond to an increase in diuresis with an increased frequency of urine transport. According to this model, several pacemaker subunits exist in the renal pelvis, the oscillation of each subunit being insufficient to trigger peristalsis; coupling processes (mainly electrical) between the subunits facilitate the synchronization and summation of activity to increase the chance of triggering a conducted peristaltic wave (Golenhofen and Hannappel, 1973; Constantinou and Yamaguchi, 1981). According to this model, the force of the total pacemaker will depend on (a) the force of each subunit and (b) the degree of synchronization between subunits. Physical stimuli, such as the distention, modulate the peristaltic frequency by changing the degree of synchronization between subunits, possibly through the generation of prostanoids (Section III.E). This model can be applied independently from the speciesdependent anatomical complexities of the calyces and renal pelvis.

Lammers et al. (1996) have analyzed the spatial and temporal variations in the pacemaker process by performing simultaneous extracellular electrophysiological recordings from a large number of sites in the whole isolated sheep renal pelvis. By reconstructing the initiation and spread of activity, they showed that the pacemaker was located at the pyelocalyceal border and never in the body of the pelvis or in the area of the pelviureteric junction. One single pacemaker was invariably responsible for a particular spread of activation and fusion of activity from two or more pacemakers did not take place. Spontaneous shifts of the pacemaker could occur from one site to another along the pyelocalyceal border. To what extent the observations of Lammers et al. (1996) in the sheep renal pelvis can be applied to other species is not known; the results contradict the hypothesis (Gosling and Dixon, 1974) that in species with multicalyceal kidneys, several pacemakers from distinct calyces could discharge simultaneously and their impulses could merge together into a single wavefront. Lammers et al. (1996) proposed that, in the intact renal pelvis, factors such as the degree of filling of a particular calyx or local stretch may determine the initiation of the pacemaker activity at a given site.

Excitation waves spread from the renal pelvis to the ureter, determining the propulsion of urine through an intermittent phasic-type contractility. Considering the ureter as a syncitium (Bozler, 1942a,b), the propagation of impulses occurs as a purely myogenic process (via electrotonic spread) at points of intimate contact between ureter muscle cells or "gap junctions" (Notley, 1968; Uehara and Burnstock, 1970; Tahara, 1990). This model predicts that the suppression of action potentials at any site of the ureter will suppress the propagation of contraction and peristalsis.

Lammers et al. (1996) analyzed the conduction of impulses in the sheep isolated renal pelvis. They observed that conduction from the site of initiation of the pacemaker current to the pyeloureteral junction is slow, inhomogeneous and contorted: multiple instances of partial or total conduction block were observed in the renal pelvis. In several cases, conduction block was related to the refractory period but in other instances no apparent relationship was evident between the occurrence of a conduction block and the length of preceding intervals. Lammers et al. (1996) speculated that a poor coupling between cells or stretch could be involved in conduction block within the renal pelvis.

A conduction block also exists between the renal pelvis and the ureter (Constantinou, 1974, 1978; Hrynczuk and Schwartz, 1975; Zawalinski et al., 1975; Constantinou and Hrynczuk, 1976); under the conditions of a normal diuresis, every pacemaker contraction of the renal pelvis does not always propagate to the ureter, suggesting that a urine flow-dependent mechanism triggers ureteral peristalsis at the pelviureteric junction. Stretching forces exerted on the pyeloureteral region by accumulating urine increase the coupling strength until they enable an incoming "pacemaker" wave of excitation to pass to the ureter (Constantinou and Yamaguchi, 1981). With increasing urine flow rates, the frequency of peristaltic contractions reaches that of the pacemaker; at this stage, further increases in urine production are accommodated by bolus volume increases, until the ureter assumes the form of an open duct (Constantinou et al., 1974).

Regional variations in the excitability and in the rate of propagation of the peristaltic wave have been described in the guinea pig, cat, and rat ureter (e.g., see Kobayashi, 1965; Tindall, 1972; Tsuchiya and Takei, 1990; Weiss, 1992). By using a three-chamber partitioned organ bath (Meini et al., 1995), we demonstrated that: (a) depolarizing stimuli applied at either end of the guinea pig ureter produce a propagated wave of excitation which travels to the other end of the organ; (b) the rate of propagation of contractions is independent of the site of application of the stimulus; (c) latent pacemakers capable of generating a propagated wave of excitation are present in all regions of the ureter; and (d) the amplitude of propagated responses is independent from the site and the nature of the depolarizing stimulus.

The findings of Meini et al. (1995) indicate also that, in principle, mechanical events are not essential for the propagation of impulses along the ureter. The main findings supporting this conclusion are as follows: (a) Bay K 8644 and glibenclamide, which produced a marked and a slight prolongation of action potential duration and enhancement of contraction, respectively (Maggi et al., 1994a,b), had no effect on the propagation of impulses in the ureter; (b) 1 µM nifedipine produced a large (>80%) inhibition of the mechanical response at the site where depolarizing stimuli were applied without affecting the intensity of the propagated response: the suppression of the propagated response by nifedipine occurred in an all-or-none manner only when the response to directly applied stimuli was also suppressed; (c) drugs that suppress the action potential by producing hyperpolarization of the membrane [cromakalim, calcitonin generelated peptide (CGRP)^b] suppressed the propagated response: both the suppression of the propagated response and its recovery upon washout of cromakalim or CGRP occur in an all-or-none manner (Meini et al., 1995).

Altogether, these findings indicated that even marked "local" changes in action potential shape/duration and contractility do not affect or impair the propagation of impulses. In other words, action potentials that are profoundly altered in their characteristics are capable of sustaining the propagation of impulses through the functional syncitium of the ureter. This conclusion does not exclude the possibility that, in physiological conditions, the stretch of the ureteral wall produced by the advancing bolus of urine may affect or modulate excitability and peristalsis.

Every single cell of the guinea pig ureter can fire an action potential in response to depolarization (Imaizumi et al., 1989a). Therefore, when applying stimuli of threshold strength, all ureter smooth muscle cells can, in principle, act as pacemakers. Action potentials produced at any site along the ureter will propagate ortho- and antidromically to determine propagated waves of excitation/contraction, i.e., peristalsis or antiperistalsis (Meini et al., 1995) (fig. 2).

View larger version (15K): [in this window] [in a new window]   Fig. 2.   Experiments illustrating the concept that latent pacemakers capable of generating a propagated wave of contraction are present at all sites of the ureter smooth muscle. The experiments were performed as described in Meini et al. (1995): the whole guinea-pig ureter was placed horizontally in a water-jacketed perspex bath for isolated organs. Two perspex partitions were used to separate the renal-, middle- and bladder-sites (R-, M- and B-site, respectively). Proximal to each partition, the ureter was pinned onto a Sylgard support. The renal and bladder ends of the ureters were connected via pulleys to isotonic transducers for recording of mechanical activity. Electrical stimuli were applied at the R-, M- or B-site by means of field stimulation. Panel a: single electrical pulses, applied at the R-, M- or B-site by field stimulation, induced propagated phasic contractions at both the R- and B-site. Note that the amplitude of evoked contractions was independent of the site of stimulation. Panel b: section of the ureter at the M site excited latent pacemakers to generate a propagated contraction at both the R- and B- site; after section, application of KCl at the M-site likewise produced a propagated wave of contraction at the R- and B- sites. Panel c: after section of the ureter at the M-site, electrical stimuli applied at the R site failed to induce a propagated contraction at the B-site; likewise, electrical stimulation at the B-site failed to induce a propagated contraction at the R-site. Electrical stimulation at the M -site evoked a propagated contraction at both B- and R-site. In all panels, the amplitude of the evoked contractions is expressed as % of the response to 80 mM KCl applied at the R- or B-site.

The activation of latent pacemakers in the ureter by chemical agents may be an important event in pathophysiological conditions because antiperistaltic waves of excitation can be generated in this way (Weiss, 1992). Latent pacemakers in the ureteral smooth muscle can be excited to fire action potentials and generate propagated contractions by mediators released from intramural nerves (neurokinin A, noradrenaline) or mast cells (histamine, serotonin) (Shuba, 1977a,b; Maggi et al., 1986; 1988a; Dodel et al., 1996; Iselin et al., 1996; Patacchini et al., 1998) and by chemical mediators normally present in the urine (bradykinin, endothelin-1) or by bacterial peptides produced in the urine during infections (Maggi et al., 1992a).

	III. Myogenic Factors Regulating Motility of Renal Pelvis and Ureter

Top Previous Next References

A. Action Potentials Recorded from the Smooth Muscle of the Pyeloureteral Tract

1. Action potentials in the renal pelvis. Zawalinski et al. (1975) reported that action potentials recorded from the renal pelvis exhibit a markedly different waveform as compared with those recorded from the ureter smooth muscle. In particular, "spontaneous" or pacemaker action potentials are recorded from the renal pelvis (fig. 1): these display a simple waveform with a slowly developing depolarization or "prepotential" preceding a slow spike. Seki and Suzuki (1990) reported that slow waves recorded from the smooth muscle of the rabbit renal pelvis are resistant to cholinergic, noradrenergic, and neuronal blockers to suggest a purely myogenic origin. Reduction of the extracellular concentration of sodium ([Na]_o) or of calcium ions ([Ca]_o) inhibited the generation of slow waves.

2. Action potentials in the ureter. The action potential of the ureter is unusually long lasting and is characterized by an initial spike followed by a plateau (fig. 3). In the guinea pig, the action potential is further characterized by the presence of multiple oscillations on the plateau (Kuriyama et al., 1967; Kuriyama and Tomita, 1970; Shuba, 1977a,b), whereas this phase is not observed in other species (fig. 3). Tetrodotoxin has no effect on the action potential of the guinea-pig ureter (Kuriyama et al., 1967). Washizu (1966) did not report any change of the action potential when extracellular Na was replaced by lithium (Li). Others reported that removal of Na abolished the plateau phase of the action potential (Kuriyama and Tomita, 1970; Shuba, 1977a; Brading et al., 1983). The removal of extracellular Ca totally eliminated the action potential (Kuriyama and Tomita, 1970; Kochemasova, 1971; Brading et al., 1983). Manganese ions (Mn) abolished the initial spike and almost totally eliminated the accompanying contraction. Verapamil markedly depresses the action potential duration and contraction and high concentration of nifedipine almost completely abolished all electrical and mechanical responses (Shuba, 1977a; Brading et al., 1983). The results of these early studies were interpreted to mean that two types of channels are involved in determining depolarization of the smooth muscle of the ureter, a "fast" Ca channel, and a "slow" channel, permeable to both Ca and Na, which would be responsible for the initial spike and for the plateau phase of the action potential, respectively.

View larger version (9K): [in this window] [in a new window]   Fig. 3.   Action potential and accompanying phasic contraction recorded by sucrose gap from the guinea-pig, human and rat ureter. In each species the action potential is characterized by an initial fast spike followed by a long lasting plateau; in the guinea-pig ureter several oscillations superimpose onto the plateau phase of the action potential.

View larger version (27K): [in this window] [in a new window]   Fig. 4.   Electrophysiological recording (lower tracing membrane potential, upper tracing mechanical activity) of the response of the guinea-pig ureter to electrical stimulation using the sucrose gap technique. Electrical stimulation produced action potentials and accompanying phasic contractions: Bay K 8644 (1 µM) enhanced the amplitude of the evoked phasic contractions. Ryanodine (100 µM) produced membrane depolarization without affecting the amplitude of evoked contractions. Reprinted with permission from Maggi et al. (1994a).

View larger version (14K): [in this window] [in a new window]   Fig. 5.   Effect of Bay K 8644 (1 µM, upper panel), ryanodine (100 µM, middle panel) or Bay K 8644 and ryanodine (1 and 100 µM, respectively, lower panel) on the electrical and mechanical response of the guinea-pig ureter to electrical stimulation. In each panel, the asterisks indicate stimulus artifact. Reprinted with permission from Maggi et al. (1994a).

B. Characterization of Ionic Currents/Channels

1. The calcium current. Lang (1989, 1990) and Imaizumi et al. (1989a,b; 1990) showed that voltage-dependent Ca channels provide the main inward current detectable in smooth muscle cells of the guinea pig ureter. The inward Ca current (I_Ca) of ureter smooth muscle exhibits a biexponential decay with a duration intermediate between that observed in rapidly spiking smooth muscles (which is remarkably shorter) and that observed in certain weakly spiking vascular smooth muscles (which is remarkably longer) (Lang, 1989).

2. Potassium (K) currents. Two main outward currents have been characterized in the guinea pig ureter (Lang, 1989; Imaizumi, 1989a, 1990; Sui and Kao, 1997c), a TEA- and charybdotoxin-sensitive Ca-dependent K current or I_K(Ca) and a voltage-dependent Ca-insensitive transient K current (also termed A-current or I_TO) which is TEA-insensitive but blocked by 4-aminopyridine (4-AP).

3. Sodium currents. It has been shown that a reduction of extracellular Na ([Na]_o) decreases the plateau phase of the action potential of the guinea pig ureter, a finding which was initially interpreted to mean that a "slow" channel, admitting both Na and Ca, could be involved in this phase, whereas a " fast" Ca channel would be responsible mainly for the initial spike (unaffected by Na replacement) (Kuriyama and Tomita, 1970; Shuba, 1977a).

C. Excitation-Contraction Coupling in the Renal Pelvis and Ureter

1. Role of extracellular calcium. Early studies on the characteristics and ionic requirements of the action potential of the guinea pig ureter had demonstrated that extracellular Ca is essential for excitation-contraction coupling and that changes in duration of the action potential are closely paralleled by changes in contractility (see Section III.A.). Indeed, a suppression of action potentials, or a pharmacological modulation (increase or decrease) of its duration by nifedipine, Bay K 8644, or charybdotoxin, are closely paralleled by equivalent changes in the amplitude and duration of spontaneous contractions of the guinea pig renal pelvis (Santicioli and Maggi, 1997).

View larger version (16K): [in this window] [in a new window]   Fig. 6.   Changes in membrane potential and tension (lower and upper tracing in each panel, respectively) recorded by single sucrose gap following 5 min superfusion with 80 mM KCl in the guinea-pig ureter (period of KCl application indicated by horizontal bars). Panel A shows the control response: note the appearance of action potentials upon start of superfusion with KCl, followed by a sustained depolarization and tonic contraction. Panel B shows the response obtained after 30 min superfusion with 10 µM H89; the tonic contraction to KCl is markedly depressed by H89; panel C shows recovery from the effect of H89, after 60 min superfusion with drug-free Krebs solution; panel D shows the effect of 30 min superfusion with 30 µM nifedipine. Reprinted with permission from Maggi et al. (1996b).

View larger version (12K): [in this window] [in a new window]   Fig. 7.   Recording of membrane potential from the guinea-pig ureter by sucrose gap: application of caffeine produces a transient hyperpolarization which is blocked by 1 µM glibenclamide and enhanced when evoked in a low K (1.2 mM) medium . Reprinted with permission from Maggi et al. (1996a).

3. Sodium/calcium exchange. The existence of a mechanism for Na/Ca exchange in smooth muscle has been a matter of discussion (see Van Breemen et al., 1979 for review). The molecular cloning of different isoforms of the Na/Ca exchanger and analysis of their presence/distribution in various tissues has positively solved the question (see Matsuda et al., 1997 for review), although the definition of the exact role of the exchanger in excitation-contraction coupling of the ureter remains uncertain.

View larger version (39K): [in this window] [in a new window]   Fig. 8.   Panel A: amplitude of EFS-evoked contractions after superfusion with normal Krebs solution (control) or low Na+, Li+-substituted Krebs solution in the presence of 1 µM Bay K 8644. Each value is mean ± s.e.m. of 7 experiments. * P < 0.05 vs control. Panel B: Effect of CPA (10 µM) on EFS-induced contractions in normal Krebs solution (empty circles) or in low Na+, Li+-substituted Krebs solution (filled circles) in the presence of 1 µM Bay K 8644. Each value is mean ± s.e.m. of 7 experiments. Panel C: Typical tracings comparing the effect of CPA on EFS-evoked contractions of the guinea-pig ureter in normal Krebs (upper panel) or in low Na+, Li+-substituted Krebs solution (middle and low tracings). All experiments were performed in the presence of 1 µM Bay K 8644. Reprinted with permission from Maggi et al. (1995b).

4. Role of intracellular pH. Intracellular pH (pH_i) is a fundamental parameter for cell physiology by regulating, among others, enzyme activity and metabolic rate. The mechanisms regulating pH_i maintain this parameter at a level considerably more alkaline than that which could be predicted on the basis of the passive distribution of hydrogen ions (H) (Roos and Boron, 1981).

As also observed in the heart, a refractory period exists in ureteral smooth muscle; after application of a conditioning stimulus, a time lag of several seconds is required for a test stimulus to induce an electrical and contractile response (Cuthbert, 1965; Kuriyama et al., 1967). The existence of a refractory period may be important for setting the maximum frequency of ureteral peristalsis and can be considered as a kind of safety factor for preventing the occurrence of antiperistaltic waves caused by the excitation of latent pacemakers.

After the discovery of the efferent role played by sensory nerves in the local regulation of pyelourereteral motility, it has been established that the refractory period of the guinea pig ureter depends, at least partially, on the inhibitory action of calcitonin gene-related peptide (CGRP) released from the peripheral endings of sensory nerves by conditioning stimulus (fig. 9; see Section V.). However, even after blockade of the CGRP inhibitory innervation (Maggi and Giuliani, 1994a), a quite long refractory period to applied depolarizing stimuli can be demonstrated, which is myogenic in origin (fig. 9). The use of in vitro capsaicin pretreatment has enabled the study of some factors regulating the myogenic component of the refractory period of the guinea pig ureter after blockade of the inhibitory CGRP innervation (Maggi et al., 1994a, 1995b).

View larger version (18K): [in this window] [in a new window]   Fig. 9.   Effect of capsaicin pretreatment or the CGRP receptor antagonist hCGRP (8-37) on the refractory period of the guinea-pig ureter to electrical stimulation. In each preparation a conditioning electrical stimulus was delivered and the duration of the refractory period was quantitated by delivering a second, test stimulus at different time intervals from the conditioning stimulus. The response to the test stimulus, if any was expressed as percent of the amplitude of the response to the conditioning stimulus. Note that control ureter the absolute refractory period exists which exceeds 10 sec; thereafter a response to the test stimulus appeared and an interstimulus interval of 100 sec was required to obtain a test response identical with the first one. Both capsaicin and hCGRP (8-37) decreased the refractory period of the guinea-pig ureter indicating that that part of this phenomenon is ascribable to the release of CGRP by sensory nerves during the conditioning stimulus which influences the excitability of the ureter smooth muscle by the test stimulus. Reprinted with permission from Maggi and Giuliani (1994a).

Because voltage-dependent Ca currents undergo Ca-dependent inactivation (Tsien, 1983) and are essential for ureteral excitation and contractility (Sections III.B.1. and III.C.1.), it may be speculated that a reduced availability of Ca channels could be involved in determining the refractory period of the ureter. This hypothesis is supported by the observation (Maggi et al., 1994a) that the L-type Ca channel agonist, Bay K 8644, increases the excitability of guinea pig ureter to depolarization as follows: (a) in the presence of Bay K 8644, the refractory period of the ureter was markedly reduced compared with the control; (b) in the presence of Bay K 8644, multiple action potentials were evoked during delivery of a prolonged train of stimuli; (c) Bay K 8644 did not affect the resting membrane potential; and (d) Bay K 8644 did not evoke spontaneous electrical or mechanical activity of the unstimulated ureter. Because the action of Bay K 8644 on L-type Ca channels is not sufficient to produce membrane depolarization of its own, or to activate action potentials, it was concluded (Maggi et al., 1994a) that Bay K 8644 reduces the refractory period of the ureter by increasing the availability of Ca channels. In other words, a decreased availability of L-type Ca channels is one of the mechanisms that regulates the myogenic component of the refractory period of the guinea pig ureter.

Moreover, intracellular Ca, by modulating the resting membrane potential, also contributes to the regulation of the excitability of the smooth muscle of guinea pig ureter. In fact, drugs which deplete the internal Ca store, ryanodine or cyclopiazonic acid, induce a sustained depolarization of the membrane of guinea pig ureter and reduce the refractory period (Maggi et al., 1994a; 1995b). The effect was ascribed to the existence of a " basal" release of Ca from the internal store that regulates the resting membrane potential via activation of Ca-dependent K channels (Imaizumi et al., 1989a). A mild elevation of extracellular K, insufficient to evoke phasic contractions of the ureter, similarly reduces the refractory period of the ureter (Maggi et al., 1994a). Notably, the effect of Bay K 8644 and ryanodine on the excitability of the guinea pig ureter and reduction of the refractory period were additive: in the presence of both drugs, multiple action potentials and phasic contractions were elicited during application of repetitive stimuli at short time intervals (Maggi et al., 1994a).

A role of prostanoids in the local regulation of motility in the upper urinary tract has been postulated. The issue is of clinical relevance because cyclooxygenase (COX) inhibitors produce pain relief during renal colic (Holmlund and Sjodin, 1978; Lundstrom et al., 1982; Oosterlink et al., 1990).

Discrepant results have been reported with regard to the effect of prostaglandins on pyeloureteral motility. Prostaglandins of the E or F series produce excitatory effects on the motility of the ureter or renal pelvis (Thulesius and Angelo-Khattar, 1985; Lundstam et al., 1985; Thulesius et al., 1986, 1987; Cole et al., 1988). However, smooth muscle hyperpolarization, relaxation, and cyclic AMP accumulation also have been reported in response to added prostaglandins to the ureter (Johns and Wooster, 1975; Vermue and Den Hertog, 1987). Because different prostanoids can be produced by the same cell type in a species- and stimulus-dependent manner, the results obtained with exogenously added prostaglandins do not necessarily reflect the physiological role of endogenous prostanoids, although the effects produced by COX inhibitors are more informative in this respect.

Several groups have shown that COX inhibitors, e.g., indomethacin, produce a profound inhibitory effect on the spontaneous or evoked motility of isolated pyeloureteral smooth muscles (Thulesius and Angelo-Khattar, 1985; Lundstam et al., 1985; Thulesius et al., 1986, 1987; Cole et al., 1988; Kimoto and Constantinou, 1991). The general conclusion emerging from these studies is that the generation of prostanoids is an important step for the local autocrine/paracrine regulation of pyeloureteral motility. Indeed a complete suppression of pyeloureteral motility by COX inhibitors has been reported in some studies (Lundstam et al., 1985; Thulesius and Angelo-Khattar, 1985; Thulesius et al., 1987; Cole et al., 1988), which suggests that prostanoid generation is mandatory for the maintenance/activation of ureteral peristalsis. The evidence originating from this approach is, however, strictly limited by the selectivity and specificity of the drugs employed. To separate the specific from nonspecific effects of COX inhibitors on pyeloureteral motility, we have compared the effect of the (S)-enantiomer of ketoprofen (active in blocking COX) and of the (R)-enantiomer (inactive in blocking COX) (Hayball et al., 1992) to the effect of indomethacin.

We found that (S)-ketoprofen produced a stereoselective concentration- and time-dependent inhibition of the spontaneous myogenic activity of the guinea pig renal pelvis (Santicioli et al., 1995a). In a sucrose gap, (S)-ketoprofen produced a time-dependent shortening of the duration of spontaneous action potentials of the guinea pig renal pelvis and reduced the amplitude and duration of the accompanying phasic contractions. However, indomethacin totally suppressed the spontaneous activity of the renal pelvis. (S)-ketoprofen had no effect on contractions of the ureter induced by depolarizing electrical pulses nor on those induced by high K, although indomethacin had a marked suppressant effect on both responses. (S)-ketoprofen had no effect on the propagation of myogenic impulses along the ureter.

Overall, our findings (Santicioli et al., 1995a) demonstrated that stereoselective COX inhibition affects pacemaker potentials and contractility in the guinea pig renal pelvis. The modulatory role of endogenous prostanoids involves an amplification of electromechanical coupling in the renal pelvis although excitability, contractility, or propagation of impulses along the ureter are almost independent of prostanoid generation. Reports of a total suppression of pyeloureteral motility by indomethacin may reflect the additive inhibitory effect exerted by specific COX inhibition and nonspecific effects on electromechanical coupling. Considering the effect of (S)-ketoprofen as a pure consequence of full COX inhibition, it would follow that modulation of action potential duration is the main mechanism through which endogenous prostanoids regulate the pacemaker potentials of the guinea pig renal pelvis. These findings suggest an amplifying or facilitatory effect of endogenous prostanoids on the L-type calcium current which sustains the pacemaker potential of the renal pelvis.

The data presented by Zhang and Lang (1994) further define the modulatory role of endogenous prostanoids in regulating the spontaneous electrical and mechanical activity of the renal pelvis. These investigators showed that indomethacin inhibits the amplitude and frequency of action potentials from " driven" cells in the guinea pig proximal renal pelvis and eventually caused a failure of driven cells to fire action potentials although the underlying pacemaker potentials were unaffected. Zhang and Lang (1994) therefore proposed that endogenous prostanoids facilitate the coupling between pacemaker and driven cells, thereby allowing the spread of electrical activity in the renal pelvis.

The reinforcement of pacemaker activity of the renal pelvis by endogenous prostanoids could be of importance for the effect of COX inhibitors on renal colic. In addition to a possible central analgesic effect, a reduction of renal blood flow and a reduced urine production are held as major mechanisms through which COX inhibitors produce pain relief in renal colic (e.g., Perlmutter et al., 1993). Assuming that endogenous prostanoids reinforce the spontaneous excitation-contraction coupling in the renal pelvis, blockade of this effect by COX inhibitors may contribute to the overall pain relief produced by these drugs in renal colic, by reducing intraureteral pressure.

For more than a century, the "myogenic theory" of ureteral peristalsis has shaped investigations on ureteral physiology: great emphasis has been given to hydrodynamic factors, such as the rate of urine flow, in determining the size and pattern of urine boluses which, in turn, affect the mechanical aspects of rhythm, rate, peristaltic amplitude, and baseline pressure (Kiil and Setekleiv, 1973; Boyarsky and Labay, 1969). Although a neurogenic contribution has been proposed from time to time, there is little doubt that neural factors play, at best, only a modulatory role on ureteral peristalsis.

The current understanding of excitation-contraction coupling in the renal pelvis and ureter has been enriched by the characterization of the ionic currents responsible for changes in electrical activity and by pharmacological studies on the mechanisms of excitation-contraction coupling. The definition of the existence of a functional capsaicin-sensitive innervation (Section E.) and its selective elimination by capsaicin pretreatment have also enabled the study of the myogenic regulation of pyeloureteral motility under more controlled conditions.

In the renal pelvis, the pacemaker activity originates from a subpopulation of specialized cells in the inner muscular layer of renal calyces and pelvis (Gosling and Dixon, 1972, 1974): these cells resemble the interstitial cells of Cajal which are the pacemaker in the smooth muscle of the intestine (Huizinga et al., 1997; Lang et al., 1998). The nature of the "clock" producing the regular firing of spontaneous action potentials by these cells remains to be established.

The spreading of the pacemaker electrical activity through the smooth muscle of the renal pelvis and its coupling with contraction both involve the activation of voltage-dependent L-type Ca channels. These channels also are recruited obligatorily for propagation of ureteral peristalsis and electromechanical coupling in the ureter smooth muscle.

I_Ca flowing through L-type Ca channels is viewed as the main (Imaizumi et al., 1989a; Lang, 1989; Sui and Kao, 1997a,b) inward current in ureteral smooth muscle. A noninactivating or slowly inactivating component of I_Ca generates a "window" current (Imaizumi et al., 1989b; Sui and Kao, 1997b) that likely is responsible for determining the plateau phase of the action potential, as well as providing a sustained activation of Ca channels during prolonged depolarizing stimuli. Phosphorylation by PKA could be especially important for enabling L-type Ca channels to sustain a prolonged depolarization (Maggi et al., 1996b). L-type Ca channels in the ureter appear to be less prone to Ca-induced inactivation than in other smooth muscles (Sui and Kao, 1997b). This characteristic, and the absence of a voltage-sensitive Ca-independent K current (Imaizumi et al., 1989a; Lang, 1989), seem to be the main factors responsible for the peculiar long duration of the action potential of ureter. Considering the potentiating effects produced by Bay K 8644, charybdotoxin, and TEA, the inactivation of L-type channels and the activation of I_K(Ca) are the main factors responsible for terminating the action potential.

The nature of the oscillations observed during the plateau of the action potential of the guinea pig ureter remains controversial (Imaizumi et al., 1989a; Sui and Kao, 1997c). It is possible that the oscillations superimposed onto the plateau of action potential recorded from single cells and from multicellular preparations have a different origin.

A controversy exists in the literature concerning the role of Na currents and the Na/Ca exchanger in regulating the action potential. Several results indicate that a Na/Ca exchanger exists in the ureter, which works in reverse mode (Ca entering the cell) in Na-loaded (ouabain-pretreated) preparations. Considering the effects produced by removing extracellular Na on the action potential, it has been proposed that the Na/Ca exchanger provides part of the depolarizing current which accounts for the long duration of the action potential. However, no direct contribution of extracellular Na to the inward current has been observed (Sui and Kao, 1997a). Functional studies also have indicated that Na/Ca exchange is likely involved in modulating the contraction/relaxation cycle of the guinea pig ureter.

Release of Ca from intracellular store(s) is apparently not essential for providing activator Ca for excitation-contraction coupling but contributes to the setting of the resting membrane potential, probably through the activation of Ca-dependent K channels. Ca reuptake into internal stores is also a mechanism involved in terminating the contractile cycle of the ureter although other mechanisms support this function after blockade of the sarcoplasmic reticulum Ca pump. In particular, three distinct mechanisms could be important (Burdyga and Magura, 1988; Maggi et al., 1994a, 1995b): (a) Ca uptake into sarcoplasmic reticulum that seems especially suited for speeding up the relaxation of phasic contractions; (b) the Na/Ca exchange mechanisms; and (c) a Ca pump at the membrane level; the latter two mechanisms being possibly more important in terminating tonic than phasic-type contractions.

	IV. Innervation of the Renal Pelvis and Ureter

Top Previous Next References

The mammalian ureter is innervated mainly by unmyelinated fibers that originate from the renal, ovarian/spermatic, and sympathetic plexuses. The lower part of the ureter may receive a pelvic innervation at least in some species (Wharton, 1932; Saria et al., 1983). The sympathetic supply to the ureter arises from T11-L1 spinal segments. At least part of these fibers synapse in the distal pole of the inferior mesenteric ganglion (Janig and McLachlan, 1987).

The dorsal root ganglia of origin of the afferent innervation of the ureter have been identified by retrograde tracing studies at the level of L2-L3 and S1-S2 in guinea pigs and at the level of T11-L3 and L6-S1 in rats (Su et al., 1986; Semenenko and Cervero, 1992). In guinea pigs, approximately 40% of total labeled cells project to both ureters (Semenenko and Cervero, 1992).

Hoyes et al. (1975a) compared the innervation of the ureter in a variety of species and reported several species-related differences with regard to the frequency of nerve profiles in different layers of the ureter. It appears that the majority of axons terminate in the mucosa and that the innervation of the muscle of the renal pelvis is somewhat larger than that of the ureter (Notley, 1968; Gosling and Dixon, 1971). In humans, the lower ureter receives a denser innervation than the upper ureter, suggesting a major contribution from the pelvis plexus (Edyvane et al., 1992).

In the mucosa, nerve fibers form networks on the luminal aspect of the muscle layer and beneath the basement membrane of the epithelial layer (Dixon and Gosling, 1971; Hoyes et al., 1975b). Varicose nerve terminals originating from the subepithelial layer terminate in the intercellular spaces between the basal cells of transitional epithelium (Notley, 1968; Hoyes et al., 1975b). The existence of a dense mucosal/subepithelial innervation lends support to the hypothesis that the majority of ureteral nerves subserve an afferent function. This is further supported by the notion that pretreatment with toxic doses of capsaicin, causes degeneration in approximately 70 to 85% of axons in the rat or guinea pig ureter (Chung et al., 1985; Kiraly et al., 1991), whereas the subepithelial nerve plexus was resistant to treatment with neurotoxic doses of 6-hydroxydopamine (Hoyes et al., 1975b).

The distribution of afferent nerves to the renal pelvis and ureter also has been assessed directly by the anterograde transport of a tracer injected in rat dorsal root ganglia T12-L1 (Marfurt and Echtenkamp, 1991). In the ureter, the afferent innervation is densest in the proximal part and decreases caudally: labeled fibers were observed in all sectional layers of the ureter (adventitia, smooth muscle, subepithelium, and epithelium). The highest density of afferent innervation was seen in the renal pelvis with a similar distribution to that described in the ureter.

The existence and distribution of ureteric ganglia is a controversial topic: Elbadawi and Schenk (1969) reported the presence of ganglia in the lower two-thirds of the cat ureter. Beatty and Gabella (1988) and Mitchell et al. (1993) reported the presence of ganglia and individual neurons along the entire length of the guinea pig ureter. Some studies have failed to detect periureteral ganglia (Notley, 1968; Dixon and Gosling, 1971) although in other studies, the presence of ganglia at the uretero-vesical junction in several species has been noted (Gosling, 1970; Wharton et al., 1981). Periadventitial ureteric ganglia are numerous in the chicken ureter (Sann et al., 1992, 1997) and are thought to modulate ureteral peristalsis in this species.

Cholinergic nerves are present in the pyeloureteral tract of several species (Schulmann, 1985; Wharton et al., 1981; Prieto et al., 1994). Acetylcholinesterase-positive (AChE⁺) nerves are present in the adventitia, smooth muscle, and mucosa of cat and guinea pig ureter, along with periadventitial AChE⁺ positive ganglion cells in the distal part (Wharton et al., 1981). In the intravesical equine ureter, AChE⁺ nerve fibers distribute along blood vessels and smooth muscle and form a subepithelial plexus in the ureteral mucosa. AChE⁺ neurons are present also in adventitial small ganglia, which suggests that part of the cholinergic innervation is derived from this local source (Prieto et al., 1994). The density of AChE⁺ nerve fibers increases from the pelvic end of the equine ureter to the bladder: the intravesical region is the most densely innervated (Prieto et al., 1994). Sann et al. (1995a) recently reported that a substantial part of cholinergic nerve fibers in the subepithelial plexus of the rat ureter are sensory in origin. They found that nerve profiles staining positively for choline acetyl transferase immunoreactivity colocalize with CGRP-like immunoreactivity.

In organ bath experiments, acetylcholine increased the contractile activity (tone/phasic activity) of the pig intravesical ureter (Hernàndez et al., 1993), the sheep uretero-vesical junction (Rivera et al., 1992), the guinea pig renal pelvis (Maggi and Giuliani, 1992), and the equine ureter (Prieto et al., 1994). In the equine intravesical ureter, acetylcholine also determined a potent endothelium-dependent relaxation of precontracted arteries, via NO generation (Prieto et al., 1994). In the dog, i.v. acetylcholine increased the peristaltic frequency and decreased the volume of urine boluses but atropine had no significant effect on peristalsis (Morita et al., 1987).

There have been reports of contractile responses produced by electrical stimulation of intramural nerves sensitive to blockade by anticholinergic drugs in the guinea pig renal pelvis (Yoshida and Kuga, 1980) and reports of significant release of acetylcholine from human, guinea pig and rabbit isolated pyeloureteral tract in response to nerve stimulation (Del Tacca, 1978). However, we have been unable to demonstrate any significant atropine-sensitive local motor response to nerve stimulation in the guinea pig isolated renal pelvis or ureter (Maggi et al., 1986; Maggi and Giuliani, 1992, 1995). Atropine did not affect spontaneous activity of rabbit isolated renal pelvis nor responses to electrical stimulation of intramural nerves in this species (Del Tacca et al., 1974, 1981).

Postganglionic noradrenergic nerves have been demonstrated in all areas of the ureter of several species (Duarte-Escalante et al., 1969; Elbadawi and Schenk, 1969; Wharton et al., 1981; Schulmann, 1985; Prieto et al., 1993). These nerves are commonly observed along the adventitia and are associated with blood vessels, in both muscle and submucosal layers. Occasional fibers distribute within the smooth muscle. In the rat ureter, dopamine--hydroxylase positive (DBH⁺) nerves provide a dense periarteriolar innervation (Sann et al., 1995b) that colocalizes with NPY-like immunoreactivity. Some DBH⁺ neurons are present in periureteric ganglia in guinea pigs (Mitchell et al., 1993).

Tyrosine hydroxylase-positive (TH⁺) nerves distribute to the smooth muscle and around arteries/arterioles of the human ureter, also in the submucosal layer (Edyvane et al., 1992). TH⁺ nerves colocalize with NPY and account for approximately 50% of total nerve profiles in the human ureter (Edyvane et al., 1994).

Noradrenaline can either stimulate or inhibit renal pelvis and ureteral contractility via - and -adrenoceptors, respectively (Hannappel and Golenhofen, 1974b; Del Tacca et al., 1981; Morita et al., 1987; Hernàndez et al., 1992). -Adrenoceptor stimulation resulted in cAMP accumulation in the guinea pig ureter (Wheeler et al., 1986). A positive inotropic effect of isoproterenol was described in the more proximal part of rabbit renal pelvis (Morita, 1986). In dogs, i.v. noradrenaline caused an increase in peristaltic frequency, an elevation of intraureteral pressure, and a decrease in bolus volume with a resultant decrease in the rate of fluid transport; isoproterenol decreased peristaltic frequency and eventually suppresses ureteral peristalsis (Morita et al., 1987).

Application of guanethidine or that of phentolamine plus propranolol did not affect the local motor responses produced by stimulation of intramural nerves in the rat or guinea pig isolated renal pelvis or ureter (Maggi and Giuliani, 1992; Maggi CA, unpublished data). However, the contractile responses produced by electrical field stimulation in the rabbit renal pelvis seems to be mediated by catecholamines acting via -adrenoceptors (Gosling and Waas, 1971; Del Tacca et al., 1974).

Two tachykinins with an established status of neurotransmitters, substance P (SP) and neurokinin A (NKA), and CGRP, are present in primary afferent nerves distributing to the mammalian renal pelvis and ureter. SP/NKA are produced from alternatively spliced forms of preprotachykinin I gene mRNA encoding for both peptides (see Maggi et al., 1993 for review). CGRP also coexists with SP/NKA in many sensory nerves; CGRP positive (CGRP⁺) nerves exist in the ureter that do not colocalize with SP/NKA (Alm et al., 1978; Hokfelt et al., 1978; Sikri et al., 1981; Hua et al., 1986a,b; 1987; Su et al., 1986; Sann et al., 1992, 1995b; Zheng and Lawson, 1997).

The cells of origin of the SP/NKA/CGRP⁺ and CGRP⁺ nerves present in the mammalian upper urinary tract reside in dorsal root ganglia (Semenenko and Cervero, 1992). The peptides synthesized at the level of the neuronal somata are then exported to the periphery. In the guinea pig, most of the DRG neurons (approximately 90%) innervating the ureter are immunoreactive for SP or CGRP and approximately 65% are immunoreactive for both peptides (Semenenko and Cervero, 1992).

The distribution of SP/NKA/CGRP⁺ nerve fibers to the mammalian ureter has been detailed in several studies (Su et al., 1986; Tamaki et al., 1992; Sann et al., 1995b; Zheng and Lawson, 1997). In the renal pelvis, the fibers run parallel to the long axis of each of the circular and longitudinal muscle layers, resulting in a lattice-like appearance of the nerve fibers. In the ureter, the fibers accumulate in the subepithelial plexus, around blood vessels, and in the muscle layer. An important feature of the SP/NKA/CGRP⁺ nerves is their sub- and intraurothelial distribution. This location may enable the sensory nerves to detect a backflow of urine into the renal pelvis and ureteral wall: the density of fibers penetrating the urothelium seems larger in the renal pelvis than in the ureter (Zheng and Lawson, 1997).

SP/CGRP⁺ nerves are also present in the chicken ureter and likely represent sensory innervation also in this species, although sensory neurons in avians are not capsaicin-sensitive (Sann et al., 1992; 1997). In the chicken ureter, SP/CGRP⁺ nerves distribute mainly in the submucosa and around periureteric ganglia.

In the human ureter SP/CGRP⁺ nerves were occasionally seen in the smooth muscle but their density is lower than that seen in other species. SP/CGRP⁺ nerves are mostly present around blood vessels and in the submucosa of the ureter but are scarce in the smooth muscle. SP/CGRP⁺ nerves account for approximately 4% and 17% of total nerve profiles in the lower and upper ureter, respectively, and a further 5% of nerve profiles is CGRP⁺ at both levels (Tainio et al., 1991; Edyvane et al., 1992, 1994).

NPY⁺ nerves have been repeatedly described in the ureter of several species (Allen et al., 1990; Edyvane et al., 1992; Prieto et al., 1997). Their density is markedly species dependent: NPY coexists with noradrenaline in some sympathetic nerves that show a perivascular localization in the rat ureter, also surrounding arterioles in the submucosa. NPY⁺ nerves in the rat ureter are sensitive to 6-hydroxydopamine treatment, indicating their noradrenergic origin (Allen et al., 1990). The density of NPY⁺ nerves markedly increases after the administration of neurotoxic doses of capsaicin. This effect could be explained by a competition existing between sensory and noradrenergic nerves for the availability of trophic factors produced by the innervated tissue that promote neuron growth/survival (Sann et al., 1995b). Some NPY⁺ neurons were reported in guinea pig periureteric ganglia (Mitchell et al., 1993). Prieto et al. (1997) described a gradient in the density of NPY⁺ nerves in the equine ureter, which is maximal at the level of the intravesical region. This region is proposed to act as a functional sphincter by facilitating urine discharge during bladder filling and by preventing vesico-ureteral reflux during micturition. Prieto et al. (1997) described the presence of numerous NPY⁺ nerves in the smooth muscle, around arteries, and within adventitial ganglia of the intravesical equine ureter. The latter also were found to contain some NPY⁺ neuronal bodies, therefore, the NPY innervation could have a local origin.

NPY⁺ nerves are seen in the smooth muscle and around all blood vessels but not in the epithelium of the human ureter (Tainio et al., 1991). Edyvane et al. (1992) reported that a vast majority (approximately 80%) of intramural nerves in the human ureter are NPY⁺. The number of NPY⁺ nerves is greater than that of noradrenergic nerves, which suggests the existence of a separate population of nonadrenergic NPY⁺ neurons. The same group subsequently showed that approximately 50% of total nerve profiles in the human ureter colocalizes TH and NPY, although a further 30% of nerve profiles colocalizes NPY and VIP but not TH (Edyvane et al., 1994). Both NPY/VIP⁺ and NPY/TH⁺ nerves distribute around blood vessels but the two populations differ with regard to their distribution to the inner and outer muscle layers, respectively (Edyvane et al., 1994).

Exogenously added NPY did not affect muscle tone of the isolated equine ureter, but markedly potentiated the contractile responses to noradrenaline, an effect apparently mediated via Y₂ receptors. NPY produced concentration-dependent vasoconstriction of ureteral resistance arteries and potentiated the response to noradrenaline via Y₁ receptors (Prieto et al., 1997). Contrary to its effect in other organs, NPY does not modulate the release of sensory neuropeptides from peripheral endings of sensory nerves in guinea pig ureter (Maggi and Giuliani, 1995).

VIP positive (VIP⁺) nerves have been detected in all layers of the cat and guinea pig ureter, and are more frequently observed in the former species (Wharton et al., 1981). A small number of VIP⁺ periadventitial ganglion cells also was observed in cats (Wharton et al., 1981; Mitchell et al., 1993). VIP⁺ nerve fibers are present in all layers of the rat ureter and show a distinct regional distribution, but are absent in the renal pelvis and upper ureter, appearing in the middle ureter and then remaining constant in the lower part of the ureter (Sann et al., 1995b). VIP⁺ nerves in the rat ureter are unaffected by capsaicin pretreatment and, owing to their distribution, were thought to originate from VIP⁺ neurons in the pelvic ganglion (Sann et al., 1995b).

VIP⁺ nerves are seen also in the smooth muscle, around blood vessels, and adjacent to the epithelium in the human ureter: approximately 40% of all nerve bundles in human ureter are VIP⁺ (Tainio et al., 1991; Edyvane et al., 1992). The major part of VIP⁺ nerves are also NPY⁺ and form a population distinct from the NPY/TH⁺ nerves (Edyvane et al., 1994; 1995).

Nerve profiles displaying immunoreactivity for nitric oxide synthase (NOS⁺) or NADPH diaphorase (NADPH⁺) have been described in the human (Smet et al., 1994; Goessl et al., 1995; Iselin et al., 1997), pig (Hernàndez et al., 1995; Iselin et al., 1997), and sheep ureter (Garcìa-Pascual et al., 1996). NOS/NADPH⁺ nerves distribute to the smooth muscle, around arteries, and in the subepithelial layers. NOS⁺ nerves colocalize with both VIP and NPY but never with TH in the human ureter (Smet et al., 1994). NADPH⁺ neurons are present also in ureterovesical ganglia in humans (Grozdanovic and Baumgarten, 1996).

NO or NO donors relaxed the smooth muscle of the pyeloureteral tract from several species and this effect appeared to be linked to an elevation of intracellular cGMP levels (Iselin et al., 1996; 1997).

However, NO synthase blockers apparently did not affect the motility of the guinea pig isolated renal pelvis or ureter (Maggi et al., 1995a, c) or that of the sheep ureter (Garcìa-Pascual et al., 1996). NO synthase inhibitors blocked the nonadrenergic noncholinergic nerve-mediated relaxations of the pig intravesical ureter, which implies a transmitter role of NO or of an NO-like substance at this level in this species (Hernàndez et al., 1995). In the latter preparation, glibenclamide, but not charybdotoxin or apamin, blocked in parallel the relaxant effect of exogenous NO and that of the endogenous nonadrenergic noncholinergic transmitter, which indicates that endogenous NO acts by opening glibenclamide-sensitive K_ATP channels (Hernàndez et al., 1997).

	V. Sensory Neuropeptides in the Pyeloureteral Complex: Release, Actions, and Receptors

Top Previous Next References

Because of the dense innervation and high content of sensory neuropeptides, the guinea pig renal pelvis and ureter have been used extensively as test objects to investigate the mechanisms of release of sensory neuropeptides (SP, NKA, and CGRP). Saria et al. (1983) first reported that the application of capsaicin causes the release of SP from the guinea pig ureter, a finding confirmed in several subsequent studies and extended to the demonstration that depolarizing stimuli, including electrical stimulation of sensory nerves, also induce the concomitant release of NKA and CGRP in the guinea pig and rat ureter (Hua et al., 1986a; Amann et al., 1988a; Santicioli et al., 1988; Dray et al., 1989; Maggi et al., 1990, 1992b,c). In addition to capsaicin and electrical stimuli, chemical agents, such as bradykinin and bacterial peptides, are effective stimulants of the release of sensory neuropeptides in the ureter and renal pelvis (Maggi et al., 1992a).

The release of sensory neuropeptides from the peripheral endings of capsaicin-sensitive primary afferent nerves is Ca dependent. Some stimuli, including depolarization by electrical pulses, use N-type voltage-dependent Ca channel for inducing secretion of sensory neuropeptides, because the response was inhibited by -conotoxin fraction GVIA (Maggi et al., 1990; Maggi and Giuliani, 1991). The release process is also induced by -latrotoxin, a component of black widow spider venom: this finding implies the involvement of the high affinity -latrotoxin receptor, neurexin Ia, in the secretion of sensory neuropeptides from afferent nerve terminals (Waterman and Maggi, 1995). However, the degree of utilization of N-type Ca channels for secretion of neuropeptides from ureteral afferent nerves is species-dependent: in the rat ureter CGRP release induced by electrical stimulation is not blocked by -conotoxin fraction GVIA (Maggi and Giuliani, 1991). Other stimuli, including capsaicin, produce the release of sensory neuropeptides via a mechanisms that is Ca dependent but does not require the activation of voltage-dependent Ca channels (Maggi, 1995 for review).

The tachykinins, SP and NKA, are powerful stimulants of pyeloureteral motility: in the renal pelvis this effect is evident as a positive chrono- and inotropic response, producing a marked potentiation of phasic contractility and, at high concentrations, an elevation of tone (Maggi et al., 1992c,d). In the guinea pig renal pelvis, both NK₁ and NK₂ receptors mediate the inotropic and chronotropic effects of tachykinins (Maggi et al., 1992d). When applied to the isolated ureter, which is electrically and mechanically quiescent, tachykinins induced the appearance of phasic contractions sustained by the firing of action potentials, not dissimilar from those evoked by direct electrical stimulation of smooth muscle cells (Hua et al., 1986a,b; Maggi et al., 1986; 1987a; 1988a; Amann et al., 1988). In the guinea pig and human ureter, the application of NKA produced a nifedipine-resistant depolarization onto which a series of nifedipine-sensitive action potentials were superimposed (Patacchini et al., 1998). In the ureter from both species, a series of phasic contractions was produced by NKA in parallel to the evoked action potentials, and these phasic contractions were suppressed by nifedipine. In the human ureter, a nifedipine-resistant tonic contraction also was evidenced (Patacchini et al., 1998).

In sharp contrast with the contractile effect of tachykinins, CGRP inhibits the motility of the isolated renal pelvis and ureter (Hua et al., 1986a,b; 1987; Maggi et al., 1987b). The effect of CGRP is especially evident in the ureter as a suppression of evoked motility: the all-or-none suppressant effect occurs because CGRP abolishes the firing of action potentials evoked either by electrical stimulation of chemical agents (see next section). A descending gradient exists in the guinea pig pyeloureteral tract regarding sensitivity to the inhibitory effect of CGRP: the ureter is extremely sensitive (Maggi et al., 1987b; Maggi and Giuliani, 1991), whereas the spontaneous activity of the renal pelvis is inhibited but not suppressed by this peptide (Maggi et al., 1992c). The sensitivity to CGRP appears in the distal region of the renal pelvis although the motility of the proximal renal pelvis is substantially unaffected by CGRP (Maggi et al., 1995a). Moreover, although the suppressant effect of CGRP in the ureter is largely linked to the activation of glibenclamide-sensitive K channels (see Section V.C.), the inhibitory effect of CGRP in the renal pelvis is glibenclamide resistant (Maggi et al., 1995a). However, cromakalim exerts a glibenclamide-sensitive suppressant effect in both the renal pelvis and ureter; therefore, part of the gradient in sensitivity to the inhibitory effect of CGRP seems to be linked to a differential coupling of CGRP receptors with distinct effector mechanisms.

In the ureter, CGRP totally suppresses the propagation of impulses and, in this way, prevents ureteral peristalsis and suppresses antiperistalsis generated by stimulation of latent pacemakers (Meini et al., 1995).

Sann et al. (1992) analyzed the distribution of SP and CGRP receptors in the guinea pig ureter: the SP binding sites (NK₁ receptor) were associated with blood vessels and the epithelium with the density order: venules > epithelium > arterioles; although CGRP binding sites were chiefly distributed on smooth muscle. This distribution agrees with the notion that tachykinin NK₁ receptors mainly mediate inflammatory reactions initiated by the stimulation of afferent nerves (Section V.D.), although CGRP mainly mediates the concomitant changes in smooth muscle activity. In the chicken ureter SP binding sites also are prominent in neurons of ureteric ganglia (Sann et al., 1992; 1997). Applied SP evoked the depolarization of a subset of neurons in the chicken ureteric ganglia and desensitization to SP reduced the responsiveness of these neurons to mechanical stimuli, suggesting that locally released SP may play a role in modulating neuronal activity in this species (Sann et al., 1997).

Peptidases limit the extent and duration of action of tachykinins and CGRP in the pyeloureteral tract: in the renal pelvis, the application of peptidase inhibitors potentiated the contractile effect of exogenous and endogenous tachykinins (Maggi et al., 1992d). In the guinea pig ureter, the inhibitory action of CGRP was reduced because of degradation by a thiorphan-sensitive peptidase, presumably neutral endopeptidase (Maggi and Giuliani, 1994b).

Since the early studies on this topic, tachykinins and CGRP have been proposed to act as transmitters producing a local modulation of motility in the pyeloureteral tract. Hua et al. (1986b) reported that the i.v. injection of capsaicin inhibits ureteric motility in anesthetized guinea pigs when administered at a low dose, whereas a biphasic response (inhibition followed by excitation) was produced when capsaicin was administered at a high dose. They also showed that i.v. administered CGRP produces inhibition although i.v. injected NKA produces excitation of ureteric motility. Hua et al. (1986b) speculated that the dual effect produced by a high dose of capsaicin may involve a differential release of these neuropeptides. Hua et al. (1986b) also showed that electrical stimulation of the inferior mesenteric ganglion produces a dual effect on ureteric motility similar to that exerted by a high dose of i.v. capsaicin. These findings were interpreted as an indication that transmitters released from sensory nerves can affect ureteral motility in vivo, probably through a local effect in the ureteral wall (Hua et al., 1986b). The evidence that sensory nerves do indeed exert a local control of ureteral motility by producing transmitter release in the ureter wall was provided by an independent study (Maggi et al., 1986). After having induced the appearance of a background phasic motility, we found that electrical field stimulation (EFS) of intramural nerves or bath application of capsaicin produced a transient inhibitory effect of the motility of the rat isolated ureter by suppressing the activity of latent pacemakers (Maggi et al., 1986). The response to EFS was abolished by tetrodotoxin, by pretreatment with high doses of capsaicin or by cold denervation, whereas it was unaffected by chronic bilateral removal of the pelvic ganglia (Maggi et al., 1986; 1987a). Overall, these data indicated that capsaicin exerts its effect in the rat ureter through a subset of intramural sensory nerves which are activated antidromically by EFS to release an inhibitory mediator, which was later identified as CGRP (Maggi et al., 1987b).

Amann et al. (1988a) showed that tachykinin and CGRP-like immunoreactivity are present in approximately similar amounts in the guinea pig ureter, whereas CGRP levels are approximately 33-fold higher than tachykinin levels in the rat ureter. Moreover, the application of capsaicin produced the corelease of tachykinin and CGRP-like immunoreactive material from guinea pig isolated ureter whereas only the release of CGRP-like immunoreactivity was detected from the rat iso lated ureter (Amann et al., 1988a). Under comparable in vitro conditions, Amann et al. (1988a) showed that the application of capsaicin exerts both excitation of latent pacemakers (ascribable to release of tachykinins) and suppression of NKA-activated latent pacemakers (ascribable to release of CGRP) in the guinea pig isolated ureter, whereas only an inhibitory effect could be demonstrated in that rat isolated ureter. Overall, the findings of Amann et al. (1988a) established that the different quality of the local response to capsaicin observed in the rat and guinea pig isolated ureter represents a true species-related difference, likely linked to the different ratios of CGRP and tachykinins stored and released from sensory nerve terminals.

A local control of motility by neuropeptides (tachykinins and CGRP) released from the peripheral endings of capsaicin-sensitive sensory nerves also was demonstrated in the guinea pig renal pelvis (Maggi and Giuliani, 1992; Maggi et al., 1992c; Patacchini et al., 1998): the prevailing motor response produced by capsaicin or EFS is the induction of a long lasting positive chrono- and inotropic effect that is tachykinin mediated. After blockade of the excitatory action of tachykinins, an unopposed transient inhibitory effect, CGRP-mediated, was disclosed.

The development of suitable receptor antagonists has enabled the conclusive identification of tachykinins and CGRP as the mediators responsible for the excitatory and inhibitory effects, respectively, observed upon stimulation of sensory nerves in the pyeloureteral tract (see Maggi, 1995 for review). In particular, tachykinin receptor antagonists blocked the local excitatory motor response produced by sensory nerve stimulation in the renal pelvis (Maggi et al., 1992c,d). The CGRP receptor antagonist, CGRP (8-37), was used to prove the involvement of endogenous CGRP in the inhibition of renal pelvis motility observed after blockade of tachykinin receptors (Maggi et al., 1992c) and in the transient suppression of latent pacemakers produced by nerve simulation in the rat and guinea pig isolated ureter (Maggi and Giuliani, 1991).

The local changes in motility produced by stimulating sensory nerves in the renal pelvis and ureter are sustained by changes in the electrical properties of smooth muscle cells. In the ureter, electrical stimulation of intramural nerves determined a transient tetrodotoxin-sensitive hyperpolarization or inhibitory junction potential (i.j.p.) (Santicioli and Maggi, 1994; fig. 10) whose latency, amplitude and duration are frequency-dependent. The i.j.p. was magnified when evoked in low-K medium and was blocked by a CGRP receptor antagonist. The acute capsaicin application determined a transient hyperpolarization of ureteral smooth muscle. The blocker of K_ATP channels, glibenclamide also blocked the EFS-induced i.j.p. indicating that endogenous CGRP acts as a K channel opener in producing its transmitter action (Santicioli and Maggi, 1994). However, EFS caused membrane depolarization and increased the frequency of action potential discharge in "pacemaker" cells and prolongation of action potential duration and reduced afterhyperpolarization in "driven" cells of the guinea pig renal pelvis (Lang et al., 1995). In nifedipine-arrested cells of the renal pelvis, nerve stimulation produced a tetrodotoxin-sensitive excitatory junction potential (Lang et al., 1995), which is presumably tachykinin mediated.

View larger version (17K): [in this window] [in a new window]   Fig. 10.   Recording of membrane potential from the guinea-pig ureter by ucrose gap: application of electrical pulses (5 Hz for 10 sec, 0.1 ms pulse width) by field stimulation determines an inhibitory junction potential (i.j.p.) mediated by CGRP released from the peripheral endings of capsaicin-sensitive primary afferent neurons. The EFS-evoked i.j.p. is abolished by tetrodotoxin (TTX, panel A) and reduced by the CGRP receptor antagonist, CGRP (8-37) (panel B). In panel C, the application of capsaicin determines a prompt and large hyperpolarization: after application of capsaicin (3 µM for 15 min before washout) the EFS -evoked i.j.p. is abolished because of desensitization of primary afferent nerves. Reprinted with permission from Santicioli and Maggi (1994).

Overall, these findings demonstrate that tachykinins and CGRP fulfill all the classical criteria required to establish the neurotransmitter role of a putative mediator. The source of the released transmitters (primary afferent nerves) poses interesting limits to the physiological and pathophysiological significance of this process, which is further complicated by the existence of regional variations whereby the action of tachykinins or that of CGRP becomes more apparent when not prevalent. The significance of this transmitter role in integrated models of ureteral peristalsis is not understood yet. A role for CGRP in the local modulation of spontaneous activity of the guinea pig renal pelvis has been recently proposed: Teele and Lang (1996, 1998) reported that application of a local stretch to the proximal renal pelvis determines an inhibitory effect on the contractility of the distal renal pelvis which is partially prevented by the CGRP receptor antagonist, CGRP (8-37), or by glibenclamide, which suggests a role of endogenous CGRP in the modulation of migrating contractions in the renal pelvis.

From the evidence reviewed in the previous sections, it seems that CGRP is the main mediator involved in the local regulation of ureteral motility: its main effect can be described as a powerful suppression of latent pacemakers of the ureter smooth muscle. Three mechanisms, not necessarily unrelated to each other, have been described to account for the smooth muscle relaxant activity of CGRP: (a) stimulation of adenylyl cyclase and elevation of intracellular cAMP; (b) generation of nitric oxide (NO); and, (c) activation of glibenclamide-sensitive potassium (K) channels (see Poyner, 1992 for review).

On this background, we have studied the mechanism of the inhibitory action of CGRP in the guinea pig ureter: to eliminate the confounding influence of endogenous CGRP from the experimental setup, our studies were performed in capsaicin-pretreated ureters, to block CGRP release from sensory nerves.

In organ bath experiments, regular phasic contractions of the ureter (twitches) (fig. 11) were obtained by applying EFS with pulses of sufficient intensity to excite latent pacemakers thus producing action potentials and accompanying twitches (Maggi and Giuliani, 1994a; Maggi et al., 1994a,b,c; 1995c; 1996a,c). When applied cumulatively, CGRP produced a concentration-dependent inhibition of the EFS-evoked twitches with an EC₅₀ of 3.3 nM (Maggi et al., 1994c). The NO synthase inhibitor, L-nitroarginine, did not affect twitches or prevent the action of CGRP (Maggi et al., 1994c). Moreover, CGRP did not elevate cGMP levels in the ureter (Santicioli et al., 1995b). Therefore, the NO/cGMP pathway cannot be held responsible for the action of CGRP. However, glibenclamide produced a profound, albeit incomplete, inhibition of the relaxant effect of CGRP (Maggi et al., 1994c). When tested on the cumulative concentration-response curve to CGRP, glibenclamide produced a pure depression of the E_max to CGRP without determining any significant rightward shift of the curve. This indicates that glibenclamide blocks an effector mechanism (K channels) downstream to CGRP receptor occupancy (Maggi et al., 1994c). Both 1 and 10 µM concentrations of glibenclamide were equally effective in reducing the E_max of the inhibitory effect of CGRP (Maggi et al., 1994c): therefore, a distinct fraction (approximately 35%) of the effect of CGRP is glibenclamide-resistant.

View larger version (37K): [in this window] [in a new window]   Fig. 11.   Tracings (upper and middle panels) and quantitative data (lower panels) showing the time course of the inhibitory effect produced by maximally effective concentrations of CGRP (0.1 µM), cromakalim (3 µM) and forskolin (10 µM) in the absence and presence of glibenclamide (1 and 10 µM). Upper panels shows the control response to the three agonists in the absence of glibenclamide. Middle panels shows the response obtained in paired ureters from the same animals in the presence of 10 µM glibenclamide. Vertical bars are % of the response to KCl. Lower panel shows the response to CGRP (left), cromakalim (middle) and forskolin (right) in the absence (circles) and presence of 1 µM (squares) or 10 µM (triangles) glibenclamide. Each value is mean ± s.e.m. of 6-16 observations.*significantly different from control values, P < 0.05. Reprinted with permission from Maggi et al. (1995c).

The existence of two distinct components in the relaxant action of CGRP, operationally defined as glibenclamide sensitive and glibenclamide resistant, was further demonstrated in experiments in which a single maximally effective concentration of CGRP (0.1 µM) was applied to the bath (fig. 11; Maggi et al., 1995c, 1996a,c; Maggi and Giuliani, 1994b). These experiments demonstrated that the efficiency of the mechanism of action of CGRP in inhibiting twitches is markedly different: in fact, in the absence of glibenclamide, CGRP produced a prompt and transient total suppression of twitches which recovered with an amplitude lower than that observed before CGRP application (fig. 11). The suppressant effect of CGRP on twitches was abolished by glibenclamide: in the presence of glibenclamide, a slowly developing partial inhibitory effect occurred in response to CGRP (approximately 30% inhibition of twitch amplitude) (Maggi et al., 1995c, 1996a; fig. 11). Notably, under comparable experimental conditions, glibenclamide totally eliminates the suppressant effect of the K_ATP channel opener, cromakalim (Maggi et al., 1995c; fig. 11).

The glibenclamide-sensitive suppressant effect of CGRP was influenced by several variables as follows: (a) when applying electrical stimuli of threshold intensity, very low concentrations of CGRP (1 to 3 nM) were sufficient to produce a transient suppression of evoked contractility (Maggi and Giuliani, 1994a), whereas low concentrations of CGRP did not suppress twitches evoked by suprathreshold stimuli (Maggi et al., 1994c); (b) the transient nature of the early suppressant effect of CGRP was not ascribable to receptor tachyphylaxis and partly involved CGRP degradation by peptidases, via a thiorphan-sensitive mechanism (Maggi and Giuliani, 1994b); however, (c) even in the presence of peptidase inhibitors, twitch suppression by CGRP was inherently transient (Maggi et al., 1994b); and, (d) upon recovery from suppression, the amplitude of twitches stabilized to approximately 60 to 70% of their original amplitude; the recovery occurred in an all-or-none manner (fig. 11).

The effect of CGRP and cromakalim on the biphasic contractile response induced by elevating [K]_o was also investigated (Maggi et al., 1994c; 1995c). KCl (80 mM) caused a biphasic contraction of the guinea pig ureter, both of which components involved the activation of L-type Ca channels (see Section III.C.1.) but can be pharmacologically differentiated on the basis of their differential sensitivities to Ca channel blockers and PKA inhibitors. CGRP exerted a profound inhibitory action on both components of the response to KCl, whereas cromakalim selectively suppressed the phasic response without affecting the tonic component of contraction (Maggi et al., 1994c, 1995c). Moreover glibenclamide effectively prevented the ability of CGRP (and cromakalim) to suppress the phasic contraction to KCl without affecting the inhibitory effect of CGRP on the tonic contraction to KCl (Maggi et al., 1994c).

These observations further support the concept that the glibenclamide-sensitive component of CGRP action is especially effective in producing a transient suppression of phasic contractions of the ureter, which are sustained by the firing of action potentials. The glibenclamide-resistant mechanism of CGRP action has a limited efficacy in inhibiting phasic contractions but can exert a profound suppression of tonic contractions sustained by prolonged depolarization.

1. CGRP, hyperpolarization and blockade of phasic contractions. The efficiency of CGRP action in producing a total suppression of phasic contractions of the ureter suggested an indirect effect on smooth muscle contractility. CGRP produced a glibenclamide-sensitive hyperpolarization of the smooth muscle of guinea pig ureter (Santicioli and Maggi, 1994) which closely resembles the hyperpolarizing and action of cromakalim (Maggi et al., 1994b). By producing hyperpolarization, CGRP prevented the firing of action potentials to depolarizing stimuli (fig. 12). In the presence of glibenclamide, CGRP shortened the action potential duration and reduced the amplitude of the accompanying contraction, but the hyperpolarization and the consequent suppressant effect were abolished (Maggi et al., 1994a). A glibenclamide-sensitive hyperpolarization was produced also by endogenous CGRP (Santicioli and Maggi, 1994) and this effect is capable to suppress action potentials and twitches produced by chemical stimulation of latent pacemakers (Maggi et al., 1987b; Maggi and Giuliani, 1991).

View larger version (18K): [in this window] [in a new window]   Fig. 12.   Action potential (recorded by sucrose gap technique) and phasic contraction produced by electrical stimulation of the guinea-pig ureter: note the hyperpolarization and transient suppression of evoked electrical and mechanical responses produced by CGRP (0.3 µM for 15 sec, applied at bars) and antagonism of the action of CGRP by glibenclamide. Note that in presence of glibenclamide CGRP produced a small inhibition of the amplitude of evoked contractions. Reprinted with permission from Maggi et al. (1994c).

2. CGRP and cAMP accumulation. In arterial and gallbladder smooth muscle, CGRP activates a class of K_ATP channels via cAMP accumulation and stimulation of PKA (Quayle et al., 1994; Zhang et al., 1994a,b; fig. 13A). Therefore we addressed the questions: (a) does CGRP stimulate adenylyl cyclase to produce cAMP accumulation in the guinea pig ureter and (b) is this effect linked, possibly via stimulation of PKA, to activation of K channels?

View larger version (20K): [in this window] [in a new window]   Fig. 13.   Schematic drawing illustrating the proposed chain of events linking the occupancy of CGRP receptors to activation of K channels in the guinea-pig ureter (see text for details). According to the available data, CGRP would activate adenylyl cyclase and produce cAMP accumulation in ureter smooth muscle cells which, in turn, activates protein kinase A (PKA). PKA could directly phosphorylate glibenclamide-sensitive K channel leading to hyperpolarization (scheme A). This chain of events has been proposed as being responsible for K channel activation by CGRP in other smooth muscles. However several considerations, detailed in the text argue against a plain relationship between cAMP accumulation and K channel activation in the guinea-pig ureter. Moreover, pharmacological evidence indicates that mobilization of Ca from the sarcoplasmic reticulum is involved in the action of CGRP: blockers of sarcoplasmic reticulum Ca mobilization prevent CGRP-induced acivation of K channels in the same manner as PKA inhibitors or cAMP antagonist do. From this evidence, reviewed in the text, an alternative scheme is proposed whereby PKA activated by cAMP elevation determines Ca release from the sarcoplasmic reticulum, producing activation of a class of Ca-sensitive K channels blocked by glibenclamide (scheme B).

1.	First, membrane-permeable and metabolically stable cAMP analogs failed to reproduce the glibenclamide-sensitive early suppressant action of CGRP, forskolin, and IBMX (Maggi et al., 1995c). A lack of mimicry of putative cAMP-mediated relaxations by cAMP analogues has been repeatedly described (see Murray, 1990 for review). The failure of cAMP analogs to reproduce the profound and transient glibenclamide-sensitive inhibition of twitches may reflect the importance of a kinetic factor in the action of CGRP, forskolin and IBMX. A common element to the latter agents is that cAMP levels are elevated within the cells, either after membrane receptor occupancy (CGRP), direct adenylyl cyclase stimulation (forskolin) or prevention of cAMP degradation (IBMX). It may be suggested that membrane permeation by the cAMP analogues is too slow to reproduce a brisk elevation of intracellular cAMP (and PKA activation), a factor that may be required for activation of KATP channels. The existence of a kinetic factor in the action of CGRP is further suggested by the transient nature of the suppressant effect produced by these agents, because the twitch amplitude recovered, after a transient suppression, despite the continued presence of forskolin (fig. 11) and IBMX (Maggi et al., 1995c).
2.	The second set of "discrepant" data arises from a kinetic comparison of the inhibitory effect of CGRP (or forskolin) versus their ability to elevate cAMP levels. Both CGRP and forskolin produced a prompt hyperpolarization of the membrane (Santicioli and Maggi, 1994; Santicioli et al., 1995b) and this " quick" effect matched the time-course of the glibenclamide-sensitive suppression of twitches. However, although an elevation of cAMP was detected at 2 min after the application of CGRP or forskolin, cAMP levels further increased with time reaching a plateau approximately 15 min after administration of either drug (Santicioli et al., 1995b). Therefore, when cAMP elevation is causally related to K channel activation along the scheme depicted in fig. 13, it seems necessary to postulate the existence of a " compensatory" mechanism that limits the duration K channel activation despite the progressive elevation of cAMP levels induced by CGRP and forskolin. One possibility is that the production of cAMP causes a localized decrease in ATP concentrations thus activating KATP channels (see also Section V.C.4.); such a mechanism could account for the transient nature of K+ channel activation by CGRP due to diffusion of ATP from the larger volume of the cytosol into the submembrane space.
3.	The third set of "discrepant" data arises from a quantitative comparison of the elevation of cAMP levels (Santicioli et al., 1995b) versus the amplitude/duration of the hyperpolarization produced by CGRP and forskolin. Forskolin was more effective in elevating cAMP than CGRP (Santicioli et al., 1995b), yet CGRP was more effective in producing a glibenclamide-sensitive inhibition of twitches (Maggi et al., 1995c). This discrepancy may have several explanations: (a) different compartmentalization of increased intracellular cAMP produced by CGRP and forskolin; (b) a K channel blocker action of forskolin itself (Krause et al., 1988; Inoue et al., 1993), auto-limiting the extent/duration of the induced hyperpolarization; (c) the existence of a depolarizing component in the action of forskolin, which was unmasked in the presence of 10 µM glibenclamide (Santicioli et al., 1995b); and (d) contrary to forskolin, which directly activates adenylyl cyclase, the action of CGRP is likely to be mediated by a G-protein coupled to CGRP receptors (see Poyner, 1992 for review); a facilitatory influence of G-proteins on KATP channel activation through an antagonism of ATP-dependent gating (Ito et al., 1994; Terzic et al., 1994) may be postulated.

3. Role of intracellular calcium in the action of calcitonin gene-related peptide. We next became interested in the possibility that mobilization/reuptake of Ca from the sarcoplasmic reticulum may be involved in causing the transient nature of K channel activation by CGRP. In fact, the sarcoplasmic reticulum Ca ATPase is a known target for cAMP-dependent phosphorylation by PKA. PKA can either increase Ca mobilization by, for example, phosphorylating the Ca-induced Ca release channel (Coronado et al., 1994 for review) or, by phosphorylating the sarcoplasmic reticulum protein, phospholamban. PKA can increase the efficiency of the pump in removing free intracellular Ca (see Suematsu et al., 1984; Komori and Bolton, 1989; Murray, 1990 for review). Moreover, the results of our studies had indicated that the sarcoplasmic reticulum may be involved in regulating the resting membrane potential of the guinea pig ureter (see Section III.C.).

View larger version (31K): [in this window] [in a new window]   Fig. 14.   Experiments illustrating the dual effect of the sarcoplasmic reticulum Ca pump inhibitor, cyclopiazonic acid (CPA) on the suppressant effect of CGRP on twitch contractions elicited by electrical stimulation of latent pacemakers in the guinea-pig ureter. The experiments demonstrate that a high concentration of CPA, assumed to empty the sarcoplasmic reticulum Ca store, blocks the ability of CGRP to suppress twitches (Panels A and B), whereas a low concentration of the drug assumed to produce a limited but incomplete inhibition of the Ca pump actually potentiates the effect of CGRP (Panels C and D). Panels A and B: Effect of 60 min exposure to 10 µM cyclopiazonic acid (CPA) on the inhibition of EFS-induced twitch contractions of guinea-pig ureter induced by 0.1 µM CGRP (Panel A) or 3 µM cromakalim (Panel C). Vertical lines are % of the maximal contraction to 80 mM KCl. Note that CGRP and cromakalim both suppress of twitches: the suppressant action of CGRP was prevented by CPA whereas the suppressant effect of cromakalim was unchanged. Panels C and D: Effect of a low concentration of cyclopiazonic acid (CPA, 1 µM for 10 min) on the inhibitory effect produced by CGRP (0.1 µM) on EFS-induced twitches of guinea-pig ureter. Electrical stimuli were delivered at a driving frequency of 1 stimulus every 60 sec (Panel C) or 15 sec (Panel D): CPA markedly prolonged the suppressant effect of CGRP at both driving frequencies. Note that when ureters were driven at a higher frequency (Panel D) the duration of the suppressant effect of CGRP is shorter than at a lower driving frequency (Panel C). Reprinted with permission from Maggi et al. (1996a).

4. Influence of "exercise" and glucose metabolism on the action of calcitonin gene-related peptide. We had noted previously (Maggi et al., 1987b) that the duration of the glibenclamide-sensitive twitches-suppressant effect of CGRP in the ureter is both transient and inversely related to the frequency of evoked ureteral contractions. Therefore, it appeared of interest to investigate this aspect of CGRP action in relation to the proposed chain of events underlying its mechanism of action at cellular level (fig. 13B). Glibenclamide-sensitive K_ATP channels are thought to provide a link between cell metabolism and excitability (see Ashcroft and Ashcroft, 1990 for review).

View larger version (27K): [in this window] [in a new window]   Fig. 15.   Tracings (panels A and B) and quantitative data (panel C) showing the inverse relationship existing between the frequency at which guinea-pig ureters were driven by EFS and intensity/duration of the inhibitory action of CGRP in the absence and presence of 1 µM glibenclamide. In panels A and B vertical bars are % of the response to KCl. The ureters were driven by EFS by applying one stimulus every 100, 60 or 15 sec for 30 min before application of CGRP: in panel A note that the suppressant action of CGRP is almost abolished after a 30 min period of driving with stimuli applied every 15 sec, whereas the suppressant effect of cromakalim is preserved. Panel C, left side, shows the intensity of peak inhibitory action of CGRP in relation to driving frequency in the absence (empty circles) and presence (filled circles) of 1 µM glibenclamide. Panel C, right side shows the inverse relationship (r = 0.997, n = 3, P < 0.01) between EFS driving frequency and duration (in min) of the suppressant effect of CGRP. In lower panels, each values is mean ± s.e.m. of 5-10 experiments. * significantly different from values obtained with 60 sec interval between the applied stimuli, P < 0.05. Reprinted with permission from Maggi et al. (1996c).

View larger version (18K): [in this window] [in a new window]   Fig. 16.   Time course of membrane hyperpolarization produced by 15 min superfusion with CGRP (0.1 µM) or cromakalim (3 µM) in resting conditions (empty circles) or within 3 min from the end of a 30 min conditioning period of " exercise" (filled circles) produced by applying electrical stimuli every 15 sec. Vertical axis shows changes (D in mV) in membrane potential (Em) recorded at various times after start of superfusion with CGRP or cromakalim. Each value is mean ± s.e.m. of 6 experiments. * significantly different from controls, P < 0.05. Reprinted with permission from Maggi et al. (1996c).

The "efferent" or "local effector" function exerted by capsaicin-sensitive afferent nerves involves the release of mediators from the peripheral endings of these primary sensory neurons (Szolcsanyi, 1984; Maggi and Meli, 1988; Maggi, 1995; Holzer, 1988). In addition to changes in smooth muscle tone, locally released sensory neuropeptides exert other effects such as vasodilatation, increase in microvascular permeability, mast cell degranulation, and recruitment of inflammatory cells. These effects are known as "neurogenic inflammation," a term that stresses the contribution of the nervous system to the initiation and maintenance of the inflammatory process. When considered as the earliest type of response that sensory nerves set into action when interacting with potentially harmful environmental stimuli, "neurogenic inflammation" can be considered a "physiological" defense response, finalized to facilitate the removal of the noxae (Maggi, 1995).

Plasma protein extravasation induced by sensory nerves has been investigated in the pyeloureteral tract level. Saria et al. (1983) demonstrated that electrical stimulation of the inferior mesenteric ganglion caused Evans blue extravasation in the rostral third of the guinea pig ureter, whereas electrical stimulation of the pelvic nerves caused Evans blue extravasation in the caudal third of the ureters. All these responses were absent in capsaicinpretreated animals which indicates the absolute requirement of the integrity of sensory nerves. Because Evans blue binds to plasma proteins, the increased tissue content of the dye observed after nerve stimulation is thought to reflect the extravascular accumulation of plasma proteins. This is an accepted quantitative index of the intensity of tissue edema/inflammation induced by the test stimulus. By using this technique, it was found that exogenously administered tachykinins, SP and NKA, produce plasma protein extravasation in the rat ureter (Maggi et al., 1987a). Abelli et al. (1989) compared the activity of different tachykinin receptor selective agonists and presented evidence indicating that NK₁ receptors produce plasma protein extravasation in the ureter. Santicioli et al. (1993) showed that plasma protein extravasation induced by SP or a selective NK₁ receptor agonist are blocked by a selective nonpeptide tachykinin NK₁ receptor antagonist and that this effect of tachykinins is independent from the generation of NO. The involvement of tachykinin NK₁ receptors in increasing microvascular permeability to plasma protein is consistent with the analysis of their distribution in venules and arterioles of the ureter (Sann et al., 1992). Finally, Nagahisa et al. (1992) showed that an NK₁ receptor antagonist blocks capsaicin-induced plasma protein extravasation in the guinea pig ureter.

Summarizing, the activation of the "efferent" function of sensory nerves is capable of producing a local inflammatory response in the ureter. It may be speculated that neurogenic inflammation occurs in the pyeloureteral tract during events that are of pathophysiological relevance for this organ. As an example, a damage of the urothelium produced by the passage of a stone can induce a backflow of urine in the ureteral wall and, by activating afferent nerves, can induce a local inflammatory response. In a similar way infections of the pyeloureteral tract, through the action of products of bacterial metabolism or indirectly, by altering urothelial permeability and inducing urine backflow, can activate neurogenic inflammation.

	VI. Pyeloureteral Reflexes

Top Previous Next References

Afferent nerves in the renal pelvis and ureter are the trigger point for the activation of several reflex responses. Although the significance of some of these reflexes is not fully understood, it is evident that noxious stimuli or "energies of activation" that are potentially noxious or tissue damaging are the most effective activators of both pyeloureteral reflexes and pain. In addition, certain reflex responses can be activated by mechanical stimuli and/or changes in the chemical composition of the urine and may be important for regulating urine production, water, and Na balance (Stella and Zanchetti, 1991 for review).

Both mechano- and chemoreceptors are present in the renal pelvis (Stella and Zanchetti, 1991). Mechanoreceptors in the renal pelvis are sensitive to changes in urine outflow pressure (Beacham and Kunze, 1969; Kopp et al., 1984). Genovesi et al. (1993) found that an increase in urine flow rate and the corresponding increases in renal pelvic pressure closely match the increase in afferent renal nerve activity, whereas the changes in renal blood flow or perfusion pressure do not correlate with changes in renal afferent nerve discharge. Genovesi et al. (1993) concluded that changes in water excretion, via changes in renal pelvis pressure, are a primary factor for regulating renal afferent nerve activity.

Chemoreceptive afferent units in the renal pelvis have been extensively investigated in rats and have been classified as R1 and R2 units (Recordati et al., 1978, 1980a,b, 1981). R1 units are unresponsive to mechanical stimuli, are silent under normal conditions, are activated during ischemia and may be located around blood vessels. R2 units exhibit a resting discharge and are sensitive to backflow of concentrated urine (nondiuretic urine). R2 units are sensitive to the chemical composition of the urine, can sense changes in the concentration of Na and K, and are also activated by chemicals normally present in the urine, such as bradykinin. Finally, R2 units (but not R1 units) are sensitive to capsaicin (Szolcsanyi, 1984).

Katholi et al. (1983) also showed that chemoreceptors in the renal pelvis but not in the ureter are activated by perfusing adenosine in the renal pelvis. The relationships existing between the activation of different populations of afferent nerves in the renal pelvis and the activation of specific reflexes is still a matter of investigation. However, R2 chemoreceptors contribute the greatest part of afferent nerve discharge to increased pressure in the renal pelvis and, therefore, work as both chemo- and mechanoreceptors (Moss and Karastoianova, 1997).

It seems that R2 chemoreceptors are implicated in sensing the concentrations of chemicals present in the renal medullary interstitial fluid and in the final urine and that their tonic activity is reduced under conditions of diuresis. Rogenes (1982) speculated that the reno-renal reflex initiated by the stimulation of R2 chemoreceptors represents a positive feedback to maintain a low rate of water and Na excretion during antidiuresis.

The studies from Kopp and coworkers provide a stimulating link between mechanical stimuli applied to the renal pelvis, the efferent function of sensory nerves (local release of sensory neuropeptides) the activation of afferent nerves and the reflex modulation of kidney function. Kopp and Smith (1991) first observed that the administration of SP into the rat renal pelvis increases the afferent discharge of renal nerves and induces a contralateral reflex diuretic and natriuretic response mimicking that produced by increasing ureteral pressure. This reno-renal reflex is abolished by pretreatment with neurotoxic doses of capsaicin (Kopp and Smith, 1993a). Moreover, the reno-renal reflex evoked by mechanoceptor stimulation, as well as that induced by infusion of SP, are selectively reduced by the intraureteral administration of the tachykinin NK₁ receptor antagonist CP 96345 (Kopp and Smith, 1993a). Gontijo and Kopp (1994) also showed that CGRP, when administered into the renal pelvis, determines a concentration-dependent increase in the contralateral afferent nerve discharge, although the perfusion with a CGRP receptor antagonist did not affect the response to mechanoreceptor stimulation. More recently, the same group reported that: (a) both bradykinin and prostaglandins activate the same reflex that is activated by mechanical stimulation of the renal pelvis; (b) endogenous prostanoids amplify the afferent discharge induced by capsaicin, SP, bradykinin, and increased ureteral pressure; and, (c) increased ureteral pressure induces an indomethacin-sensitive release of SP in the rat renal pelvis, which suggests that the mechanically evoked reno-renal reflex requires the release of SP and activation of NK₁ receptors, via generation of prostanoids (Kopp and Smith, 1993b; Kopp et al., 1996, 1997).

Therefore, an inhibitory reno-renal reflex (i.e., stimulation of afferent reduces the sympathetic outflow to the kidney, resulting in diuresis and natriuresis) can be activated by noxious stimulation applied to the renal pelvis: this response can be activated during pathological conditions that increase the intraureteral pressure such as pyelonephritis or presence of renal stones, and when normal constituents of the urine (KCI, hyperosmolarity, low pH, or bradykinin) or products of bacterial metabolism penetrate through the urothelium to stimulate sensory nerves. At the renal pelvis level, a generation of prostanoids and a local release of sensory neuropeptides could both be involved in initiating this response. Recently Kopp et al. (1998) reported that the natriuretic reno-renal reflex initiated by distension of the renal pelvis is impaired in spontaneously hypertensive rats (SHR) and suggested that this deficit could contribute to the Na retention observed in these animals. When addressing the mechanisms of this deficit, they observed that the increase in renal afferent activity produced by administration of phorbol esters, to activate protein kinase C, was equally evident in SHR and control rats. Moreover, the release of prostaglandins provoked by distension of the renal pelvis was also evident in both SHR and controls rats. In contrast, distension of the renal pelvis evoked a smaller release of SP in SHR as compared with control rats. Finally, exogenous SP failed to increase renal afferent activity in SHR animals (Kopp et al., 1998). The observed changes suggest that an alteration in the sensory SP innervation at both pre- and postjunctional level may account for the impairment of the reno-renal reflex observed in SHR animals.

Cervero and Sann (1989) presented a detailed analysis of the stimuli capable of exciting afferent nerves in the guinea pig ureter. They found that the majority of afferent nerves at this level have conduction velocities in the C-fiber range. On the basis of their mechanosensitivity, afferent nerves in the guinea pig ureter have been divided in two main groups: (a) U1 units (9% of all mechanosensitive units) that respond to contractions of the ureter and do not show ongoing activity or afterdischarge to mechanical stimulation; U1 units have a low threshold to intraluminal distension and respond after a short latency to mechanical stimuli; and (b) U2 units (91% of all mechanosensitive units) that do not respond to contractions of the ureter, have low frequency spontaneous activity (< 2.5 Hz) and exhibit a prolonged afterdischarge to mechanical stimuli; U2 units respond after a long latency (> 3 sec) to increases of intraluminal pressure in the range of 5 to 30 mmHg (Cervero and Sann, 1989). The movement of an intraluminal glass bead under the receptive field evoked strong responses of U2 units. Cervero and Sann (1989) speculated that U1 units are tension receptors placed in series with smooth muscle and monitor the normal peristalsis of the ureter, whereas U2 units are likely involved in signaling noxious events. U2 units are not excited by active contraction although they can be excited by compression or distension; U2 units also were reported to be chemosensitive, being excited by K, bradykinin, and capsaicin, much like the polymodal nociceptors commonly observed in other viscera.

Sann (1998) recently readdressed the issue of chemosensitivity of U1 and U2 units in the guinea pig ureter. He found that various chemical stimuli such as urine (> 800 mosm/liter), bradykinin, or capsaicin are capable of stimulating both U1 and U2 units: however, the evoked excitation of U1 units closely follows the phasic contractions induced by applied chemical stimuli whereas the induced excitation of U2 units does not show a close relationship with induced contractions. Sann (1998) reported that an elevation of extracellular K mimics the excitatory effect of urine, whereas the application of hyperosmolar solutions or the application of Na or urea did not match the excitatory effect produced by urine. Moreover, U2 units displayed sensitization to applied mechanical stimuli after chemical stimulation.

Afferent units with characteristics of U1 and U2 units as described in guinea pig ureter also were detected in the chicken ureter. In this species, a third population of afferent nerves also has been detected (U-G units) which are thought to code electrical activity arising from intramural ganglion cells which project to adjoining viscera (colo-rectum, cloaca) (Hammer et al., 1993).

Relatively little information is available concerning the nature of reflex responses initiated by stimulation of ureteral afferents. Amann et al. (1988b) showed that repetitive electrical stimulation of ureteric nerves evoked a slow excitatory postsynaptic potential (sEPSP) in neurons of the guinea pig inferior mesenteric ganglion (IMG); the sEPSP was abolished by pretreatment with capsaicin that itself depolarized IMG ganglion cells. Distension of the ureter or intraureteric application of capsaicin also produced a depolarization of capsaicin-sensitive IMG neurons (Amann et al., 1988b). Superfusion of the IMG with exogenous tachykinins depolarized ganglion cells and partly blunted the ganglionic sEPSP to electrical stimulation, suggesting a mediator role of these neuropeptides (Amann et al., 1988b,c). Intense (putatively noxious) stimulation of afferent nerves also induces transient changes in cardiovascular function that are interpreted as pseudoaffective equivalents of visceral pain (see Section VII.).

	VII. Ureteral Pain

Top Previous Next References

Pain is the only conscious sensation evoked by stimulation of the ureter in humans: pinching, cutting, heating, cooling of the outer surface of the renal parenchyma do not elicit any sensation whereas distension of the kidney, pelvis, or ureter or electrical stimulation of the renal pelvis or ureter results in intense pain (see McLellan and Goodell, 1942; Ammons, 1989; Cervero, 1994 for review)

Studies in vitro (Cervero and Sann, 1989; Hammer et al., 1993; Sann, 1998) have shown that the ureter is innervated by a large population of afferents with high activation thresholds (mean = 34 mmHg in guinea pigs) that do not respond to peristalsis and show long afterdischarges to mechanical stimuli (U2 units). U2 units are obvious candidates for transmitting the information producing pain because the pain threshold in humans is reached at a mean ureteral pressure of approximately 30 mmHg (Risholm, 1954) and pseudoaffective responses are evoked in animals at ureteral pressures in the range of 25 to 30 mmHg (Beacham and Kunze, 1969; Roza and Laird, 1995; Sann, 1998). This response is inhibited by drugs that are effective in relieving pain of renal colic (Laird and Cervero, 1996). As an example, Sann (1998) showed that increases of intraluminal pressure or application of capsaicin to the guinea pig ureter caused a transient increase in systemic blood pressure, heart rate, and changes in respiratory frequency, all indicative pseudoaffective pain responses. A smaller population of units with a low threshold for activation (mean = 8 mmHg) that respond to peristaltic contractions (U1 units) also were found that presumably monitor peristalsis (Cervero and Sann, 1989; Hammer et al., 1993). The latter may be involved in regulating reflexes (see above) but also can encode stimuli in the noxious range and may contribute to the genesis of ureteral pain.

By checking the induction of changes in systemic blood pressure (pseudoaffective response) as an index of effective painful stimulation, Matsumoto et al. (1996) used the fos-labeling technique to identify neurons in the rat spinal cord that receive noxious afferent input from the ureter. They found that after unilateral stimulation of the ureter, fos⁺ cells are present bilaterally at L1-L2 and L6-S1 spinal levels: fos⁺ cells are found in the superficial medial and lateral dorsal horn and in the sacral parasympathetic nucleus. More recently, Avelino et al. (1997) developed a novel model of acute ureteral occlusion in the rat to study fos expression in the dorsal spinal cord. They found that fos expression peaks at L1-L2 and involves neurons in laminae I, IV-V, VII, and X. Labeling in lamina I was mainly ipsilateral and was ascribed to a nociceptive input giving rise to conscious pain. Labeling in the deeper laminae was observed both ipsi- and contralaterally and was suggested to be associated with the activation of autonomic reflexes. Capsaicin pretreatment abolished the induced fos labeling at all levels of the spinal cord.

A few studies have investigated the responses of spinal dorsal horn neurons to noxious stimulation of the ureter: Laird et al. (1996) found that approximately 42% of rat dorsal horn neurons in T12-L1 spinal segments with an ipsilateral somatic receptive field are also excited by noxious distension of the ureter (80 mmHg for 30 sec) in rats. For intraluminal pressure values > 20 mmHg, these neurons showed an increased firing with long latencies and afterdischarges. Moreover, all neurons with ureter afferent input also had a somatic nociceptive input: stimulation of the ureter produces changes in the somatic receptive field area, indicating a degree of plasticity in the ureteric nociceptive pathway (Laird et al., 1996).

Animal models of experimental ureteral calculi have been developed. The complete obstruction of the ureter provoked restlessness and other behavioral alterations in conscious sheep (Moriel et al., 1990), but did not evoke referred hyperalgesia or overt signs of spontaneous pain behavior. Electrical stimulation of the ureter induces pain and muscular hyperalgesia in rats. Giamberardino et al. (1988) reported that vocalization electric thresholds in muscles of the lower back were significantly reduced after ipsilateral painful electrical stimulation of the ureter. Moreover, the implantation of an experimental calculus in rats results in referred hyperalgesia and "crises" of abdominal stretching, indicative of spontaneous visceral pain (Giamberardino et al., 1990; 1995a). Interestingly, a ligature of the ureter did not produce hyperalgesia on its own, indicating that ureteral occlusion as such is not the cause of pain. The combined administration of an antimuscarinic agent and of a COX inhibitor completely eliminated the referred hyperalgesia induced by artificial calculi in rats (Giamberardino et al., 1995b). Laird et al. (1997) showed that partial obstruction induced by placement of an artificial calculus in the upper ureter produces a marked increase in the amplitude of ureteral contractions along with a slight reduction in their frequency and a decrease in the baseline pressure. However, total obstruction or ligature of the ureter totally abolished ureteral motility. The results suggested that increased motility caused by a stone likely contributes to the development and maintenance of visceral pain and of the referred hyperalgesia that persists even after elimination of the stone. The increased amplitude of ureteral contractions is thought to reflect an amplifying effect of stretch: this could involve, in principle, a direct effect on smooth muscle, the production of prostanoids, the release of tachykinins from afferent nerves or the passage of urine into the ureteral wall because of a damage of the mucosa (Laird et al., 1997). However, the decrease in baseline pressure observed below the obstruction could involve the smooth muscle relaxant action of CGRP (Laird et al., 1997).

Giamberardino et al. (1996) recorded changes in cell activity in the ipsilateral spinal cord (T11-T12) in rats with hyperalgesia of the obliquus externus muscle induced from artificial calculi of the ipsilateral upper ureter. They found a significantly higher number of spontaneously active cells indicating central sensitization in this model of referred hyperalgesia from ureteral calculi. Changes in the characteristics of spinal neurons receiving ureter afferent input after implantation of an experimental ureteral stone were also reported by Roza et al. (1998): these investigators observed that the presence of an ureteral stone increases the background activity and the number of ureter-driven cells in the dorsal horn of the rat spinal cord. Moreover, they reported that in rats with experimental calculi there are neurons which respond to both ureter distension and innocuous somatic stimuli, which are not usually found in normal rats: Roza et al. (1998) speculated that this kind of change may account for referred hyperalgesia induced inpatients with a ureteral stone by innocuous somatic stimuli.

On the background of the existence of a prominent capsaicin-sensitive innervation of the renal pelvis and ureter, Bultitude (1995) attempted the intraureteric instillation of capsaicin in humans, aiming to produce pain relief in patients suffering from loin pain/hematuria syndrome. A preliminary study in a small number of patients resulted in a lasting (> 2 months) relief of pain symptoms without evidence for adverse effects on renal function (Bultitude, 1995). In a subsequent study, saline distension of the renal pelvis was used as a standard painful stimulus to provide an objective measurement of the effect of intraureteric capsaicin administration (Allan et al., 1997). In that study, capsaicin pretreatment was found to produce a significant and lasting effect on distension-induced pain.

	VIII. Pathophysiological Significance of the Sensory Innervation of the Pyeloureteral Complex

Top Previous Next References

As evidenced in this review, the capsaicin-sensitive sensory innervation of the pyeloureteral complex can activate several local and reflex responses which affect the production of urine and modify ureteral peristalsis. As discussed in Section VI, it appears that the threshold energy of mechanical stimuli for activating capsaicin-sensitive sensory nerves largely exceeds that produced during normal ureteral peristalsis. Various chemical agents that are adequate stimuli for exciting these afferent nerves are not normally present in their interstitium but, being present into the urine, could come in contact with capsaicin-sensitive afferent nerve endings when the urothelium were damaged or when its permeability characteristics were altered.

When released by appropriate stimuli, tachykinins and CGRP profoundly affect the excitability and contractility of pyeloureteral smooth muscle by stimulating specific receptors: the differential expression of receptors/coupling mechanisms determines specialized motor patterns which are suited to facilitate the expulsion/removal of potentially harmful substances present in the urine (Maggi et al., 1988b; 1992b).

The prevailing response observed upon stimulation of sensory nerves in the renal pelvis is a reinforcement of the spontaneous chrono- and inotropic activity driven by natural pacemakers: this could increase the frequency of ureteral peristalsis and facilitate urine transport to the bladder by increasing the coupling of oscillators in the renal pelvis and the efficiency of impulse transmission at the pyeloureteral junction. On a whole, this effect could facilitate the orthodromic transport/elimination of irritants coming in contact with the urothelium of the renal pelvis.

Excitatory responses can be produced by tachykinins in the ureter by exciting latent pacemakers. It is clear, however, that the prevailing effect produced by stimulating neuropeptide release in the ureter is a suppression of latent pacemakers, an effect mediated by CGRP through its K channel opener action. Endogenous CGRP, released from sensory nerves, provides a neurogenic contribution to the refractory period of ureteral smooth muscle and, in this way, contributes to set the maximal frequency of ureteral peristalsis (Maggi and Giuliani, 1994a). Moreover, endogenous CGRP can suppress the propagation of impulses along the ureteral smooth muscle: this effect involves the activation of glibenclamide-sensitive K_ATP channels (Meini et al., 1995). CGRP provides a neurogenic "brake" producing a local suppression of the action potential and blocking the propagation of ortho- or antidromic peristaltic waves in the ureteral smooth muscle.

The observation that the ability of CGRP to suppress latent pacemakers in smooth muscle is inversely related to the driving frequency of stimulation has important consequences for the understanding of the role of this neuropeptide in the local regulation of peristalsis. Stimuli producing CGRP release (presence of stone, infections, breakdown of urothelial barrier and intramural penetration of chemicals present in the urine) also have the ability to concomitantly excite latent pacemakers in the smooth muscle, giving rise to antiperistaltic waves and urine backflow toward the kidney. The concomitant excitation of sensory nerves, producing a local release of CGRP, blocks this event through K channel activation. However, the local suppression of excitability by CGRP would also block the orthodromic propagation of the peristaltic wave. An increased frequency of peristalsis is the first mechanism for increasing urine transport during stimulation of diuresis (Weiss, 1992). The transient and local conduction block produced by CGRP appears to be suited to prevent antiperistalsis especially during low frequency of ureteral peristalsis, when an antiperistaltic wave would have a greater chance to propel urine back toward the kidney. The CGRP-mediated local conduction block may become less and less important when the frequency of peristalsis is high: in that case, the antiperistaltic wave of excitation would have a high chance to collide with orthodromically propagated waves of excitation. We propose that the inverse relationship existing between the frequency of ureteral contractions and the intensity/duration of CGRP-induced blockade via K channels enables CGRP to locally block the excitability of the ureter in relation to varying physiological needs of urine transport. Moreover, by producing local suppression of latent pacemakers and preventing antiperistalsis CGRPcan protect the kidney from ascending infections.

The myogenic properties of the pyeloureteral smooth muscle, as reviewed in the first part of this article, seem largely sufficient to account for normal ureteral peristalsis. However, the local efferent and afferent function of capsaicin-sensitive sensory nerves are likely to affect ureteral peristalsis under pathophysiological conditions such as those occurring during a passage of a stone or a bacterial infection.