Integrative Erectile Biology: the Role of Signal Transduction and Cell-To-Cell Communication in Coordinating Corporal Smooth Muscle Tone and Penile Erection

Integrative erectile biology: the role of signal transduction and cell-to-cell communication in coordinating corporal smooth muscle tone and penile erection

GJ Christ, S Richards and A Winkler

Urology Research Laboratory, Albert Einstein College of Medicine, 1300 Morris Park Avenue, Bronx, NY 10461, USA

The contractility of corporal smooth muscle plays a critical role in the entire erectile process in man. Moreover, in the absence of severe vascular disease, or congenital or other structural abnormalities/malformations, relaxation of the corporal smooth muscle is both necessary and suf®cient to elicit a sustained erection. As such, understanding the initiation, maintenance and modulation of corporal smooth muscle tone is an absolute prerequisite to the improved understanding, diagnosis and treatment of erectile dysfunction. Despite this fact, identi®cation of both the precise mechanistic basis by which endogenous and exogenous vasomodulators exert their effects on individual corporal smooth muscle cells, and moreover, the process by which these signals are spread among the diverse array of parenchymal cells in the paired corpora, remains somewhat of a physiological enigma. Therefore, the goal of this report is two-fold: ®rst, to review current knowledge of the regulation of corporal smooth muscle tone at the cellular and molecular level; and second, to outline a cogent explanation for the rapid and syncytial integration of the effects of diverse stimuli among corporal smooth muscle cells in the human penis.

Keywords: intercellular communication; corporal smooth muscle; signal transduction; integrative biology; syncytial tissue triad; erection

The physiology of erection smooth muscle relaxation, blood ¯ow is preferentially shifted from the cavernous arteries and nutrient arterioles through the helicine vessels to 1 As recently described for approximately 23 h each the sinusoidal spaces. Relaxation of the corporal day the penis of most adult men is in the ¯accid smooth muscle signi®cantly increases the compli- state. It is only during rapid eye movement dream ance of the trabecular meshwork, and ensures the sleep, or sexual or re¯ex stimulation that the penis maximally effective transmission of systolic blood becomes erect. This rapid shift of physiologic status, pressure required to occlude blood ¯ow through the from ¯accid to erect, is a result of the complex small, subtunical emissary veins. During erection integration of responses occurring throughout the the blood ¯ow in the penis increases from a penile neurovascular network. In this scenario, large relatively low level of 5 ml/min, with a concomitant volumes of blood can be rapidly shifted from, or intracorporal pressure of 5±8 mm Hg, to a much stored in, the penis within a few seconds. The higher blood ¯ow of 50±80 ml/min and a corre- majority of smooth muscle cells of the penis lie sponding pressure roughly equal to the mean outside the cavernous and helicine arteries of the systolic blood pressure. Upon achievement of penis, and are enmeshed in a collagen-containing rigidity, blood is actually stored in the sinusoids at trabecular network of cavernous.2±4 During ¯accidity a very low ¯ow rate, and thus, the penis becomes the corporal smooth muscle exists in a partially fully erect. As such, the corporal smooth muscle contracted state, and the blood is supplied to the cells serve a capacitive;5 that is, when they are corpora via the paired cavernous arteries. As such, contracted the penis is a low pressure, low ¯ow blood ¯ows preferentially from the cavernous organ with blood preferentially shunted from the arteries, through the small nutrient vessels of the erectile spaces, and conversely, when they are trabecular network, to the emissary veins and relaxed the blood is fed into the sinusoidal spaces, subsequently out of the corpora. Following corporal creating a high pressure, closed system. The aforementioned description emphasizes the critical importance of the corporal smooth muscle Correspondence: Dr GJ Christ. cells in modulating erection, detumescence and Received and accepted 10 February 1997 ¯accidity in the human penis, and moreover, high- Integrative erectile biology GJ Christ et al 70 lights the necessity for understanding the initiation, neural activation, (2) Corporal and arterial smooth intracellular transduction and parenchymal spread muscle activation (intracellular signal transduction), of the largely locally restricted neural signals that and, (3) Integration of the response among the vast are a hallmark of the erectile process. As illustrated array of smooth muscle cells via intercellular in Figure 1, coordination of the erectile process can communication through gap junctions. These three be broken into three general phases: (1) Autonomic phases will be considered in order below.

Figure 1 Flow chart depicting the salient features of integrative erectile biology. As illustrated, erectile function has been divided into three broad phases. Phase 1 involves activation of the autonomic nervous effector pathway, by any one of a variety of known mechanisms, resulting in release of neurotransmitter/neuromodulator from synaptic varicosities around blood vessels and throughout the corporal parenchyma. Phase 2 consists of two distinct parts. Phase 2A consists of diffusion of neurotransmitter to the smooth muscle membrane bound receptor sites (or in the case of NO7, diffusion across the smooth muscle cell membrane), and activation of the corresponding second messenger pathway (namely IP3/DAG, cAMP, cGMP, etc.). Phase 2B consists of the effects of the increased second messenger levels on intracellular events; namely these would be effects on junctional and especially nonjunctional ion channels (K and Ca2), but of course phosphorylation/activation of proteins/enzymes that effect other regulatory pathways might also be important (for example myo®lament calcium sensitivity or other calcium sequestration/extrusion or release mechanisms; see text). Phase 3 involves spread of the cellular activation process to adjacent, coupled smooth muscle cells via intercellular communication through gap junctions. Ultimately resulting in synchronous tissue activation; in this particular case, coordinated contraction or relaxation of corporal and arterial smooth muscle. Integrative erectile biology GJ Christ et al 71 Phase 1. The autonomic nervous system (ANS): crease corporal and arterial smooth muscle tone; a primary site of signal origination in corporal (2) parasympathetic nonadrenergic/noncholinergic and arterial smooth muscle nerves which directly decrease arterial and corporal smooth muscle tone (primarily nitric oxide and VIP;4,6 and (3) an additional parasympathetic choli- The complex physiologic cascade responsible for nergic pathway which indirectly modulates tone by corporal erectile function requires precise modula- altering sympathetic input. tion of central and peripheral neural pathways, as well as the integrity of the penile vasculature. Anatomically, neural control of penile erection can Characteristics of the autonomic be divided into four broad categories: (1) sympa- neuromuscular junction in corporal tissue thetic, (2) parasympathetic, (3) sensory, and (4) somatic. Of course the integrity of all four pathways is required for normal penile erection. However, the The distribution of autonomic nerves in the corpora importance of the somatic and sensory pathways to consists primarily of perivascular and parenchymal regulation of the erectile process have been dis- nerves. Both of these are easily demonstrable in the cussed in much detail in recent reviews,2±4 and will corpora, although the former appears to predomi- not be further discussed here. As such, this report nate. With respect to the latter, the bulk of erectile will focus solely on the role of the autonomic tissue is made up of smooth muscle bundles, within nervous system in modulating corporal and arterial which are widely distributed effector nerve ®bers. smooth muscle tone. As illustrated in Figure 2, near effector tissues, the terminal axons become varicose; that is, the last few axons are devoid of schwann cells. The essential Autonomic innervation of corporal smooth muscle feature of the terminal axon is the neuromuscular junction, where neurotransmitter is released from varicosities during nerve impulse conduction. The As described above, the autonomic nervous system junction between terminal varicosities and smooth consists of two anatomically distinct pathways the muscle cells is highly variable in both con®guration sympathetic and parasympathetic pathways, respec- and distance. Thus, smooth muscle cells in close tively. Pre-ganglionic parasympathetic nerves are proximity to nerve varicosities are likely to be myelinated and originate from the somata of cranial directly affected by the released neurotrans- and sacral nerves. Post-ganglionic parasympathetic mitter; while more distal cells within smooth nerves which are non-myelinated, and synapse at or muscle bundles are likely to be activated near the tissues which they innervate. Sympathetic indirectly by propagation of electrical or chemical pre-ganglionic nerves are also myelinated, and activity through gap junction channels3,7 (see text originate from nuclei which reside in the thoracic below). and upper lumbar spinal column. The cell bodies of post-synaptic sympathetic nerves primarily originate in the sympathetic trunk; sympathetic, post- Synthesis/disposition of prominent autonomic ganglionic nerves are far longer than their para- neurotransmitters in the corpora sympathetic counterparts. The pelvic origin of penile innervation lies in visceral efferent parasympathetic pre-ganglionic There are clearly a plethora of neurotransmitters ®bers from the second through fourth sacral roots, present in the human corpora, and this has recently and the sympathetic post-ganglionic ®bers from been reviewed in much detail. However, for the sake thoracolumbar ganglion T12-L2. Parasympathetic of clarity and simplicity, this report will focus on a input from the pelvic splanchnic nerves, and detailed presentation of only a few prominent sympathetic input via the super®cial hypogastric neurotransmitters. These are, adrenergic (norepi- nerves help form the pelvic plexus. The dual nephrine), cholinergic (acetylcholine) and non- autonomic innervation from the pelvic plexus sends adrenergic, non-cholinergic (NANC; nitric oxide visceral branches to the corpora cavernosa, termed and vasoactive intestinal polypeptide). the cavernous nerves. Both of these anatomical classi®cations, can in turn, be further subdivided according to the neuro- Adrenergic and cholinergic neurotransmitters transmitter/neuromodulator released into the corporal milieu. In general then, with respect to the control of corporal and arterial smooth muscle tone, The primary neurotransmitter of cholinergic nerves the most physiologically relevant autonomic effector is acetylcholine (Ach), whose synthesis takes place pathways thus characterized are: (1) sympathetic in the autonomic neuronal cell body and the adrenergic constrictor nerves which directly in- peripheral axon terminal. Ach is stored in vesicles Integrative erectile biology GJ Christ et al 72 and released by exocytosis into the neuromuscular Non-adrenergic, non-cholinergic (NANC) cleft. Ach is hydrolyzed to choline, which is taken neurotransmitters up at the endplate. The captured choline is thought to be important in maintaining continued neurotransmitter synthesis.8 Noradrenaline is stored in Non-adrenergic, non-cholinergic nerves use multi- granular vesicles together with dopamine B-hydro- ple transmitter substances, the two most prominent xylase and chromogranin A. Noradrenaline is also in the corpora appear to be vasoactive intestinal released by exocytosis. Following activation at the polypeptide (VIP)4,9 and nitric oxide (NO7).4,10 With post-synaptic muscle membrane, noradrenaline is respect to the former, as reviewed elsewhere,4 re-incorporated at the nerve terminal, if not de- several immunohistochemical studies have demon- graded by MAO or COMT.8 strated the presence of VIP in autonomic nerves in Integrative erectile biology GJ Christ et al 73 the smooth muscle and penile blood vessels of the phenomenon of co-transmission, i.e. the release of human corpus cavernosum. With respect to the more than one neurotransmitter/neuromodulator latter, NO7 is produced by the enzyme NOS (nitric from a particular neuron is well recognized, and oxide synthase, which is distributed in NANC likely to be very important; although it certainly nerves throughout the corpora. NO7 is produced represents a very poorly characterized phenomenon as the enzymatic by product of molecular oxygen with respect to penile erection3. As such, it is (O2) and L-arginine, which, as mentioned above, is conceivable that co-transmission and/or alterations catalyzed by NOS. NO7 is an interesting neuromo- in co-transmission could turn out to be an important dulator/neurotransmitter in that it is highly lipo- regulatory site for erectile physiology and dysfunc- phillic, with a very short half-life ( 5 s), and a tion. For example, certain substances may act via concomitantly limited radius of diffusion.11 their own receptors/effectors to produce synergistic Moreover, NO7 is synthesized `on demand' and actions with other more principal neurotransmitters, is not known to be stored by any mechanism in for example as has been proposed for VIP and NO7). any appreciable amounts in NANC nerves, or else- In fact, it has been shown that the number of where. VIPergic nerves providing vasodilatory control of penile smooth muscle are reduced or lost entirely in diabetic neuropathy associated with erectile dysfunction.12 As such, it is very important to empha- Release of neurotransmitters at the autonomic size that one must look beyond the neuromuscular neuromuscular junctions (NMJs) junction to see both the macro and microcellular effects of altered neurotransmission, and their concomitant effects on erectile vascular physiol- The neuromuscular junction is one site at which the ogy/function.3,13 There is no doubt that this is delicate balance between vasoconstricting and va- clearly a fruitful area for future research. sodilating agents is readily apparent. Adrenergic A related phenomenon is neuromodulation. This and cholinergic nerve terminals are in close proxi- represents another important aspect of erectile mity to each other and the smooth muscle cells they neurophysiology that occurs at the NMJ, and has innervate; although as described above, the majority heretofore received little recognition or investiga- of varicosities do not form `classical' synaptic clefts tion3. Speci®cally, alterations in the amount of with the adjacent smooth muscle cells. In addition, neurotransmitter/neuromodulator released by neu- the wide conformational variability of the neuro- rons, as well as the timing of neurotransmitter muscular cleft is likely indicative of the versatility release/binding/activity at the neuromuscular junc- of the autonomic neuromuscular junction. The tion may have a dramatic effect not only on adjacent

Figure 2 Schematic depiction of the general geometry/topology of the autonomic neuromuscular junction in corporal smooth muscle. Panel A. As illustrated, nerve terminals in the corporal parenchma, as well as around blood vessels, become unmyelinated and varicose. These varicosities are thought to be the site of release of neurotransmitter in this and other peripheral tissues. As illustrated, varicosities coursing through a group of smooth muscle cells give the appearance of `beads on a string'. Release of neurotransmitter will result in graded activation of smooth muscle cells in direct relation to their proximity to the varicosity (regardless of whether the relevant pathway is extracellular or intercellular; although as discussed in the text, there is compelling evidence in favor of the latter). Although not depicted, the cells are interconnected by gap junctions, so that direct activation of membrane receptors on cell #1, would result in increases in intracellular second messenger levels, the effects of which, could be transmitted by the intercellular pathway to cells 2, 3 and 4. Obviously, it is dif®cult to illustrate a 3-D effect in two dimensions, thus, one should keep in mind that a similar scenario applies for all possible planes of diffusion (for example on the Z axis in and out of the page). Also note that since cytoplasmic diffusion is approximately one order of magnitude greater than junctional permeability, intercellular diffusion will be faster along the long axis of the smooth muscle cells, than across the width of the smooth muscle cells (namely anisotropic). Such anisotropic diffusion may have important implications to erectile physiology/dysfunction, depending on the particular arrangement of smooth muscle cells within the corpora. Panel B shows a schematic diagram of how activation of one cell (namely the far left cell #1) could lead to recruitment of adjacent, but not directly activated cells via the passive intercellular spread of second messenger molecules/ions. Such a mechanism is important to permit coordination of corporal smooth muscle tone; note that this would be true whether this cellular activation results from the release of a locally restricted neuronal or endothelial signal, or from the intracorporal injection of a vasorelaxant (see text). Note that when intercellular diffusion is truly passive, it will result in a second messenger gradient across an array of smooth muscle cells; with the cells nearest the activated cell `seeing' much higher second messenger concentrations than more distally removed cells. As such, the `sensitivity' or corporal smooth, muscle cells to these intracellular messengers (namely the amount or concentration of second messenger required for cellular activation) will be a very important determinant of the size of the `active vascular unit' (see text). Panel C illustrates the predicted intercellular diffusion pro®les on the same cells depicted in B (that is, the numbers on panel C correspond to same cells numbered in panel B). These intercellular diffusion pro®les of junction permeant second messenger molecules were calculated by mathematical modeling techniques.35 Brie¯y, for this calculation, cell 1 is assumed to have some arbitrary, but constant, concentration of second messenger molecules due to persistent receptor activation by agonist or neurotransmitter. Thus, the model depicts the expected passive diffusion of second messenger molecules from cell one to as many as ®ve cells distantly removed. The arrows in the diffusion pro®le emphasize the differential intercellular second messenger concentrations expected across the junctional membrane of contiguous cells, while the dashed lines depict the anticipated concentration gradients across the width of the cell. The junctional permeability used for this comparison was 561074 cm/sec, a junctional permeability characteristic of well-coupled corporal smooth muscle cells expressing the gap junction protein connexin43. Integrative erectile biology GJ Christ et al 74 neurons, but also on the corporal smooth muscle Regulation of contraction of corporal smooth muscle cell. There are two obvious examples of the by intracellular calcium levels importance of neuromodulation to erectile function. The ®rst is nitric oxide, whose synthesis via the nitric oxide synthase enzyme (see text below) has Contraction of corporal smooth muscle may be not only more than one source cell (namely neurons elicited by activation of a membrane receptor and endothelial cells) but perhaps additionally, both subsequent to occupancy of the receptor by sub- constitutive and inducible components as well.14 A stances found in the adjacent plasma or by neuro- second relevant example of the potential physiolo- transmitters released from autonomic nerves ®bers in gical relevance of neuromodulation is the hypoth- the region of the smooth muscle cell. As illustrated in esis advanced by Christ and coworkers,15,16 in which Figure 3, the precise mechanism of action is endothelin-1 (ET-1), also synthesized by more than presumably similar to that demonstrated for other one cell type (smooth muscle cells, as well as smooth muscles; although it should be emphasized endothelial cells) is proposed to augment the effects that all aspects of this have not been unequivocally of underlying adrenergic tone via its direct effects on established in corporal smooth muscle. Activation of corporal smooth muscle cells; moreover, there is the appropriate membrane receptor leads, through a likely to be a reciprocal relationship between ET-1 G protein-dependent mechanism, to the cleavage of and the nitric oxide levels. As such, neuromodula- phosphatidylinosotil bisphosphate (PIP2), liberating tion may be an especially important consideration to inositoltrisphosphate (IP3) and diacylglycerol (DAG). erectile physiology. The former is responsible for liberation of intracellular calcium from the sarcoplasmic reticulum, while the latter leads to activation of protein kinase C, and subsequent phosphorylation events; it is important to note that phosphorylation events are thought to be linked to both alterations in membrane ion channel activity (both K and Ca channels; see below), and Phase 2A. Activation of corporal and arterial perhaps alterations in myo®lament calcium sensitiv- smooth muscle ity; the latter would ensure that so long as the agonist signal remains, even when the intracellular calcium Control of corporal smooth muscle contraction and concentrations return to near resting levels, the smooth muscle cell would remain in the contractile relaxation state. More speci®cally, activation of both the a1- Ultimately, activation of corporal smooth muscle adrenergic and ETA-receptor subtypes (by pheny- cells, either contraction or relaxation, is mediated by lephrine (PE) and endothelin-1 (ET-1), respectively) the rise and fall, respectively, in the free intracel- in human corporal smooth muscle have been lular calcium levels; although alterations in myo®la- documented to elicit a rapid and transient elevation ment calcium sensitivity have also been proposed.17 in the intracellular calcium concentration in cul- As such, both dynamic alterations in (namely tured human corporal smooth muscle cells;15,20 the erection and detumescence), as well as sustained latter, but not the former, has been shown to be (tonic) contraction (namely ¯accidity) of, corporal coupled to increased IP3 metabolism (Christ et al, smooth muscle are ultimately controlled, in large unpublished observations). Moreover, as the mea- part, by the effects of various endogenous (namely sured increases in the peak amplitude of the neural or hormonal) and exogenous (intracavernous intracellular calcium transient in the cultured injection of vasoactive compounds) substances on smooth muscle cells occur over the same concentra- the free intracellular calcium ion3,7,15,16,18±20 (see tion range as the increase in the magnitude of the Figure 1 for details). Still, in the partially contracted steady-state contractile response,15,16 this has been state of the corporal smooth muscle cell, as in other taken as further evidence of the major importance of vascular tissues,17 small ¯uxes of Ca2 ion presum- intracellular calcium to the modulation of corporal ably continually enter the cell through L-type smooth muscle tone elicited by these two endogen- voltage dependent Ca channels on the cell mem- ous agonists. brane. Subsequently, the calcium ion is distributed/ stored within the sarcoplasmic reticulum, cyto- Regulation of corporal smooth muscle relaxation plasm and the nucleus of the cell. In light of the importance of intracellular calcium homeostasis to the control of corporal smooth muscle tone, and Functional (physiological) antagonism of intracel- thus, penile erection, we review the effects, again, of lular calcium levels/myo®lament calcium sensitiv- prominent vasomodulators only, on intracellular ity is the primary mechanism responsible for calcium levels, and thus, corporal smooth muscle relaxation of smooth muscle. In this regard, there tone and erectile capacity. are two main intracellular messenger molecules that Integrative erectile biology GJ Christ et al 75

Figure 3 Signal transduction in human corporal smooth muscle. Illustrated are the salient features of Phase 2 (see Figure 1) in the coordination of erectile function, for some of the relevant vasomodulators. As illustrated, activation of membrane receptors and/or intracellular effectors results in alterations in nonjunctional ion channel function. More speci®cally, K ion ¯ux is controlled by three main effector pathways, the ®rst two are the cAMP/PKA (PKA; protein kinase A) and cGMP/PKG (protein kinase G) pathways which are 7 activated by, for example, PGE1 and NO , respectively, and while these pathways apparently modulate the activity of the maxi-K channel; their effects on the KATP channel have not been documented, but seem likely. A third pathway is the potassium channel modulators, which exert their relaxant actions on corporal smooth muscle, at least in part, by altering the activity of the KATP channel. Because of the disposition of K and Ca2 ions in the intra- and extracellular spaces (see Figure 4), the opening of Ca2 channels leads to the in¯ux of Ca2 down it's electrochemical gradient and subsequent intracellular depolarization, while the opening of K channels lead to ef¯ux of K from the cell down its electrochemical gradient, and subsequent intracellular hyperpolarization. The effects of these reciprocal pathways on membrane potential and the level of corporal smooth muscle tone are exerted, at least in part, through modulation of the intracellular calcium concentration; with increases in intracellular calcium associated with contraction, and decreases in intracellular calcium associated with relaxation. Although not shown, the cAMP/PKA and cGMP/PKG pathways are thought to have an inhibitory effect on Ca2 channel activity (namely decreased open probability); presumably through phosphorylation effects, as well as the effects of membrane potential changes. Where (): denotes a positive or stimulatory effect; (7) denotes a negative or inhibitory 7 effect; ?: denotes an unknown action. PIP2 phosphatidylinositol bisphosphate; DAG: diacylglycerol; IP3: inositoltrisphosphate; NO : nitric oxide; NTG: the nitrate donor nitroglycerine; ET-1: endothelin-1; PE: phenylephrine; L-type Ca2: L-type, voltage-dependent calcium channel. Please note that this diagram is meant to re¯ect only the most likely scenario for these intracellular events, based on our observations,3,7,11,15,16,27±31,46,47 and extensive work in other vascular tissues.63±65 modulate the continuous transmembrane calcium activation of protein kinase A and G, respectively; ¯ux through voltage-dependent calcium channels although recent in vitro data suggests that there may that is so critical to the sustained contraction of well be cross-talk between cAMP and cGMP effector corporal smooth muscle (see Figs. 3 and 4). These pathways (namely cAMP could activate G kinase, are, respectively, the intracellular concentrations of and vice versa), and/or the presence of an addi- the second messenger molecules cyclic adenosine tional, as yet unidenti®ed kinase.21±24 The physiolo- monophosphate (cAMP) and cyclic guanosine gical actions of increased intracellular second monophosphate (cGMP). messenger levels are summarized in Figure 3 and With respect to the former, increases in intracel- described below. lular cAMP levels represents the primary mechanism by which most intracavernous injection agents mediate their effects, for example, prostaglandin E1 Phase 2B. Alterations in the activity of K and papaverine. With respect to the latter, increases channels and their effects on corporal smooth in intracellular cGMP levels are thought to be the muscle tone primary mechanism by which NO7 released from NANC nerves exerts its relaxant actions, subsequent to activation of soluble guanylate cyclase. Increases The most widely utilized form of intracavernous in either cAMP or cGMP levels, in turn, result in pharmacotherapy relies on receptor-mediated in- Integrative erectile biology GJ Christ et al 76 creases in intracellular cAMP levels to elicit relaxa- voltage-dependent calcium channels (smooth mus- tion of corporal and arterial smooth muscle, and cle cell relaxation erection). With respect to the thus, produce an erection. While there is arguably latter, elevations in intracellular cAMP2 and perhaps no absolute separation between receptor-mediated cGMP as well, have a dual effect. They may increases in intracellular second messenger levels, contribute not only to direct alterations in K channel and alterations in corporal smooth muscle tone, it is activity, but they might also affect the disposition worth drawing such a distinction. If for no other and integrity of intracellular calcium stores and/or reason, than for the simple fact that modern the sensitivity of contractile myo®laments to in- molecular techniques have made it possible to tracellular calcium. All of these actions will, of pharmacologically modulate the targets of receptor- course, promote smooth muscle relaxation. mediated second messenger increases, independent In this regard, potassium channels exhibit greater of activation of membrane receptors. Such ap- diversity than any other nonjunctional ion channel. proaches are currently being used therapeutically. Furthermore, the KATP and maxi-K potassium For example, papaverine is a well known phospho- channel subtypes are thought to be the most diesterase enzyme inhibitor (phosphodiesterases physiologically relevant K channels expressed in represent a family of enzymes responsible for the corporal smooth muscle. The salient features of degradation of intracellular cAMP and cGMP), that these channels are described below. has been effectively used as an intracavernous agent. A more novel approach has been the use of the 25,26 diterpine plant alkaloid forskolin, which directly The maxi-K channel subtype activates the adenylate cyclase enzyme to produce increased intracellular cAMP levels; again, independent of receptor activation, although synergistic The maxi-K channel (KCa) subtype, is similar to the with it. While a more detailed discussion of these calcium-sensitive K channel present in many methods/techniques is clearly beyond the scope of smooth muscle cell types.7,27,28 It is clearly the this report, suf®ce it to say, that alterations in the predominant K channel subtype found in corporal activities of various protein kinases and/or ion smooth muscle, and is responsible for approxi- channels is now possible, and the possibility of mately 90% of the observed whole cell outward K genetic therapies for the treatment of erectile currents seen in cultured human corporal smooth dysfunction are clearly on the horizon. However, muscle cells. The maxi-K channel appears to have the major draw back to any such pharmacological/ signi®cant activity near and even below the mean genetic treatments is that while they do have some resting membrane potential suggesting that it may be selectivity for their end organ effects/target sites, an important modulator of membrane potential; in there is still much room for improvement with corporal smooth muscle cells. Moreover, patch respect to their absolute speci®city of action. clamp experiments conducted on cultured corporal smooth muscle cells indicate that the activity of the maxi-K channel is increased by both the addition of NO donors (namely nitroglycerine; Christ et al, Potassium channels as a convergence point for unpublished observations), as well as addition of modulating corporal smooth muscle tone and PGE1 and the Schwarz/Pharma inclusion complex, 29 erectile capacity PGE1-a-CD. In addition, the increase in maxi-K channel activity produced by agents such as PGE1, is correlated with a signi®cant alteration in the peak Potassium channels have been shown to play a amplitude of the ET-1 induced intracellular calcium fundamental role in both the physiologic and transient seen in fura-2 loaded cultured corporal pathophysiologic regulation of smooth muscle tone. smooth muscle cells.30 Speci®cally, pre-incubation These channels act as a convergence point for of cultured human corporal smooth muscle cells modulation of a number of receptor and effector with PGE1 resulted in an approximately 40% systems. In addition to both affecting basal tissue decrease in the peak amplitude of ET-1 induced tone, as well as the initiation and termination of a calcium transient. The observed decrease in the ET- contractile stimulus, potassium channel activity 1-induced intracellular calcium transient is a re¯ec- effects corporal tissue relaxation, particularly with tion of the inhibition of transmembrane calcium respect to responses that are dependent on hyper- ¯ux; once again highlighting the close association polarization. As illustrated in Figure 4, any intra- between K channel activity and the activity of cellular second messenger effect that is associated voltage-dependent Ca channels. with increased K channel activity/mean open time Recent electrophysiological studies on cultured will result in an increased potassium ion ef¯ux from human corporal smooth muscle cells28 have also the cell interior. This in turn causes cellular suggested that alterations to the function/regulation/ hyperpolarization which reduces calcium ion in¯ux structure/activity of the maxi-K channel may be through a decreased activity/mean open time of characteristic of organic disease per se. Such a Integrative erectile biology GJ Christ et al 77

Figure 4 Illustration of the inter-relationship among the known junctional and nonjunctional ion channels that regulate corporal smooth muscle tone; compare with Figure 3. As illustrated, all of the physiologically relevant second messenger molecules are freely diffusible across the junctional membrane between cells. There exists an important and functionally antagonistic relationship between the activity of Ca nd K channels on corporal smooth muscle. Moreover, the degree of intercellular communication among these cells ensures their partial cytoplasmic continuity. Thus, alterations in ions/second messengers can be quite readily conveyed to adjacent smooth muscle cells. supposition is consistent with other recent in vitro validate the possibility that genetic therapies that experiments conducted on cultured human corporal result in alterations in post-receptor mechanisms smooth muscle cells, which suggested that organic might well represent an attractive therapeutic disease per se may also be associated with altera- approach to the treatment of erectile dysfunction. tions in intracellular calcium mobilization/homeostasis.30 Finally, studies conducted on isolated human corporal tissue strips are also consistent The K channel subtype with the hypothesis that the activity of the maxi-K ATP channel is an important modulator of both phenylephrine (PE)- and ET-1-induced contractile re- The KATP channel subtype is another important sponses in human corporal smooth muscle.31 modulator of corporal smooth muscle tone. Electro- As such, it can be seen that the important physiological studies have indicated that the KATP contribution of the maxi-K channel subtype to the channel subtype in human corporal smooth muscle regulation of corporal smooth muscle tone has been is apparently similar to the metabolically-regulated proven by multiple pharmacologic/physiologic stu- K channels found in other smooth muscle cell dies. Perhaps most importantly of all, transfection of types.7,27 Moreover, pharmacological studies have the human maxi-K cDNA hSlo in rats in vivo,32 and documented that there is a signi®cant (glibencla- humans in vitro,33 have documented that increased mide-sensitive) decrease in the sensitivity of pre- expression of the human maxi-K channel is asso- contracted corporal tissue strips excised from ciated with physiologically relevant alterations in diabetic patients to relaxation by structurally di- corporal smooth muscle cell function. These results verse K channel modulators (namely pinacidil, Integrative erectile biology GJ Christ et al 78 U83757, P1075), when compared to equivalently vasoactive stimuli among corporal smooth muscle precontracted rings of corporal tissue obtained from cells seems relevant. This discussion has been saved nondiabetic patients.34 Consistent with these obser- for last, as a thorough understanding of integrative vations, precincubation of fura-2 loaded cultured aspects of corporal biology is very dependent on human corporal smooth muscle cells with pinacidil detailed knowledge of the mechanistic basis for results in an apparently near complete ablation of signal origination/transduction in corporal smooth transmembrane calcium ¯ux in response to addition muscle. That being said, a model of integrative of ET-1. Such observations provide further evidence corporal tissue function has recently been devel- for the importance of K channels not only in the oped.35 The explicit underlying assumption of this modulation of corporal smooth muscle tone, but model is that neuronal innervation (Phase 1), in also, as putative targets of disease. concert with myogenic intracellular signal transduction mechanisms (Phase 2A and 2B) and intercellular communication through gap junctions (Phase Phase 3. Integrative erectile biology: The role of 3), forms a `syncytial tissue triad' that plays a major role in erectile response generation/modulation intercellular communication in coordinating (Figure 5). The salient features of this model are corporal smooth muscle tone and penile described below. erection

Initiation and propagation of locally restricted How is coordination of corporal smooth muscle signals in the corpora responses accomplished?

In light of the aforementioned discussion, a review Clearly much has been learned in the past decade of the principles governing the propagation of concerning the mechanistic basis for the modulation

Figure 5 Schematic diagram depicting the proposed model for integrative erectile biology, known as the syncytial tissue triad. The concept advanced is that the contraction and relaxation responses among the vast network of largely inexcitable corporal smooth muscle cells (nominally the ®nal effectors of erectile capacity) are primarily cordinated by a triumvirate of mechanisms (see text): (1) Innervation: the characteristics of the effector neural pathways in the penis that are responsible for initiating both erection and detumescence (namely innervation density, ®ring rate, etc.); (2) Propagation: those processes (namely intercellular communication through gap junctions) that permit cells not directly activated by a locally restricted neural or hormonal signal to contribute to the ensuing tissue response (namely erection); and (3) Transduction: those intracellular processes that occur following cellular activation (namely intracellular signal transduction pathways). Modi®ed from (2). Integrative erectile biology GJ Christ et al 79 of corporal smooth muscle tone. However, the smooth muscle. As such, regenerative electrical process(es) responsible for achieving the rapid and events (namely action potentials), or the synchroni- syncytial smooth muscle responses required for zation of such events by pacemaker cells seem an normal erectile function are less well understood. unlikely scenario for coordinating corporal smooth As described elsewhere,13,27 several contingencies muscle responses. exist for the syncytial activation of corporal smooth muscle: (1) Direct neuronal innervation of every or almost every smooth muscle cell; (2) The presence of electrically excitable smooth muscle cells such that syncytial tissue responses could be achieved by regenerative electrical events triggered by fewer The importance of intercellular communication neuromuscular junctions; (3) The passive movement through gap junctions to the local control of penile of intercellular second messenger molecules/ions erection through gap junctions from directly activated cells to distally removed coupled cells, subsequent to activation of only a small fraction of smooth muscle. It is precisely because neither neuronal innervation, These possibilities are considered below. nor electrical excitability are suf®cient, by them- selves, to account for the syncytial nature of corporal smooth muscle activity, that we have turned out attention to the study of gap junction- Neuronal innervation of the corpora mediated intercellular communication. As reviewed below, the anatomical, structural, biochemical, Clearly, as described above, the diversity of neuro- pharmacological and electrophysiological character- transmitters present within the corpora re¯ects the istics of connexin43-mediated intercellular commu- complexity of the neuronal innervation of the nication have been shown to be suf®cient to account corpora. While from an anatomical standpoint, there for the proposed role of intercellular communication has clearly been no systematic investigation of in the physiology of erection in vivo. Thus, it seems innervation patterns in the human corpora, it seems likely that intercellular diffusion of second messen- certain that the number of smooth muscle cells far ger molecules/ions through gap junctions, serves a primary role as a modulator of corporal smooth exceeds the number of neuronal varicosities (see 3,7,13,20±42 above). As such, the bulk of human erectile tissue muscle tone. The general anatomical and apparently consists of bundles of smooth muscle physiological properties of gap junctions are re- cells, with the autonomic innervation consisting of viewed below. widely distributed nerve ®bers rather than intricate neuronal innervation of individual smooth muscle cells. Therefore, it seems unlikely that ANS innervation of the corporal smooth muscle is suf®cient to provide syncytial activation of the corporal smooth Gap junction structure and function cells in the absence of other mechanisms.

Gap junctions represent a diverse family of related Electrical excitability of corporal smooth muscle proteins, known as connexins, that are named according to their molecular weights.43±45 Six, presumably homologous, membrane spanning connex- A review of some aspects of the electrical properties ins join to form a connexon, or hemichannel, in the of human corporal smooth muscle cells in vitro that cell membrane. The union of these connexons have recently been described seems prudent. Cor- across the extracellular space, provides an aqueous poral smooth muscle cells are known to express a intercellular channel and, thus, partial cytoplasmic variety of membrane ion channels, including potas- continuity between adjacent cells (see Figure 6). sium channels,27,28 calcium channels,7 and perhaps Moreover, at areas of membrane apposition between chloride channels (Christ et al, unpublished obser- adjacent corporal smooth muscle cells, these inte- vations) as well. Although rigorous characterization gral membrane proteins are known to coalesce to of all aspects of the behavior of these nonjunctional form junctional plaques consisting of hundreds of membrane ion channels is still lacking, there is individual channels. This has been documented in certainly no compelling in vitro data that indicate human corporal smooth muscle both in vitro and in that corporal smooth muscle cells possess any ionic vivo.35 While 12 distinct mammalian connexins capacity for regenerative electrical events (namely have been identi®ed, connexin43 appears to be the action potentials). Moreover, there is also no ®rm predominant gap junction protein expressed in evidence, at either the tissue or cellular level, for the human corporal smooth muscle, and thus, it is the presence of pacemaker cells in human corporal only connexin considered here. Integrative erectile biology GJ Christ et al 80

Figure 6 Diagrams illustrating the composition (panel A) and mechanism (panel B) of intercellular communication through gap junctions. Panel A is a cartoon depicting the salient features of two representative gap junction channels, and some relevant junction permeant second messenger molecules/ions. Note that when many of these intercellular channels coalesce in the same area of the cell membrane they give rise to the structural basis for the junctional plaque; that is, an area of close apposition of adjacent cell membranes, with a small intercellular `gap'. Panel B shows how gap junctions unite adjacent cells to permit the intercellular passage of second messenger molecules/ions.

Conductance and permeability of contributes to its physiological relevance in vivo.47 connexin43-derived gap junction channels In fact, it is precisely these electrophysiological properties that ensure that the permeability of connexin43-derived gap junction channels is suf®- Electrophysiological studies have documented that cient to permit the intercellular passage of the connexin43-derived gap junctions can express mul- majority of physiologically relevant current carrying tiple conductance states, ranging from 30±100 pS ions and second messenger molecules. Thus it is not (namely predominantly 30, 60 and 90 pS). Further- surprising that inositol trisphosphate and/or cal- more, the distribution of these conductance states cium ions, for example, have been shown to diffuse has been shown to be modulated by some of the between cultured corporal smooth muscle cells.20 As same second messenger molecules described above such, intercellular solute diffusion among coupled (for example activation of protein kinase A or C; see cells has been postulated to have a major role in the Ref. 45 for details). More recently, however, it has integration and coordination of responses in diverse been shown that the 100 pS conductance is the most vascular tissues,41,48±53 including corporal smooth dominant, and additionally, that the high open muscle.20,37±41,46,47 Sophisticated electrophysiologi- probability ( > 90%) and characteristically long cal studies as well as mathematical analyses of such mean open time (1±5 s), of this channel undoubtedly data have been utilized to elucidate important Integrative erectile biology GJ Christ et al 81 features of junctional communication in the coordi- nervous activity.60 Therefore, it seems certain that nation of corporal tissue tone.48 While a detailed the extracellular diffusion pathway is likely to be description of these studies is outside the more much more restrictive than the intercellular path- focused aims of this report, suf®ce it to say that way (see Refs. 11 and 54). when taken together with the results of recent In addition, all of our modeling to date has been anatomical, molecular, immunochemical and phar- based on the assumption that intercellular diffusion macological and clinical studies, it appears that the is a passive process, when in fact, there is quite physiological pro®le of gap junctions in vitro is convincing experimental evidence for the presence 2 more than suf®cient to confer the requisite proper- of regenerative intercellular IP3 and/or Ca waves ties required of these channels in vivo. in some cell types,61,62 including cultured human corporal smooth muscle cells.20 By de®nition, this would be an `active' process, resulting in greater signal spread (namely larger space constants; see Physiological relevance of intercellular diffusion of below for de®nition) of second messenger mole- second messenger molecules/ions through gap cules/ions among corporal smooth muscle cells. junctions among corporal smooth muscle cells

If every cell could be directly activated by neuro- The àctive vascular unit': an indicator of the extent transmitter or drug diffusion through the extracel- of signal spread in the human corpora lular space, the intercellular pathway would be arguably physiologically redundant. Thus, in recent publications we have utilized mathematical model- Irrespective of the exact physical nature of the ing techniques to address this issue. Our studies intercellular pathway, it is clear that the passive/ have evaluated the anticipated intercellular diffu- active intercellular spread of second messenger sion of second messenger molecules and current molecules is of great import to the extent of spread carrying ions through connexin43-derived gap junc- of locally restricted neural/hormonal signals among tions in vascular tissues,11 and compared these the vast array of largely inexcitable corporal smooth estimates to reasonable expectations for the diffu- muscle cells. For example, let us consider as an sion of similar size molecules in the extracellular àctive vascular unit' a group of corporal smooth space.53±55 The details concerning these methods are muscle cells that are within one space constant of a described elsewhere.11,46,53±55 In short, the modeling directly activated corporal smooth muscle cell. studies indicate that if one assumes that intercel- Speci®cally, this àctive vascular unit' would in- lular diffusion is a passive process (namely obeying clude any cell, in any plane of diffusion (namely Fick's law of diffusion), then the pro®le for inter- 1-D, 2-D or 3-D) in which the intercellular signal cellular transit of second messenger molecules originating from the directly stimulated cell (either among corporal smooth muscles can approximate, an electrical or a chemical signal) maintains 37% but cannot exceed, the predicted diffusion pro®le of its original value. Taking into account the for drug or neurotransmitter in the extracellular likelihood that the anatomy of the human corpora space (see Ref. 54, 55). This is an important point, favors anisotropic intercellular diffusion (that is, because it indicates that in theory the intercellular intercellular diffusion would not be expected to be and extracellular passive diffusion pathways are equivalent in all possible directions), as well as the roughly equivalent. However, there is a compelling characteristics (width of 5±10 mm; length of 100 mm) amount of clinical and experimental data on the of smooth muscle cells in the corpora, the àctive human corpora, and in other vascular tis- vascular unit' would clearly be very dependent on sues,3,11,13,20,37±42,55 all of which point to the fact that corporal geometry and the extent of neural innerva- the physiologically important pathway is likely to be tion. It is important to point out that because gap the intercellular pathway. junctional permeability is an order of magnitude Perhaps this conclusion is not surprising in light less than the cytoplasmic diffusion coef®cient, of the fact that there are many aspects of the in situ second messenger molecule diffusion or current tissue environment that act to severely restrict the ¯ow will always be much greater along the long axis extracellular diffusion of neurotransmitters. For of the smooth muscle cell than perpendicular to the example, one must take into account tissue tortuos- long axis of the cell. Therefore, depending on the ity factors (namely bundles of smooth muscle cells precise orientation of the corporal smooth muscle interspersed in a heavily collagenous extracellular cells in the extracellular matrix, the number of cells matrix; see Ref. 57), variations in the distribution of in the àctive vascular unit' could range from a varicosities and their concomitant neuromuscular minimum of tens of cells to a maximum of hundreds distances,57±59 all of which are further exacerbated of cells.13,35 Thus, neurons, endothelial cells or by the relatively short half-lives and conspicuously intracorporally injected drugs (for example prosta- small amount of neurotransmitter released during glandin E1), can all be regarded as point sources for Integrative erectile biology GJ Christ et al 82 the local activation of groups of corporal smooth seems especially relevant in light of the increased muscle cells. Furthermore, the size of the àctive ef®cacy and greater availability of contemporary vascular unit' would be impacted upon by any age- intracavernous injection modalities, and more re- or disease-related alteration that resulted in signi®- cently of projected oral therapies. All of these are cant changes in the number of nerve terminals, thought to cause erection by virtue of their ability to endothelial cells, or corporal smooth muscle cells (for either increase the production of (namely prosta- example age-related decreases in smooth muscle, glandin E1), or prolong the effects of (papaverine, neuronal or endothelial content), or in the amount/ sildena®l, milrinone) second messenger molecules. composition of the extracellular matrix (for example Presumably, the ability of microgram quantities of the presence of signi®cant ®brosis leading to greater vasoactive compounds injected into one corpora (for separation of the smooth muscle cells, commensu- example PGE1) to elicit rapid and syncytial erectile rately smaller areas of cellular apposition, and thus, responses, is directly related, in large part, to the decreased intercellular coupling). presence of the intercellular coupling pathway that exists among corporal smooth muscle cells. The `syncytial tissue triad': the mechanistic basis for Thus, the ef®cacy of any drug or therapy will be integrative erectile physiology directly related to its ability to take advantage of the ubiquitously distributed gap junction channels to reduce the activity of Ca channels, or increase the In summary then, it is clear that no one mechanism activity of the K channels present in corporal is suf®cient to account for all aspects of corporal smooth muscle. Those therapeutic modalities that smooth muscle function. In this scenario, the ANS, do so in the most ef®cient and least invasive manner intracellular signal transduction mechanisms and will be the most widely utilized by the medical intercellular communication through gap junctions specialists and the consuming public. As described all act in concert to ensure the rapid and syncytial in detail elsewhere1, one attractive way to do so is corporal smooth muscle response required for by the introduction of genes responsible for the normal erectile function. Mathematical modeling synthesis of proteins that regulate the most impor- analyses that taken into account such diverse factors tant facets of corporal smooth muscle tone, eg as innervation density, geometry of innervation potassium channels. (namely one, two, three dimensional), second messenger diffusion/metabolism, and the activity of the autonomic nervous system,56±59 as well as the Acknowledgements distribution, conductance and permeability of the intercellular pathway also support this interpreta- We gratefully acknowledge the helpful comments tion. Thus it seems likely that the participation of and suggestions of Dr Arnold Melman during the gap junctions in erectile physiology/disease is most preparation of this manuscript. The authors are also directly related to the `plasticity' that the presence of grateful to Dr Saul Melman for assistance with these ubiquitously distributed channels are likely to Figure 3. This work was supported by NIH USPHS confer upon corporal smooth muscle cells. Analo- grants DK42027 & 46379. gous to the situation in the heart, where the presence of gap junctions is very important to the prevention of conduction defects, the presence of References gap junctions in the penis would appear to provide a `safety factor' for normal erectile function. Conver- sely, with respect to erectile dysfunction, the innate 1 Christ GJ, Melman A. Molecular studies of human corporal smooth muscle: Implications for the understanding, diagnosis, redundancy provided by gap junctions ensure that and treatment of erectile dysfunction. Molecular Urology 1997; multiple and/or severe derangements would need to 1: 47±56. occur in order to disrupt erectile function. Drawing 2 Lerner SE, Melman A, Christ GJ. A review of erectile another analogy from a distinct literature, the dysfunction: new insights and more questions. J Urol 1993; presence of gap junctions in the human corporal 149: 1246±1255. 3 Christ GJ. The penis as a vascular organ. In: Melman A (ed). ensures that, as in the hypothesis of cancer forma- Impotence. Urologic Clinics of North America. W.B. Saunders tion, `multiple hits' would be required to cause Company: Philadelphia, 1995, pp 727±745. erectile dysfunction in this multi-layered system. 4 Andersson KE, Wagner G. Physiology of penile erection. Physiol Rev 1995; 75: 191±236. 5 Saenz De Tejada I et al. Trabecular smooth muscle modulates Summary and conclusions: why is studying the the capacitor tone of the penis: studies on a rat model. Am J regulation and coordination of corporal smooth Physiol 1994; 5 (2): H1590±H1595. muscle tone so important? 6 Bush PA et al. Nitric oxide as a potent relaxant of human and rabbit corpus cavernosum. J Urol 1992; 147: 1650±1655. 7 Christ GJ, Brink PR, Melman A, Spray DC. The role of gap junctions and ion channels in the modulation of electrical and As discussed above, understanding the regulation chemical signals in human corpus cavernosum smooth and coordination of corporal smooth muscle tone muscle. Int J Impot Res 1993; 5: 77±96. Integrative erectile biology GJ Christ et al 83 8 Burnstock G. Autonomic neuromuscular junctions: current calcium transients in cultured human corporeal tissue. J Urol developments and future directions. J Anat 1986; 146: 1±30. 1996; 155: 622A. 9 Masato T et al. Co-existence of nitric oxide synthase, tyrosine 31 Spektor M, Rosenbaum R, Melman A, Christ G. Further hydroxylase and vasoactive intestinal peptide in human demonstration of the physiological relevance of potassium penile tissueÐa triple histochemical and immunohistochem- channels to contraction of human corporeal smooth muscle in ical study. J Urol 1995; 153: 530±534. vitro. J Urol 1997; 157: 149. 10 Holmquist F, Hedlund H, Andersson KE. L-NG-nitro-arginine 32 Christ GJ et al. Intracavernous injection of the hSlo maxi-K inhibits non-adrenergic, non-cholinergic relaxation of human cDNA: A test of the utility of gene therapy for the treatment of isolated corpus cavernosum. Acta Physiol Scand 1991; 141: erectile dysfunction. 1996 IJIR Abstract A15. 441±443. 33 Christ GJ et al. Expression of potassium channels and 11 Christ GJ, Brink PR, Ramanan SV. Dynamic gap junctional potassium channel genes in human corporal smooth muscle. communication: a delimiting model for tissue responses. J Urol 1997; 157: 414. Biophys J 1994; 67: 1335±1344. 34 Giraldi A et al. Differential relaxation of human corpus 12 Lincoln JR et al. Changes in VIPergic, cholinergic and cavernosum smooth muscle by potassium channel openers: adrenergic innervation of human penile tissue in diabetic evidence that relaxation is both agonist dependent and altered and non-diabetic impotent males. J Urol 1987; 137: 1053±1059. by diabetes mellitus. Int J Impot Res 1995; 10: 1±10. 13 Christ GJ et al. Gap junctions in vascular tissues. Circ Res 35 Christ GJ. The syncitial triadÐa model for understanding how 1996; 79 (4): 1±16. gap junctions participate in the local control of penile 14 Garban H et al. Correction of erectile dysfunction by induction erection. World J Urol 1997; 15: 36±44. of nos in the penis. IJIR Abstract 1996; A03. 36 Campos de Carvalho AC et al. Gap junctions formed of 15 Zhao W, Christ GJ. Endothelin-1 as a putative modulator of connexin 43 interconnect smooth muscle cells of human erectile dysfunction. II. Calcium mobilization in cultured corpus cavernosum. J Urol 1993; 149: 1568±1575. human corporal smooth muscle cells. J Urol 1995; 154: 1571± 37 Christ GJ. Minireview: modulation of alpha-adrenergic con- 1579. tractility in isolated vascular tissues by heptanol; a functional 16 Christ GJ, Lerner SE, Kim DC, Melman A. Endothelin-1 as a demonstration of the potential importance of intercellular putative modulator of erectile dysfunction: I. Characteristics of communication to vascular response generation. Life Sci 1995; contraction of isolated corporal tissue strips. J Urol 1995; 153: 56: 709±721. 1998±2003. 38 Christ GJ et al. Intercellular communication through gap 17 Somlyo AP, Somlyo AV. Signal transduction and regulation in junctions: a potential role in pharmacomechanical coupling smooth muscle. Nature 1994; 372 (17): 231±236. and syncytial tissue contraction in vascular smooth muscle 18 Christ GJ, Maayani S, Valcic M, Melman A. Pharmacologic isolated from human corpus cavernosum. Life Sci 1991; 49: studies of human erectile tissue: characteristics of sponta- 195±200. neous contractions and alterations in alpha-adrenoreceptor 39 Christ GJ et al. Gap junctions modulate tissue contractility and responsiveness with age and disease in isolated tissues. Br J alpha adrenergic agonist ef®cacy in isolated rat aorta. J Pharmacol 1990; 101: 375±381. Pharmacol Exp Ther 1993; 266: 1054±1065. 19 Christ GJ, Stone B, Melman A. Age dependent alterations in 40 Christ GJ, Brink PR, Ramanan SV. The impact of innervation the ef®cacy of phenylephrine-induced contractions in vascu- density on syncytial responses in tissues containing gap lar smooth muscle isolated from corpus cavernosum of junctions. FASEB J 1995; Abstract 9: A913. impotent men. Can J Physiol Pharmacol 1991; 69: 909±913. 41 Christ GJ, Rehman J. Neurophysiology. In: Krane R, Sirosky 20 Christ GJ, Moreno AP, Melman A, Spray DC. Gap junction MB, Goldstein I (eds). Male Sexual Dysfunction. Churchill mediated intercellular diffusion of Ca2 in cultured human Livingstone: London, 1997 (in press). corporal smooth muscle cells. Am J Physiol 1992; 263: 373± 42 Christ GJ, Brink PR, Ramnan SV. Neuronal innervation, signal 383. transduction and erection. J Urol 1996; 155: 620A. 21 Lincoln TM, Cornwell TL, Taylor AE. cGMP-dependent 43 Bennett MVL et al. Gap Junctions: new tools new answers, protein kinase mediates the reduction of Ca2 by cAMP in new questions. Neuron 1991; 6: 305±320. vascular smooth muscle cells. Am J Physiol 1990; 258: C399± 44 Dermietzel R, Kessler JA, Hertzberh EL, Spray DC. Gap C407. Junctions between cultured astrocytes: Immunocytochemical, 22 Lincoln TM, Cornwell TL. Towards an understanding of the molecular and electrophysiological analysis. J Neuroscience mechanism of action of cyclic AMP and cyclic gmp in smooth 1993; 11: 1421±1432. muscle relaxation. Blood Vessels 1991; 28: 129±137. 45 Hall JE, Zampighi GA, Davis RM (eds). Gap Junctions, 23 Cornwell TL, Arnold E, Boerth NJ, Lincoln TM. Inhibition of Progress in Cell Research, Vol. 3. Elsevier Science Publishers: smooth muscle cell growth by nitric oxide and activation of Amsterdam, 1993. cAMP-dependent protein kinase by cGMP. Am J Physiol 1994; 46 Moreno AP et al. Gap junctional communication between 267: C1405±C1413. human corpus cavernosum smooth muscle cells in culture: 24 Jiang H, Colbran JL, Francis SH, Corbin JD. Direct evidence for Gating behavior and single channel events. Am J Physiol 1993; cross activation of cGMP-dependent protein kinase by cAMP 264: C80±C92. in pig coronary arteries. J Biol Chem 1991; 267: 1015±1019. 47 Brink RP, Ramnan SV, Christ GJ. Human connexin 43 gap 25 Palmer LS et al. Characterization of Camp accumulation in junction channel gating: evidence for mode shifts and/or cultured human corpus cavernosum smooth muscle cells. J heterogeneity. Am J Physiol 1996; 271: C321±331. Urol 1994; 152: 1308±1314. 48 Barr L, Dewey MM. Intercellular connection between smooth 26 Cahn D, Melman A, Valcic M, Christ GJ. Forskolin: A muscle cells: the nexus. Science 1962; 137: 670±672. promising new adjunct to intracavernous pharmacotherapy. J 49 Barr L, Berger W, Dewey MM. Electrical transmission at the Urol 1996; 155 (5): 1789±1794. nexus between smooth muscle cells. J Gen Physiol 1968; 52: 27 Christ GJ, Spray DC, Brink PR. Characterization of K currents 347±368. in cultured human corporal smooth muscle cells. J Androl 50 Johansson B, Ljung B. Spread of excitation in the smooth 1991; 14: 319±328. muscle of the rat portal vein. Acta Physiol Scand 1967; 70: 28 Fan SF, Brink PR, Melman A, Christ GJ. An analysis of the 312±322. Maxi-K channel in cultured human corporal smooth muscle 51 Johansson B, Ljung B. Role of Myogenic propagation in cells. J Urol 1995; 153: 818±825. vascular smooth muscle response to vasomotor nerve stimula- 29 Zhang P, Christ GJ, Brink PR, Ney P. PGE induced alterations tion. Acta Physiol Scand 1968; 73: 501±510. in maxi-k channel activity in cultured human corporal smooth 52 Johansson B, Somlyo AP. Electrophysiological and excita- muscle cells. 1996 AUA Abstract 678A. tion-contraction coupling. In: Bohr DF, Somylo AP, Sparks 30 Zhao W, Christ GJ, Ney P. PGE-1 alters endothelin-1 induced HV (eds). Handbook of Physiology: Section 2: the Cardiovas- Integrative erectile biology GJ Christ et al 84 cular System. The Williams and Wilkins Company: Balti- sion in arteries and arterioles. Physiol Rev 1989; 69: 546± more, 1980, pp 301±323. 604. 53 Christ GJ et al. Pharmacomechanical coupling, signal trans- 60 Ganon F, Msghina M, Stjarne L. Kinetics of noradrenaline duction and intracellular communication in human vascular released by sympathetic nerves. Neuroscience 1993; 56: 535± smooth muscle. FASEB J 1996; Abstr 10: A69. 538. 54 Christ GJ et al. Gap junctions modulate tissue contractility and 61 Cornell-Bell AH, Finkbeneiner SM, Cooper MS, Smith SJ. alpha adrenergic agonist ef®cacy in isolated rat aorta. J Glutamate induces calcium waves in cultured astrocytes: long Pharmacol Exp Ther 1993; 266: 1054±1065. range glial signaling. Science 1990; 247: 470±473. 55 Brink PR, Ramnan SV. A model for the diffusion of ¯uorescent 62 Lue TF, Tanagho EA. Physiology of erection and pharmaco- probes in the septate giant axon of the earthworm. Biophys J logic management of impotence. J Urol 1993; 137: 829. 1985; 48: 299±309. 63 Kuriyama H, Kitamura K, Nabata H. Pharmacological and 56 Burnstock G, Relevic V. Neural Endothelial Interactions in the physiological signi®cance of ion channels and factors that Control of Local Vascular Tone. R.G. Landes Company: Austin modulate them in vascular tissues. Pharmacol Rev 1995; 47: Texas, 1993. 391±573. 57 Kenakin T. In: Pharmachologic Analysis of Drug Receptor 64 Sperelakis N, Kuriyama H (eds). In: Ion Channels of Vascular Interaction. Raven Press Ltd: New York, 1994. Smooth Muscle Cells and Endothelial Cells. Elsevier: New 58 Gibbons IL, Morris JL, Furness JB, Costa M. In: Burnstock G, York, 1991. Grif®th SG (eds). Nonadrenergic Innervation of Blood Vessels. 65 Nelson MT, Quayle JM. Physiological roles and properties of CRC Press: Boca Raton, FL, 1988. potassium channels in arterial smooth muscle. Am J Physiol 59 Hirst GDS, Edwards FR. Sympathetic neuroeffector transmis- 1995; 268: C799±C822.