Gap Junctions and Ion Channels: Relevance to Erectile Dysfunction

Gap junctions and ion channels: relevance to erectile dysfunction

GJ Christ1*

1Departments, of Urology and Physiology and Biophysics, Albert Einstein College of Medicine, 1300 Morris Park Avenue, Bronx NY 10461, USA

The corporal myocyte is a critical determinant of erectile capacity whose functional integrity, in the vast majority of impotent patients, is suf®cient to guarantee its relevance as a therapeutic target. As with numerous other smooth muscle cell types, ion channels are important modulators of corporal smooth muscle tone=contractility. As such, the transmembrane ¯ow of ions (ie Ca2, K and Cl7) plays an important role in modulating membrane potential and contractile status in individual human corporal smooth muscle cells, while intercellular ion ¯ow ensures the functionality of myocyte cellular networks. The integral membrane proteins that selectively regulate many aspects of these critical transmembrane (eg K and Ca2 channels) and intercellular (eg gap junctions) ionic movements have been identi®ed. To date, the large conductance calcium-sensitive K channel (ie KCa), the metabolically regulated K channel (ie 2 KATP), and the L-type voltage-dependent Ca channel appear to be the most physiologically relevant nonjunctional ion channels. With respect to intercellular ionic=solute=second messenger movement, connexin43-derived gap junction channels are widely recognized as an obligatory component to normal integrative erectile biology. The presence of an intercellular pathway ensures that individual cellular alterations are carefully orchestrated in the rapid and syncytial fashion required for normal erectile function. This report reviews the known details concerning junctional and nonjunctional ion channels in human corporal tissue, and illustrates how one particular application of this knowledge, that is, preclinical studies utilizing low ef®cacy gene therapy (ie low transfection ef®ciency) with the KCa channel has further con®rmed the physiological relevance and therapeutic potential of gap junctions and ion channels to erectile physiology=dysfunction. International Journal of Impotence Research (2000) 12, Suppl 4, S15± S25.

Keywords: transmembrane potential; smooth muscle tone; corporal myocyte; KCa channel; KATP channel; K channel gene therapy

The importance of the corporal myocyte to an erectile response if the relaxing stimulus is erectile function, disease and improved therapy suf®cient. The documented clinical success of commonly used intracavernous agents in more than 70 ± 80% of patients with erectile dysfunction is The corporal myocyte is a major regulator of erectile testimony to the adequacy of corporal myocyte capacity, and as such, understanding modulation of integrity in many impotent men,2,16 and more recent corporal myocyte contractility provides great oppor- clinical studies indicate that this number may be tunities for the improved therapy of impotence. This pushed even higher using more ef®cacious muscle supposition has been veri®ed by numerous basic relaxants.17 While these observations do not dimin- and clinical studies,1±17 which have clearly shown ish the critical contribution of central nervous that in the absence of severe vascular disease, and system (CNS) pathways to modulation of erectile barring traumatic injury, congenital or other struc- physiology=dysfunction,18 ± 21 they do provide a tural defects, relaxation of the corporal myocyte is cogent scienti®c rationale for further studies of necessary and suf®cient for erection. The histo- corporal smooth muscle contractility. That is, logical=anatomical correlate of these documented regardless of the therapeutic target (ie central or facts is that the majority of impotent men possess peripheral), the endpoint is still the same, namely, adequate corporal smooth muscle integrity to ensure improved corporal smooth muscle relaxation.

*Correspondence: GJ Christ, Departments of Urology and Physiology and Biophysics, Laboratory of Molecular and The ionic equilibrium of the corporal myocyte Integrative Urology, Room 716S, Forchheimer Building, Albert Einstein College of Medicine, 1300 Morris Park Avenue, Bronx, NY 10461, USA. Enzymes, pumps and membrane carriers establish Email: [email protected] and maintain the electrochemical gradients that are Gap junctions and ion channels in ED GJ Christ S16 critical to cellular homeostasis and function in all established and maintained, but rather on their cells, including myocytes.22 The distribution and relationship to transcellular ionic movements. More concentration of ions (ie K,Ca2 and Cl7) across speci®cally, the importance of the passive electro- the cell membrane determine the resting membrane chemical diffusion of ions across the corporal potential, and thereby set the boundary conditions myocyte plasma membrane from the intracellular for the magnitude and direction of current ¯ow to the extracellular space, and vice versa, will be through ion selective transmembrane pores. This discussed with respect to their relevance to both report will not focus on how these gradients are erectile physiology and erectile disease (Figure 1).

Figure 1 Summary of the ionic equilibria and regulatory mechanisms that control corporal smooth muscle tone. Panel (A) shows the electrochemical gradients which direct the ¯ow of ions into and out of the corporal smooth muscle cell. Panel (B) shows a schematic depiction of some of the primary mechanisms thought to modulate corporal smooth muscle cell tone. To illustrate this important point we have arbitrarily divided the myocyte regulatory mechanisms into three distinct components. Note that K channels signi®cantly impact modulation of smooth muscle tone by virtue of their ability to affect both transmembrane calcium ¯ux and intracellular calcium homeostasis=levels. Because of their direct and indirect impact on the factors that govern the degree of smooth muscle contractility, they are an ideal target for modulation of smooth muscle tone. SR: sarcoplasmic reticulum; RyR: ryanodine receptor; SMPP: smooth muscle myosin phosphatase; MLCK: myosin light chain kinase. PKA=C=G: protein kinase A=C=G, etc.; DAG: diacylglycerol; IP3: inositol trisphosphate. Gs=Gq: heterotrimeric G proteins coupled to adenylate cyclase and phospholipase C, respectively. Rho A: monomeric G protein. ?- denotes the fact that this pathway is uncon®rmed in the human tissues used in this proposal. The information summarized in Panel (B) was derived from references 43 ± 61.

International Journal of Impotence Research Gap junctions and ion channels in ED GJ Christ S17 Ion channels are a convergence point for modulating is documented.29 Splice variants of the a subunit myocyte contractility of Slo have been detected in other tissues,74 ± 81 indicating the possibility of transcriptional control of K channel function in the human corpora as Despite the overt complexity of the bifurcating signal well. However, the potential physiological relevance transduction cascades initiated by the wide range of of alternatively spliced hSlo transcripts to erectile extant neural and hormonal signals, it is clear that physiology and function remains to be determined. nonjunctional ion channels represent a critical con- Also worthy of note is the fact that the a-subunit of vergence point for modulating human corporal the maxi-K channel forms a functional channel smooth muscle tone. Recent molecular and biophysi- when expressed by itself, although the activity of cal studies on both freshly isolated tissue strips and the a-channel is signi®cantly modulated by the individual myocytes from the corpora are beginning presence of the b-subunit. to reveal some of the key characteristics of these In this regard, the b-subunit represents the important molecular targets,23 ± 33 as well as their regulatory component of the maxi-K channel, and putative physiological roles in modulating erectile is present in a 1:1 stoichiometry with the a subunit capacity. These characteristics are discussed later. (Figure 2A). When expressed alone, the b-subunit does not form a functional channel, and as such, the perceived role of the b-subunit, in most tissues, is The importance of K channels to the modulation of largely that of modulation of a subunit activity= corporal myocyte contractility sensitivity to, eg Ca2.79 ± 89 Published studies document the presence of two distinct b-subunits, 78 the b1 and b2-subunits, respectively; although at K channels are major modulators of membrane least one other b-subunit is likely to exist. The b1- potential and contractile force in myocytes, and subunit transcript is apparently most highly ex- therefore, they have important therapeutic poten- pressed in muscle, with transcript levels varying tial.34 ± 41 Although genes for numerous K channels rather dramatically among other tissues (eg brain, 87 (more than 30) have already been identi®ed, this liver, lymphatic tissue) A b2-subunit has also report will focus on only two: (1) the large-conduc- recently been cloned from chromaf®n cells and tance calcium sensitive K channel (ie the KCa or hippocampal neurons, and appears to have a K maxi-K channel); and (2) the metabolically-gated K current inactivating role not seen in muscle cells. In channel (ie KATP). Our discussion is limited to these contrast to the a-subunit, there are no documented two K channel subtypes because there is an splice variants of the b-subunit. Preliminary data extensive literature documenting modulation of their from our laboratory indicate that the b1-subunit is activity=function by numerous drugs, neurotransmit- present in human corporal tissue (Day et al, ters, neuromodulators and hormones, following both unpublished observations, 2000). receptor- and nonreceptor-mediated cellular activa- Consistent with the molecular studies, patch tion,34 ± 72 and furthermore, because our published clamp experiments also document the presence data documents the presence and relevance of these of the KCa channel in both cultured and freshly K channel subtypes in the human corpora. isolated human corporal smooth muscle cells. Speci®cally, the voltage-dependence and calcium- sensitivity of the outward whole cell currents, as The maxi-K or K channel well as the measured unitary conductance ( 180 ± Ca 200 pS) are typical of those reported for the maxi-K channel in many cell types.24,25,32 Perhaps more By far, the KCa channel is the best described ion importantly, activation of the maxi-K channel is channel in human corporal smooth muscle. As with known to play an important role in the relaxation many other voltage-dependent K channels, the KCa response to prostaglandin E1, the most commonly channel is formed by a multimeric complex com- used intracavernosally injected=intraurethrally ap- prising both a and b subunits (Figure 2A). The a- plied therapeutic agent for the treatment of erectile subunit of the KCa is the product of a single slo gene dysfunction. That is, PGE1-induced production of mapped to human chromosome 10q23.1.73 The cAMP in cultured human corporal myocytes90 and 73 cDNA has been cloned, and is refered to as hSlo. activation of KCa currents in cultured and freshly Like other voltage-sensitive K channels, the a or isolated human corporal myocytes occurs over the pore forming subunit of the KCa channel, is a same range of PGE1 concentrations that relax tetrameric assembly of homologous subunits. Of isolated corporal tissue strips.91 At the single interest, hSlo represents the human homologue of channel level, the effects of PGE1 are rountinely the slowpoke (ie hslo)K channel gene originally manifested as a 3 ± 5 fold increase in the open cloned from the slowpoke locus of mutant Droso- probability of the KCa channel. Consistent with such phila melanogaster fruit ¯ys (refered to as dSlo). The observations, more recent experiments have also presence of the a subunit in human corporal tissue shown that the relaxation responses of human

International Journal of Impotence Research Gap junctions and ion channels in ED GJ Christ S18

Figure 2 Structural cartoons depicting the subunit composition and arrangement of the KCa (A), KATP (B) and gap junction (C) channels known to be present in human corporal smooth muscle cells. More speci®cally, panel (A) shows the tetrameric arrangement characteristic of the two subunits that comprise the maxi-K channel (ie a-and b-subunits). Again, note that although not drawn to scale, it is clear that the b-subunit, the regulatory subunit, is much smaller than the a-subunit. Panel (B) illustrates the tetrameric arrangement characteristic of each of the two subunits (ie Kir (a-subunit) and SUR (b-subunit); see text) that comprise the smooth muscle KATP channel. In this case, the size of the SUR component (the regulatory subunit, but not necessarily drawn to scale) is greater than that of the Kir subunit. Panel (C) shows the hexameric arrangement characteristic of Cx43-derived gap junction channels. Each individually numbered monomer is referred to as a connexin, in this case, Cx43. The six monomeric connexins form a connexon or hemichannel in one cell, and their union across the extracellular space (see Figure 4) provides partial cytoplasmic continuity between adjacent cells.

corporal tissue strips to PGE1, as well as nitrogly- composed of a tetrameric assembly of inward cerine (NTG) are signi®cantly attenuated by selec- recti®er K channel subunits (ie Kir; the a-subunit) tive blockade of the KCa channel with 1 mM TEA that exist in a 1:1 stoichiometry with a sulfonylurea (Christ, Spektor and Rodriquez, 1999, unpublished receptor subunit (ie SUR, the b-subunit and target of observations). Taken together, such observations orally active sulfonylurea compounds such as provide compelling evidence for the mechanistic glibenclamide) (Figure 2B).62 ± 72 Unlike the pore importance of the KCa channel to relaxation re- forming region of the KCa channel, the tetrameric Kir sponses produced by physiologicaly relevant stimuli assembly is not known to form a functional channel for erection (ie cAMP and cGMP pathways) (see in the absence of the SUR subunit. The Kir gene Figures 1 and 3). family has been divided into six subfamilies based on the sequence identity of the cloned channel cDNAs. Major subfamilies of 1.0, 2.0 and 6.0 have been identi®ed. The Kir2x and Kir6x are arguably The KATP channel the most relevant to the KATP channels found in smooth muscle, and thus far, only Kir6.2 is known

The KATP channel is a also an octameric complex of to be present in human corporal smooth muscle, and a- and b-subunits. The pore forming region is will be further discussed herein.26,29

International Journal of Impotence Research Gap junctions and ion channels in ED GJ Christ S19

Figure 3 Illustration of some primary membrane receptor=effector pathways known to modulate human corporal smooth muscle tone. A simpli®ed depiction of the major control mechanisms responsible for corporal smooth muscle contraction (top left) and relaxation (top right). This schematic depicts the mechanistic basis for the integral role(s), both direct and indirect, played by K channels (top panels, left and right) and gap junctions (bottom panel) in the modulation of contraction and relaxation responses in myocytes. As shown, the responses of the individual smooth muscle cells, as depicted here and in Figure 1, must be integrated into syncytial corporal smooth muscle response. The normal functioning of the corporal myocyte cellular networks are critical to ensure the rapid and coordinated contraction and relaxation responses that are a prerequisite to normal erectile function.

At least two SUR genes have been identi®ed, to relax corporal smooth muscle from animals and SUR1 and SUR2, respectively. Splice variants of man, and furthermore, to elicit erections in monkeys SUR2 provide for SUR2A, SUR2B and SUR2C and humans when injected intracavernosally.59,60 subtypes. There are currently no data available Consistent with these pharmacological observations, concerning the disposition of the SUR subunit(s) patch clamp studies on freshly isolated human in human corporal myocytes, but this area is corporal myocytes have identi®ed two ATP- currently under investigation. Notably, the re- inhibited, glibenclamide sensitive unitary currents, combination of the Kir6.2 subtype with SUR2B of 20 and 60 pS, respectively.26 In addition, a isoforms in vitro yields a KATP-like channel, with a pinacidil-induced, glibenclamide-sensitive whole pharmacological and electrophysiological pro®le cell outward current has been demonstrated on both 23 consistent with the putative KATP channel recently cultured and freshly isolated human corporal reported in freshly isolated human corporal myocytes.26 Our observations indicate that pinacidil myocytes (ie Kir6.2).26 increases the open probability of both the 20 and 60 Studies on isolated tissue strips have con®rmed pS channels. These observations clearly indicate the ability of several K channel activators (eg that the KATP channel subtype is an important pinacidil, cromakalim, lemakalim and nicorandil) modulator of human corporal smooth muscle tone.

International Journal of Impotence Research Gap junctions and ion channels in ED GJ Christ S20 Intercellular communication through gap junctions The distribution, regulation and biophysical characteristics of Cx43-derived gap junction channels in human corporal smooth muscle are ideally Gap junctions represent yet another ion channel consituted for their required role in coordinating gene family found in human corporal smooth contraction and relaxation responses among the muscle. In contrast to the tetrameric nature of K corporal myocytes. Speci®cally, they have a rela- channels, the pore forming subunit of gap junctions, tively large conductance ( 100 ± 120 pS), a rela- refered to as a connexon or hemichannel, are formed tively long mean open time (0.5 ± 5 s), a much by hexamers of connexins (Figure 2C).92 ± 95 The shorter mean closed time, and therefore, a corre- union across the extracellular space of adjacent spondingly high open probability of 90%.98 ± 100 hemi-channels from cell pairs provide the aqueous Moreover, gap junctions are relatively nonselective intercellular pathway for cytosolic solute diffusion ion channels, that is, in contrast to K channels, gap (Figure 4). Rafts of hundreds to thousands of these junctions discriminate poorly based on charge as individual intercellular channels form the structural re¯ected by an 0.5 ± 0.8 anion=cation selectivity basis for the gap junctional plaques reported in ratio.94 Gap junctions are permeable to most sub- human corporal smooth muscle.96 While more than stances with molecular weights less than 1 kD. In 12 mammalian connexins have been identi®ed, the short then, they are ubiquitously distributed, much diversity of gap junction channels pales in compar- more likely to be open than closed, and freely ison to the diversity of K channels. Gap junctions permeable to virtually all of the major modulators of are named according to their predicted molecular corporal smooth muscle tone. It is clear that they are weights, which are known to range from 26 to idealy suited for their critical role in modulating 56 kD.92 ± 96 There are no known splice variants of erectile capacity=function (Figure 4). the connexin proteins, and furthermore, there are no reported regulatory subunits associated with The concept of the `Syncytial Tissue Triad' the connexin proteins. Finally, a single connexin transcript, that is, connexin43 (ie Cx43) seems to account for the vast majority of intercellular It is now clear that nonjunctional ion channels communication in the human corpora.96 ± 101 regulate the myocyte cellular response, while gap

Figure 4 A cartoon of the organization of, and structural basis for, intercellular communication in human corporal smooth muscle. These structures form the anatomic backbone of the intercellular pathways shown in Figure 3, and therefore, are critical to ensure the proper functioning of the corporal myocyte network.

International Journal of Impotence Research Gap junctions and ion channels in ED GJ Christ S21 junction channels coordinate the composite corpor- tissue triad' in modulating tissue responses and al tissue response. However, how these two compo- erectile capacity have recently been codi®ed.103,104 nents are coordinated with the central and The contributions of these three distinct compo- peripheral neural control of erection adds yet nents to normal erectile physiology and function another layer of complexity. To address this issue, corresponds to intuition, but are nicely veri®ed by and obtain an even more complete mechanistic the preclinical success of Ion Channel Therapy (ie understanding of integrative erectile biology, we K channel gene therapy) for the treatment of have developed a conceptual framework refered to impaired erectile capacity in the male rat. as the `Syncytial Tissue Triad', which takes the additional step of incorporating the function=activity of the nervous system into erectile physiology `Low ef®ciency gene transfer' and the restoration of (Figure 5).4,102 Brie¯y, the `Syncytial Tissue Traid' compromised erectile capacity in the aged and posits that coordination of corporal smooth muscle STZ-diabetic rat tone, and erectile capacity, occurs at the intersection of the dynamic interplay of the following three components: (1) `The signal', which re¯ects the As previously documented for hSlo gene therapy in neuronal innervation density, geometry of innerva- retired breed rats,28,31 incorporation and expression tion, ®ring rate, etc. (2) `Signal transduction', the of hSlo cDNA has also been shown to restore the intracellular signal transduction mechanisms, diabetes-related decline in the nerve-stimulated which re¯ect the response of the corporal myocyte intracavernosal pressure (ICP) response.105,106 Pre- to changes in the extracellular hormonal mileu, and sumably, in both instances, low level expression of a (3) `Signal Spread', which re¯ects the absolute human gene with high sequence homology to the rat importance of intercellular communication through sequence accounts for the apparent absence of any gap junctions, because the corporal smooth muscle detectable immune response (Christ et al, unpub- cells are electrically inexcitable (ie not capable of lished observations). Moreover, as previously de- regenerative electrical events). Theoretical treat- scribed,28 the mechanistic basis for the ef®cacy of ments of the role of gap junctions and the `syncytial hSlo gene therapy is most probably related to the

Figure 5 Schematic depiction of the `Syncytial Tissue Triad'.

International Journal of Impotence Research Gap junctions and ion channels in ED GJ Christ S22 increased hyperpolarizing ability provided by the of corporal myocyte tone (see Figures 1 ± 3), and as over-expression of K channels on a fraction of the such, can be activated independent of any given rat corporal myocytes. Cx43-derived intercellular receptor=effector=second messenger pathway. How- communication ensures that the àugmented' hyper- ever, because K channels represent an important polarizing signals are spread throughout the corpora convergence point for the modulation of corporal resulting in ìncreased' sensitivity of the myocyte to smooth muscle tone, their activation will produce nerve stimulation; note that the increased sensitivity smooth muscle relaxation regardless of whether they occurs whether or not there are any other myogenic are directly (ie K channel modulators; see later) or changes. Such observations bode well for the indirectly (ie via second messenger pathways and therapeutic possibilities of gene therapy for the protein kinase phosphorylation of consensus amino treatment of erectile dysfunction, and moreover, acids sites) activated. Second, because 106 ±108 provide an important validation of the concept ions=sec can ¯ow through an open K channel, it refered to as the `Syncytial Tissue Triad'. is clear that relatively small alterations (ie increases) in their activity will produce a physiologically relevant degree of corporal smooth muscle cell hyperpolarization, and thus, relaxation. Third, The importance of ion channel modulation of because the corporal myocytes are interconnected corporal contractility, rather than pharmacological by a network of small gap junction assemblies, not ablation of corporal contractility every cell would have to èxperience' an increase in K channel activity in order for signi®cant smooth muscle relaxation to occur. Fourth, because K In summary, it is clear from the published clinical channels nominally modulate, but do not ablate, and experimental literature that relatively subtle corporal smooth muscle contraction, they are not alterations in the balance between the effects of likely to result in priapism. In conclusion then, it endogenous vasoactive substances on the corporal seems that K channel gene therapy approaches, smooth muscle cell are involved in the etiology of alone or in combination with K channel modulat- erectile failure in a majority of impotent men.1±15 ing drugs, would provide an attractive therapeutic Equally clear from our recently published pre- strategy for the treatment of erectile dysfunction. clinical studies on the rat model in vivo is that a relatively subtle increase in the hyperpolarizing ability of the corporal myocytes (as evidenced by the fact that RT-PCR was required to detect the References presence of a human gene expressed in the rat corporal smooth muscle cell) is suf®cient to restore the erectile capacity that is compromised by both 1 Christ GJ. The penis as a vascular organ: the importance of corporal smooth muscle tone in the control of erection. age- and diabetes-related changes in corporal func- Urologic Clinics of North America 1995; 22: 727 ± 745. tion.28,105,106 This exciting preclinical observation 2 Melman A, Christ GJ. The vascular physiology of penile stands in stark contrast to the traditional pharma- erection and pharmacotherapy erectile dysfunction. In: cological approach to the treatment of erectile Frishman WH, Sonnenblick EH, eds. Cardiovascular Phar- macotherapeutics. McGraw-Hill: NY, 1997, pp 1221 ± 1229. dysfunction, which was largely based on the 3 Christ GJ. Erection. The Encyclopedia of Reproduction. empirical observation that potent vasorelaxants, Academic Press: NY, 1998, pp 38 ± 49. such as prostaglandin E1, can overwhelm the 4 Christ GJ. The `syncytial tissue triad': a model for under- contractile status of the corporal myocyte to induce standing how gap junctions participate in the local control of erections following intracavernous (ie Caverject1) penile erection. World J Urol 1997; 13: 36 ± 44. 5 Christ GJ, Richards S, Winkler A. Integrative erectile biology: or intraurethral (MUSE) administration. It is now the role of signal transduction and cell-to-cell communica- clear that a new age, one of more subtle and selective tion in coordinating corporal smooth muscle tone and penile oral therapy for the treatment of erectile dysfunction erection. Int J Impot Res 1997; 9: 1 ± 16. has arrived. This fact is re¯ected by the availability 6 Taub HC, Lerner SE, Melman A, Christ GJ. Relationship of ViagraTM, and further evidenced from the appar- between contraction and relaxation in human and rabbit corpus cavernosum. Urology 1993; 42: 698 ± 704. ent success of other compounds in late stage clinical 7 Lerner SE, Melman A, Christ GJ. A review of erectile trials, such as Uprima. However, even these treat- dysfunction: new insights and more questions. J Urol 1993; ments are dependent on the presence of suf®ciently 149: 1246 ± 1255. intact receptor=effector=second messenger pathways 8 Wagner G. Erection. In: Wagner G, Green R, eds. Impotence: Physiological, Psychological and Surgical Diagnosis and for the manifestation of their actions=ef®cacy. Treatment. Plenum Press: New York, 1981, pp 25 ± 36. With regard to this, and in light of all of the 9 Saenz De Tajada I et al. Impared neurogenic andendothe- aforementioned discussion, it seems logical that lium-mediated relaxation of penile smooth muscle from treatments directed to modulation of ion channels, diabetic mem with impotence. New Engl J Med 1989; 320: K channels in particular, may have several intrin- 1025. 10 Saenz De Tejada I et al. Trabecular smooth muscle modulates sic advantages over currently available forms of the capacitor function of the penis studies on a rabbit model. therapy. First, K channels are primary modulators Am J Physiol 1991; 260: H1590 ± H1595.

International Journal of Impotence Research Gap junctions and ion channels in ED GJ Christ S23 11 Wagner G, Saenz De Tejada I. Update on male erectile 35 Jackson WF. Ion channels and vascular tone. Hypertension dysfunction. Br Med J 1998; 316: 678 ± 682. 2000; 35: 173 ± 178. 12 Krane RJ, Goldstein I, Saenz De Tejada I. Medical progress: 36 Lawson K. Is there a therapeutic future for `potassium impotence. New Engl J Med 1989; 321: 1648 ± 1659. channel openers'? Clinical Science 1996; 91: 651 ± 663. 13 Lue TF, Tanagho EA. Physiology of erection and pharmaco- 37 Lawson K. Potassium channel activation: a potential ther- logic management of impotence. J Urol 1987; 137: 829 ± 836. apeutic approach? Pharmacol Ther 1996; 70: 39 ± 63. 14 Andersson K-E, Wagner G. Physiology of penile erection. 38 Nelson MT, Quayle JM. Physiological roles and properties of Physiological Reviews 1995; 75: 191 ± 225. potassium channels in arterial smooth muscle. Am J Physiol 15 Christ GJ. The control of corporal smooth muscle tone, the 1995; 268: C799 ± C822. coordination of penile erection and the etiology of erectile 39 Ashcroft FM. Ion Channels and Disease: Channelopathies. dysfunction: the devil is in the details. J Sex Educat Ther Academic Press: NY, 2000. 1998; 23: 187 ± 193. 40 Gribkoff V, Starrett JE Jr, Dworetzsky SI. The pharmacology 16 Linet IO et al. Ef®cacy and safety of intracavernosal and molecular biology of large-conductance calcium- alprostadil in men with erectile dysfunction. The alprostadil activated (BK) potassium channels. In: Adv Pharmacol, study group. New Engl J Med 1996; 334: 873 ± 877. Academic Press: NY, 1997, pp 319 ± 347. 17 Mulhall JP et al. Intracavernosal forskolin: role in manage- 41 Starrett JE, Dworetzky SI, Gribkoff VK. Modulators of large- ment of vasculogenic impotence resistant to standard 3-agent conductance calcium-activated potassium (BK) channels as pharmacotherapy. J Urol 1997; 158: 1752 ± 1758. potential therapeutic targets. Curr Pharmac Design 1996; 2: 18 Mc Vary KT, Rathnau CH, McKenna KE. Sexual dysfunction 413 ± 428. in the diabetic BB=WOR rat: a role of central neuropathy. Am 42 Dorschner H et al. Stoichiometry of sulfonylurea-induced J Physiol 1997; 272: R259 ± R267. ATP-sensitive potassium channel closure. Mol Pharmacol 19 McKenna KE. Central control of penile erection. Int J Impot 1999; 55: 1060 ± 1066. Res 1998; 10(Suppl): S25 ± S34. 43 Nelson M et al. Relaxation of arterial smooth muscle by 20 Bernabe J, Rampin O, Sachs BD, Guiliano F. Intracavernous calcium sparks. Science 1995; 270: 633 ± 637. pressure during erection in rats: an integrative approach 44 Quayle JM, Nelson MT, Standen NB. ATP-Sensitive and based on telemetric recording. Am J Physiol 1999; 276: inwardly rectifying potassium channels in smooth muscle. R441 ± R449. Physiol Rev 1997; 77: 1165 ± 1232. 21 Nofzinger EA. Sexual dysfunction in patients with diabetes 45 Jaggar JH, Stevenson AS, Nelson MT. Voltage dependence of mellitus: the role of a `central' neuropathy. Semin Clin Ca2 sparks in intact cerebral arteries. Am J Physiol 1998; Neuropsychiatry 1997; 2: 31 ± 39. 274: C1755 ± C1761. 22 Speralakis N. Origin of resting membrane potentials. In: 46 Bielefeldt K, Jackson MB. Phosphorylation and dephos- Speralakis N, (ed). Cell Physiology Source Book, 2nd edition. phorylation modulate a Ca2-activated K channel in Academic Press: NY, 1998, pp 178 ± 206. rat peptidergic nerve terminals. J Physiol 1994; 475: 23 Christ GJ, Spray DC, Brink PR. Characterization of K currents 241 ± 254. in cultured human corporal smooth muscle cells. J Androl- 47 Chung S et al. Protein kinase A activity closely associated ogy 1993; 14: 319 ± 328. with a reconstituted calcium-activated potassium channel. 24 Fan SF, Brink PR, Melman A, Christ GJ. An analysis of the Science 1991; 253: 560 ± 562. 2 Maxi-K (KCa) channel in cultured human corporal smooth 48 Heppner TJ, Bonev AD, Nelson MT. Ca -activated K muscle cells. J Urol 1995; 153: 818 ± 825. channels regulate action potential depolarization in urinary 25 Lee SW et al. Prostaglandin E1 activates the large conduc- bladder smooth muscle. Am J Physiol 1997; 273: C110 ± C117. tance KCa channel in human corporal smooth muscle. Int J 49 Quast U, Guillon JM, Cavero I. Cellular pharmacology of Impot Res 1999; 11: 179 ± 188. potassium channel openers in vascular smooth muscle 26 Lee SW, Wang H-Z, Christ GJ. Characterization of ATP- [Review]. Cardiovasc Res 1994; 28: 805 ± 810. sensitive potassium channels in human corporal smooth 50 Robertson BE, Schubert R, Hescheller J, Nelson MT. cGMP- muscle cells. Int J Impot Res 1999; 11: 189 ± 199. dependent protein kinase activates Ca-activated K channels 27 Christ GJ, Richards S, Winkler A. Integrative erectile biology: in cerebral artery smooth muscle cells. Am J Physiol 1993; the role of signal transduction and cell-to-cell communica- 265: C299 ± C303. tion in coordinating corporal smooth muscle tone and penile 51 Somlyo AP, Wu X, Walker LA, Somlyo AV. Pharmacome- erection. Int J Impot Res 1997; 9: 1 ± 16. chanical coupling: the role of calcium, G-proteins, kinases 28 Christ GJ et al. Transfection of rat corporal tissue in vivo with and phosphatases. Rev Physiol Biochem Pharmcol 1999; 134: hSlo cDNA is associated with physiologically relevant 202 ± 234. alterations in erectile function. Am J Physiol 1998; 275: 52 Somlyo AP, Somlyo AV. Signal transduction and regulation H600 ± H608. in smooth muscle. Nature 1994; 372: 213 ± 236. 29 Christ GJ et al. Ion channels and gap junctions: their role in 53 Somlyo AP, Somlyo AV. From pharmacomechanical coup- erectile physiology, dysfunction and future therapy. Mol ling to G-proteins and myosin phosphatase. Acta Physiol Urology 1999; 3: 61 ± 74. Scand 1998; 164: 437 ± 448. 30 Christ GJ, Melman A. Molecular studies of human corporal 54 Garland CJ, Angus JA. Pharmacology of Vascular Smooth smooth muscle: implications for the understanding, diag- Muscle. Oxford University Press: New York, 1996. nosis, and treatment of erectile dysfunction. Mol Urol 1997; 55 Sleight RG, Lieberman MA. Signal transduction. In: Sper- 1: 45 ± 54. elakis N, (ed). Cell Physiology Source Book. Academic Press: 31 Christ GJ. A new frontier: gene therapy for erectile dysfunc- NY, 1998. tion. In: Morales A, (ed). Current Concepts in the Treatment 56 Sperelakis N, Kuriyama H. Ion Channels of Vascular Smooth of Erectile Dysfunction, Martin Dunitz Publishers: London, Muscle Cells and Endothelial Cells. Elsevier: NY, 1991. 1998, pp 209 ± 230. 57 Catterall WA. Molecular properties of a superfamily of 32 Wang H-Z, Lee SW, Christ GJ. Electrophysiological evalua- plasma membrane cation channels. Curr Opinion Cell Biol tion of K channels in enzymatically dissociated myocytes 1994; 6: 607 ± 615. from human and rat corpus cavernosum. Int J Impot Res 58 Noack T, Noack P. Multiple types of ion channels in 2000; 12: (in press). cavernous smooth muscle. World J Urol 1997; 15: 45 ± 49. 33 Fan S-F, Christ GJ, Melman A, Brink PR. A stretch-sensitive 59 Andersson K-E. Clinical pharmacology of potassium channel chloride channel in corpora cavernosa myocytes. Int J Impot openers. Pharmacol Toxicol 1992; 70: 244 ± 254. Res 1999; 11:1±7, 60 Benevides MD et al. Intracavernosal (IC) injection of a 34 Standen NB, Quayle JM. K channel modulation in arterial potassium channel opener to treat erectile dysfunction. J Urol smooth muscle. Acta Physiol Scand 1998; 164: 549 ± 557. 1999; 161(Suppl): 212: (Abstract #812).

International Journal of Impotence Research Gap junctions and ion channels in ED GJ Christ S24 61 Bradley KK et al.Kir2.1 encodes the inward recti®er 83 Toro L, Wallner M, Meera P, Tanaka Y. Maxi-KCa, a unique potassium channel in rat arterial smooth muscle cells. member of the voltage-gated K channel superfamily. News J Physiol (Lond.) 1999; 515: 639 ± 651. Physiol Sci 1998; 13: 112 ± 117. 62 Aguilar-Bryan L et al. Toward understanding the assembly 84 Tseng-Crank J et al. Cloning, expression, and distribution of 2 and structure of KATP channels. Physiol Rev 1998; 78: 227 ± aCa -activated K channel-subunit from human brain. Proc 245. Natl Acad Sci 1996; 93: 9200 ± 9205. 63 Babenko AP, Aguilar-Bryan L, Bryan J. A view of SUR=Kir 85 Xie J, McCobb DP. Control of alternative splicing of 6.x, KATP channels. Ann Rev Physiol 1998; 60: 667 ± 687. potassium channels by stress hormones. Science 1998; 280: 64 Erginel-Unaltuna N, Yang WP, Blanar MA. Genomic organi- 443 ± 446. zation and expression of KCNJ8=Kir6.1, a gene encoding a 86 Knaus HG et al. Distribution of high-conductance Ca2- subunit of an ATP-sensitive potassium channel. Gene 1998; activated K channels in rat brain: targeting to axons and 211: 71 ± 78. nerve terminals. J Neuroscience 1996; 16: 955 ± 963. 65 Inagaki N et al. Reconstitution of IKATP: an inward recti®er 87 Jiang Z, Wallner M, Meera P, Toro L. Human and rodent subunit plus the sulfonylurea receptor. Science 1995; 270: maxi-K channel b-subunit genes: cloning and characteriza- 1166 ± 1170. tion. Genomics 1999; 55: 57 ± 67. 66 Inagaki N et al. Cloning and functional characterization 88 Saito M, Nelson C, Salkoff L, Lingle CJ. A cysteine-rich of a novel ATP-sensitive potassium channel ubiquitously domain de®ned by a novel exon in a Slo variant in rat adrenal expressed in rat tissues, including pancreatic islets, chromaf®n cells and PC12 cells. J Biol Chem 1997; 272: pituitary, skeletal muscle, and heart. J Biol Chem 1995; 270: 11710 ± 11717. 5691 ± 5694. 89 Tanaka Y et al. Molecular constituents of maxi KCa channels 67 Inagaki N, Inazawa J, Seino S. cDNA sequence, gene in human coronary smooth muscle: predominant a b structure, and chromosomal localization of the human subunit complexes. J Physiol 1997; 502: 545 ± 557. ATP-sensitive potassium channel, uKATP-1, gene (KCNJ8). 90 Palmer L et al. Characterization of cAMP accumulation in Genomics 1995; 30: 102 ± 104. cultured human corpus cavernosum smooth muscle cells. 68 Isomoto S et al. A novel sulfonylurea receptor forms with J Urol 1994; 152: 1308 ± 1314. BIR (Kir6.2) a smooth muscle type ATP-sensitive K channel. 91 Cahn DJ, Melman A, Valcic M, Christ GJ. Forskolin: a J Biol Chem 1996; 271: 24321 ± 24324. promising new adjunct to intracavernous pharmacotherapy. 69 Seino S. ATP-sensitive potassium channels: a model of J Urol 1996; 155: 1789 ± 1794. heteromultimeric potassium channel=receptor assemblies. 92 Gilula NB. Gap Junction-Mediated Intercellular Signaling in Ann Rev Physiol 1999; 61: 337 ± 362. Health and Disease. John Wiley & Sons: NY, 1999. 70 Stoffel M et al. Cloning of rat KATP-2 channel and decreased 93 Brink PR, Valiunas V, Christ GJ. Homotypic, heterotypic and expression in pancreatic islets of male zucker diabetic fatty heteromeric gap junction channels. In: Perachia C, (ed). rats. Biochem Biophys Res Comm 1995; 212: 894 ± 899. Current Topics in Membranes, Vol 49. Academic Press: NY, 71 Dorschner H et al. Stoichiometry of sulfonylurea-induced 2000, pp 43 ± 60. ATP-sensitive potassium channel closure. J Pharmacol Exp 94 Veenstra RD. Ion permeation through connexin gap junction Ther 1999; 55: 1060 ± 1066. channels: effects on conductance and selectivity. In: Perachia 72 Schwanstecher M et al. Potassium channel openers require C, (ed). Current Topics in Membranes, Vol 49. Academic ATP to bind to and act through sulfonylurea receptors. EMBO Press: NY, 2000, pp 95 ± 130. J 1998; 17: 5529 ± 5535. 95 Christ GJ et al. Gap junctions in vascular tissues: evaluating 73 McCobb DP et al. Human calcium-activated potassium the role of intercellular communication to the modulation of channel gene expressed in vascular smooth muscle. Am J vasomotor tone. Circ Res 1996; 79: 631 ± 646. Physiol 1995; 269: H767 ± H777. 96 Campos de Carvalho AC et al. Gap junctions formed of 74 Ferrer J, Wasson J, Salkoff L, Permutt MA. Cloning of human connexin43 interconnect smooth muscle cells of the human pancreatic islet large conductance Ca2-activated K channel corpus cavernosum. J Urol 1993; 149: 1568 ± 1575. (hSlo) cDNAs: evidence for high levels of expression in 97 Serels S et al. Molecular studies of connexin 43 (Cx43) pancreatic islets and identi®cation of a ¯anking genetic expression in isolated corporal tissue strips and cultured marker. Diabetologia 1996; 39: 891 ± 898. corporal smooth muscle cells. Int J Impot Res 1998; 10: 75 Atkinson NS, Robertson GA, Ganetsky B. A component of 1±9. calcium-activated potassium channels encoded by the 98 Moreno AP, Campos de Carvalho AC, Christ GJ, Spray DC. Drosophila slo locus. Science 1991; 253: 551 ± 555. Gap junctional communication between human corpus 76 Tseng-Crank J et al. Cloning, expression, and distribution of cavernosum smooth muscle cells in culture: gating functionally distinct Ca2-activated K channel isoforms behavior and single channel events. Am J Physiol 1993; 264: from human brain. Neuron 1994; 13: 1315 ± 1330. C80 ± C92. 77 Butler A et al. mSlo, a complex mouse gene encoding `maxi' 99 Christ GJ, Brink PR. An analysis of the presence and calcium-activated potassium channels. Science 1993; 261: physiological relevance of subconducting states of connex- 221 ± 224. in43-derived gap junction channels in cultured human 78 Wallner M, Meera P, Toro L. Molecular basis of fast corporal vascular smooth muscle cells. Circ Res 1999; 84: inactivation in voltage and Ca2-activated K channels. A 797 ± 803. transmembrane b-subunit homolog. Proc Natl Acad Sci 1999; 100 Brink PR, Ramanan SV, Christ GJ. Human connexin43 gap 96: 4137 ± 4142. junction channel gating: evidence for mode shifts and=or 79 Jones EMC, Gray-Keller M, Fettiplace R. The role of Ca2- heterogeneity. Am J Physiol 1996; 271: C321 ± C331. activated K channel spliced variants in the tonotopic 101 Christ GJ et al. Intercellular communication through gap organization of the turtle cochlea. J Physiol 1999; 518: junctions: a potential role in pharmacomechanical coupling 653 ± 665. and syncytial tissue contraction in vascular smooth muscle 80 Rosenblatt KP, Sun ZP, Heller S, Hudspeth AJ. Distribution isolated from the human corpus cavernosum. Life Sci 1991; of Ca2-activated K channel isoforms along the tono- 49: PL-195 ± 200. topic gradient of the chicken's cochlea. Neuron 1997; 19: 102 Christ GJ. The `syncytial tissue triad': the mechanistic basis 1061 ± 1075. for integrative tissue physiology. Einstein Quarterly J Biol 81 Ramanathan K et al. A molecular mechanism for electrical Med 1999; 15: 179 ± 191. tuning of cochlear hair cells. Science 1999; 283: 215 ± 217. 103 Ramanan SV, Brink PR, Christ GJ. Neuronal innervation, 82 Adelman JP et al. Calcium-activated potassium channels intracellular signal transduction and intercellular coupling: a expressed from cloned complementary DNAs. Neuron 1992; model for syncytial tissue responses in the steady state. 9: 209 ± 216. J Theor Biol 1998; 193: 69 ± 84.

International Journal of Impotence Research Gap junctions and ion channels in ED GJ Christ S25 104 Christ GJ, Brink PR, Ramanan SV. Dynamic gap junctional 106 Day NS et al. hSlo transcript is locally restricted after a single communication: a delimiting model for tissue responses. intracavernous injection of naked hSlo=pcDNA. FASEB J Biophys J 1994; 67: 1335 ± 1344. 1999; 13: A356. 105 Christ GJ et al.K channel gene therapy is associated with increased nerve-stimulated intracavernous pressure (ICP) responses despite diminished neuronal innervation following 3 ± 4 months of experimental diabetes in rats. FASEB J 1999; 13: A192.

Appendix somebody a pill, they may not have an erection right away; but if they also have sexual stimuli and thus, a more responsive apparatus because the channels are Open discussion following Dr Christ's presentation more likely to be open than closed, that extra signal may really ramp up the open probability of the channels. You increase the hyperpolarizing capacity Dr Heaton: You presented the story of ion channels of the tissue and see a much better erection. that exist in every cell of the body, so what The second possibility is that there is splice characterizes the speci®city of the channels that variance and channel isoforms and also other forms we're going to ®nd in the penis that allow you to of beta subunits. Combinations and variations in the manipulate those channels? Is there a sentinel expression of isoforms in different tissues may characteristic of potassium channels that makes them have a different key for different parts of the provide opportunities to develop drugs for speci®c regions of the channel subunit so that they're only body? activating the channel in certain tissues. From our own experience using the in vivo rat Dr Christ: We have also questioned whether there model, we can ®nd concentrations of `dirty' K a tissue-speci®c role for the way the channel ATP and maxi K channel openers that we can inject operates. For example, suppose that in corpus intravenously into rats without affecting blood cavernosum, the open probability of the potassium pressure but signi®cantly enhance the nerve stimu- channel is 1=10 of 1% and in every other cell type in lated intercavernous pressure-response, and we can the body, it's 10 times that. Then, suppose a K block it with selective K channel blockers. channel modulator given systemically activated this channel and pick a concentration at which you could increase its activity threefold. If you go from Dr Heaton: In a nutshell, the chances of ®nding 1=10 of 1% to 3=10 of 1%, you may notice a huge penile-speci®c isoforms within the domain of K difference in the corpora. But in a cell type that channels is at least as good and maybe better than it already has an open probability of 5 or 10%, at the is within the domain of phosphodiesterases. same concentration of modulator, you won't notice any difference at all. So, one possibility when you discuss tissue-speci®c functions of potassium chan- Dr Christ: Correct. Because with ViagraTM, you nels, is that they spend so much time in the closed only see the signal under certain conditions, even state in one organ, that any increase in opening is though that isoform is distributed throughout the going to affect function of that organ. If you give body.

International Journal of Impotence Research