HOME HELP FEEDBACK SUBSCRIPTIONS ARCHIVE SEARCH TABLE OF CONTENTS		QUICK SEARCH:   [advanced] Author: Keyword(s): Year:  Vol:  Page:

This Article Abstract Full Text (PDF) Services Email this article to a friend Similar articles in this journal Similar articles in PubMed Alert me to new issues of the journal Download to citation manager Citing Articles Citing Articles via Google Scholar Google Scholar Articles by Ryan, J. S. Articles by Kelly, M. E. M. Search for Related Content PubMed PubMed Citation Articles by Ryan, J. S. Articles by Kelly, M. E. M.

Adrenergic regulation of calcium-activated potassium current in cultured rabbit pigmented ciliary epithelial cells

Jennifer S. Ryan *, Qian-Ping Tao * and Melanie E. M. Kelly *¹

Received 5 January 1998; accepted after revision 6 May 1998.

	ABSTRACT

Top Abstract Introduction Methods Results Discussion References

1. The effects of adrenergic agonists on K⁺ currents were studied in cultured rabbit pigmented ciliary epithelial (PCE) cells.

2. Outward K⁺ current (I_K) was reduced by tetraethylammonium chloride, the Ca²⁺-activated K⁺ (K(Ca)) channel blocker iberiotoxin (IbTX), or Ca²⁺-free external Ringer solution. The calcium ionophore ionomycin increased an IbTX-sensitive I_K in PCE cells.

3. The adrenergic agonists adrenaline and phenylephrine increased I_K in PCE cells. The induced current was blocked by IbTX and the ₁-antagonist prazosin, suggesting that adrenergic agonists activate I_K(Ca) via ₁-adrenoreceptors.

4. Internal dialysis of D-myo-inositol 1,4,5-trisphosphate (IP₃) increased I_K, whilst pre-incubation of PCE cells with thapsigargin or the phospholipase C (PLC) inhibitor U-73122 reduced phenylephrine-induced increases in I_K(Ca). Adrenergic increases in I_K(Ca) were mediated by a pertussis toxin-insensitive G protein.

5. These results demonstrate that I_K(Ca) channels in rabbit PCE cells are coupled to ₁-adrenergic receptors and a PLC/IP₃ signalling pathway. Activation of these channels may modulate fluid secretion by the ciliary epithelium.

	INTRODUCTION

Top Abstract Introduction Methods Results Discussion References

The ciliary epithelium of the eye is a secreting epithelium comprising two different epithelial layers, a non-pigmented ciliary epithelial (NPCE) cell layer, whose basolateral membrane faces the posterior and vitreal spaces of the vertebrate eye, and a pigmented ciliary epithelial (PCE) cell layer, whose basal surface faces the stromal side. These two cell layers are coupled via gap junctions at their apical membranes (Raviola & Raviola, 1978; Cole, 1984). The formation of aqueous humour in the eye occurs by ultrafiltration across the capillary walls of the ciliary body and by electrolyte secretion across the ciliary epithelium. The rate and quantity of aqueous humour production is an important determinant of intraocular pressure and is subject to autonomic modulation (Cole, 1984).

Recent transport data from intact and dispersed ciliary epithelial tissue suggest that the PCE cells have solute uptake properties and are coupled to the NPCE cells, which have solute efflux properties (Weiderholt et al. 1991; Edelman et al. 1994). Thus, ions and water pass from PCE cells to NPCE cells via gap junctions and are secreted at the basolateral NPCE cell membrane as aqueous humour, an isotonic solution composed primarily of water, Na⁺, Cl^- and HCO₃^-. The mechanisms by which ion transport and fluid secretion are regulated in coupled PCE and NPCE cells still remains unclear, although it is established that vectorial transport of ions and solutes between the coupled epithelial cells requires the co-ordinated interaction of membrane-localized transporters and intracellular signalling pathways (Jacob & Civan, 1996).

Studies of Ca²⁺ mobilization and heterocellular signal transfer in the rabbit ciliary epithelium, however, has revealed that various agonists such as adrenergic and muscarinic agonists can increase Ca²⁺ in NPCE cells of intact ciliary processes in a synergistic manner (Farahbaksh & Cilluffo, 1994). More recently, studies comparing the Ca²⁺ response in isolated ciliary body epithelial (CBE) cells and the intact ciliary body have demonstrated that isolated PCE and NPCE cells differ in their responses to various agonists with respect to Ca²⁺ mobilization. For example, isolated PCE cells respond to ₁-adrenoceptor (₁-AR) agonists but show little sensitivity to ₂-adrenoceptor (₂-AR) agonists, whilst the opposite response profile is found in NPCE cells. These experiments demonstrated distinct rectificatory behaviour in the ciliary epithelial syncitium with synergistic agonist-induced increases in Ca²⁺ in NPCE cells transferred to coupled PCE cells while the reverse does not occur (Schutte & Wolosin, 1996; Suzuki et al. 1997). Further investigations are now required to identify the different cellular ion pathways responsive to agonist-induced increases in [Ca²⁺]_i in both PCE and NPCE cells. This information will further clarify the mechanisms contributing to vectoral fluid secretion in the ciliary body epithelium.

In the present study we have examined the effects of adrenergic stimulation on K⁺ currents in isolated rabbit PCE cells. Our results demonstrate the presence of a Ca²⁺-activated K⁺ current in rabbit PCE cells that is regulated by ₁-AR stimulation. Furthermore, we confirm that the signalling pathway linking ₁-AR to Ca²⁺-activated K⁺ (K_Ca) channels involves a PTX-insensitive G protein coupled to D-myo-inositol 1,4,5-trisphosphate (IP₃)-sensitive Ca²⁺ stores. Agonist activation of K⁺ channels in PCE cells would permit membrane repolarization and increase the driving force for salt secretion across the ciliary epithelium.

	METHODS

Top Abstract Introduction Methods Results Discussion References

Cell preparation and identification

Eyes were enucleated from 5- to 10-week-old pigmented rabbits (Reimens, Ontario, Canada) in accordance with the Association for Research in Vision and Opthalmology (ARVO) statement for the Use of Animals in Ophthalmic and Vision Research. Rabbits were killed with 0·3 ml kg^-1 of Euthanol (sodium pentobarbitone; MTC pharmaceuticals, Cambridge, Ontario, Canada) by intravenous injection into the marginal ear vein. Enucleated eyes were placed in sterile Dulbecco's phosphate-buffered saline (D-PBS; Gibco) and the ciliary body was freed from the iris by an incision between the iridal and ciliary processes. The epithelium was then dissected from the underlying stroma and treated for 30-40 min with D-PBS containing collagenase (1·5 mg ml^-1; Sigma) and pronase (1 mg ml^-1; Boehringer Manneheim). Epithelial pieces were triturated gently to yield single PCE and NPCE cells and small tissue explants. The cell suspension was centrifuged for 5 min and then washed with enzyme free D-PBS. Cells were seeded onto glass coverslips and incubated at 37°C in Dulbecco's modified Eagle's medium (DMEM) containing 5 or 10 % fetal calf serum and placed in an atmosphere of 5 % CO₂-95 % O₂. Cells were maintained in primary culture until use (up to a maximum of 14 days for PCE cells). PCE cells were distinguished from NPCE cells by the presence of pigment granules. NPCE cells tended to have limited survival under the culture conditions used and after 5 days ciliary epithelial cultures consisted primarily of PCE cells (Cilluffo et al. 1986).

Solutions and drugs

Cells attached to glass coverslips were placed in a shallow recording chamber and positioned on the stage of a Nikon inverted microscope. The chamber was superfused (1-2 ml min^-1) with standard low-Cl^- physiological solution composed of (mM): sodium aspartate, 100; NaCl, 30; KCl, 5; CaCl₂, 1; MgCl₂, 1; Na₂HCO₃, 10; Hepes, 10; and glucose 10. For low-Ca²⁺ solutions, extracellular Ca²⁺ was replaced by 0·2 mM CaCl₂ and 1·5 mM EGTA. The extracellular calcium concentration with this solution was estimated to be < 10 nM (software provided by A. French, Department of Physiology, Dalhousie University, Halifax, Nova Scotia, Canada). All external solutions were continuously bubbled with 5 % CO₂-95 % air and adjusted to pH 7·4 with NaOH. Drugs and ligands were applied by bath superfusion or by pneumatic pressure ejection from a micropipette. Test solutions were applied for a minimum of five and usually for ten complete (1 ml) bath exchanges. For application of test substances by pressure ejection, micropipettes ( 2 mm in diameter) were positioned 50-100 µm from the cell and 14-34 kPa (2-5 lb in^-2) pressure applied to the back of the micropipette using a Picospritzer II (General Valve Corp., Fairfield, NJ, USA). Standard electrode filling solution for whole-cell recordings was composed of (mM): potassium aspartate, 110; KCl, 30; MgCl₂, 1; Hepes, 20; EGTA, 1; CaCl₂, 0·4; Mg-ATP, 1; and Na₂-GTP, 0·1; adjusted to pH 7·3 with KOH. Free internal [Ca²⁺] was estimated to be < 100 nM. In some experiments intracellular Ca²⁺ was buffered to < 10 nM by inclusion of 10 mM BAPTA in the electrode solution. Solution osmolarity was measured by freezing point depression (Osmette A, Fisher Scientific, Nepean, Ontario, Canada). The osmolarity of external solutions was 330 mosmol l^-1 and the osmolarity of internal solutions was 320 mosmol l^-1. Experiments were conducted at room temperature (21-22°C).

The potassium channel blockers (tetraethylammonium chloride (TEA) and iberiotoxin (IbTX)), thapsigargin (TG), the phospholipase C (PLC) inhibitor U-73122 and the ₁-adrenoceptor antagonist prazosin (PZ) were added to the extracellular solution and superfused during electrophysiological recording. The adrenergic agents adrenaline and phenylephrine (PHe) and the Ca²⁺ ionophore ionomycin were applied by pneumatic pressure injection. D-myo-Inositol 1,4,5-trisphosphate (IP₃) and guanosine 5'-O-(2-thiodiphosphate) (GDPS) were included in the intracellular recording solution. Pertussis toxin (PTX) was added to the culture medium 24 h prior to electrophysiological recording. Thapsigargin, IP₃, U-73122 and PTX were purchased from Calbiochem. Iberiotoxin was purchased from Peninsula Laboratories Inc. (Belmont, CA, USA). All other chemicals were purchased from Sigma.

Electrophysiological recording techniques

Ionic currents in isolated epithelial cells were studied using whole-cell, tight-seal, patch-clamp recording methods (Hamill et al. 1984). The recording conditions used have been described previously (Tao et al. 1994; Poyer et al. 1996). Patch electrodes were pulled from borosilicate glass with an external diameter of 1·5 mm and an internal diameter of 1·1 mm (Sutter Instrument Co., Novato, CA, USA) using a two stage vertical microelectrode puller (Narishige model PP83, Tokyo). Electrodes had resistances of 3-5 M when filled with internal solution and were fire-polished and coated with beeswax to reduce capacitance. The reference electrode used was a sealed electrode-salt bridge combination (Dri-ref2; World Precision Instruments, Sarasota, FL, USA). Offset potentials were nulled using the amplifier circuitry before seals were made on the cells. Liquid junction potentials (LJPs) arising between the bath and the electrode were measured experimentally and defined as the potential of the bath solution with respect to the pipette solution (Barry & Lynch, 1991). For whole-cell recording, the membrane potential of the cell, V_m, was calculated as V_m = Vp - LJP (where V_p is pipette potential). To confirm experimentally generated measurements, LJPs were also calculated using a software program (JPCalc, version 2.00; P. H. Barry, Sydney, Australia). All the data and current-voltage relationships shown have been corrected for LJPs, which were 2 mV for standard low-Cl^- external Ringer solutions and potassium aspartate pipette solution and 5·4 mV for low-Cl^-/low-Ca²⁺ external and BAPTA intracellular solutions.

Membrane potential and ionic currents were recorded with an Axopatch-1D amplifier (Axon Instruments). Currents were filtered with a 4-pole low-pass Bessel filter and digitized at a sampling frequency of 5-10 kHz using pCLAMP software (Axon Instruments). Current and voltage were displayed on a Kikuzui 5040 oscilloscope and on a Gould TA240 chart recorder and stored on computer disc. Cell capacitance values were obtained from the capacitance compensation circuitry on the amplifier and ranged between 10 and 48 pF (26 ± 0·8; n = 108). Measures of series resistance were obtained from the amplifier and were always less than 15 M. Using the amplifier circuitry, 80 % series resistance compensation was usually employed.

Data are represented as means ± S.E.M. and were analysed using Student's unpaired t test unless otherwise stated and were considered significant at P < 0·05.

	RESULTS

Top Abstract Introduction Methods Results Discussion References

Whole-cell currents in PCE cells

Figure 1 shows typical whole-cell current recordings made from cultured PCE cells with 110 mM potassium aspartate electrode solution and standard low-Cl^- Ringer solution. The mean membrane potential measured in current clamp upon break-in was -42 ± 1 mV (n = 33). Under voltage clamp cells were held at a holding potential (V_h) of -62 mV and the membrane potential was stepped in 20 mV increments to potentials negative and positive to the holding potential. At potentials positive to -62 mV, an outward K⁺ current was observed in all PCE cells recorded from (Fig. 1A). Figure 1B shows the I-V relationship for the cell shown in Fig. 1A. For this and subsequent current-voltage relations current was obtained by averaging the current amplitude measured over 50 ms prior to the cessation of depolarizing step commands. The outward K⁺ current activates at potentials positive to -62 mV and increases with increasing depolarization. In 31 % of PCE cells recorded from, voltage steps to hyperpolarizing potentials between -130 and -80 mV also evoked an inward K⁺ current. In the representative cell shown in Fig. 1C, the hyperpolarization-activated inward current showed some time-dependent decay at the more negative potential of -130 mV. A slowly inactivating outward K⁺ current is also apparent in this cell as the membrane potential was depolarized positive to -50 mV. The I-V relationship for the currents shown in Fig. 1C (Fig. 1D) indicates an inwardly rectifying current which reverses at around -62 mV and an outward rectifying K⁺ current which activates positive to -50 mV and increases with depolarization. These currents are similar to K⁺ currents described previously in both rabbit and bovine PCE cells (Jacob, 1988; Fain & Farahbakhsh, 1989). The ionic selectivity of the outward current was investigated by examining time- and voltage-dependent relaxations (tail currents) at potentials between -82 and 0 mV, following activation of outward current by a 100 ms voltage step to +20 mV. Analysis of the tail currents in 5 mM [K⁺]_o indicated that these currents reversed direction at -76 ± 4 mV (n = 5), which approaches the value of -84 mV calculated for the K⁺ equilibrium (EK) under our recording conditions (data not shown).

		View larger version [in this window] [in a new window]
Figure 1. Whole-cell K+ currents in cultured PCE cells Representative whole-cell currents elicited by a series of step depolarizations from -122 to +58 mV from a holding potential of -62 mV. Depolarization-induced outwardly rectifying K+ current (A) was observed in all PCE cells examined, while 31 % displayed both inward and outward rectifying K+ current (C). Dashed line represents zero current in this and subsequent figures. B and D, current-voltage relations measured from the traces shown in A and C, respectively.

In 66 % (58/83) of the cells recorded from, the outward K⁺ current appeared 'noisy' at more depolarized potentials (+10 to +58 mV) suggesting the presence of large-conductance channels (Tao & Kelly, 1996). A representative cell is shown in Fig. 2A. The voltage protocol used is indicated above the current traces, with the cell held at a Vh of -62 mV and stepped for 500 ms from -82 to +58 mV. A 2 min superfusion of the K⁺ channel blocker TEA (1 mM) reversibly reduced outward current. Figure 2B shows the I-V plot for the currents represented in Fig. 2A. In this cell, outward K⁺ current is almost completely blocked by 1 mM TEA with a 68 % (65 ± 10 %; n = 3) reduction in I_K measured at +58 mV and a 65 ± 18 % (n = 3) reduction at +18 mV. The TEA-sensitive current, obtained by digitally subtracting currents measured in the presence of TEA from control currents, indicates an outwardly rectifying K⁺ current which activates positive to -50 mV.

		View larger version [in this window] [in a new window]
Figure 2. Ca2+-activated K+ current PCE cells Whole-cell current traces recorded from two representative PCE cells and elicited by voltage steps from -82 to +58 mV from a holding potential of -62 mV. Outward current was reduced in the presence of 1 mM TEA (A) and 10 nM of the maxi-K+ KCa channel blocker IbTX (C). B and D, current-voltage relations measured from the currents recorded in A and C, respectively, in the absence () and presence () of TEA (B) or IbTX (D). The TEA- or IbTX-sensitive current () was obtained by subtracting current records measured in the presence of TEA or IbTX, respectively, from control currents.

A similar decrease in the outward current was also observed when the scorpion toxin IbTX was externally applied. IbTX has been demonstrated to block selectively the large-conductance or maxi-K_Ca channels in a number of cell types (Giangiacomo et al. 1992, 1993) including retinal pigment epithelial cells (Tao & Kelly, 1996). Figure 2C and D shows the whole-cell current recording and I-V plot for a representative cell before and after exposure to 10 nM IbTX for 5 min. IbTX reduced I_K measured at +58 mV by 71 % (69 ± 11 %, n = 4) and at +18 mV by 69 ± 5 % (n = 4). The IbTX-sensitive K⁺ current (Fig. 2D) in this cell, as above, was outwardly rectifying with a steep increase in slope conductance at potentials depolarized to +20 mV. Superfusion of PCE cells for 15 min in nominally Ca²⁺-free external solution also decreased the outward current (69 ± 17 % at +18 mV and 76 ± 7 % at +58 mV; n = 4, data not shown). The TEA, IbTX and Ca²⁺ sensitivity of the outward current further suggests the presence of maxi-K_Ca channels in rabbit PCE cells.

Further confirmation of the presence of K_Ca channels in PCE cells was obtained using the calcium ionophore ionomycin to increase [Ca²⁺]_i. Figure 3A, shows an I-V plot from a representative cell with 110 mM potassium aspartate internal solution and standard low-Cl^- Ringer solution before and after exposure to 10 µM ionomycin. Puffer application of ionomycin increased the outward current. The ionomycin-sensitive current was outwardly rectifying and activated at potentials positive to -60 mV. In the cell shown, puffer application of ionomycin increased outward K⁺ current by 69 % at +58 mV. In nine additional cells recorded from, ionomycin significantly increased the outward current measured at +58 mV from 15 ± 2 to 35 ± 5 pA pF^-1 (P < 0·05; n = 9) (Fig. 3B). The ionomycin-induced increase in the outward current was blocked by pretreatment of the cells for 10 min with 10 nM IbTX (n = 4).

		View larger version [in this window] [in a new window]
Figure 3. Activation of IK(Ca) by ionomycin in PCE cells A, current-voltage plot elicited by voltage steps from -82 to +58 mV from a holding potential of -62 mV in a representative PCE cell before () and after a 40 s pressure application of the Ca2+ ionophore ionomycin (10 µM; ). The ionomycin-sensitive current () was obtained by subtracting current records measured in the presence of ionomycin from control currents. B, mean + S.E.M. current measured at +58 mV for control current (n = 13) and current recorded following ionomycin application in the absence (10 µM Ion; n = 9) and presence of IbTX (Ion + 10 nM IbTX; n = 4). All currents were normalized for cell capacitance. * P < 0·05 relative to control.

Adrenergic agonists activate I_K(Ca) in PCE cells

Calcium-activated channels in other cell types are targets for modulation by a number of transmitters and neuromodulators. Since ₁-adrenoreceptors have been demonstrated to increase [Ca²⁺]_i in rabbit PCE cells (Schutte et al. 1996; Suzuki et al. 1997), we examined the actions of the adrenergic agonist adrenaline on whole-cell K⁺ currents in PCE cells. Figure 4A shows whole-cell currents recorded by stepping from -102 to +58 mV in 20 mV increments for 500 ms from Vh = -62 mV, in standard extracellular solution in the absence (control) and presence of adrenaline (100 µM). In this cell, puffer application of 100 µM adrenaline substantially increased outward K⁺ current activated at depolarized potentials. In addition, the I-V plot for the currents shown in Fig. 4A (Fig. 4B) demonstrates that adrenaline shifted the activation potential for the outward current from -47 to -57 mV (-33 ± 5 to -52 ± 5 mV; n = 11) and increased whole-cell current at -52 and -42 mV by 8 and 13 pA pF^-1, respectively, compared with control (14 ± 6 and 21 ± 9 pA pF^-1; n = 11). Figure 4C demonstrates that adrenaline substantially increased the current measured at the more depolarized potential of +58 mV with current increasing from 23 ± 6 to 50 ± 11 pA pF^-1 (117 % increase, P < 0·05; Student's paired t test). In another four cells tested application of adrenaline produced a 280 ± 96 % increase (P < 0·05) in I_K measured at +58 mV from V_h = -62 mV. In these cells, a second application (5 min after the first application of adrenaline) of 100 µM adrenaline in the presence of 10 nM IbTX failed to increase I_K (12 ± 7 %) in comparison with the first adrenaline application in the absence of toxin (P > 0·05), confirming that the current activated by adrenaline was an IbTX-sensitive I_K(Ca) (Fig. 4D).

		View larger version [in this window] [in a new window]
Figure 4. Activation of IK(Ca) by adrenaline A, whole-cell currents elicited from voltage steps from -102 to +58 mV from a holding potential of -62 mV, recorded in the absence (Control) and presence of 100 µM adrenaline (Adr). B, current-voltage relation for the cell in A in the absence () and presence of adrenaline (). C, mean + S.E.M. current amplitude measured at +58 mV and normalized for cell capacitance, in 16 cells before (control) and after 40 s pressure application of 100 µM adrenaline (Adr). * P < 0·05 relative to control. D, mean percentage increase in IK measured at +58 mV after exposure of 4 cells to 100 µM adrenaline in the absence (Adr) and presence of IbTX (Adr + 10 nM IbTX). * P < 0·05 relative to Adr.

Activation of I_K(Ca) by adrenergic agonists was via an ₁-receptor mediated pathway. Figure 5A shows current records recorded before and after application of the selective ₁-adrenoreceptor agonist phenylephrine (PHe) in a representative PCE cell with 110 mM potassium aspartate electrode solution and standard low-Cl^- Ringer solution. Exposure of the cell to PHe (100 µM) increased depolarization-activated outward current. Figure 5B shows the I-V plots for mean current values measured in ten cells before (control) and after exposure to 100 µM PHe. In these cells, outwardly rectifying current activated positive to -60 mV and was increased at all potentials positive to -20 mV in the presence of 100 µM PHe. The increase in outward current by PHe was significant (P < 0·05) compared with control for currents measured at +38 and +58 mV. In a total of fifty-five cells tested, PHe at a doses of 10 and 100 µM increased the outward current measured at +58 mV (Fig. 5C). In the presence of 10 µM PHe, outward K⁺ current increased from 20 ± 4 to 30 ± 10 pA pF^-1 (n = 5), whilst in the presence of 100 µM PHe, the increase in I_K was from 20 ± 2 to 41 ± 4 pA pF^-1 (P < 0·05). The increase in I_K was reproducible following a repeat exposure of the cells to PHe (from 24 ± 11 to 36 ± 17 pA pF^-1; P < 0·05, n = 4, data not shown).

		View larger version [in this window] [in a new window]
Figure 5. Activation of IK(Ca) by 1-adrenoceptors A, whole-cell currents recorded from a representative PCE cell. Currents were measured from a holding potential of -62 mV in response to voltage steps from -82 mV to +58 mV in the absence (control) and presence of 100 µM of the selective 1-agonist phenylephrine (PHe). B, current-voltage relation for mean ± S.E.M. current amplitude measured in 10 representative PCE cells before () and after a 40 s exposure to 100 µM PHe (). Whole-cell currents were elicited and by voltage steps from -122 to +58 mV from a holding potential of -62 mV and have been normalized for cell capacitance. C, mean + S.E.M. current amplitude measured at +58 mV before (Control; n = 5) and after 40 s pressure application of 10 µM (n = 5) or 100 µM PHe (n = 50). D, bar graph shows mean + S.E.M. percentage current measured at +58 mV for the same 5 cells before and after 40 s application of 100 µM PHe in the absence (Control; Phe) or presence of 100 µM prazosin (PZ; PZ + PHe). * P < 0·05 relative to control.

Further verification that PHe actions were mediated via an ₁-receptor were obtained when the selective ₁-AR antagonist prazosin (PZ) was added to the superfusate (Fig. 5D). In five cells tested, 100 µM PHe significantly increased outward current measured at the end of a 500 ms voltage step to +58 mV (V_h = -62 mV) from 15 ± 4 to 26 ± 5 pA pF^-1 (P < 0·05; Student's paired t test). In the presence of 100 µM of PZ, PHe failed to produce a significant increase in outward current (P > 0·05; Student's paired t test). Control current recorded at +58 mV in presence of PZ, in the same five cells previously exposed to an initial application of PHe, was 22 ± 6 pA pF^-1 and this value only increased slightly to 25 ± 6 pA pF^-1 following exposure to a second 40 s application of PHe in the presence of PZ. Thus the actions of PHe in activating I_K(Ca) are mediated via activation of an ₁-AR coupled pathway.

Signalling pathways involved in ₁-adrenoreceptor activation of I_K(Ca)

In many cells, the actions of ₁-AR agonists are associated with a rise in [Ca²⁺]_i as a result of agonist/receptor activation of phospholipase C-mediated hydrolysis of phosphatidylinositol 4,5-bisphosphate and liberation of IP₃. Inositol phosphate then mobilizes Ca²⁺ from intracellular storage sites by interacting with a specific receptor (Berridge, 1993; Minneman, 1993). We examined whether the IP₃ pathway was involved in PHe-induced activation of I_K(Ca) in PCE cells by either dialysing the cell cytosol with IP₃ to activate I_K(Ca), or by exposing PCE cells to thapsigargin and nominally zero-Ca²⁺ Ringer solution and looking for attenuation of the actions of PHe. Thapsigargin mobilizes Ca²⁺ from intracellular stores by inhibiting Ca²⁺ sequestration into IP₃-sensitive pools (Foskett et al. 1991).

Whole-cell current traces are shown from a representative PCE cell immediately after break-in and after 10 min dialysis with 10 µM IP₃ in the standard pipette solution (Fig. 6A). The cell was held at -62 mV and stepped from -122 to +58 mV in 20 mV increments for 500 ms to activate outward current. The mean outward current measured at +58 mV was significantly increased from 19 ± 7 to 51 ± 15 pA pF^-1 (P < 0·05 Student's paired t test; n = 11) after 10 min internal dialysis with IP₃. Current amplitude was not significantly different when cells were dialysed for 10 min with internal recording solution in the absence of IP₃ (data not shown). Pre-incubation of PCE cells with 5 µM thapsigargin (TG) for 30 min in nominally Ca²⁺-free Ringer solution to empty endoplasmic reticular Ca²⁺ stores, blocked the PHe-mediated increase in IK(Ca) (n = 5; Fig. 6B). In contrast, in PCE cells from the same culture which were not treated with thapsigargin, PHe significantly increased I_K(Ca) from 20 ± 8 to 53 ± 17 pA pF^-1 (P < 0·05; n = 5). These data support a role for IP₃-induced intracellular calcium release in mediating the effects of PHe in activating I_K(Ca).

		View larger version [in this window] [in a new window]
Figure 6. Involvement of IP3 and intracellular Ca2+ release in IK(Ca) activation A, representative whole-cell currents produced from voltage steps between -122 and +58 mV from a holding potential of -62 mV immediately after obtaining whole-cell configuration (break-in) and after 10 min dialysis with 10 µM IP3 in the pipette. B, mean + S.E.M. current amplitude measured at +58 mV and normalized for cell capacitance in 5 PCE cells treated for 30 min with 5 µM thapsigargin before (TG) and after 40 s application of 100 µM PHe (TG + PHe), and in 5 cells in the absence of thapsigargin before (Control) and after PHe application (PHe). * P < 0·05 relative to control.

To verify further the involvement of an IP₃ pathway in the ₁-AR increase in I_K(Ca) we exposed cells to the PLC enzyme inhibitor U-73122 (Hildebrandt et al. 1997). Figure 7A shows a PHe-induced increase in I_K(Ca) in a representative PCE cell before and after 20-30 min of superfusion with 10 µM of the PLC inhibitor U-73122. The cell was stepped from V_h = -62 mV to +58 mV for 500 ms. In this cell and three other cells tested, PHe-mediated increases in I_K(Ca) were effectively blocked following exposure of the cells to U-73122 compared with control (P > 0·05; Student's paired t test) (Fig. 7B). Thus, these results together with the findings on the effects of internal dialysis of IP₃ and pre-incubation with thapsigargin/zero-Ca²⁺ on PCE cells, confirm the involvement of an IP₃ signalling pathway in mediating the PHe-induced increases in I_K(Ca).

		View larger version [in this window] [in a new window]
Figure 7. Inhibition of PLC inhibits PHe-induced increases in IK(Ca) A, representative whole-cell currents from a single PCE cell in response to a 500 ms step depolarization to +58 mV from a holding potential of -62 mV. Current is shown before and after pressure application of 100 µM PHe in the same cell prior to incubation with U-73122 (Control; PHe) and after 20-30 min incubation with 10 µM U-73122 (U-73122; + PHe). B, PHe-mediated increase in the mean + S.E.M. current amplitude normalized for cell capacitance, measured at +58 mV in 4 cells before (Control; PHe) and after incubation with U-73122 (U-73122; U-73122 + PHe) for 20-30 min. * P < 0·05 relative to control.

The role of G proteins in I_K(Ca) activation by PHe

In many cell types ₁-ARs are coupled to PTX-insensitive heterotrimeric G proteins of the G_q class (Summers & McMartin, 1993). To confirm G protein involvement in mediating PHe increases in I_K(Ca), the standard 100 µM GTP in the pipette was replaced with 2 mM of the G protein blocker GDPS. Phenylephrine significantly increased I_K(Ca) measured at +58 mV from 32 ± 11 to 98 ± 21 pA pF^-1 (n = 5) with GTP in the pipette (Fig. 8A). However, in another four cells tested on the same day from the same culture after 10 min dialysis with 2 mM GDPS in the pipette, 100 µM PHe failed to increase significantly outward current measured at +58 mV over the control value (24 ± 4 vs. 30 ± 5 pA pF^-1; P > 0·05). This result confirms G protein involvement in the PHe-induced activation of I_K(Ca) in PCE cells.

To determine whether the ₁-AR modulation of I_K(Ca) involved a PTX-insensitive G protein, we then pretreated PCE cells for 12 h with 500 ng ml^-1 of PTX, a dose previously shown to be effective in blocking PTX-sensitive G protein action in isolated ocular epithelial cells (Poyer et al. 1996). The toxin inactivates G proteins of the G_i/G_o/G_z class by catalysing the ADP ribosylation of the -subunit (Ui, 1990). Whole-cell recordings were made from control cells (not exposed to PTX) as well as from PTX-treated cells in standard intracellular and extracellular solutions. All cells were from the same culture and were plated at the same time prior to PTX treatment. Figure 8B shows that in four control cells without PTX pretreatment, PHe (100 µM) significantly increased I_K measured at +58 mV (V_h = -62 mV) from 10 ± 3 to 27 ± 3 pA pF^-1 (Student's paired t test; P < 0·05) following a 40 s exposure to the agonist. In another four cells pretreated with PTX, a similar increase in outward current was observed following PHe application with current increasing from 20 ± 8 to 48 ± 5 pA pF^-1 (Student's paired t test; P < 0·05). These data suggest that the G protein involved was not PTX sensitive and therefore was not of the G_i/G_o/G_z class.

		View larger version [in this window] [in a new window]
Figure 8. G proteins are involved in PHe-mediated IK(CA) increase A, bar shows mean + S.E.M. current amplitude measured at +58 mV with 0·1 mM GTP or 2 mM GDPS in the pipette before (GTP; GDPS) or after 40 s application of 100 µM PHe (GTP + PHe; GDPS + PHe). B, mean + S.E.M. current amplitude at +58 mV in 4 untreated cells before (Control) and after PHe application (+PHe) and in 4 cells pretreated with 500 ng ml-1 of PTX before (PTX-treated) and after PHe application (PTX + PHe). Data have been normalized for cell capacitance. * P < 0·05 relative to control.

	DISCUSSION

Top Abstract Introduction Methods Results Discussion References

In this study we have identified a calcium-activated potassium current in isolated rabbit pigmented ciliary epithelial (PCE) cells that is activated by adrenergic agonists. Furthermore, we have demonstrated that I_K(Ca) activation is mediated by ₁-adrenoceptors coupled to a PTX-insensitive G protein(s) and a PLC/IP₃ signalling pathway. To our knowledge this is the first report of ₁-adrenergic modulation of a K⁺ current in mammalian ciliary epithelial cells.

Previous studies in isolated PCE cells have identified a number of voltage dependent ionic currents. TTX-sensitive Na⁺ current (Fain & Farahbakhsh, 1989), T- and L-type Ca²⁺ current (Jacob, 1991a; Farahbakhsh et al. 1994), Cl^- currents, a non-selective cation current (Jacob & Civan, 1996) and both outwardly and inwardly rectifying K⁺ currents (Fain & Farahbaksh, 1989; Jacob, 1991b; Stelling & Jacob, 1991) have all been reported. Under our recording conditions we observed similar currents in isolated rabbit PCE cells. All cells exhibited outwardly rectifying K⁺ current and 33 % had inwardly rectifying K⁺ current. Consistent with observations in bovine PCE cells (Jacob, 1991b), we also found that a component of the outward current in rabbit PCE cells was sensitive to Ca²⁺.

To date, all the K_Ca channels identified are dependent upon Ca²⁺ for their activation. These channels have been subcategorized as big (BK or maxi-K⁺) channels, intermediate (I_K) channels or small (SK) channels on the basis of their single channel conductance, and their sensitivity to voltage and specific blockers (reviewed by McManus, 1991). We identified the I_K(Ca) current that we observed in rabbit PCE cells as arising from activation of a BK or maxi-K⁺ type of channel based on the sensitivity of this current to voltage, external Ca²⁺ concentration and blockade by the selective maxi-K⁺ channel blocker IbTX (Giangiacomo et al. 1992, 1993). In cultured rabbit PCE cells between 1 and 7 days in culture, 36-87 % of the depolarization activated outward current was identified as IbTX-sensitive I_K(Ca). The large variability in the amount of I_K(Ca) between cells may reflect heterogeneity in K⁺ channel expression (Fain & Farahbakahsh, 1989; Stelling & Jacob, 1991) or Ca²⁺ channel expression (Jacob, 1991a) as has been reported for mammalian ciliary epithelial cells and also for retinal pigment epithelial cells (Tao et al. 1994; Tao & Kelly, 1996).

The CBE has been shown to be a site for modulation by a variety of agonists, including adrenergic agonists. Both and agonists have been demonstrated to modulate aqueous humour dynamics including the formation, secretion and outflow of aqueous humour (Ross & Drance, 1970; Chiou, 1983). Studies in the isolated rabbit iris CBE have also demonstrated that both - and -AR agonists alter ionic movement across the epithelium, as measured by a decrease in CBE short circuit current (Isc) in the presence of these agonists (Krupin et al. 1991; Shi et al. 1996). More recently, increases in intracellular Ca²⁺ levels have been demonstrated in both PCE and NPE cells of isolated rabbit CBE in response to -AR agonists (Schutte & Wolosin, 1996; Shutte et al. 1996; Suzuki et al. 1997). This Ca²⁺ mobilization showed refractory behaviour with responses transferred only from NPE to PCE cells, not the reverse.

In isolated PCE cells we have shown that the non-selective adrenergic agonist adrenaline and the selective ₁-AR agonist phenylephrine increase outward current. The adrenergic stimulated outward current was identified as I_K(Ca) based on the observations that it was mimicked by the Ca²⁺ ionophore ionomycin and blocked by the maxi-K⁺ channel blocker IbTX. Our finding that the ₁-AR antagonist prazosin blocks PHe-mediated increases in I_K(Ca) further supports the involvement of an ₁-AR subtype in mediating this response.

Radioligand binding studies have identified and localized a number of the adrenergic receptor subtypes present in CBE (Mittag & Tormay, 1985; Wax & Molinoff, 1987; Jin et al. 1994). Adrenergic receptors comprise three major subfamilies, ₁, ₂ and (Reviewed by Summers & McMartin, 1993). A number of closely related subtypes of the ₁ subfamily (_1A, _1B, _1C and _1D) and the ₂ subfamily (_2A/D, _2B and _2C) have been identified on the basis of their sequence homology, pharmacological profile and signal transduction mechanisms (Watson & Arkinstall, 1994; MacKinnon et al. 1994). Both subfamilies of -AR have been identified on the rabbit CBE (Mallorga et al. 1988; Jin et al. 1994) and more recent studies have suggested that in rabbit CBE ₁-ARs are localized on the PCE cells, whilst ₂-ARs, specifically _2A-ARs, are localized to the NPE cells (Schutte et al. 1996; Suzuki et al. 1997). Our studies are consistent with these findings in that the phenylephrine-mediated increase in I_K(Ca) is mediated via an ₁-receptor subtype in rabbit PCE cells.

In most cell types, ₁-ARs couple to a PTX-insensitive G_q protein leading to the activation of PLC and the initiation of the IP₃/DAG cascade (Summers & McMartin, 1993). This signalling pathway culminates in rises in intracellular Ca²⁺ which are thought to occur as a result of IP₃-mediated release of Ca²⁺ from internal stores and/or the influx of extracellular Ca²⁺ via voltage dependent and non-voltage dependent Ca²⁺ channels (Minneman, 1988; Summers & McMartin, 1993). In contrast, ₂-ARs are thought to predominately couple to PTX-sensitive G_i/G_o proteins and inhibition of adenylate cyclase (MacKinnon et al. 1994).

Our findings that increases in I_K(Ca) could be mimicked by internal dialysis with synthetic IP₃ and blocked either when PLC was inhibited or IP₃-sensitive intracellular Ca²⁺ stores were depleted with thapsigargin confirmed involvement of the IP₃ signalling pathway in mediating phenylephrine's effect on I_K(Ca) in rabbit PCE cells. These findings combined with our demonstration that phenylephrine couples to a PTX-insensitive G protein demonstrate that in rabbit PCE cells ₁-adrenoceptors acting via a PLC/IP₃ pathway lead to the activation of I_K(Ca).

Studies in bovine PCE cells have shown that I_K(Ca) is activated near the resting membrane potential of these cells via influx of Ca²⁺ through T-type calcium channels (Jacob, 1991a). In contrast, activation of voltage-dependent K⁺ current (I_K(V)) occurred at potentials depolarized by about 40 mV from rest and thus outside the membrane potential range that these non-excitable cells normally enter. This led to the suggestion that under normal physiological conditions, activation of I_K(Ca) in PCE cells may play a more important role in contributing to ion transport across the CBE than I_K(V) (Jacob, 1991b). Our findings also indicate that I_K(Ca) may be activated at or near the resting membrane potential in rabbit PCE cells via release of receptor-coupled Ca²⁺ stores. In addition, the negative shift in the activation potential for the outward K⁺ current in PCE cells in the presence of adrenaline is consistent with data reported on maxi-I_K(Ca) currents in rabbit retinal pigment epithelial cells (Tao & Kelly, 1995). These data indicate that elevations in [Ca²⁺]_i increase the probability that I_K(Ca) channels open at membrane potentials close to the physiological resting potential of these cells. Once activated, I_K(Ca) would lead to K⁺ exit and PCE cell membrane hyperpolarization. Potential changes in PCE cells are relayed to NPE cells via electrically coupled gap junctions between the cells. Thus, activation of I_K(Ca) in PCE cells would then lead to NPE membrane hyperpolarization and, via alterations in transepithelial potential, provide an increased driving force for Cl^- secretion into the aqueous humour via anion channels on coupled NPE cells (Jacob, 1991b; Jacob & Civan, 1996).

In conclusion, we have demonstrated that I_K(Ca) in rabbit PCE cells can be activated by ₁-ARs coupled to release of Ca²⁺ from IP₃-sensitive Ca²⁺ stores. In vivo , transmitters and hormones could modulate I_K(Ca) activity providing a mechanism for both paracrine and autocrine regulation of aqueous humour secretion across the ciliary epithelium.

	REFERENCES

Top Abstract Introduction Methods Results Discussion References

Watson, S. & Arkinstall, S. (1994). The G-Protein Linked Receptor Facts Book, pp 32-54. Academic Press, San Diego, CA, USA.

Barry, P. & Lynch, J. (1991). Liquid junction potentials and small cell effects in patch-clamp analysis. Journal of Membrane Biology 121, 107-117.
Berridge, M. J. (1993). Inositol trisphosphate and calcium signalling. Nature 361, 315-325	[Medline]
Chiou, G. C. Y. (1983). Effects of 1 and 2 activation of adrenergic receptors on aqueous humor dynamics. Life Science 32, 1699-1704.
Cilluffo, M. C., Fain, M. J., Fain, G. L. & Brecha, N. C. (1986). Culture of rabbit ciliary body epithelium. Investigative Ophthalmology and VisualScience 27, suppl., 322.
Cole, D. A. (1984). Ocular fluids. In The Eye, vol 1A. 3rd edn, ed. Davson, H., pp. 269-390. Academic Press, New York.
Edelman, J. L., Sachs, G. & Adorante, J. S. (1994). Ion transport asymmetry and functional coupling in bovine pigmented and nonpigmented ciliary epithelial cells. American Journal of Physiology 266, C1210-1221	[Medline]
Fain, G. L., & Farahbakhsh, N. A. (1989). Voltage-activated currents recorded from rabbit pigmented ciliary body epithelial cells in culture. The Journal of Physiology 417, 83-103.
Farahbaksh, N. A. & Cilluffo, M. C. (1994). Synergistic effect of adrenergic and muscarinic receptor activation on [Ca2+]i in rabbit ciliary body epithelium. The Journal of Physiology 77, 215-221.
Farahbakhsh, N. A., Cilluffo, M. C., Chronis, C. & Fain, G. L. (1994). Dihydropyridine-sensitive Ca2+ spikes and Ca2+ currents in rabbit ciliary body epithelial cells. Experimental Eye Research 58, 197-206	[Medline]
Foskett, J. K., Roifman, C. M. & Wong, D. (1991). Activation of calcium oscillations by thapsigargin in parotid acinar cells. Journal of Biological Chemistry 266, 2778-2782	[Abstract]
Giangiacomo, K. M., Garcia, M. L. & McManus, O. B. (1992). Mechanism of iberiotoxin block of the large-conductance calcium-activated potassium channel from bovine aortic smooth muscle. Biochemistry 31, 6719-6727	[Medline]
Giangicomo, K. M., Sugg, E. E., Garcia-Calco, M., Leonard, R. J., McManus, O. B., Kaczorowski, G. I. & Garcia, M. L. (1993). Synthetic charybdotoxin-iberiotoxin chimeric peptides define toxin binding sites in calcium-activated and voltage dependent potassium channels. Biochemistry 32, 2363-2370.	[Medline]
Hamill, O. P., Marty, A., Neher, E., Sakmann, B. & Sigworth, F. J. (1981). Improved patch-clamp techniques for high resolution current recording from cell and cell-free membrane patches. Pflügers Archiv 391, 85-100	[Medline]
Hildebrandt, J. P., Plant, T. D. & Meves, H. (1997). The effects of bradykinin on K+ currents in NG108-15 cells treated with U73122, a phospholipase C inhibitor, or neomycin. British Journal of Pharmacology 120, 841-850	[Medline]
Jacob, T. J. C. (1988). Inward and outward currents in isolated epithelial cells of the bovine ciliary body. The Journal of Physiology 407, 103P.
Jacob, T. J. C. (1991a). Identification of a low-threshold T-type calcium channel in bovine ciliary epithelial cells. American Journal of Physiology 261, C808-813	[Medline]
Jacob, T. J. C. (1991b). Two outward K+ currents in bovine pigmented ciliary epithelial cell: IK(Ca) and IK(V). American Journal of Physiology 262, C1055-1062.
Jacob, T. J. C. & Civan, M. M. (1996). Role of Ion channels in Aqueous humor formation. American Journal of Physiology 271, C703-720	[Medline]
Jin, Y., Verstappen, A. & Yorio, T. (1994). Characterization of alpha 2-adrenoceptor binding sites in rabbit ciliary body membranes. Investigative Ophthalmology and Visual Science 35, 2500-2508	[Abstract]
Krupin, T. E., Wax, M. B., Carre, D. A., Moolchandani, J. & Civan, M. M. (1991). Effects of adrenergic agents on transepithelial electrical measurements across the isolated iris-ciliary body. Experimental Eye Research 53, 709-716	[Medline]
MacKinnon, A. C., Spedding, M. & Brown, C. M. (1994). 2-Adrenoceptors: more subtypes but fewer functional differences. Trends in Pharmacological Sciences 15, 119-123.	[Medline]
McManus, O. B. (1991). Calcium-activated potassium channels: Regulation by calcium. Journal of Bioenergetics and Biomembranes 23, 537-559	[Medline]
Mallorga, P., Buisson, S. & Sugrue, M. F. (1988). Alpha 1-adrenoceptors in the albino rabbit ciliary process. Journal of Ocular Pharmacology 4, 203-214.	[Medline]
Minneman, K. P. (1993). 1-Adrenergic receptor subtypes, inositol phosphates and sources of cell Ca2+. Pharmacological Reviews 40, 87-115.
Mittag, T. W. & Tormay, A. (1985). Adrenergic receptor subtypes in rabbit iris-ciliary body membranes: classification by radioligand studies. Experimental Eye Research 40, 239-24	[Medline]
Poyer, J. F., Ryan, J. S. & Kelly, M. E. M. (1996). G Protein activation of a nonspecific cation current in cultured rat retinal pigment epithelial cells. Journal of Membrane Biology 153, 13-26	[Medline]
Raviola, G. & Raviola, E. (1978). Intercellular junctions in the ciliary epithelium. Investigative Ophthalmology and Visual Science 17, 958-981	[Abstract]
Ross, R. A. & Drance, S. M. (1970). Effects of topically applied isoproterenol on aqueous dynamics in man. Archives of Ophthalmology 83, 39-46.	[Medline]
Schutte, M., Diadori, A., Wang, C. & Wolosin, J. M. (1996). Comparative adreno-cholinergic Ca2+ pharmacology in the dissociated rabbit ciliary epithelial cell layers. Investigative Ophthalmology and Visual Science 37, 212-220	[Abstract]
Schutte, M. & Wolosin, J. M. (1996). Ca2+ mobilization and interlayer signal transfer in the heterocellular bilayered epithelium of the rabbit ciliary body. The Journal of Physiology 496, 25-37	[Abstract]
Shi, X.-P., Zamudio, A. C., Candida, O. A. & Wolosin, J. M. (1996). Adreno-cholinergic modulation of junctional communication between the pigmented and nonpigmented layers of the ciliary body epithelium. Investigative Ophthalmology and Visual Science 37, 1037-1046	[Abstract]
Stelling, J. W. & Jacob, T. J. C. (1991). The inward rectifier K+ current underlies oscillatory membrane potential behaviour in bovine pigmented ciliary epithelial cells. The Journal of Physiology 458, 439-456.	[Abstract]
Summers, R. J. & McMartin, L. R. (1993). Adrenoceptors and their second messenger systems. Journal of Neurochemistry 60, 10-20	[Medline]
Suzuki, Y., Nakano, T. & Sears, M. (1997). Calcium signals from intact rabbit ciliary epithelium observed with confocal microscopy. Current Eye Research 16, 166-175.	[Medline]
Tao, Q.-P. & Kelly, M. E. M. (1996). Calcium-activated potassium current in cultured rabbit retinal pigment epithelial cells. Current Eye Research 15, 237-246.	[Medline]
Tao, Q.-P., Rafuse, P. E. & Kelly, M. E. M. (1994). Potassium currents in rabbit retinal pigment epithelial cells. Journal of Membrane Biology 141, 123-138	[Medline]
Ui, M. (1990). Pertussis toxin as a valuable probe for G-protein involvement in signal transduction. In ADP-Ribosylating Toxins and G Proteins. Insights into Signal Transduction. ed. Moss J. & Vaughan, M., pp. 45-77. American Society for Microbiology Press, Washington, DC.
Wax, M. B. & Molinoff, P. B. (1987). Distribution of beta-adrenergic receptors in human iris-ciliary body. Investigative Ophthalmology and Visual Science 28, 420-430	[Abstract]
Wiederholt, M., Helbig, H. & Korbmacher, C. (1991). Ion transport across the ciliary epithelium: lessons from cultured cells and proposed role of the carbonic anhydrase. In Carbonic Anhydrase, ed. Botre, F., Gross, G. & Storey, B. T., pp 232-244.

Acknowledgements

The authors wish to thank Christine Jollimore for her technical assistance and Mr Robert Gilbert for valuable discussions and suggestions. This work was supported by the Medical Research Council of Canada, grant no. MT-13484 and a National Science and Engineering Research Council of Canada Post-Graduate Scholarship B (PGSB) studentship to J. S. R.

Corresponding author

M. E. M. Kelly: Department of Pharmacology, Faculty of Medicine, Sir Charles Tupper Medical Building, Dalhousie University, Halifax, Nova Scotia, Canada B3H 4H7.

Email: mkelly{at}is.dal.ca

This Article Abstract Full Text (PDF) Services Email this article to a friend Similar articles in this journal Similar articles in PubMed Alert me to new issues of the journal Download to citation manager Citing Articles Citing Articles via Google Scholar Google Scholar Articles by Ryan, J. S. Articles by Kelly, M. E. M. Search for Related Content PubMed PubMed Citation Articles by Ryan, J. S. Articles by Kelly, M. E. M.