The Effect of Surface Charge on the Voltage-Dependent Conductance Induced in Thin Lipid Membranes by Monazomycin

Differences in the behavior of phosphatidylethanolamine (PE) and phosphatidylglycerol (PG) thin lipid membranes treated with monazomycin are shown to be due to the negative surface charge on PG membranes. We demonstrate that shifts of the conductance-voltage (g-V) characteristic of PG films produced by changes of univalent or divalent cation concentrations result from changes of the membrane surface potential on one or both sides. In particular, if divalent cations are added to the aqueous phase not containing monazomycin, the resulting asymmetry of the surface potentials results in an intramembrane potential difference not recordable by electrodes in the bulk phases. Nevertheless, this intramembrane potential difference is "seen" by the monazomycin, and consequently the g-V characteristic is shifted along the voltage axis. These changes are accounted for by diffuse double layer theory. Thus we find it unnecessary to invoke specific binding of Mg⁺⁺ or Ca⁺⁺ to the negative charges of PG membranes to explain the observation that concentrations of these ions some 100-fold lower than that of the univalent cation present produce large shifts of the g-V characteristic. We suggest that analogous shifts of g-V characteristics in axons produced by changes of divalent cation concentration are also best explained by diffuse double layer theory.     

Inactivation of monazomycin-induced voltage-dependent conductance in thin lipid membranes. II. Inactivation produced by monazomycin transport through the membrane

At sufficiently large conductances, the voltage-dependent conductance induced in thin lipid membranes by monazomycin undergoes inactivation. This is a consequence of depletion of monazomycin from the membrane solution interface, as monazomycin crosses the membrane to the opposite (trans) side from which it was added. The flux of monazomycin is directly proportional to the monazomycin-induced conductance; at a given conductance it is independent of monazomycin concentration. We conclude that when monazomycin channels break up, some or all of the molecules making up a channel are deposited on the trans side. We present a model for the monazomycin channel: approximately five molecules, each spanning the membrane with its NH3+ on the trans side and an uncharged hydrophilic (probably sugar) group anchored to the cis side, form an aqueous channel lined by--OH groups. The voltage dependence arises from the flipping by the electrical field of molecules lying parallel to the cis surface into the "spanned state;" the subsequent aggregation of these molecules into channels is, to a first approximation, voltage independent. The channel breakup that deposits monomers on the trans side involves the collapsing of the channel in such a way that the uncharged hydrophilic groups remain in contact with the water in the channel as they close the channel from behind. We also discuss the possibility that inactivation of sodium channels in nerve involves the movement from one side of the membrane to the other of the molecules (or molecule) forming the channel.     

Inactivation of monazomycin-induced voltage-dependent conductance in thin lipid membranes. I. Inactivation produced by long chain quaternary ammonium ions

The voltage-dependent conductance induced in thin lipid membranes by monazomycin undergoes inactivation upon the introduction of quaternary ammonium ions (QA) having a long alkyl chain (e.g. dodecyltrimethylammonium [C12]) to the side containing monazomycin. That is, in response to a step of voltage the conductance rises to a peak and then falls to a much lower steady-state value. We demonstrate that the basis of this phenomenon is the ability of QA to pass through the stimulated membrane and bind to the opposite surface. As a consequence, the surface potential on that side becomes more positive, thus reducing the voltage across the membrane proper and turning off the monazomycin-induced conductance. Because the flux of QA through the membrane increases linearly with conductance, we believe that these ions pass through the monazomycin channels. QA permeability increases with alkyl chain length; remarkably, in spite of its much larger size, C12 is about 150 times more permeant than K+. It appears, therefore, that there is a hydrophobic region of the cahnnel that favors the alkyl chain; we propose that this region is formed by the hydrophobic faces of the monazomycin channels in lipid bilayers to QA inactivation of potassium channels in the squid giant azon, and suggest that there may be a common structural feature for the two channels. It is possible that some of the inactivation phenomena in excitable cells may arise from local field changes not measurable by the recording electrodes.     

Monazomycin-induced single channels. I. Characterization of the elementary conductance events

Monazomycin (a positively charged, polyene-like antibiotic) induces voltage-dependent conductance changes in lipid bilayer membranes when added to one of the bathing solutions. These conductance changes have generally been attributed to the existence of channels spanning the membrane. In this article we characterize the behavior of the individual conductance events observed when adding small amounts of monazomycin to one side of a lipid bilayer. We find that there are several apparent channel types with one or sometimes two amplitudes predominating. We find further that these fairly similar amplitudes represent two different states of the same fundamental channel entity, presumed to be the monazomycin channel. The current-voltage characteristics of these channels are weakly hyperbolic functions of applied potential. The average lifetimes are essentially voltage independent (between 50 and 400 mV). The average channel intervals, on the other hand, can be strongly voltage dependent, and we can show that the time-averaged conductance of a membrane is proportional to the average channel frequency.     

Hv1 Proton Channel Opening Is Preceded by a Voltage-independent Transition

The voltage sensing domain (VSD) of the voltage-gated proton channel Hv1 mediates a H⁺-selective conductance that is coordinately controlled by the membrane potential (V) and the transmembrane pH gradient (ΔpH). Allosteric control of Hv1 channel opening by ΔpH (V-ΔpH coupling) is manifested by a characteristic shift of approximately 40 mV per ΔpH unit in the activation. To further understand the mechanism for V-ΔpH coupling in Hv1, H⁺ current kinetics of activation and deactivation in excised membrane patches were analyzed as a function of the membrane potential and the pH in the intracellular side of the membrane (pH_I). In this study, it is shown for the first time to our knowledge that the opening of Hv1 is preceded by a voltage-independent transition. A similar process has been proposed to constitute the step involving coupling between the voltage-sensing and pore domains in tetrameric voltage-gated channels. However, for Hv1, the VSD functions as both the voltage sensor and the conduction pathway, suggesting that the voltage independent transition is intrinsic to the voltage-sensing domain. Therefore, this article proposes that the underlying mechanism for the activation of Hv1 involves a process similar to VSD relaxation, a process previously described for voltage-gated channels and voltage-controlled enzymes. Finally, deactivation seemingly occurs as a strictly voltage dependent process, implying that the kinetic event leading to opening of the proton conductance are different than those involved in the closing. Thus, from this work it is proposed that Hv1 activity displays hysteresis.     

Nystatin-induced liposome fusion. A versatile approach to ion channel reconstitution into planar bilayers.

A simple method is described for promoting and detecting fusion of liposomes with planar bilayer membranes. Liposomes containing ergosterol are doped with the pore-forming antibiotic nystatin, and the planar bilayer is kept ergosterol-free. Under these conditions, when a transbilayer salt gradient is applied, liposomes added to the high-salt side of the bilayer elicit the appearance of abrupt conductance jumps of 5-300 pS. The increase in conductance is transient, decaying back to baseline on the order of 10 s. Each of these "spikes" represents the fusion of a single liposome with the bilayer, resulting in the simultaneous insertion of many nystatin channels. Relaxation of the conductance back to baseline occurs because ergosterol, required for the integrity of the nystatin pore, diffuses away into the sterol-free planar bilayer after liposome fusion. When Torpedo Cl- channels are reconstituted into liposomes containing ergosterol and nystatin, fusion spikes are observed simultaneously with the appearance of Cl- channels. This method allows the calculation of the density of functional ion channels in a preparation of proteoliposomes containing reconstituted channel protein.     

The Ion Permeability Induced in Thin Lipid Membranes by the Polyene Antibiotics Nystatin and Amphotericin B

Characteristics of nystatin and amphotericin B action on thin (<100 A) lipid membranes are: (a) micromolar amounts increase membrane conductance from 10^-8 to over 10^-2 Ω^-1 cm^-2; (b) such membranes are (non-ideally) anion selective and discriminate among anions on the basis of size; (c) membrane sterol is required for action; (d) antibiotic presence on both sides of membrane strongly favors action; (e) conductance is proportional to a large power of antibiotic concentration; (f) conductance decreases ~10⁴ times for a 10°C temperature rise; (g) kinetics of antibiotic action are also very temperature sensitive; (h) ion selectivity is pH independent between 3 and 10, but (i) activity is reversibly lost at high pH; (j) methyl ester derivatives are fully active; N-acetyl and N-succinyl derivatives are inactive; (k) current-voltage characteristic is nonlinear when membrane separates nonidentical salt solutions. These characteristics are contrasted with those of valinomycin. Observations (a)–(g) suggest that aggregates of polyene and sterol from opposite sides of the membrane interact to create aqueous pores; these pores are not static, but break up (melt) and reform continuously. Mechanism of anion selectivity is obscure. Observations (h)–(j) suggest—NH₃⁺ is important for activity; it is probably not responsible for selectivity, particularly since four polyene antibiotics, each containing two—NH₃⁺ groups, induce ideal cation selectivity. Possibly the many hydroxyl groups in nystatin and amphotericin B are responsible for anion selectivity. The effects of polyene antibiotics on thin lipid membranes are consistent with their action on biological membranes.     

Hv1 Proton Channel Opening Is Preceded by a Voltage-independent Transition

The voltage sensing domain (VSD) of the voltage-gated proton channel Hv1 mediates a H⁺-selective conductance that is coordinately controlled by the membrane potential (V) and the transmembrane pH gradient (ΔpH). Allosteric control of Hv1 channel opening by ΔpH (V-ΔpH coupling) is manifested by a characteristic shift of approximately 40 mV per ΔpH unit in the activation. To further understand the mechanism for V-ΔpH coupling in Hv1, H⁺ current kinetics of activation and deactivation in excised membrane patches were analyzed as a function of the membrane potential and the pH in the intracellular side of the membrane (pH_I). In this study, it is shown for the first time to our knowledge that the opening of Hv1 is preceded by a voltage-independent transition. A similar process has been proposed to constitute the step involving coupling between the voltage-sensing and pore domains in tetrameric voltage-gated channels. However, for Hv1, the VSD functions as both the voltage sensor and the conduction pathway, suggesting that the voltage independent transition is intrinsic to the voltage-sensing domain. Therefore, this article proposes that the underlying mechanism for the activation of Hv1 involves a process similar to VSD relaxation, a process previously described for voltage-gated channels and voltage-controlled enzymes. Finally, deactivation seemingly occurs as a strictly voltage dependent process, implying that the kinetic event leading to opening of the proton conductance are different than those involved in the closing. Thus, from this work it is proposed that Hv1 activity displays hysteresis.     

Monazomycin-induced single channels. II. Origin of the voltage dependence of the macroscopic conductance

The voltage dependence of the conductance induced induced in thin lipid membranes by monazomycin is shown here to be caused by voltage- dependent variations in the frequency of channel openings. We also experimentally demonstrate certain interesting properties of the channel activity that are predicted by a chemical kinetic model (Muller and Peskin, 1981), which successfully describes the macroscopic conductance. We conclude that two parallel mechanisms--one autocatalytic, the other simple mass action--exist that allow monazomycin to enter (or leave) the membrane so that the monazomycin molecules can be in a position to form channels.     

Ionic mechanisms of nonlinearity of the retinal horizontal cell membrane

Changes in ionic conductance lying at the basis of nonlinearity of the current-voltage characteristic curve of the cell (nonsynaptic) membrane of horizontal cells were studied in experiments on the goldfish and turtle retina. All measurements were made during blocking of synaptic transmission by bright light or Co⁺⁺. An increase in the K⁺ concentration led to depolarization and to a reduction of the steepness of the hyperpolarization branch of the current-voltage curve, whereas a decrease in K⁺ had the opposite effect. Changes in the Cl^– or Na⁺ concentrations had no significant effect on membrane potential or on the shape of the current-voltage curve. The principal potential-forming ion in the horizontal cells is thus K⁺; conductance for Cl^– is absent or very low, and conductance for Na⁺ also is evidently small. In the presence of Ba⁺⁺ (2–5 mM) the steepness of the hyperpolarization branch of the current-voltage curve was increased and the whole curve became more linear. It is concluded that nonlinearity of the current-voltage curve of the horizontal cell membrane is due mainly to potential-dependent potassium channels, whose conductance increases during hyperpolarization; this increase in conductance is blocked by Ba⁺⁺. An increase in the Ca⁺⁺ concentration to 20 mM led to an increase in steepness of the depolarization branch of the current-voltage curve, suggesting that depolarization increases membrane conductance for Ca⁺⁺.Institute for Problems in Information Transmission, Academy of Sciences of the USSR, Moscow. Translated from Neirofiziologiya, Vol. 13, No. 5, pp. 531–539, September–October, 1981.     

Mitogen-activated Ca++ channels in human B lymphocytes

Two complementary experimental methods have been used to examine mitogen-induced transmembrane conductances in human B cells using the Daudi cell line as a model for human B cell activation. Spectrofluorometry was used to investigate mitogen-induced changes in [Ca⁺⁺]_i and transmembrane potential. Activation of human B cells with anti-μ antibodies resulted in a biphasic rise in [Ca⁺⁺]_i, the second phase being mediated by the influx of extracellular Ca⁺⁺. Ca⁺⁺ influx was inhibited by high [K⁺]e, suggesting that this influx was transmembrane potential sensitive. Membrane currents of Daudi cells were investigated using voltage clamp techniques. Before mitogenic stimulation, the cells were electrically quiet. Within several minutes of the addition of anti-μ antibodies to the bath solution, inward currents were observed at negative voltages. Whole-cell currents changed instantly with voltage steps and were transmembrane potential sensitive in that at potentials more positive than ?40 mV no currents were detectable. A similar conductance was also activated by the introduction of IP₃ into the intracellular solution, suggesting that IP₃ generation after surface IgM crosslinking is involved in the activation of this conductance. Both anti-μ and IP³ induced currents were blocked by 1 mM La⁺⁺⁺, which is known to block Ca⁺⁺ channels. These results strongly support the presence of membrane Ca⁺⁺ channels in human B cells that function in the early stages of activation. Changes in transmembrane potential appear to be important in regulating Ca⁺⁺ influx. These mechanisms work in concert to regulate the level of [Ca⁺⁺]_i during the early phases of human B cell activation. © 1993 Wiley-Liss, Inc.     

Single channel conductance at lipid bilayer membranes in presence of monazomycin

The single channel conductance in the presence of the antibiotic monazomycin at lipid bilayer membranes was measured. It was found to be 3 times smaller than the single channel conductance of gramicidin A under the same conditions. The conductance of the monazomycin pore is voltage independent.     

Coupling of water and potassium ions in K+ channels of the tonoplast of Chara

Electrokinetic measurements, of streaming potential, were carried out on an excised inside-out patch of the vacuolar membrane of Chara corallina. A water activity gradient was imposed across the patch membrane containing a single K⁺ channel by addition of sorbitol to one side. Two different K⁺ channels were found in the tonoplast. Their open channel conductance was investigated as a function of KCl concentration. They had a maximal open channel conductance of 247 and 173 pS, and an apparent affinity (K_M) of 116 and 92 mM, respectively. Single-channel zero-current potentials were determined in the presence of an osmotic gradient, and dilution artifacts were corrected for by addition of valinomycin to the bath. Our results suggest that 29 water molecules were coupled to the transport of one K⁺ ion in the large conductance K⁺ channel which has a pore radius of ~1.5 nm.     

The Initiation of Spike Potential in Barnacle Muscle Fibers under Low Intracellular Ca++

Electrical properties of the muscle fiber membrane were studied in the barnacle, Balanus nubilus Darw. by using intracellular electrode techniques. A depolarization of the membrane does not usually produce an all-or-none spike potential in the normal muscle fiber even though a mechanical response is elicited. The intracellular injection of Ca⁺⁺-binding agents (K₂SO₄ and K salt of EDTA solution, K₃ citrate solution, etc.) renders the fiber capable of initiating all-or-none spikes. The overshoot of such a spike potential increases with increasing external Ca concentration, the increment for a tenfold increase in Ca concentration being about 29 mv. The threshold membrane potential for the spike and also for the K conductance increase shifts to more positive membrane potentials with increasing [Ca⁺⁺]_out. The removal of Na ions from the external medium does not change the configuration of the spike potential. In the absence of Ca⁺⁺ in the external medium, the spike potential is restored by Ba⁺⁺ and Sr⁺⁺ but not by Mg⁺⁺. The overshoot of the spike potential increases with increasing [Ba⁺⁺]_out or [Sr⁺⁺]_out. The Ca influx through the membrane of the fiber treated with K₂SO₄ and EDTA was examined with Ca⁴⁵. The influx was 14 pmol per sec. per cm² for the resting membrane and 35 to 85 pmol per cm² for one spike. From these results it is concluded that the spike potential of the barnacle muscle fiber results from the permeability increase of the membrane to Ca⁺⁺ (Ba⁺⁺ or Sr⁺⁺).     

Characterization of mechanosensitive channels in Escherichia coli cytoplasmic membrane by whole-cell patch clamp recording

Whole-cell patch clamp recordings were done on giant protoplasts of Escherichia coli. The pressure sensitivity of the protoplasts was studied. Two different unit conductance mechanosensitive channels, 1100 ± 25 pS and 350 ± 14 pS in 400 mm symmetric KCl solution, were observed upon either applying positive pressure to the interior of the cells or down shocking the cells osmotically. The 1100 pS conductance channel discriminated poorly among the monovalent ions tested and it was permeable to Ca²⁺ and glutamate^?. Both of the two channels were sensitive to the osmotic gradient across the membrane; the unit conductances of the channels remained constant while the mean current of the cell was increased by increasing the osmotic gradient. Both of the channels were voltage sensitive. Voltage-ramp results showed that the pressure sensitivity of protoplasts was voltage dependent: there were more channels active upon depolarization than hyperpolarization. The mech anosensitive channels were reversibly blocked by gadolinium ion. Also they could reversibly be inhibited by protons. Mutations in two of the potassium efflux systems, KefB and KefC, did not affect the channel activity, while a null mutation in the gene for KefA changed the channel activity significantly. This indicates a potential modulation of these channels by KefA.     

Calcium entry in rabbit corneal epithelial cells: Evidence for a nonvoltage dependent pathway

We performed experiments to elucidate the calcium influx pathways in freshly dispersed rabbit corneal epithelial cells. Three possible pathways were considered: voltage-gated Ca⁺⁺ channels, Na⁺/Ca⁺⁺ exchange, and nonvoltage-dependent Ca⁺⁺-permeable channels. Whole cell inward currents carrying either Ca⁺⁺ or Ba⁺⁺ were not detected using voltage clamp techniques. We also used imaging technology and the Ca⁺⁺-sensitive ratiometric dye fura 2 to measure changes in intracellular Ca⁺⁺ concentration ([Ca]_i). Bath perfusion with NaCl Ringer's solution containing the calcium channel agonist Bay-K-8644 (1 m), or Ni⁺⁺ (40 m), a blocker of many voltage-dependent calcium channels, did not affect [Ca⁺⁺]_i. Membrane depolarization with a KCl Ringer's bath solution resulted in a decrease in [Ca⁺⁺]_i. These results are inconsistent with the presence of voltage gated Ca⁺⁺ channels. Nonvoltage gated Ca⁺⁺ entry, on the other hand, would be reduced by membrane depolarization and enhanced by membrane hyperpolarization. Agents which hyperpolarize via stimulation of K⁺ current, such as flufenamic acid, resulted in an increase in ratio intensity. The cells were found to be permeable to Mn⁺⁺ and bath perfusion with 5 mm Ni⁺⁺ decreased [Ca⁺⁺]_i suggesting that the Ca⁺⁺ conductance was blocked. These results are most consistent with a nonvoltage gated Ca⁺⁺ influx pathway. Finally, replacing extracellular Na⁺ with Li⁺ resulted in an increase in [Ca⁺⁺]_i if the cells were first Na⁺-loaded using the Na⁺ ionophore monensin and ouabain, a Na⁺-K⁺-ATPase inhibitor. These results suggest that Na⁺/Ca⁺⁺ exchange may also regulate [Ca⁺⁺] in this cell type.The authors are grateful to Chris Bartling for expert technical assistance with the imaging experiments, Helen Hendrickson for cell preparation, and Jonathon Monck for helpful discussions regarding imaging technology. This work was supported by National Institutes of Health grants EYO3282, EYO6005, DK08677, and an unrestricted award from Research to Prevent Blindness.     

Voltage-dependent channels found in the membrane fraction of corn mitochondria

Transmembrane channels have been found in the membrane fraction of corn (Zea mays W64AN) mitochondria that exhibit a remarkable resemblance to the voltage dependent anion-selective channels (VDAC) located in the outer membrane of animal (Rattus norvegicus), protist (Paramecium aurelia), and fungal (Neurospora crassa) mitochondria. The channels in corn were demonstrated to be essentially identical to VDAC channels in three characteristic properties: (a) single channel conductance magnitude, (b) weak anion selectivity, and (c) nature of voltage dependence. These findings led us to conclude that the channels present in corn mitochondria are VDAC channels. This discovery may have repercussions concerning the regulation and function of higher plant mitochondria, and the causation of higher plant excitability.     

Permeability and Electrical Properties of Thin Lipid Membranes

We present and discuss the permeability and electrical properties of thin lipid membranes, and the changes induced in these properties by several agents added to the aqueous phases after the membranes have formed. The unmodified membrane is virtually impermeable to ions and small "hydrophilic" solutes, but relatively permeable to water and "lipophilic" molecules. These properties are consistent with those predicted for a thin film of hydrocarbon through which matter is transported by dissolving in the membrane phase and then diffusing through it. The effect of cholesterol in reducing the water and "lipophilic" solute permeability is attributed to an increase of the "viscosity" of the hydrocarbon region, thus reducing the diffusion coefficient of molecules within this phase. The selective permeability of the membrane to iodide (I^-) in the presence of iodine (I₂) is attributed to the formation of polyiodides (perhaps I₅ ^-), which are presumed to be relatively soluble in the membrane because of their large size, and hence lower surface charge density. Thus, I₂ acts as a carrier for I^-. The effects of "excitability-inducing material" and the depsipeptides (particularly valinomycin) on ion permeability are reviewed. The effects of the polyene antibiotics (nystatin and amphotericin B) on ion permeability, discussed in greater detail, are the following: (a) membrane conductance increases with the 10th power of nystatin concentration; (b) the membrane is anion-selective but does not discriminate completely between anions and cations; (c) the membrane discriminates among anions on the basis of size; (d) membrane conductance decreases extraordinarily with increasing temperatures. Valinomycin and nystatin form independent conductance pathways in the same membrane, and, in the presence of both, the membrane can be reversibly shifted between a cation and anion permeable state by changes in temperature. It is suggested that nystatin produces pores in the membrane and valinomycin acts as a carrier.     

Reconstitution in planar lipid bilayers of a voltage-dependent anion-selective channel obtained from paramecium mitochondria

Summary We have incorporated into planar lipid bilayer membranes a voltage-dependent, anion-selective channel (VDAC) obtained fromParamecium aurelia. VDAC-containing membranes have the following properties: (1) The steady-state conductance of a many-channel membrane is maximal when the transmembrane potential is zero and decreases as a steep function of both positive and negative voltage. (2) The fraction of time that an individual channel stays open is strongly voltage dependent in a manner that parallels the voltage dependence of a many-channel membrane. (3) The conductance of the open channel is about 500 pmho in 0.1 to 1.0m salt solutions and is ohmic. (4) The channel is about 7 times more permeable to Cl^– than to K⁺ and is impermeable to Ca⁺⁺. The procedure for obtaining VDAC and the properties of the channel are highly reproducible.VDAC activity was found, upon fractionation of the paramecium membranes, to come from the mitochondria. We note that the published data on mitochondrial Cl^– permeability suggest that there may indeed be a voltage-dependent Cl^– permeability in mitochondria.The method of incorporating VDAC into planar lipid bilayers may be generally useful for reconstituting biological transport systems in these membranes.     

Hormonal and Neurotransmitter Regulation of Ca++ Influx Through Voltage-Dependent Slow Channels in Cardiac Muscle Membrane

The voltage-and time-dependent slow channels in the myocardial cell membrane are the major pathway by which Ca⁺⁺ ions enter the cell during excitation for initiation and regulation of the force of contraction of cardiac muscle. These slow channels behave kinetically as if their gates open, close, and recover more slowly than those of the fast Na⁺ channels; in addition, the slow channel gates operate over a less negative (more depolarized) voltage range. Tatrodotoxin does not block the slow channels, whereas the calcium antagonistic drugs, Mn⁺⁺, Co⁺⁺, and La⁺⁺⁺ ions do. The slow channels have some special properties, including their functional dependence on metabolic energy, their selective blockade by acidosis, and their regulation by cyclic AMP level. Because of their regulation by cyclic AMP, it is proposed that either the slow channel protein or an associated regulatory protein must be phosphorylated in order for the channel to be made available for voltage activation during excitation. That is, the dephosphorylated channel would be electrically silent.

The requirement for phosphorylation allows the extrinsic control of the slow channels and Ca⁺⁺ influx by neurotransmitters, hormones, and autacoids that affect the cyclic nucleotide levels.