Voltage-dependent potassium channels in the amphibian lens membranes: evidence from radiotracer and electrical conductance measurements

Electrical signaling in control of ocular cell behaviors

Epithelia of the cornea, lens and retina contain a vast array of ion channels and pumps. Together they produce a polarized flow of ions in and out of cells, as well as across the epithelia. These naturally occurring ion fluxes are essential to the hydration and metabolism of the ocular tissues, especially for the avascular cornea and lens. The directional transport of ions generates electric fields and currents in those tissues. Applied electric fields affect migration, division and proliferation of ocular cells which are important in homeostasis and healing of the ocular tissues. Abnormalities in any of those aspects may underlie many ocular diseases, for example chronic corneal ulcers, posterior capsule opacity after cataract surgery, and retinopathies. Electric field-inducing cellular responses, termed electrical signaling here, therefore may be an unexpected yet powerful mechanism in regulating ocular cell behavior. Both endogenous electric fields and applied electric fields could be exploited to regulate ocular cells. We aim to briefly describe the physiology of the naturally occurring electrical activities in the corneal, lens, and retinal epithelia, to provide experimental evidence of the effects of electric fields on ocular cell behaviors, and to suggest possible clinical implications.

The lens: local transport and global transparency

The perception of the lens changed remarkably during the career of David Maurice. The early view was that it was an inert sack of protein that assisted the cornea in focusing light on the retina. As investigators looked more carefully, more and more complexity was revealed and today we know the lens is a living, dynamic organ that carries out a host of biochemical and physiological processes necessary for homeostasis. We have worked on the lens over this period and have provided a small part of the data on lens physiology. This paper is an overview of our own contributions, in the context of the ever evolving view of the lens. Given this is a brief tribute to the career of David Maurice, there is not enough space nor is it appropriate to provide a complete review of all the work that has contributed to this evolving

Effects of Ca2+ on rabbit translens short-circuit current: evidence for a Ca2+ inhibitable K+ conductance

PURPOSE

To characterize the effects of medium Ca2+ levels on rabbit lens electrical properties. Early studies with wholly submerged lenses had shown that Ca2+ removal from the bath resulted in an increased Rb+ efflux, a consequence of an increased Na+ Permeability and lens depolarization.

METHODS

Lenses were bathed with Ussing-type chambers under short-circuited conditions, an arrangement in which the translens short-circuit current (Isc) is carried out across the posterior lens surface mainly by an influx of Na+, and across the anterior face largely by a K+ efflux.

RESULTS

Under the present conditions in which the effects of Ca2+ were characterized unilaterally, the above established effects could only be ascribed to the posterior surface. When Ca2+ removal was limited to the anterior face, the Isc increased from 11.87 +/- 1.17 to 17.04 +/- 1.52 microA/cm2 (means +/- SE's, n = 18; an accompanying translens resistance (Rt) decrease of 0.23 +/- 0.049 K omega.cm2 was also recorded). Conversely, increasing the control, anterior-bath [Ca2+] from 1.8 to 3.6 mM reduced the K+ efflux-dependent Isc from 10.54 +/- 1.09 to 8.93 +/- 1.02 (n = 10, with an Rt increase of 0.11 +/- 0.013). These changes were reversible Na(+)-independent, and fully inhibited by the presence of K+ channel blockers (quinidine or Ba2+). Inhibitions of the Ca2+ effects were also obtained with strontium, a Ca2+ surrogate. The Isc was less responsive to changes in the Ca2+ content of the posterior bath. Removal of the cation caused a gradual 1.65 +/- 0.72 microA/cm2 increase (n = 9, with an Rt decrease of 0.090 +/- 0.021 K omega.cm2). In the absence of posterior Na+, Ca2+ withdrawal resulted in highly variable responses, with some specimens exhibiting salient current increases, suggesting that an outwardly directed, posterior efflux of an anion could also have been affected. During the course of this study it was consistently observed that the removal of Na+ from the anterior bath led to an Isc decrease of 2.62 +/- 0.22 microA/cm2 (n = 32, with an Rt increase of 0.35 +/- 0.029 k omega.cm2). This change occurred in both the presence of ouabain and the absence of Ca2+, suggesting that it did not result from an inhibition of the Na(+)-K+ pump current nor from a reversal in putative Na+/Ca2+ exchange activity. Small Isc increases upon anterior Na+ withdrawal (1.68 +/- 0.17, n = 7), consistent with Na+ efflux from the lens, could only be observed with K+ channels inhibited with Ba2+. Also congruent with the observations of a relatively limited anterior Na+ permeability, was the finding that the induction of nonspecific cation channels with amphotericin B reduced the Isc by following Na+ from the anterior bath to enter the lens. Thus, changes in lens Isc can differentiate changes in K+ permeability across the native anterior epithelium from changes in Na+ permeability.

CONCLUSIONS

Overall, these results suggest that lens Ca2(+)-mobilizing agents (e.g. acetylcholine) could trigger the inhibition of epithelial K+ conductance(s) by the direct action of Ca2+ on K+ channels.

Altered lens short-circuit current in adult cataract-prone Dahl hypertensive rats

We assessed components of lenticular short-circuit current in adult hypertensive Dahl salt-sensitive rats (DS) during chronic control (0.4% sodium) versus high (3% sodium) dietary NaCl intake begun at the age of 4 weeks until rats were studied. We also evaluated the influence of barium, a potassium channel blocker, and ouabain, a specific inhibitor of Na+, K(+)-ATPase activity, by adding them to the anterior lens surface, thus measuring barium-sensitive, ouabain-sensitive, and barium- and ouabain-in-sensitive short-circuit currents. During control NaCl intake, short-circuit current in DS and their control group, Dahl salt-resistant rats (DR), did not differ significantly. DS were subclassified into cataract-prone rats and rats unlikely to develop cataracts on the basis of their initial pressor response to the change from a normal to high NaCl diet during the first weeks of age. Although only transparent lenses were studied, total lens short-circuit current was already markedly decreased in the cataract-prone subgroup compared with DS unlikely to develop cataracts and control DR. This was in sharp contrast to the increase in short-circuit current previously reported in Sprague-Dawley rats and now observed in control DR in response to high dietary NaCl. The decrease in lens short-circuit current in cataract-prone rats was associated with lower absolute values of barium- and ouabain-sensitive short-circuit currents as well as with low barium- and ouabain-insensitive short-circuit current. Although the barium- and ouabain-sensitive components of the short-circuit current were similar in DS unlikely to develop cataracts and DR, the barium- and ouabain-insensitive component of the short-circuit current was lower in DS unlikely to develop cataracts than values in DR. Interestingly, this component of lens short-circuit current also increased in DR during chronic high NaCl, whereas the opposite change occurred in cataract-prone DS and DS unlikely to develop cataracts. Thus, the barium- and ouabain-insensitive short-circuit current may be a mechanism that protects the normal lens from developing cataracts. Possible candidates for this short-circuit current component are voltage-dependent potassium channels, calcium-activated potassium channels, or both. Our studies show altered lens short-circuit current in response to high NaCl intake in cataract-prone DS and suggest the possibility of altered lens potassium transport during sustained hypertension but before loss of lens transparency.

Acetylcholine modulation of the short-circuit current across the rabbit lens

Rabbit lenses were bathed within a bicameral Ussing-type chamber under short-circuit conditions. In this situation the short-circuit current (Isc) reflects, across the anterior aspect, the presence of anteriorly facing K+ conductance(s) plus the Na(+)-K+ pump current. Across the posterior surface the Isc is primarily carried by the movement of Na+ from the posterior bathing solution to the lens. Addition of acetylcholine (ACh) to the posterior hemichamber did not affect the translens electrical parameters; but, its introduction to the anterior bath at 1 microM immediately reduced the Isc from 8.91 +/- 1.47 to 5.84 +/- 1.28 microA cm-2 and increased the translens resistance from 1.50 +/- 0.08 to 1.59 +/- 0.09 K omega cm2 (+/- S.E.S; P < 0.05 as paired values, n = 25 lenses). The suppressed Isc gradually recovered and reached 75% of the control value 5 min after the introduction of the neurotransmitter. In six cases the recovery was nearly complete (> or = 95% of control) within this time. The preaddition of 0.1 microM atropine prevented an effect by 1 microM ACh. When atropine was added within 1 min of ACh, the suppressed Isc immediately recovered. The ACh-elicited Isc suppression was averted in lenses pre-exposed to either K+ channel blockers (quinidine or barium) or to the endoplasmic reticular Ca(2+)-ATPase inhibitor thapsigargin (Tg: 0.1 microM), which in itself produced Isc inhibitions similar to those seen with ACh under control conditions. Similarly comparable were the ACh-evoked Isc inhibitions garnered upon introduction of the agonist to lenses bathed in the absence of extracellular Ca2+. In these cases, however, the Isc recovered fully within 2-3 min. This condition also revealed that the anterior removal of medium Ca2+ increased the Isc by about 50%, a completely reversible phenomenon; Ca2+ restoration in the presence of the Ca2+ channel blocker, nifedipine (10 microM), blunted markedly the reversal to the control Isc. Overall, these results suggest that ACh receptor activation induces the release of intracellularly stored Ca2+, which in turn leads to the temporary deactivation of a K+ conductance(s); in addition, secondary Ca2+ inflow may further extend the observed inhibition. During this study, the Isc of about 30% of the lenses used spontaneously oscillated (common duration of 30 min, with a mean peak frequency of 0.76 +/- 0.32 cycle min-1 and mean amplitude of 4.07 +/- 2.65 microA cm-2; +/- S.D.S, n = 24). Experiments attempted to determine the sensitivity of the oscillatory activity to ACh. Tg, nifedipine, and the phorbol ester PMA.(ABSTRACT TRUNCATED AT 400 WORDS)

Increased ion traffic through non-specific cation pathways in the ageing human lens. Evidence from radiotracer fluxes studies

86Rb efflux studies were carried out on normal human lenses in vitro. The data confirmed previous studies showing that 86Rb efflux increases with age. Removal of Ca2+ from the lens perifusate increased 86Rb efflux at all ages. The fractional increase above baseline was highest in the younger lenses, while the net increase of the 86Rb efflux induced by a Ca(2+)-free medium increased with age. This study supports the idea that Ca(2+)-sensitive, non-specific cation channels are present in the human lens and that their contribution to membrane permeability increases as the lens ages.

Diamide alters membrane Na+ and K+ conductances and increases internal resistance in the isolated rat lens

The voltage and conductance of the isolated rat lens were measured using a two-internal-microelectrode technique and the potassium permeability was calculated by applying Goldman theory to 86Rb efflux data. The SH oxidizing agent diamide induces a multiphasic response in lens voltage, conductance and potassium permeability. The initial response (less than or equal to 30 min) to 1 mM diamide consists of a small depolarization (approximately 10 mV) of membrane potential accompanied by a significant decrease in conductance. The 86Rb efflux and permeability data also show an initial decrease. As this initial response is abolished by TEA (20 mM) it is probably due to an inactivation of voltage-sensitive potassium channels. After 30 min exposure to 1 mM diamide both the electrical conductance and the rate of depolarization increase. The 86Rb permeability also increases. Since the conductance increase is abolished by replacing Na+ by methyl glucamine and as it is reduced by amiloride (10(-4) M) the second phase is probably due to the activation of nonspecific cation channels. The third phase is only apparent after prolonged (approximately 12 hr) incubation in 1 mM diamide and consists of a marked increase in the bulk resistance component of the lens impedance. It is suggested that this component arises from an increase in the resistance of the fibre cell gap junctions. This cellular uncoupling may be due to calcium entering the lens through the nonspecific cation channels.

p-chloro-mercuriphenyl sulphonate activates a quinine-sensitive potassium conductance in frog lens

1. The effects of the sulphydryl-complexing reagent p-chloro-mercuriphenyl sulphonate (pCMPS) on membrane voltage and electrical conductance were studied on the isolated frog lens. 2. At low concentrations (0.1-50 microM) pCMPS induced a rapid and graded hyperpolarization of the lens membrane potential which saturated at -97 mV. 3. The lens conductance also showed a graded increase, but the initial changes were apparent only at concentrations above 1 microM. 4. Decreasing the external potassium concentration from 2.5 to 0.5 mM had little effect on the membrane potential in the absence of pCMPS, but increased the voltage from -97 to -110 mV when pCMPS was present. 5. Quinine (300 microM) had no effect when added in control solution, but depolarized the membrane potential and decreased the conductance when added to a pCMPS-treated preparation. 6. These data suggest that pCMPS activates voltage-sensitive potassium channels that are quiescent at the frog resting potential in control solution. 7. At pCMPS concentrations greater than or equal to 100 microM, the initial hyperpolarization is followed by a marked but slow depolarization of the membrane potential and a further increase in lens conductance. These data suggest that non-specific cation channels are activated in this case. 8. Cysteine (5 mM) added to a pCMPS-treated lens leads to a rapid recovery of membrane potential and conductance to near their resting values whether the lens had previously been exposed to low or high concentrations of pCMPS. 9. All of these changes in lens voltage and conductance occurred without apparent alteration in the lens internal sulphydryl content.

Fresh and cultured human lens epithelial cells: an electrophysiological study of cell coupling and membrane properties

Microelectrode studies of fresh human and rabbit lens epithelia revealed stable membrane potentials [VR (human) = -36 mV; VR (rabbit) = -45 mV] and low input resistances [Ri (human) = 10 M omega; Ri (rabbit) = 20 M omega]. Coupling studies, using two voltage microelectrodes, demonstrated that the low input resistance of the fresh epithelial tissue was due to electrotonic coupling, which was found to be extremely labile and sensitive to perfusion of the apical (fibrefacing) surface of the epithelium. The intercellular coupling could be stabilized by raising the calcium concentration of the perfusate. Studies performed on confluent monolayers of cultured human lens epithelial (HLE) cells demonstrated a membrane potential (VR = -33 mV) and input resistance (Ri = 29 M omega) similar to their fresh counterparts. The intercellular coupling of these cells was found to be much more robust. Ultrastructural studies revealed that the apical junction of cultured HLE cells was less complex than that found in fresh tissue, the latter exhibiting multiple interdigitations and folds. The cultured monolayer was dissociated into single cells by a variety of methods and the membrane properties of individual cells were studied. Single cells were found to have a lower membrane potential (-20 to -25 mV) and an input resistance in the range 110-170 M omega, depending on the method of dissociation. Channel blocking and ion replacement studies revealed significant conductance pathways for potassium, sodium and chloride and a cell-attached patch clamp investigation revealed three distinct channel types. Of the two channels with inward currents at the resting potential, one, with a conductance of 25 pS, is identified as a non-selective cation channel, and the other, with a conductance of 14 pS and reversal potential of - 14 mV, is a possible candidate for a chloride channel but has yet to be characterized. A third channel with an outward current at the resting potential is identified as a potassium channel with a conductance of 49 pS. A link between epithelial uncoupling and certain types of cataract is proposed.

Oxidative cross-linking of fodrin parallels a membrane conductance increase in the mammalian lens

An oxidative cross-linking of the lens spectrin-like protein fodrin was induced by incubating WKY-rat lenses in the presence of the SH-reagent diamide. The oxidation of fodrin was paralleled by an increase in lens membrane conductance. The time relationship between these two events as well as the reversibility of both, achieved by incubating the lens in the presence of dithiothreitol, indicate that normal permeability characteristics of the lens membranes require the integrity of the membrane attached cytoskeleton.

Membrane and junctional properties of the isolated frog lens epithelium

The isolated frog lens epithelium can be maintained intact in both appearance and electrical properties for more than 24 hours. The mean resting membrane potential was -80 mV and the cells were depolarized by both high potassium and low calcium Ringer's solution in a manner very similar to that of the whole lens. The epithelial cells were found to be well coupled using both electrical and dye-injection techniques. Electrical coupling was measured using separate current-injection and voltage-measuring electrodes and the relationship between the induced voltage and distance from the current-passing electrode could be well fitted by a Bessel Function solution to the cable equation. The values obtained from the fit for the membrane and internal resistances were 1.95 omega m2 and 25 omega m, respectively. Exposure to octanol (500 microM) or low external Ca2+ (less than 1 microM) failed to disrupt significantly the intercellular flow of current. There was evidence to suggest that raised intracellular calcium does, however, uncouple the cells. Dye coupling was investigated by microinjecting Lucifer Yellow CH into single epithelial cells. Diffusion into surrounding cells was rapid and, in control medium, occurred in a radially symmetrical manner. In contrast to the electrical coupling data, dye transfer appeared to be blocked by exposure to 500 microM octanol and was severely restricted on perfusing with low external calcium. Differences between the electrical and dye-coupling experiments indicate either that there are two types of junction within the cell and only the larger type, permeable to Lucifer Yellow, is capable of being uncoupled or that there is only one large type of junction which can be partially closed by uncoupling agents.

Equatorial potassium currents in lenses

Earlier work with the vibrating probe demonstrated the existence of outward potassium currents at the equator and inward sodium currents at the optical poles of the lens. By adding microelectrodes to the system, it is possible to relate steady currents (J) to the potential difference (PD) measured with a microelectrode. By injecting an outward current (I), it is possible to determine resistances and also the PD at which the steady outward potassium current becomes zero (PDJ = 0). At this PD the concentration gradient for potassium efflux and the electrical gradient for potassium influx are balanced so that there is no net flow of potassium across the membranes associated with the production of J. The PDJ = 0 for 18 rat lenses was 86 mV and that for 12 frogs lenses was -95 mV. This agrees with the potassium equilibrium potential and provides strong evidence to support the view that the outward equatorial current, J, is a potassium current. With the injection of outward current, I, the PD becomes more negative, the outward equatorial current, J, decreases, and the inward current at the optical poles increases. This suggests that there are separate electrical loops for K+ and Na+ that are partially linked by the Na, K-pump. Using Ohm's law, it is possible to calculate the input resistance (R = delta PD/I), the resistance related to the production of J (RJ = delta PD/delta J), and the effect of the combined resistances (delta J/I). The driving force for J can be estimated (PDJ = 0-PD). The relationships among currents, voltages and resistance can be used to determine the characteristics of the membranes that are associated with the outward potassium current observed at the equator. The effects of graded deformation of the lens were determined. The effects were reversible. The sites of inward and outward currents were not altered. Following deformation, the equatorial current, J, increased, and the PD became less negative. The PDJ = 0 remains the same so the ratio of K+ concentrations across the membrane responsible for J is unchanged. Therefore, the decrease in PD is ascribed to an increase in Na+ permeance with a resultant increase in driving force accounting for the increase in J.

Membrane permeability characteristics of perfused human senile cataractous lenses

Human cataractous lenses were removed by the cryoprobe technique and were maintained for up to 24 hr in a solution of similar ionic composition to human aqueous humour. The bimodal distribution of internal sodium concentrations was similar to that previously reported for unincubated human lenses. Lenses with lower total and free sodium contents had relatively higher membrane potentials and they lost 86Rb at a slower rate than lenses with high internal sodium. The 86Rb efflux in these lenses was stimulated four-fold by removing external calcium. The efflux was reduced by increasing external calcium, but was increased during a small (60 mosmol) hyperosmotic shock. A similar hyperosmotic shock also surprisingly increased 86Rb efflux. Lenses with increasing internal sodium (and calcium) levels showed an increasing rate of loss of 86Rb and the stimulation by calcium removal was progressively diminished. The efflux from lenses with disturbed ion levels was also relatively insensitive to changes in external osmolarity and to increasing external potassium. Lenses with raised sodium concentrations also had an increased inulin space. Frog, rat and rabbit lenses were also exposed to the same range of stimuli and only frog lenses responded to the low calcium solution with more than a four-fold increase in efflux rate. Although only a two- to four-fold increase in efflux rate was obtained from rabbit lenses exposed to Ca-free conditions, this was the only type of animal lens so far tested that, like the human lens, responded to both hyperosmotic and isosmotic shocks with an increase in efflux rate. All three species of mammalian lenses responded with an increase in efflux rate when exposed to a hyperosmotic test solution while in the frog, the efflux rate from the lens decreased. The glucose efflux from human cataractous lenses was inhibited by cytochalasin B in a similar manner to the efflux from rat and frog lenses. It was concluded, therefore, that the cryoprobed human lens can be kept for a limited period in a relative simple artificial aqueous humour solution. The potassium permeability characteristics of low sodium cataracts remained relatively intact and showed a unique response (relative to lenses from other animals) when exposed to various stimuli that are known to be potentially cataractogenic.

The localization of transport properties in the frog lens

The selectivity of fiber-cell membranes and surface-cell membranes in the frog lens is examined using a combination of ion substitutions and impedance studies. We replace bath sodium and chloride, one at a time, with less permeant substitute ions and we increase bath potassium at the expense of sodium. We then record the time course and steady-state value of the intracellular potential. Once a new steady state has been reached, we perform a small signal-frequency-domain impedance study. The impedance study allows us to separately determine the values of inner fiber-cell membrane conductance and surface-cell membrane conductance. If a membrane is permeable to a particular ion, we presume that the conductance of that membrane will change with the concentration of the permeant ion. Thus, the impedance studies allow us to localize the site of permeability to inner or surface membranes. Similarly, the time course of the change in intracellular potential will be rapid if surface membranes are the site of permeation whereas it will be slow if the new solution has to diffuse into the intercellular space to cause voltage changes. Lastly, the value of steady-state voltage change provides an estimate of the lens' permeability, at least for chloride and potassium. The results for sodium are complex and not well understood. From the above studies we conclude: (a) surface membranes are dominated by potassium permeability; (b) inner fiber-cell membranes are permeable to sodium and chloride, in approximately equal amounts; and (c) inner fiber-cell membranes have a rather small permeability to potassium.

Response of the intracellular potentials of cultured bovine lens cells to ions and inhibitors

Micropuncture of bovine lens epithelial cells cultured on plastic culture dishes gave values for the plasma membrane voltage (V) which remained stable for periods of up to several hours. The value of V was mainly in the range -30 to -45 mV, mean value -36.9 +/- 0.5 mV (S.E.M., n = 188). Raising extracellular [K+] from 5 to 40 mM depolarized V by 10 +/- 3 mV. The extent of this depolarization increased with increasing steady-state V. Barium (2 mM) caused a rapid, reversible depolarization of 7.9 +/- 1.2 mV. In the presence of Ba2+, the response to 40 mM K was reduced to 3.6 +/- 1.1 mV. Ouabain (10(-5) M) caused a fast depolarization by 5.3 +/- 1.2 mV. Exposure to calcium-free EGTA-Ringer's depolarized V reversibly by 19.5 +/- 5.0 mV. In Ca-free medium, the depolarization induced by 40 mM K was reduced to 3.2 +/- 2.4 mV. Whereas in control Ringer's sodium conductance (as measured by exposure to a 10 mM [Na]-Ringer's) is small as compared to potassium conductance, it increased markedly in Ca-depleted medium. Amiloride (10(-4) and 10(-3) M) had no effect on this Na conductance. An increase in the relative conductance for sodium was also elicited by Ba2+ (2 mM). Extracellular acidification led to a depolarization, alkalinization to a hyperpolarization. The extent of this effect was virtually equal in the absence or presence of HCO3-, excluding a significant pathway for bicarbonate. No evidence for a significant chloride conductance could be obtained.

86-Rb efflux in normal and cataractous human lenses

86Rb efflux has been studied in normal lenses and in human senile cataracts. The rate constant (Ki) of the efflux gradually increases in cataractous lenses with progression of lens damage. Efflux experiments run in the presence of BaC12 suggest that a progressive activation of BaC12 inhibitable efflux routes occurs in cataractous lenses. In the final stages of opacification the ineffectiveness of BaC12 enriched or Ca++ free media on the efflux suggests that a direct disruption of the lens membranes has occurred.

Specificity of glucose transport inhibitors in the frog lens

The unidirectional fluxes (influx and efflux) of glucose across frog lens membranes were investigated using the radio-labelled sugar analogue 3-O-methyl-D-[U-14C]glucose. The effect of various inhibitors of sugar transport on the movement of 3-O-methylglucose was studied and the specificity of the inhibition was estimated by carrying out concomitant measurements of lens sodium content. The movement of 3-O-methylglucose into the lens was rapid, and 50% equilibration occurred within 5 hr. The influx was reduced in the presence of phloretin, phloridzin, ouabain, iodoacetate and cytochalasin B, but only in the case of the latter was there no concomitant change in lens sodium content. Only cytochalasin B can therefore be regarded as a specific inhibitor of glucose transport. The efflux of 3-O-methylglucose was followed after 16 hr incubation with [14C]3-O-methylglucose. The efflux kinetics had a double exponential form and the half-time of the slower phase was 165 mins. The efflux of the slow phase was found to be sensitive to the presence of inhibitors. Phloretin and cytochalasin B produced the most marked reduction in efflux rate, but again only in the latter case was the effect reversible. Bidirectional transport of glucose therefore occurs in the lens and movement in both directions is reduced by cytochalasin B, which appears to be the only inhibitor of transport so far studied that does not disturb lens ion levels.