- Email: [email protected]
The role of divalent cations in controlling amphibian lens membrane permeability; The mechanisms of toxic cataracts
Eccp,. Eye Res. (1983)
595
36; 595-606
The Role of Divalent Cations in Controlling Amphibian Lens Membrane Permeability; The Mechanisms of Toxic Cataracts T.J.C.
JACOBANDG.DUNCAN
School of Biological Sciences, University of East Anglia, Norwich NR4 7TJ, U.K. (Received 17 November
1982 and accepted 17 January
1983, London)
The effects of a range of divalent ions on lens sodium and potassium permeability characteristics were studied in calcium competition and replacement experiments. Resting voltage and conductance were measured and also voltage-independent conductance. Strontium and manganese were the only divalent ions able to maintain, in the absence of calcium: both sodium and potjassium permeability at or near the control level. R’either cobalt nor magnesium had any effect on lens voltage of conductance in the presence of calcium, but neither of these ions could maintain lens permeability properties in the absence of calcium. Cadmium and barium had little effect on sodium permeability, but the former increased potassium permeability while the latter reduced it. Barium was the only divalent studied that appeared to inactivate voltage-sensitive potassium channels in the presence of calcium. Nickel, zinc and copper increased both sodium and potassium permeability in the presence of calcium and so they are likely to be particularly damaging to the lens. Copper was extremely toxic since it was able to overturn the regulatory influence of calcium when it was present in concentrations as low as IO-” M. Key words: membrane; divalent; cataract; toxic; calcium.
1. Introduction It has recently been shown that calcium controls both the sodium and potassium permeability of the lens (Jacob and Duncan, 1981). When the calcium level in the medium surrounding the lens is reduced; the membrane potential depolarises as a result of the sodium permeability increase. There is also a marked change in the slope of the membrane conductance relationship, indicating that a concomitant decrease in the voltage-dependent potassium permeability has occurred. Reducing external calcium also causes marked disturbances in the ion and water balance of the lens (Owers and Duncan, 1979; Delamere and Paterson, 1981) and hypocalcaemia has in fact’ long been associated with cataract (Bellows, 1944; Duke-Elder, 1969). However, it has recently been pointed out that the presence of trace amounts of certain divalent ions, e.g. cobalt and zinc (Hollwich, Boateng and Kolk, 1975) also initiate cataract and these divalent ions might well be expected to interfere with the regulation of sodium and potassium permeability by calcium ions. Recently Bernardini, Peracchia and Venosa (1981) have shown that calcium is involved in healing-over in the rat lens after damage and they suggested that an increase in intracellular calcium, following injury, switches the membrane gap junction from a low to a high resistance state (Bernardini and Peracchia, 1981). The healing-over process is in fact greatly reduced on incubation in calcium-free media. The membrane effects of calcium that have been observed in other tissues (e.g. in nerve) are often interpreted in terms of a fixed negative charge model (Gilbert and Ehrenstein, 1969) in which the calcium ions may be adsorbed on to fixed charges at the membrane surface. This would change the profile of the diffuse double layer at 0014.4835/83/040595
+ 11 toa.oo/o
0 1983 Academic
Press Inc. (London)
Limited 01 -9
.596
T. J. C!. JACOB
the membrane removal
of
surface calcium
binding constants hence the properties In
this
study
or
AP;D
and hence alter
the transmembrane
the
competing
addition
would
of
be expected
to alter
of ion channels. we shall be particularly
ions to substitute
for calcium
in controlling
2. Materials Theoretical From
G. DUNCAN
polyvalent
membrane
electrical cations
charge
field. with
configurations
interested in the ability of various lens membrane permeability. and
Thus t’he differing
and divalent
Methods
considerations the
Goldman
potential
equation: E = $
In (A/B),
A = K, + aNa,
+ /3Cli,
B = Ki+czNai+/3CI,, the subscripts i and o refer to lens interior and bathing medium respectively. CL and p are permeability ratios such that a = PNa/PK and /? = PcL/PK. E is the resting potential; R, T and F have their usual meanings. It can be calculated that for the amphibian lens the ratio d/B = 0.044 (Jacob and Duncan, 1981). It follows that an increase in potassium permeability (PK) will cause the potential (E) to hyperpolarise; an increase in sodium permeability will cause a depolarisation and any change in chloride permeability (Pcl) will have little effect since chloride is distributed at equilibrium. The conductance-voltage relationship of the lens membranes reveals two important elements of conductance (see Fig. 2). First, the baseline conductance (g,,) at hyperpolarised potentials which represents the passive leak of sodium, potassium and chloride and, secondly, the exponential rising phase which has been attributed to the opening of voltage-sensitive potassium channels (Patmore and Duncan, 1979). Experimental Preparation. Small acclimatised frogs (Ranapipien.s) were used for these experiments. After pithing, the whole eye was removed, equatorially bisected and the lens lifted free with a glass loop. Solutions. The lens was placed anterior face down in a Perspex perifusion chamber, where it was bathed in a Ringer solution of a similar composition to the aqueous humour. This solution was modified for particular experiments. The basic composition of the control Ringer was as follows: NaCl, 111.5 mM; KCl, 2.5 mM; MgCl,, 1.2 mM; Hepes, 5 mM; n-glucose, 5.5 mM; CaCl,, 2 mM. For calcium replacement experiments, an equimolar amount of the test ion was substituted for the calcium, except in the case of copper, where a range of lower concentrations was used. In competition experiments, the test ion was simply added to the control solution. The osmolarity of the control solution was 209 mosmol which exactly matched the measured osmolarity of the aqueous humour. The test solutions were also checked and found to be within + 1 o/0 of the control value. Electrical recordings. A two-internal microelectrode method was used for measuring lens potential and conductance and the voltage and current-measuring circuits are described in detail elsewhere (Jacob and Duncan 1980).
3. Results A range
of divalent
for the control
ions
of sodium
were tested and potassium
for
their ability permeability
both to compete and to substitute
with calcium for calcium in
this role. They were therefore added to the normal medium both in the presence and absence of calcium (2 mM). The divalent cations tested fall into certain distinct groups,
DIVALEST
05
-601
099
CATIOK
CONTROL
0.6 0.66 l-34 0.74 I-12 0.72 0.97 0.69 0.72
Ionic
radius
OF
LENS
PERMEABILITY
597
-601
(A)
FIG. 1, The effects of divalent cations on the lens resting potential and slope conductance (a) in 2 IIIMcalcium containing solutions or (b) in calcium-free solutions. The origin for the voltage axis represents the resting potential in control solution ( - 77% f 0.7 mV; n = 29) and positive displacements represent a depolarisation of potential. The origin for conductance, measured in the control solution, was 27 +@I S m-*; ?z = 29. In competition experiments, the appropriate ion was added to the control solution, and the volta.ge and conductance measured after 15 miu. In replacement experiments, the lens was perfused first with the control solution and then with the replacement solution where calcium was omitted and the appropriate divalent ion added. The data were again obtained at the end of a 15 min exposure. The effect of a calcium-free solution alone is shown in the first column of(b). The data represent the mean of at least three lenses in each case. Sate that in all of the experiments 2 mlw-concentrations of the test ions were used except in the case of copper, which was added at a concentration of 10 px.
each with its characteristic effect on lens potential and conductance (Fig. 1). The addition of 2 mnl-calcium to the control Ringer (2 mM-Ca2+) causes only minimal changes in the lens resting potential and conductance [Fig. l(a)], whereas omitting calcium altogether depolarises the lens potential by 25 mV within 15 min [Fig. i(b)]. The resting conductance also increases by 50 %. These data therefore act as controls for competition a.nd substitution experiments. When added in the presence of calcium, strontium, manganese, cobalt and magnesium induce only minor changes in potent,ial and conductance [Fig. l(a)]. When added in the absence of calcium, only strontium and manganese prevent a significant depolarisation in voltage. Cobalt is a relatively good substit’ute for calcium but magnesium fails to prevent a very large depolarisation in voltage and increase in conductance. Nickel, zinc and copper are very potent calcium antagonists as they depolarise the lens potential and increase conductance in the presence of calcium. The effect of copper was particularly marked and the data shown in Fig. 1 were obtained with a concentration of 10e5 1~. All other ions were added at a concentration of 2 mM. Cadmium is unique in that it induces a massive increase in conductance and yet produces a hyperpolarisation in the lens potential [Fig. 1 (a)] and SO in terms of equation (1) it appears to act mainly on lens potassium channels. It still permits calcium access to the membrane sodium channels, so the voltage does not depolarise when calcium
598
T. J. C. JACOB
AND
6.
DUlYCAS
is present. Cadmium may also assist in the regulation of the sodium channel permeability as a very large depolarisation does not occur when calcium is removed [Fig. l(b)]. The regulatory influence of barium is almost a mirror image to that of cadmium since it induces a depolarisation in the presence of calcium and a decrease in membrane conductance. Hence, while the presence of cadmium leads to an increase in potassium permeability, the presence of barium initiates a decrease. When calcium is removed from the solution, only a small further depolarisation is observed [Fig. l(b)], indicating that barium regulates sodium permeability in the absence of calcium. Magnesium, which has the smallest ionic radius of any of the ions tested, did not compete with calcium [Fig. l(a)] nor did it to any great extent subst,itute for calcium in the control of membrane potential and conductance [Fig. l(b)]. In order to investigate the mechanism of action of the various divalent ions in more detail, the membrane conductance of the lens was measured at different potentials (see Jacob and Duncan, 1980, for experimental details). In this way both the linear and non-linear conductances of the lens can be assessed separately. The former conductance (in some preparations called ‘leakage conductance ‘) arises mainly from the diffusion of sodium and potassium across the lens membranes and through the extracellular space of the lens while the latter conductance mainly arises from voltage-sensitive potassium channels in the surface membranes (Duncan, Patmore and Pynsent, 1981; Jacob and Duncan, 1981). Figure 2 shows the slope conductance of the lens a.s a function of the lens membrane potential both in control medium (0) and in the presence of barium (A). Barium obviously competes with calcium for the control of the linear (go) and non-linear conductances. The effect on g, can be interpreted as a reduction in potassium permeability since the voltage depolarises in barium (4). Barium also reduces the extent to which the conductance
Voltcge
FIG. 2. The slope
conductance solution (O), in the presence (a). The arrows indicate the
(mV)
of the lens as a function of lens membrane potential of 2 mw-barium with calcium (A) and in 2 mM-barium resting voltage in each solution.
measured without
in control calcium
DIVALENT
CATION
CONTROL
OF
LENS
PERMEABILITY
599
is sensitive to voltage and hence barium also appears to inactivate membrane voltage-sensitive potassium channels. Removing calcium, but maintaining barium at 2 mM (a) does not lead to any great change in the conductance-voltage relationship and hence barium has the ability to replace calcium in the control of sodium permeability. In the presence of barium the shape of the voltage-clamp current transient assumes a time-dependency (Fig. 3) similar to that observed in calcium-free solution (Jacob and Duncan, 1981).
FIG. 3. Current transients (upper trace) obtained in response to 15 mV command clamp pulses. The initial spike represents charging membrane capacitance and in a linear network system, this should be followed by an exponential decline to a plateau value (see fig. 4 in Jacob and Duncan, 1981). However, in the case of the depolarising command, the spike is followed by an increasing phase. representing a time-dependent conductanoe change. This phase is absent from hyperpolarising commands.
The addition of 2 mM-calcium, -magnesium, -strontium or -manganese in the presence of 2 mm-calcium had little effect either on the resting potential and conductance [Fig. l(a)] or on the conductance-voltage relationship of the lens. However, of these ions only strontium and manganese have the ability to replace calcium in the control of lens voltage and conductance [Fig. l(b); Fig. 41. The slight increase in conductance and hyperpolarisation of lens voltage that occurs can be explained by a small increase in lens potassium permeability in the presence of these ions. There is no evidence, however, for a change in sodium conductance. Although cobalt, like manganese and strontium, does not compete with calcium, it cannot replace calcium to the same extent. A small depolarisation of membrane potential accompanies the increase in membrane conductance [Fig. l(b)] and hence in this case a small increase in sodium conductance occurs in the absence of calcium. The conductance-voltage curves were, however, essentially similar in form to those obtained with strontium in the absence of calcium (Fig. 4). Magnesium is interesting as it appears to be the only divalent cation tested that does not compete with calcium [Fig. l(a)] and yet does not replace calcium in the control of sodium and potassium permeability. The conductance-voltage curves obtained in the presence of 2 mM-
T. J. C. JBC’OB dh’U G. DCiY\‘C’BS
600
.
Voltage
FIG. 1. The effect (A), control (0).
on
the lens conductance-voltage
(mV)
curve
of replacing
calcium
with
2 mwstrontium
magnesium [and the absence of calcium had essentially the same form as those obtained in calcium-free solution (Jacob and Duncan, 1981)]. The addition of 2 m&r-cadmium to the bathing medium caused an immediate hyperpolarisation of approximately 6 mV [Fig. l(a)], accompanied by an 80 y. increase in membrane conductance within 20 min. Return to control solution initiated a biphasic voltage response. Within the first few minutes a depolarisation of approximately 10 mV was observed and this was followed by a slow hyperpolarising phase until the initial control voltage was obtained. The resistance, however, recovered monophasically to a value near the original control level. The removal of calcium from the cadmium Ringer caused a depolarisation of approximately 8 mV, so that the initial control value is not only reached; but slightly exceeded [Fig. l(b)]. The conductance increases still further and hence the observed depolarisation is probably due to an increase in sodium permeability. In the presence of calcium, cadmium therefore acts to increase membrane potassium permeability and, as the increase appears to be the same throughout the voltage range (Fig. 5), cadmium probably does not interfere with membrane voltage-sensitive channels. Cadmium also does not appear to interfere with calcium regulation of sodium permeability. In the absence of calcium, cadmium still acts to increase potassium permeability and it can to a large extent substitute for calcium in the control of sodium permeability. However, the control is not complete as there is a small increase in conductance and depolarisation of membrane potential when calcium is removed from the test solution containing cadmium. Nickel, zinc and copper had similar effects in that they all depolarised the membrane potential and increased membrane conductance when added to the control medium (Fig. 6). Again the conductance was increased throughout t.he voltage range and hence mainly the linear elements of the sodium and potassium conductances are involved.
DIVALENT
CATION
OF
LENS
PERMEABILITY
601
.
c t
CONTROL
,
;;i,:, -100
,
Voltage
I -910
,
-60
-00
--L
I
?I
I
-90
ICI
I
-70
CrnV)
Voltage
FIG. 5
FIG.
-50
, -30
( mV )
6
FIG. 5. The effect on the lens conductance-voltage curve of adding 2 mx-cadmium (A) to the control solution (0). Note that in this case, the data were obtained after 6 min since the conductance increase obtained after 15 min is so large [Fig. l(a)] that the conductance-voltage curve cannot be mapped out over the whole range. FIG. solution
6. The (0).
effect
on the
lens
conductance-voltage
curve
of adding
2 miwnickel
(A)
to the
control
60, control Gi u-2 g -
8
. ! . ..‘a.
50 .
40
,Fj 2
30
5
20 t
*
B 07
.
F
!O 01 0
.
copper ( 10-5M
. 1. .
.
l* I IO
I 20
I 30
I 40
I 50 Time
I 60
I 70
I 80
I 90
I 100
(min)
FIG. 7. The effect on the lens conductance of adding 10 pwcopper. Note that in this case the lens voltage was clamped at the resting value ( - 80 mV) and the conductance measured by superimposing small test command voltage pulses (see Jacob and Duncan. 1980, for further details).
T. J. C. JACOB
602
AND 6. DUNCAS
Only [Fig.
copper, however, was effective in the micromolar range of concentrations 71. Figure 7 shows the time-course of the conductance increase caused by 10,LhM CL?+, with the voltage clamped at - 80 mV. There is a very rapid increase, from 118 x 10v5 to 46 x 10e5 S within 30 min, and this is followed by a gradual but incomplete recovery within the course of the experiment (100 min). Zinc, when added to the control Ringer at a concentration of 2 mM, caused a depolarisation of 138 + 2.0 mV (n = 3) after 30 min. The conductance, when measured at the resting potential, was found to have increased from 13.8+ 2.0 to 62.7 + 8.2 x 10P5 S (n = 3). However, when the conductance is measured over a range of voltages, the increase was restricted to those potentials in the hyperpolarising direction. This fact, taken together with the observed depolarisation, implies an increase in sodium permeability alone. The exponentially rising phase of the conductance remains unchanged as in Fig. 8. When calcium is removed from the Zn2+ Ringer, there is a further depolarisation and increase in resting conductance [Fig. l(b)]. The conductance-voltage relationship now shows a form chajracteristic of that observed in calcium-free solution (Jacob and Duncan, 1981). The decrease in the slope of the curve indicates that an inactivation of membrane voltage-sensitive potassium channels has occurred.
-80
-60
Voltoge
FIG. 8. The solution (0).
effect
on the conductance-voltage
-40
-20
(mV 1
curve
of adding
lO par-copper
(A)
to the control
4. Discussion Jacob and Duncan (1981) have shown that calcium controls both sodium and potassium permeability and it appears now that the control occurs at two separate membrane sites as certain cations are able to influence the two permeabilities separately. For example, cadmium increases potassium permeability in the presence of calcium, without altering the sodium conductance, while barium decreases potas-
DIVALENT
CATIOK
COSTROL
OF
LESS
PERMEABILITY
603
sium permeability without apparently influencing sodium movement [Fig. l(a)]. Nickel, zinc and copper, however, probably increase both sodium and potassium conductance in the presence of calcium and the effects of these ions are likely to be particularly damaging. When the lens conductance is mapped out as a function of voltage, the potassium permeability can be seen to have both linear and non-linear components (Figs 2, 4, 5, 6 and 8). In the presence of calcium, only barium appears to influence both the linear and non-linear conductances, reducing them both. These data extend the previous work of Delamere, Duncan and Paterson (1980) and Delamere, Paterson, Duncan and-Holmes (1980) who found that barium reduced the 42K efflux from the lens and also reduced the extent to which the membrane conductance was sensitive to voltage. They did not, however, map the conductance out over a very wide range. Strontium and manganese are the only divalent ions able to perform the role of calcium in controlling sodium and potassium permeability [Fig. 1 (b)], perhaps because their ionic radii are similar to that of calcium (Fig. 1). However, when the conductance is mapped out over the whole voltage range (Fig. 4) it can be seen from the overall conductance increase that the substitution is not perfect. It is quite possible that divalent cations achieve their effect by interaction with lens membrane surface charge. The most obvious consequence of the mechanism whereby the ions affect the surface potential by screening rather than by specific binding is that equimolar substitution of Ca 2+ by another divalent ion should produce no change in membrane properties (Schauf, 1975). This appears to be true only for Sr2+ and Mn2+ and thus the other ions must achieve their relative effects by binding to the membranes. The other two ions tested with relatively large ionic radii, namely cadmium and barium, had opposite, but specific, effects on potassium conductance. The ionic radius of barium is in fact very similar to that of potassium (1.36 A) and it is possible that barium enters the lens via the K channel and operates a block of outward K-currents from the inside, accounting for the reduced conductance at depolarised potentials (Fig. 4). Barium is a powerful blocker of potassium conductance in frog skeletal muscle (Standen and Stanfield, 1978), Aplysia neurones (Adams and Gage, 1980) and non-excitable membranes such as frog gastric mucosa (Pacific0 et al., 1969). Cobalt has been found to cause a cataract in rabbits and rats (when administered in vivo) resembling that induced by alloxan (Alagna and d’Aquino, 1956). However, when presented to the amphibian lens in vitro, no changes in membrane characteristics were observed in the short term [Fig. l(a)]. Cobalt and manganese were able to substitute, to a large extent, for calcium [Fig. l(b)] in the control of lens permeability, whereas in other tissues they have inhibitory effects on conductance (Hagiwara and Takahashi, 1967; Adams and Gage, 1980). However, in both of the tissues involved, the inhibitory effect is a consequence of Co2+ and Mn2+ blocking voltage-sensitive cc&&n channels and these probably either do not exist in the lens membranes or do not significantly contribute to the control of conductance. Similarly, Cd2+ inactivates potassium currents in vertebrate neurones (Adams, Constanti; Brown and Clark, 1982) and LimuEus photoreceptors (Lisman, Fain and Swan, 1978) where again a primary blocking of a calcium channel is involved. In the lens, Cd2+ increases potassium conductance probably by binding to the surface membrane directly (Fig. 5). Zinc, nickel and copper ions are all recognised as cataractogenic (Heydenreich, 1966; Hollwich et al., 1975) and all produce similarly disruptive effects on the lens, namely a depolarisation of lens membrane potential and increase in conductance (Figs 1 and 6). However, the copper-induced changes occur at much lower concentrations than
T. J. C. JACOB
604
AND
G. DCNCAS
the others (Fig. 7) and indeed at concentrations near the physiological range. The copper concentration in normal human plasma is ZO,UM, of which about 90% is tightly bound to proteins (Henkin, Schulman, Schulman and Bronze&, 1973; Sarka,r and Kruck, 1966). The normal concentration of copper in the aqueous humour is 2 ,UM, but this can increase to 40 ,/AM in Wilson’s disease (Gerhard, 1966). Coulter, Oliver, Whitener, Collie and Engelke (1978) have shown that the cultured rat lens is extremely sensitive to copper in the bathing medium and have pointed out that changes in aqueous humour amino acids may play a role in copper toxicity as they carry out most of the chelating in the aqueous humour. Since copper can overturn the regulatory influence of calcium when it is present in concentrations as low as 10e6 M [Fig. i(a)], it is not surprising that Bellows and Bellows (1975) recommend surgical removal of intralenticular copper-containing particles. For any other particulate body, surgical removal from the lens is never recommended. The electrophysiological methods described here provide a means not only of investigating the mechanism of copper toxicity in great detail, but it also provides an accurate and rapid method of screening certain drugs (Coulter et al., 1978) that may be able to overcome the toxic effects. REFERENCES Sdams, D. J. and Gage, P. W. (1980). Divalent ion currents and the delayed potassium conductance in an Aplysia neurone. J. Physiol. 304, 297-313. Adams, P. R., Constanti, A., Brown, D. A. and Clark, R. B. (1982). Intracellular Ca2+ activates
a fast voltage-sensitive
K+ current
in vertebrate
(London) 296, 746-9. Alagna, 0. and D’Aquino, S. (1956). Augenveranderungen
sympathetic
neurones.
Nature
durch Kobaltchlorid. Arch. Ottal. 60, 5-29. Bellows, J. G. (1944). Cataract and Abnormalities of the Lens. C. V. Mosby, St. Louis, MO. Bellows, J. G. and Bellows, R. T. (1975). Traumatic cataract. In Cataract and Abnormalities of the Lens (Ed. Bellows, J. G.). Pp. 265-72. Grune and Stratton, N.Y. Bernardini, G. and Peracchia, C. (1981). Gap junction crystallization in lens fibres after an increase in cell calcium. Invest. OphthaZmoZ. Vis. Sei. 21, 291-9. Bernardini, G., Peracchia, C. and Venosa, R. A. (1981). Healing-over in rat crystalline lens. J. Physiol. 320, 187-92. Coulter, J. B., Oliver, 6. S., Whitener, C. M., Collie, 6. J. and Engelke, J. A. (1978). Toxic effects of copper on cultured rat lenses. Exp. Eye Res. 26, 547-54. Delamere, N. A., Duncan, G. and Paterson, C. A. (1980). Characteristics of voltage-dependent conductance in the membranes of a non-excitable tissue: the amphibian lens. J. Physiol. 308, 49-59. Delamere, N. A. and Paterson, C. A. (1981). Hypocalcaemic cataract. In Mechanisms of Cataract Formation in the Human Lens (Ed. Duncan, G.). Academic Press, London. Delamere, N. A., Paterson, C. A., Duncan, G. and Holmes, D. L. (1980). Relative roles ofCaZf, Sr2+, Ba2+ and Mg2+ in controlling lens permeability characteristics. Cell Calcium 1,81-90. Duncan, G., Patmore, L. afnd Pynsent, P. B. (1981). Impedance of the amphibian lens. J. Physiol. 312, 17-27. Duke-Elder, S. (1969). System of Ophthalmology XI. C. V. Mosby, St. Louis, MO. Gerhard, J. P. (1966). Study of the copper of aqueous humor. DOG. OphthaEmoZ. 20, 104410. Gilbert, D. L. and Ehrenstein, G. (1969). Effect of divalent cations on potassium conductance of squid axons: determination of surface charge. Biophys. J. 9, 447-63. Hagiwara, S. and Takahashi, K. (1967). Surface density of calcium ions and calcium spikes in the barnacle fibre membrane. J. gen. Physiol. 50, 583-601. Henkin, R. I., Schulman, J. D., Schulman, C. B. and Bronzert, D. 8. (1973). Changes in total non-diffusible anddiffusible plasmazinc and copper during infancy. J. Pediatr. 82,831-7. Heydenreich, A. (1966). Chemisch-toxische Schaden der Augen. IiZin. Monatsbl. Augenheilkd. 149. 145-60.
DIVALENT
CATION
CONTROL
OF
LEKS
PERMEABILITY
606
Hollwich, F., Boateng, A. and Kolck; B. (1975). Toxic cataract in Cataract and Abnormalities of the Lens (Ed. Bellows; J. G.). Pp. 23943. Grune and Stratton N.Y. Jacob, T. J. C. a,nd Duncan, G. (1980). Osmotic influences on lens membrane characteristics. Exp. Eye Res. 31, 50512. Jacob, T. J. 6. and Duncan, G. (1981). Calcium controls both sodium and potassium permeability of lens membranes. Exp. Eye Res. 33, 8&93. Lisman, J., Fain, G. and Swan, M. (1978). Voltage-dependent conductances in Limulus ventral photoreceptors. Biol. Bull. 155, 4534. Owers, J. and Duncan, G. (1979). The viability of the bovine lens in organ culture. Exp. Eye Res. 28, 73S42. Pacifico, A. D., Schwartz, M.; Mackrell, T. N., Spangler, S. G., Sanders, S. S. and Rehm, W. S. (1969). Reversal by potassium of an effect of barium on the frog gastric mucosa. A,m. J. Physiol. 216, 536-41. Patmore, L. and Duncan, G. (1979). A TEA-sensitive component in the conductance of a non-excitable tissue (the amphibian lens). Exp. Eye Res. 28, 349-52. Sarkar, B. and Kruck, T. P. A. (1966). Copper-amino acid complexes in human serum. In The Biochemistry of Copper (Eds Peisach, J., Aisen, P. and Blumberg, W.). Pp. 183-196. Academic Press, N.Y. 9chauf, C. L. (1975). The interactions of calcium with Myxicola giant axons and a description in terms of a simple surface charge model. J. Physiol. 248, 3, 613-25. Standen, N. B. and Stanfield, P. R. (1978). A potential and time-dependent blockade of inward rect.ification in frog skeletal muscle fibres by barium and strontium ions. J. Physiol. 280, 169-91.
595
36; 595-606
The Role of Divalent Cations in Controlling Amphibian Lens Membrane Permeability; The Mechanisms of Toxic Cataracts T.J.C.
JACOBANDG.DUNCAN
School of Biological Sciences, University of East Anglia, Norwich NR4 7TJ, U.K. (Received 17 November
1982 and accepted 17 January
1983, London)
The effects of a range of divalent ions on lens sodium and potassium permeability characteristics were studied in calcium competition and replacement experiments. Resting voltage and conductance were measured and also voltage-independent conductance. Strontium and manganese were the only divalent ions able to maintain, in the absence of calcium: both sodium and potjassium permeability at or near the control level. R’either cobalt nor magnesium had any effect on lens voltage of conductance in the presence of calcium, but neither of these ions could maintain lens permeability properties in the absence of calcium. Cadmium and barium had little effect on sodium permeability, but the former increased potassium permeability while the latter reduced it. Barium was the only divalent studied that appeared to inactivate voltage-sensitive potassium channels in the presence of calcium. Nickel, zinc and copper increased both sodium and potassium permeability in the presence of calcium and so they are likely to be particularly damaging to the lens. Copper was extremely toxic since it was able to overturn the regulatory influence of calcium when it was present in concentrations as low as IO-” M. Key words: membrane; divalent; cataract; toxic; calcium.
1. Introduction It has recently been shown that calcium controls both the sodium and potassium permeability of the lens (Jacob and Duncan, 1981). When the calcium level in the medium surrounding the lens is reduced; the membrane potential depolarises as a result of the sodium permeability increase. There is also a marked change in the slope of the membrane conductance relationship, indicating that a concomitant decrease in the voltage-dependent potassium permeability has occurred. Reducing external calcium also causes marked disturbances in the ion and water balance of the lens (Owers and Duncan, 1979; Delamere and Paterson, 1981) and hypocalcaemia has in fact’ long been associated with cataract (Bellows, 1944; Duke-Elder, 1969). However, it has recently been pointed out that the presence of trace amounts of certain divalent ions, e.g. cobalt and zinc (Hollwich, Boateng and Kolk, 1975) also initiate cataract and these divalent ions might well be expected to interfere with the regulation of sodium and potassium permeability by calcium ions. Recently Bernardini, Peracchia and Venosa (1981) have shown that calcium is involved in healing-over in the rat lens after damage and they suggested that an increase in intracellular calcium, following injury, switches the membrane gap junction from a low to a high resistance state (Bernardini and Peracchia, 1981). The healing-over process is in fact greatly reduced on incubation in calcium-free media. The membrane effects of calcium that have been observed in other tissues (e.g. in nerve) are often interpreted in terms of a fixed negative charge model (Gilbert and Ehrenstein, 1969) in which the calcium ions may be adsorbed on to fixed charges at the membrane surface. This would change the profile of the diffuse double layer at 0014.4835/83/040595
+ 11 toa.oo/o
0 1983 Academic
Press Inc. (London)
Limited 01 -9
.596
T. J. C!. JACOB
the membrane removal
of
surface calcium
binding constants hence the properties In
this
study
or
AP;D
and hence alter
the transmembrane
the
competing
addition
would
of
be expected
to alter
of ion channels. we shall be particularly
ions to substitute
for calcium
in controlling
2. Materials Theoretical From
G. DUNCAN
polyvalent
membrane
electrical cations
charge
field. with
configurations
interested in the ability of various lens membrane permeability. and
Thus t’he differing
and divalent
Methods
considerations the
Goldman
potential
equation: E = $
In (A/B),
A = K, + aNa,
+ /3Cli,
B = Ki+czNai+/3CI,, the subscripts i and o refer to lens interior and bathing medium respectively. CL and p are permeability ratios such that a = PNa/PK and /? = PcL/PK. E is the resting potential; R, T and F have their usual meanings. It can be calculated that for the amphibian lens the ratio d/B = 0.044 (Jacob and Duncan, 1981). It follows that an increase in potassium permeability (PK) will cause the potential (E) to hyperpolarise; an increase in sodium permeability will cause a depolarisation and any change in chloride permeability (Pcl) will have little effect since chloride is distributed at equilibrium. The conductance-voltage relationship of the lens membranes reveals two important elements of conductance (see Fig. 2). First, the baseline conductance (g,,) at hyperpolarised potentials which represents the passive leak of sodium, potassium and chloride and, secondly, the exponential rising phase which has been attributed to the opening of voltage-sensitive potassium channels (Patmore and Duncan, 1979). Experimental Preparation. Small acclimatised frogs (Ranapipien.s) were used for these experiments. After pithing, the whole eye was removed, equatorially bisected and the lens lifted free with a glass loop. Solutions. The lens was placed anterior face down in a Perspex perifusion chamber, where it was bathed in a Ringer solution of a similar composition to the aqueous humour. This solution was modified for particular experiments. The basic composition of the control Ringer was as follows: NaCl, 111.5 mM; KCl, 2.5 mM; MgCl,, 1.2 mM; Hepes, 5 mM; n-glucose, 5.5 mM; CaCl,, 2 mM. For calcium replacement experiments, an equimolar amount of the test ion was substituted for the calcium, except in the case of copper, where a range of lower concentrations was used. In competition experiments, the test ion was simply added to the control solution. The osmolarity of the control solution was 209 mosmol which exactly matched the measured osmolarity of the aqueous humour. The test solutions were also checked and found to be within + 1 o/0 of the control value. Electrical recordings. A two-internal microelectrode method was used for measuring lens potential and conductance and the voltage and current-measuring circuits are described in detail elsewhere (Jacob and Duncan 1980).
3. Results A range
of divalent
for the control
ions
of sodium
were tested and potassium
for
their ability permeability
both to compete and to substitute
with calcium for calcium in
this role. They were therefore added to the normal medium both in the presence and absence of calcium (2 mM). The divalent cations tested fall into certain distinct groups,
DIVALEST
05
-601
099
CATIOK
CONTROL
0.6 0.66 l-34 0.74 I-12 0.72 0.97 0.69 0.72
Ionic
radius
OF
LENS
PERMEABILITY
597
-601
(A)
FIG. 1, The effects of divalent cations on the lens resting potential and slope conductance (a) in 2 IIIMcalcium containing solutions or (b) in calcium-free solutions. The origin for the voltage axis represents the resting potential in control solution ( - 77% f 0.7 mV; n = 29) and positive displacements represent a depolarisation of potential. The origin for conductance, measured in the control solution, was 27 +@I S m-*; ?z = 29. In competition experiments, the appropriate ion was added to the control solution, and the volta.ge and conductance measured after 15 miu. In replacement experiments, the lens was perfused first with the control solution and then with the replacement solution where calcium was omitted and the appropriate divalent ion added. The data were again obtained at the end of a 15 min exposure. The effect of a calcium-free solution alone is shown in the first column of(b). The data represent the mean of at least three lenses in each case. Sate that in all of the experiments 2 mlw-concentrations of the test ions were used except in the case of copper, which was added at a concentration of 10 px.
each with its characteristic effect on lens potential and conductance (Fig. 1). The addition of 2 mnl-calcium to the control Ringer (2 mM-Ca2+) causes only minimal changes in the lens resting potential and conductance [Fig. l(a)], whereas omitting calcium altogether depolarises the lens potential by 25 mV within 15 min [Fig. i(b)]. The resting conductance also increases by 50 %. These data therefore act as controls for competition a.nd substitution experiments. When added in the presence of calcium, strontium, manganese, cobalt and magnesium induce only minor changes in potent,ial and conductance [Fig. l(a)]. When added in the absence of calcium, only strontium and manganese prevent a significant depolarisation in voltage. Cobalt is a relatively good substit’ute for calcium but magnesium fails to prevent a very large depolarisation in voltage and increase in conductance. Nickel, zinc and copper are very potent calcium antagonists as they depolarise the lens potential and increase conductance in the presence of calcium. The effect of copper was particularly marked and the data shown in Fig. 1 were obtained with a concentration of 10e5 1~. All other ions were added at a concentration of 2 mM. Cadmium is unique in that it induces a massive increase in conductance and yet produces a hyperpolarisation in the lens potential [Fig. 1 (a)] and SO in terms of equation (1) it appears to act mainly on lens potassium channels. It still permits calcium access to the membrane sodium channels, so the voltage does not depolarise when calcium
598
T. J. C. JACOB
AND
6.
DUlYCAS
is present. Cadmium may also assist in the regulation of the sodium channel permeability as a very large depolarisation does not occur when calcium is removed [Fig. l(b)]. The regulatory influence of barium is almost a mirror image to that of cadmium since it induces a depolarisation in the presence of calcium and a decrease in membrane conductance. Hence, while the presence of cadmium leads to an increase in potassium permeability, the presence of barium initiates a decrease. When calcium is removed from the solution, only a small further depolarisation is observed [Fig. l(b)], indicating that barium regulates sodium permeability in the absence of calcium. Magnesium, which has the smallest ionic radius of any of the ions tested, did not compete with calcium [Fig. l(a)] nor did it to any great extent subst,itute for calcium in the control of membrane potential and conductance [Fig. l(b)]. In order to investigate the mechanism of action of the various divalent ions in more detail, the membrane conductance of the lens was measured at different potentials (see Jacob and Duncan, 1980, for experimental details). In this way both the linear and non-linear conductances of the lens can be assessed separately. The former conductance (in some preparations called ‘leakage conductance ‘) arises mainly from the diffusion of sodium and potassium across the lens membranes and through the extracellular space of the lens while the latter conductance mainly arises from voltage-sensitive potassium channels in the surface membranes (Duncan, Patmore and Pynsent, 1981; Jacob and Duncan, 1981). Figure 2 shows the slope conductance of the lens a.s a function of the lens membrane potential both in control medium (0) and in the presence of barium (A). Barium obviously competes with calcium for the control of the linear (go) and non-linear conductances. The effect on g, can be interpreted as a reduction in potassium permeability since the voltage depolarises in barium (4). Barium also reduces the extent to which the conductance
Voltcge
FIG. 2. The slope
conductance solution (O), in the presence (a). The arrows indicate the
(mV)
of the lens as a function of lens membrane potential of 2 mw-barium with calcium (A) and in 2 mM-barium resting voltage in each solution.
measured without
in control calcium
DIVALENT
CATION
CONTROL
OF
LENS
PERMEABILITY
599
is sensitive to voltage and hence barium also appears to inactivate membrane voltage-sensitive potassium channels. Removing calcium, but maintaining barium at 2 mM (a) does not lead to any great change in the conductance-voltage relationship and hence barium has the ability to replace calcium in the control of sodium permeability. In the presence of barium the shape of the voltage-clamp current transient assumes a time-dependency (Fig. 3) similar to that observed in calcium-free solution (Jacob and Duncan, 1981).
FIG. 3. Current transients (upper trace) obtained in response to 15 mV command clamp pulses. The initial spike represents charging membrane capacitance and in a linear network system, this should be followed by an exponential decline to a plateau value (see fig. 4 in Jacob and Duncan, 1981). However, in the case of the depolarising command, the spike is followed by an increasing phase. representing a time-dependent conductanoe change. This phase is absent from hyperpolarising commands.
The addition of 2 mM-calcium, -magnesium, -strontium or -manganese in the presence of 2 mm-calcium had little effect either on the resting potential and conductance [Fig. l(a)] or on the conductance-voltage relationship of the lens. However, of these ions only strontium and manganese have the ability to replace calcium in the control of lens voltage and conductance [Fig. l(b); Fig. 41. The slight increase in conductance and hyperpolarisation of lens voltage that occurs can be explained by a small increase in lens potassium permeability in the presence of these ions. There is no evidence, however, for a change in sodium conductance. Although cobalt, like manganese and strontium, does not compete with calcium, it cannot replace calcium to the same extent. A small depolarisation of membrane potential accompanies the increase in membrane conductance [Fig. l(b)] and hence in this case a small increase in sodium conductance occurs in the absence of calcium. The conductance-voltage curves were, however, essentially similar in form to those obtained with strontium in the absence of calcium (Fig. 4). Magnesium is interesting as it appears to be the only divalent cation tested that does not compete with calcium [Fig. l(a)] and yet does not replace calcium in the control of sodium and potassium permeability. The conductance-voltage curves obtained in the presence of 2 mM-
T. J. C. JBC’OB dh’U G. DCiY\‘C’BS
600
.
Voltage
FIG. 1. The effect (A), control (0).
on
the lens conductance-voltage
(mV)
curve
of replacing
calcium
with
2 mwstrontium
magnesium [and the absence of calcium had essentially the same form as those obtained in calcium-free solution (Jacob and Duncan, 1981)]. The addition of 2 m&r-cadmium to the bathing medium caused an immediate hyperpolarisation of approximately 6 mV [Fig. l(a)], accompanied by an 80 y. increase in membrane conductance within 20 min. Return to control solution initiated a biphasic voltage response. Within the first few minutes a depolarisation of approximately 10 mV was observed and this was followed by a slow hyperpolarising phase until the initial control voltage was obtained. The resistance, however, recovered monophasically to a value near the original control level. The removal of calcium from the cadmium Ringer caused a depolarisation of approximately 8 mV, so that the initial control value is not only reached; but slightly exceeded [Fig. l(b)]. The conductance increases still further and hence the observed depolarisation is probably due to an increase in sodium permeability. In the presence of calcium, cadmium therefore acts to increase membrane potassium permeability and, as the increase appears to be the same throughout the voltage range (Fig. 5), cadmium probably does not interfere with membrane voltage-sensitive channels. Cadmium also does not appear to interfere with calcium regulation of sodium permeability. In the absence of calcium, cadmium still acts to increase potassium permeability and it can to a large extent substitute for calcium in the control of sodium permeability. However, the control is not complete as there is a small increase in conductance and depolarisation of membrane potential when calcium is removed from the test solution containing cadmium. Nickel, zinc and copper had similar effects in that they all depolarised the membrane potential and increased membrane conductance when added to the control medium (Fig. 6). Again the conductance was increased throughout t.he voltage range and hence mainly the linear elements of the sodium and potassium conductances are involved.
DIVALENT
CATION
OF
LENS
PERMEABILITY
601
.
c t
CONTROL
,
;;i,:, -100
,
Voltage
I -910
,
-60
-00
--L
I
?I
I
-90
ICI
I
-70
CrnV)
Voltage
FIG. 5
FIG.
-50
, -30
( mV )
6
FIG. 5. The effect on the lens conductance-voltage curve of adding 2 mx-cadmium (A) to the control solution (0). Note that in this case, the data were obtained after 6 min since the conductance increase obtained after 15 min is so large [Fig. l(a)] that the conductance-voltage curve cannot be mapped out over the whole range. FIG. solution
6. The (0).
effect
on the
lens
conductance-voltage
curve
of adding
2 miwnickel
(A)
to the
control
60, control Gi u-2 g -
8
. ! . ..‘a.
50 .
40
,Fj 2
30
5
20 t
*
B 07
.
F
!O 01 0
.
copper ( 10-5M
. 1. .
.
l* I IO
I 20
I 30
I 40
I 50 Time
I 60
I 70
I 80
I 90
I 100
(min)
FIG. 7. The effect on the lens conductance of adding 10 pwcopper. Note that in this case the lens voltage was clamped at the resting value ( - 80 mV) and the conductance measured by superimposing small test command voltage pulses (see Jacob and Duncan. 1980, for further details).
T. J. C. JACOB
602
AND 6. DUNCAS
Only [Fig.
copper, however, was effective in the micromolar range of concentrations 71. Figure 7 shows the time-course of the conductance increase caused by 10,LhM CL?+, with the voltage clamped at - 80 mV. There is a very rapid increase, from 118 x 10v5 to 46 x 10e5 S within 30 min, and this is followed by a gradual but incomplete recovery within the course of the experiment (100 min). Zinc, when added to the control Ringer at a concentration of 2 mM, caused a depolarisation of 138 + 2.0 mV (n = 3) after 30 min. The conductance, when measured at the resting potential, was found to have increased from 13.8+ 2.0 to 62.7 + 8.2 x 10P5 S (n = 3). However, when the conductance is measured over a range of voltages, the increase was restricted to those potentials in the hyperpolarising direction. This fact, taken together with the observed depolarisation, implies an increase in sodium permeability alone. The exponentially rising phase of the conductance remains unchanged as in Fig. 8. When calcium is removed from the Zn2+ Ringer, there is a further depolarisation and increase in resting conductance [Fig. l(b)]. The conductance-voltage relationship now shows a form chajracteristic of that observed in calcium-free solution (Jacob and Duncan, 1981). The decrease in the slope of the curve indicates that an inactivation of membrane voltage-sensitive potassium channels has occurred.
-80
-60
Voltoge
FIG. 8. The solution (0).
effect
on the conductance-voltage
-40
-20
(mV 1
curve
of adding
lO par-copper
(A)
to the control
4. Discussion Jacob and Duncan (1981) have shown that calcium controls both sodium and potassium permeability and it appears now that the control occurs at two separate membrane sites as certain cations are able to influence the two permeabilities separately. For example, cadmium increases potassium permeability in the presence of calcium, without altering the sodium conductance, while barium decreases potas-
DIVALENT
CATIOK
COSTROL
OF
LESS
PERMEABILITY
603
sium permeability without apparently influencing sodium movement [Fig. l(a)]. Nickel, zinc and copper, however, probably increase both sodium and potassium conductance in the presence of calcium and the effects of these ions are likely to be particularly damaging. When the lens conductance is mapped out as a function of voltage, the potassium permeability can be seen to have both linear and non-linear components (Figs 2, 4, 5, 6 and 8). In the presence of calcium, only barium appears to influence both the linear and non-linear conductances, reducing them both. These data extend the previous work of Delamere, Duncan and Paterson (1980) and Delamere, Paterson, Duncan and-Holmes (1980) who found that barium reduced the 42K efflux from the lens and also reduced the extent to which the membrane conductance was sensitive to voltage. They did not, however, map the conductance out over a very wide range. Strontium and manganese are the only divalent ions able to perform the role of calcium in controlling sodium and potassium permeability [Fig. 1 (b)], perhaps because their ionic radii are similar to that of calcium (Fig. 1). However, when the conductance is mapped out over the whole voltage range (Fig. 4) it can be seen from the overall conductance increase that the substitution is not perfect. It is quite possible that divalent cations achieve their effect by interaction with lens membrane surface charge. The most obvious consequence of the mechanism whereby the ions affect the surface potential by screening rather than by specific binding is that equimolar substitution of Ca 2+ by another divalent ion should produce no change in membrane properties (Schauf, 1975). This appears to be true only for Sr2+ and Mn2+ and thus the other ions must achieve their relative effects by binding to the membranes. The other two ions tested with relatively large ionic radii, namely cadmium and barium, had opposite, but specific, effects on potassium conductance. The ionic radius of barium is in fact very similar to that of potassium (1.36 A) and it is possible that barium enters the lens via the K channel and operates a block of outward K-currents from the inside, accounting for the reduced conductance at depolarised potentials (Fig. 4). Barium is a powerful blocker of potassium conductance in frog skeletal muscle (Standen and Stanfield, 1978), Aplysia neurones (Adams and Gage, 1980) and non-excitable membranes such as frog gastric mucosa (Pacific0 et al., 1969). Cobalt has been found to cause a cataract in rabbits and rats (when administered in vivo) resembling that induced by alloxan (Alagna and d’Aquino, 1956). However, when presented to the amphibian lens in vitro, no changes in membrane characteristics were observed in the short term [Fig. l(a)]. Cobalt and manganese were able to substitute, to a large extent, for calcium [Fig. l(b)] in the control of lens permeability, whereas in other tissues they have inhibitory effects on conductance (Hagiwara and Takahashi, 1967; Adams and Gage, 1980). However, in both of the tissues involved, the inhibitory effect is a consequence of Co2+ and Mn2+ blocking voltage-sensitive cc&&n channels and these probably either do not exist in the lens membranes or do not significantly contribute to the control of conductance. Similarly, Cd2+ inactivates potassium currents in vertebrate neurones (Adams, Constanti; Brown and Clark, 1982) and LimuEus photoreceptors (Lisman, Fain and Swan, 1978) where again a primary blocking of a calcium channel is involved. In the lens, Cd2+ increases potassium conductance probably by binding to the surface membrane directly (Fig. 5). Zinc, nickel and copper ions are all recognised as cataractogenic (Heydenreich, 1966; Hollwich et al., 1975) and all produce similarly disruptive effects on the lens, namely a depolarisation of lens membrane potential and increase in conductance (Figs 1 and 6). However, the copper-induced changes occur at much lower concentrations than
T. J. C. JACOB
604
AND
G. DCNCAS
the others (Fig. 7) and indeed at concentrations near the physiological range. The copper concentration in normal human plasma is ZO,UM, of which about 90% is tightly bound to proteins (Henkin, Schulman, Schulman and Bronze&, 1973; Sarka,r and Kruck, 1966). The normal concentration of copper in the aqueous humour is 2 ,UM, but this can increase to 40 ,/AM in Wilson’s disease (Gerhard, 1966). Coulter, Oliver, Whitener, Collie and Engelke (1978) have shown that the cultured rat lens is extremely sensitive to copper in the bathing medium and have pointed out that changes in aqueous humour amino acids may play a role in copper toxicity as they carry out most of the chelating in the aqueous humour. Since copper can overturn the regulatory influence of calcium when it is present in concentrations as low as 10e6 M [Fig. i(a)], it is not surprising that Bellows and Bellows (1975) recommend surgical removal of intralenticular copper-containing particles. For any other particulate body, surgical removal from the lens is never recommended. The electrophysiological methods described here provide a means not only of investigating the mechanism of copper toxicity in great detail, but it also provides an accurate and rapid method of screening certain drugs (Coulter et al., 1978) that may be able to overcome the toxic effects. REFERENCES Sdams, D. J. and Gage, P. W. (1980). Divalent ion currents and the delayed potassium conductance in an Aplysia neurone. J. Physiol. 304, 297-313. Adams, P. R., Constanti, A., Brown, D. A. and Clark, R. B. (1982). Intracellular Ca2+ activates
a fast voltage-sensitive
K+ current
in vertebrate
(London) 296, 746-9. Alagna, 0. and D’Aquino, S. (1956). Augenveranderungen
sympathetic
neurones.
Nature
durch Kobaltchlorid. Arch. Ottal. 60, 5-29. Bellows, J. G. (1944). Cataract and Abnormalities of the Lens. C. V. Mosby, St. Louis, MO. Bellows, J. G. and Bellows, R. T. (1975). Traumatic cataract. In Cataract and Abnormalities of the Lens (Ed. Bellows, J. G.). Pp. 265-72. Grune and Stratton, N.Y. Bernardini, G. and Peracchia, C. (1981). Gap junction crystallization in lens fibres after an increase in cell calcium. Invest. OphthaZmoZ. Vis. Sei. 21, 291-9. Bernardini, G., Peracchia, C. and Venosa, R. A. (1981). Healing-over in rat crystalline lens. J. Physiol. 320, 187-92. Coulter, J. B., Oliver, 6. S., Whitener, C. M., Collie, 6. J. and Engelke, J. A. (1978). Toxic effects of copper on cultured rat lenses. Exp. Eye Res. 26, 547-54. Delamere, N. A., Duncan, G. and Paterson, C. A. (1980). Characteristics of voltage-dependent conductance in the membranes of a non-excitable tissue: the amphibian lens. J. Physiol. 308, 49-59. Delamere, N. A. and Paterson, C. A. (1981). Hypocalcaemic cataract. In Mechanisms of Cataract Formation in the Human Lens (Ed. Duncan, G.). Academic Press, London. Delamere, N. A., Paterson, C. A., Duncan, G. and Holmes, D. L. (1980). Relative roles ofCaZf, Sr2+, Ba2+ and Mg2+ in controlling lens permeability characteristics. Cell Calcium 1,81-90. Duncan, G., Patmore, L. afnd Pynsent, P. B. (1981). Impedance of the amphibian lens. J. Physiol. 312, 17-27. Duke-Elder, S. (1969). System of Ophthalmology XI. C. V. Mosby, St. Louis, MO. Gerhard, J. P. (1966). Study of the copper of aqueous humor. DOG. OphthaEmoZ. 20, 104410. Gilbert, D. L. and Ehrenstein, G. (1969). Effect of divalent cations on potassium conductance of squid axons: determination of surface charge. Biophys. J. 9, 447-63. Hagiwara, S. and Takahashi, K. (1967). Surface density of calcium ions and calcium spikes in the barnacle fibre membrane. J. gen. Physiol. 50, 583-601. Henkin, R. I., Schulman, J. D., Schulman, C. B. and Bronzert, D. 8. (1973). Changes in total non-diffusible anddiffusible plasmazinc and copper during infancy. J. Pediatr. 82,831-7. Heydenreich, A. (1966). Chemisch-toxische Schaden der Augen. IiZin. Monatsbl. Augenheilkd. 149. 145-60.
DIVALENT
CATION
CONTROL
OF
LEKS
PERMEABILITY
606
Hollwich, F., Boateng, A. and Kolck; B. (1975). Toxic cataract in Cataract and Abnormalities of the Lens (Ed. Bellows; J. G.). Pp. 23943. Grune and Stratton N.Y. Jacob, T. J. C. a,nd Duncan, G. (1980). Osmotic influences on lens membrane characteristics. Exp. Eye Res. 31, 50512. Jacob, T. J. 6. and Duncan, G. (1981). Calcium controls both sodium and potassium permeability of lens membranes. Exp. Eye Res. 33, 8&93. Lisman, J., Fain, G. and Swan, M. (1978). Voltage-dependent conductances in Limulus ventral photoreceptors. Biol. Bull. 155, 4534. Owers, J. and Duncan, G. (1979). The viability of the bovine lens in organ culture. Exp. Eye Res. 28, 73S42. Pacifico, A. D., Schwartz, M.; Mackrell, T. N., Spangler, S. G., Sanders, S. S. and Rehm, W. S. (1969). Reversal by potassium of an effect of barium on the frog gastric mucosa. A,m. J. Physiol. 216, 536-41. Patmore, L. and Duncan, G. (1979). A TEA-sensitive component in the conductance of a non-excitable tissue (the amphibian lens). Exp. Eye Res. 28, 349-52. Sarkar, B. and Kruck, T. P. A. (1966). Copper-amino acid complexes in human serum. In The Biochemistry of Copper (Eds Peisach, J., Aisen, P. and Blumberg, W.). Pp. 183-196. Academic Press, N.Y. 9chauf, C. L. (1975). The interactions of calcium with Myxicola giant axons and a description in terms of a simple surface charge model. J. Physiol. 248, 3, 613-25. Standen, N. B. and Stanfield, P. R. (1978). A potential and time-dependent blockade of inward rect.ification in frog skeletal muscle fibres by barium and strontium ions. J. Physiol. 280, 169-91.