Age- and Disease-Related Changes of Calcium Channel–Mediated Currents in Human Müller Glial Cells

Abstract

purpose. To determine whether the expression of voltage-gated Ca²⁺ channels in human Müller glial cells changes during normal aging and in cells from patients with proliferative vitreoretinopathy (PVR).

methods. Müller cells were enzymatically isolated from retinas of healthy donors and from excised retinal pieces of patients with PVR, and the whole-cell, voltage-clamp technique was used to characterize the current densities of transient, low-voltage–activated calcium channels and of sustained, high-voltage–activated calcium channels, respectively. To obtain maximal currents through both channel types, Na⁺ ions were used as the charge carrier.

results. During normal aging, Müller cells developed a hypertrophy, as indicated by an increase of the cell membrane capacitance. The mean membrane capacitance of cells from aged donors (≥ 60 years old) was elevated by 25% compared with cells from younger donors. The hypertrophy was not accompanied by a changed density of low-voltage–activated currents, whereas the density of the high-voltage–activated currents was enhanced by 76%. The density of the high-voltage–activated currents increased in correlation with the increase of the cell membrane capacitance and with the age of the donors. In the case of PVR, Müller cells displayed a strong hypertrophy accompanied by a downregulation of both current types by approximately 65%.

conclusions. Both normal aging and PVR cause a gliotic reactivity of human Müller cells, as indicated by their hypertrophy. The type of reactivity, however, differs between the two conditions. Normal aging is accompanied by an increased expression of voltage-gated Ca²⁺ channels, whereas in PVR Ca²⁺ channel expression is decreased.

Müller glial cells from the human retina were previously reported to change their membrane permeability significantly during several diseases of the eye, for example, during proliferative vitreoretinopathy (PVR). The dominant type of ion channels in Müller cell membranes of healthy human donors, the inwardly rectifying K⁺ channel, is significantly downregulated or even absent in cells from patients with PVR, ^{1 2} whereas voltage-gated fast Na⁺ currents are strongly upregulated in their densities. ³ The downregulation of the K⁺ channels is accompanied by a significant depolarization of Müller cells from patients with PVR compared with cells from healthy donors. ^{1 2} Moreover, it has been found that Ca²⁺-activated K⁺ channels of big conductance (BK) show a significantly higher activity (increased open probability and increased current amplitude) at the resting membrane potential in cells from patients with PVR than in cells from healthy donors. ² It was speculated that the enhanced BK channel activity in cells from PVR retinas is due to both the depolarization of the cells and to an increased intracellular Ca²⁺ level. ² Such an increased intracellular Ca²⁺ concentration may be generated by several different mechanisms, among others by an enhanced activity of voltage-insensitive cation channels or of voltage-gated Ca²⁺ channels; both types of channels were previously described to be activated by certain growth factors such as the basic fibroblast growth factor. ^{4 5} Moreover, both the BK channels ⁶ and the voltage-gated Ca²⁺ channels ⁵ have been implicated in the maintenance of growth factor–induced proliferative activity of cultured human Müller cells. Therefore, the aim of the present study was to investigate whether the expression of voltage-gated Ca²⁺ channel–mediated currents is changed in cells from patients with PVR. In comparison to possible diseased-induced alterations, the age-dependent changes of the Ca²⁺ channel currents were investigated. 

In cultured human Müller cells, the presence of at least two different types of Ca²⁺ channel currents was described according to their activation kinetics: transient low-voltage–activated (LVA) and long-lasting (L-type) high-voltage–activated (HVA) currents. ^{5 7} In the majority of freshly isolated human Müller cells, however, no resolvable Ca²⁺ channel currents were observed when the extracellular solution contained Ca²⁺ ions at a concentration of 2 mM. ⁸ Even when the external Ca²⁺ concentration was elevated to 60 mM, the amplitude of the Ca²⁺ currents did not elevate significantly. When Ba²⁺ ions (50 mM) were used as the charge carrier, the currents through Ca²⁺ channels still remained relatively small. To achieve a reliable recording of voltage-gated Ca²⁺ channels in freshly isolated cells, the amplitude of the currents was maximized by using a particular feature of these channels, that is, to carry fluxes of monovalent cations when divalent cations are largely absent in the extracellular solution. ^{9 10 11} Therefore, the Na⁺ currents through Ca²⁺ channels were recorded in extracellular solutions that lacked divalent cations. ⁸  

Methods

All tissue was used in accordance with applicable laws and with the Declaration of Helsinki. The use of human material was approved by the ethics committee of the Leipzig University Medical School. Postmortem eyes from organ donors with no reported history of eye disease (mentioned in the text as “healthy donors”) were supplied within 12 and 24 hours after death. Retinal tissue from patients was obtained from vitreoretinal surgery between 2 and 6 hours after the tissue was excised. Retinal pieces were removed from eyes when partial retinectomies were necessary to relieve traction due to PVR. Müller cells were isolated using papain- and DNase-containing solutions as described previously. ¹  

Electrophysiological Recordings

Whole-cell records ¹² were made at room temperature (22–25°C) using an EPC 7 amplifier (List Electronics, Darmstadt, Germany) and the TIDA 5.72 computer program (HEKA Elektronik, Lambrecht, Germany). The sampling rate was 20 kHz; high frequencies > 4 kHz were cut off. The series resistance (13–16 MΩ) was compensated by 30%. Patch pipettes were pulled from borosilicate glass and had resistances between 3 and 7 MΩ. 

The Ca²⁺ channel–mediated currents of monovalent cations were evoked by step protocols (V_h −40 mV, depolarizing voltage steps from a 500 ms prepulse). For activation of the sum currents, depolarizing voltage steps were applied to voltages between −100 and +20 mV, with an increment of 10 mV, after prepulses to −120 mV. For selective HVA current activation, steps were applied after prepulses to −70 mV. The LVA currents were determined as difference of both records. Cell membrane capacitance was measured by the integral of the uncompensated charging transient in response to a voltage step from −80 to −90 mV. For recording the capacitive artifact, the sampling rate was 30 kHz, and the frequencies above 10 kHz were cut off. 

Measurements of time-dependent changes of the Ca²⁺ channel–mediated currents showed that after disruption of the membrane, dialysis of the cell interior was completed within 1 minute, and thereafter, the current amplitudes of the LVA currents remained stable for at least 10 minutes Thus, any run-down of the LVA channel–mediated currents was considered to be negligible within this period. HVA currents, however, continuously decreased in their amplitudes during the recording period. Therefore, the current amplitudes were measured between the first and the third minute after disruption of the cell membrane. In the mean, the peak amplitude of the HVA currents decreased by 14.4 ± 2.4% between the first and the third minute after establishing the whole-cell configuration. There was no significant difference in the time course of the run-down among cells from different donor populations investigated. 

Solutions

Ca²⁺ currents were recorded using a bath solution containing (in mM) 113 NaCl, 2 CaCl₂, 1 MgCl₂, 10 HEPES, and 11 glucose (pH 7.4 adjusted with Tris-base). For recording Ba²⁺ currents, NaCl was replaced by 50 mM BaCl₂ and 63 mM tetraethylammoniumchloride (TEACl). Na⁺ currents through Ca²⁺ channels were recorded in a bath solution that contained (in mM) 113 NaCl, 10 HEPES, 11 glucose, and 1 EGTA (pH 7.4). The pipette solution consisted of (in mM) 10 NaCl, 130 CsCl, 1 CaCl₂, 2 MgCl₂, 10 EGTA, and 10 HEPES (pH 7.2), resulting in approximately 13 nM free Ca²⁺. Nimodipine and flunarizine were from Calbiochem (Bad Soden, Germany); tetrodotoxin was from Alomone Laboratories (Jerusalem, Israel). All other substances were from Sigma (Deisenhofen, Germany). Lipophilic drugs were dissolved in dimethylsulfoxide (Sigma) as vehicle; control records showed that the vehicle alone had no effects on the currents. Drugs were applied by changing the perfusate within the recording chamber. 

Data Presentation

The traces were not leak-subtracted. When calculations were made, the leak current was determined at a voltage step from −100 to −110 mV and was subtracted. Statistical analysis (unpaired Student’s t-test, regression analysis) and curve fits were made using the Prism program (Graphpad Software, San Diego, CA). To determine the age- and disease-dependent changes of currents, 5 to 13 cells per donor were recorded; in the further statistical analysis, only the mean values of the cells from each donor were used. The data are expressed as means ± SD. 

Results

Currents through Ca²⁺ Channels

In K⁺-free bath solution containing Ca²⁺ (2 mM) and Mg²⁺ ions (1 mM), the vast majority of freshly isolated human Müller cells investigated (∼85%) displayed no resolvable inwardly directed currents when the membranes were stepped to depolarized potentials (Fig. 1A ). However, after removing the divalent cations from the bath solution, the cells become permeable to monovalent cations, and large inwardly directed, voltage-dependent currents were observed. Because these currents were found to be largely sensitive to extracellular application of the Ca²⁺ channel blockers flunarizine (1 μM) and nimodipine (10 μM) but virtually insensitive to exposure of tetrodotoxin (10 μM) as described recently, ⁸ it is concluded that these currents reflect ion fluxes through voltage-gated, Ca²⁺ channels, as previously shown in a variety of other cell types. ^{9 10 11} The ionic nature of the inwardly directed currents was studied by applying Na⁺-free external solution. After equimolar replacement of the Na⁺ ions in the bath solution by choline, the inwardly directed currents disappeared, and only outwardly directed “leak currents” were observed, probably representing currents of Na⁺ and Cs⁺ ions from the pipette solution (Fig. 1B) . Thus, Na⁺ ions flow through voltage-dependent, Ca²⁺ channels when the external solution is largely free of divalent cations. 

In cells from three donors, however, it was possible to record very small inwardly directed currents when the external solution contained 2 mM Ca²⁺. The mean voltage-dependence of Ca²⁺ channel–mediated currents in eight of these cells is shown in Figure 1C . The currents were recorded using the following ions as charge carrier: Na⁺ (113 mM), Ca²⁺ (2 mM), and Ba²⁺ (50 mM). The activation of the currents shifted toward more positive voltages the more divalent cations were present in the bath solution, likely as a consequence of the alteration of membrane surface charges. The peak Na⁺ currents were found at −37 mV, the peak Ca²⁺ currents were at −5 mV, and the peak Ba²⁺ currents were at +15 mV. The peak Ca²⁺ current density was very small (0.4 pA/pF) compared with the peak Na⁺ current density (7.1 pA/pF) and also to the peak Ba²⁺ current density (2.0 pA/pF). 

The currents through Ca²⁺ channels were composed of two components: a transient LVA component, and a long-lasting HVA component. Both current components could be observed when Na⁺, Ca²⁺, or Ba²⁺ ions were used as charge carrier and could be separated by changing the prepulse potentials (for Na⁺ ions: −120 and −70 mV, respectively), as indicated by the example shown in Figure 2A . The transient Na⁺ current component had a mean activation threshold of −81.7 ± 3.8 mV and peaked at −46.6 ± 6.1 mV (n = 10; Fig. 2B ). The noninactivating, sustained current component activated at −61.4 ± 5.7 mV and showed a maximum at −25.6 ± 6.0 mV. Both current components were displaced to more positive membrane potentials (by approximately 20 mV) when Ca²⁺ ions were used as the charge carrier (Fig. 2C) and by approximately 40 mV when Ba²⁺ ions were used as the charge carrier (Fig. 2D) . 

Cell Membrane Capacitance

Both normal aging and PVR were accompanied by changes in the membrane capacitance of Müller cells. As indicated in Figure 3 , Müller cells from healthy aged donors (older than 60 years) displayed a mean membrane capacitance that was significantly greater (by 25.6%) than that of Müller cells from healthy donors younger than 60 years (P < 0.01). Müller cells from PVR retinas displayed an even stronger increase of their membrane capacitance, independent of the age of the donors. As the membrane capacitance is dependent on the surface area of a cell, we conclude that Müller cells from both aged and diseased retinas display a hypertrophy. 

Age-Dependent Changes of Ca²⁺ Channel Currents

For the evaluation of age- and disease-dependent alterations of voltage-gated Ca²⁺ channels, Na⁺ currents were recorded in divalent cation-free external solution. As shown in Figure 4A , the density of the sum currents increased slightly, but not significantly in the course of aging. Though the LVA currents of younger and older donors were comparable in their densities, the HVA current density was significantly enhanced with aging, with a mean increase by 76.4% in cells from donors older than 60 years compared with cells from younger donors (P < 0.05). The HVA current density increased in correlation with the cell membrane capacitance (r = 0.66, n = 15, P < 0.01) and in correlation with the age of the donors (r = 0.59, n = 15, P < 0.05). The specific increase of the HVA currents during aging resulted in an enhancement of the ratio between the peak HVA and LVA currents. Although in cells from younger donors this ratio was 0.54, in cells from older donors it was 0.88 (P < 0.05). The ratio was found to increase in correlation with the age of the donors (r = 0.62, n = 15, P < 0.05; Fig. 4B ). The age-dependent increase of the HVA current density was not accompanied by significant changes of the voltage dependence of the activation threshold or of the peak current (Fig. 4C) . 

PVR-Related Alterations of Ca²⁺ Channel Currents

In cells from PVR retinas, both the LVA and the HVA currents were found to be strongly depressed (Fig. 4A) . The mean LVA current was reduced by 63.0% in cells from PVR retinas compared with cells from healthy donors of the same age (P < 0.001). The mean HVA current was decreased by 66.7% (P < 0.001). This decrease of the peak HVA current occurred independent on the increase of the cell membrane capacitance, that is, no inverse correlation was found between both parameters (not shown). Because of the similar reduction of both types of currents, the ratio between the peak HVA and LVA currents was found to be in the same range as in cells from healthy donors of the same age (0.88 in cells from older patients; Fig. 4B ). The decrease of the HVA current density was not accompanied by alterations of the voltage dependence of the activation threshold or of the peak current (Fig. 4C) . Figure 1B illustrates records of the sum and of the HVA currents in one cell of a PVR retina. Generally, Müller cells from PVR retinas displayed a strongly increased expression of fast transient Na⁺ currents, as described earlier. ³ When prepulses to −120 mV were applied, fast transient Na⁺ currents were recorded as fast inwardly (downwardly) directed currents that time-dependently inactivated after a latency of approximately 15 to 25 ms. When prepulses to −70 mV were applied, no fast transient Na⁺ currents (and no LVA channels) were activated; thus, the records reflect only the activity of HVA Ca²⁺ channels (Fig. 1B , right side). 

To rule out that the enhanced expression of fast transient, tetrodotoxin-sensitive Na⁺ channels may artificially decrease the amplitude of Na⁺ currents through Ca²⁺ channels, tetrodotoxin was tested in 11 cells from 3 diseased retinas. The amplitudes of the LVA and HVA currents remained unaffected when tetrodotoxin (10 μM) was added to the bath solution, whereas the currents through fast Na⁺ channels were greatly decreased (not shown). ^{3 8} Moreover, in the case of two diseased retinas, the peak current densities in cells displaying fast Na⁺ currents were compared with those in cells lacking fast Na⁺ currents (the lack of fast Na⁺ currents was proven during exposure to Ca²⁺- and Mg²⁺-containing bath solution). The sum currents and the HVA currents in both cell populations displayed similar densities; for the sum currents, the values were 2.5 ± 1.1 pA/pF for 10 cells lacking fast Na⁺ currents and 2.7 ± 0.7 pA/pF for 7 cells displaying fast Na⁺ currents (not significant). It is thus concluded that the decrease of the Ca²⁺ channel–mediated currents is independent on the increase of the fast transient Na⁺ currents and is also no artifact caused by the recording conditions. 

Discussion

At physiological Ca²⁺ and Mg²⁺ concentrations in the bath solution (2 and 1 mM, respectively), we found no resolvable currents through Ca²⁺ channels in the vast majority of the cells investigated. In the remaining cells, the currents were too small to be reliably studied in any detail (Fig. 1C) . This corresponds to previous findings on cultured astrocytes where Ca²⁺ currents were usually undetectable. Addition of agents that increase intracellular cAMP ¹³ or coculturing with neurons ¹⁴ were necessary to record Ca²⁺ currents in cortical astrocytic cultures. ¹⁵ The difficulty to demonstrate Ca²⁺ currents may also be a reason for the controversies about the functional significance of voltage-gated Ca²⁺ channels in astrocytes in situ, for example, in hippocampal astrocytes. ^{16 17} Ca²⁺ channels in freshly isolated human Müller cells were electrophysiologically detected by two methods: either by Na⁺ currents in divalent cation-free bath solutions or by Ba²⁺ currents. Using Na⁺ currents, measurable Ca²⁺ currents were detected in every investigated cell. Using Ba²⁺ ions, the currents were relatively small. Therefore, to determine maximal currents through Ca²⁺ channels, Na⁺ ions were used as the charge carrier. 

Reactive astrocytes of the brain upregulate the expression of L-type Ca²⁺ channels, ¹⁸ and an elevation of the intracellular Ca²⁺ concentration is crucial for the induction of gliosis. ¹⁹ We now show that Müller cells from old donors display a hypertrophy that is accompanied by a specific increase of the density of HVA channels (Fig. 4A) . In addition to the hypertrophy, the increased density of HVA channels may indicate that Müller cells from old donors may undergo a slight but demonstrable reactivity. An age-dependent volume increase was already described for Müller cells of the rat. ²⁰ Hypertrophy was also observed in Müller cells of an animal model of age-related retinal degeneration. ²¹  

By contrast, in Müller cells of patients with PVR the voltage-gated Ca²⁺ channel–mediated currents are downregulated. The downregulation of Ca²⁺ channels and the accompanying upregulation of fast transient Na⁺ channels described previously ³ may indicate that Müller cells undergo different types of reactivity during normal aging and during PVR, respectively. The increased activity of Ca²⁺-activated K⁺ (BK) channels found in Müller cells during PVR ² has been ascribed to the more positive resting membrane potential compared with that in cells from healthy donors. ^{1 2} However, it cannot be ruled out that a higher Ca²⁺ influx through voltage-gated Ca²⁺ channels participates in the stimulation of BK channel activity because the more positive and apparently unstable membrane potential should increase the opening probability of voltage-gated channels. The reason for the downregulation of voltage-gated Ca²⁺ channels in Müller cells during PVR is unclear. One may assume that a downregulation of these channels may protect the Müller cells against cytotoxic intracellular Ca²⁺ overload. The increased expression of fast Na⁺ channels may serve to support the activity of the Na⁺/K⁺-ATPase ²² and, therefore, to maintain a more negative membrane potential. 

At the present time, the functional role(s) of Ca²⁺ channels in Müller cells remains a matter of speculation. We suggest that these channels may conduct Ca²⁺ ions only when, simultaneous to the membrane depolarization, certain second messengers (after stimulation by growth factors or by certain neurotransmitters) alter the channel conformation. Probably, these channels are normally impermeable for Ca²⁺ to prevent an excess Ca²⁺ entry into glial cells when the extracellular K⁺ concentration is elevated during regular neuronal activity. Na⁺ currents through Ca²⁺ channels are unlikely to occur under physiological conditions. However, there are some indications that under certain pathophysiological conditions; for example, after lipid peroxidation, voltage-gated Ca²⁺ channels may become permeable for Na⁺ ions. ²³  

Conclusions

In conclusion, both normal aging and PVR are accompanied by a hypertrophy of Müller glial cells in human retinas. Hypertrophy is one of the characteristic indicators of reactive glial cells. The type of reactivity, however, is different at the two conditions. During normal aging, the increase of the membrane area is accompanied by an increased expression of voltage-gated Ca²⁺ channels, especially of HVA channels. During PVR, the newly synthesized plasma membrane apparently does not contain Ca²⁺ channels, but contains a higher density of voltage-gated Na⁺ channels. ³ The reason for this switch from Ca²⁺ channel expression to Na⁺ channel expression during PVR remains to be elucidated. 

 Supported by grants from the Bundesministerium für Bildung, Forschung und Technologie (BMB+F), Interdisciplinary Center for Clinical Research at the University of Leipzig (Grant 01KS9504, Project C5), and by Grant Re 849/8–1 from the Deutsche Forschungsgemeinschaft, Bonn, Germany.

 Submitted for publication January 26, 2000; revised March 27, 2000; accepted April 11, 2000.

 Commercial relationships policy: N.

 

Figure 1.

 

Ca²⁺ channel–mediated, whole-cell currents in freshly isolated human Müller cells. (A) Example of a whole-cell record in a cell from a healthy donor (60 years old). When the K⁺-free bath solution contained Ca²⁺ (2 mM) and Mg²⁺ ions (1 mM), no resolvable inwardly directed currents were observed. After removing the divalent cations from the bath solution and adding 1 mM EGTA, large inwardly directed, voltage-dependent currents were evoked by depolarizing voltage steps (from −100 to +10 mV, increment 10 mV) after prepulses to −120 mV. (B) Example of whole-cell records in one cell from a patient (72 years old). The inwardly directed currents were evoked by depolarizing voltage steps to test potentials between −100 and +20 mV (increment 10 mV) after prepulses to −120 mV (left traces) or to −70 mV (right traces). The currents were recorded in extracellular solutions containing Na⁺ ions (top traces) and lacking Na⁺ ions (bottom traces), respectively. (C) Mean voltage dependence of the peak Ca²⁺ channel–mediated, whole-cell currents in eight other cells from healthy donors, depending on the charge carrier used: Na⁺, Ca²⁺, and Ba²⁺ ions. Depolarizing voltage steps were applied after prepulses to −120 mV. To record Na⁺ currents, the external solution contained (in mM) 113 NaCl, 10 HEPES, 11 glucose, and 1 EGTA. To record Ca²⁺ currents, a bath solution was used that contained (in mM) 113 NaCl, 2 CaCl₂, 1 MgCl₂, 10 HEPES, and 11 glucose. The Ba²⁺ currents were recorded using a bath solution containing (in mM) 50 BaCl₂, 63 TEACl, 2 CaCl₂, 1 MgCl₂, 10 HEPES, and 11 glucose (pH 7.4 adjusted with Tris-base).

Figure 2.

 

Human Müller cells express two different types of currents through voltage-gated Ca²⁺ channels. (A) Example of inwardly directed Na⁺ currents in one cell from a healthy donor (60 years old). Depolarizing voltage steps (increment 10 mV) were applied from prepulses to −120 mV (left traces) and to −70 mV (middle traces), respectively. Voltage steps after prepulses to −70 mV evoked a sustained, noninactivating, HVA current. The difference between both records reveals the presence of a transient LVA current (right traces). (B) Mean peak current density–voltage (I-V) curves of the inwardly directed Na⁺ currents evoked by depolarizing voltage steps after prepulses to −120 and to− 70 mV, respectively. The difference between both curves shows the mean peak current density–voltage curve of the transient LVA current. Mean of cells from 10 healthy donors. (C) Mean I-V curves of the inwardly directed Ca²⁺ currents evoked by depolarizing voltage steps after prepulses to −120 and to −60 mV, respectively. Mean of 11 cells from 3 healthy donors. (D) Mean I-V curves of the Ba²⁺ currents evoked by depolarizing voltage steps after prepulses to −80 and to −30 mV, respectively. Mean of five cells. In (A) and (B), divalent cation-free bath solution containing 1 mM EGTA was used. In (C), the bath solution contained 2 mM Ca²⁺ and 1 mM Mg²⁺. In (A), (B), and (C), the bath solutions contained 113 mM NaCl.

Figure 3.

 

Age- and disease-related changes of the mean membrane capacitance of human Müller cells. Cells were obtained from retinas of healthy donors and of patients with PVR. The data were divided into two groups according to the age of the donors. Numbers in parentheses indicate the numbers of donors. Significant differences of ^•• P < 0.01 and ^••• P < 0.001. n.s., not significant.

Figure 4.

 

Age- and disease-related changes of Na⁺ currents through voltage-gated Ca²⁺ channels in human Müller cells. (A) Mean densities of the peak currents through Ca²⁺ channels. The sum currents were recorded by applying depolarizing voltage steps from prepulses to −120 mV; the HVA currents were recorded at depolarizing steps after prepulses to −70 mV. The LVA currents were determined as difference of both records. Numbers in parentheses indicate the numbers of donors. The points indicate significant differences of ^• P < 0.05 and ^••• P < 0.001. n.s., not significant. (B) Scatter plot of the ratio between peak HVA and LVA currents vs. the age of the donors. Each symbol represents the mean value of currents recorded in cells from one donor. (C) Mean voltage dependence of the density of peak HVA currents recorded in cells from healthy and diseased human donors.