Calcium-dependent efflux of K+ ions from stimulated T lymphocytes of young and old mice

ELSEVIER SCIENTIFI( PUBLISttERS IRI~LAND

Mechanisms of Ageing and Development 71 (1993) 111-129

Calcium-dependent efflux of K ÷ ions from stimulated T lymphocytes of young and old mice Jacek M. Witkowski* Department of Histology, Medical Academy of Gdansk, Gdansk, Poland (Received 8 January 1993; revision received 14 June 1993; accepted 17 June 1993)

Abstract

The changes in the rate of efflux of potassium cations from ionomycin-, A23187-, and Concanavalin A (Con A)-stimulated lymphocytes from young and old mice were studied with an ion-selective electrode in a 'half-micro' system. It was found that, for both types of stimuli, the maximal rate of Ca2+-dependent efflux of K + from young T cells was more than twice that attained by old cells. A bimodality of the curve illustrating change of efflux rate was observed, indicating possible existence of two types of Ca2÷-dependent conductances. The kinetics of change of the effiux rate was similar for young and old cells stimulated with calcium ionophores, but differed in the cells stimulated with Con A. There, a lag period between mitogen addition and onset of measurable K + efflux was absent in the case of old T cells, suggesting that function of other (Ca-independent?) mechanisms of K ÷ efflux during the mitogen stimulation may also be changed there. The measured efflux of K ÷ was only partially dependent on the extracellular Ca 2÷. Also, it was quantitatively different in a 'physiological' medium containing 140 mM Na ÷, as compared to a sodium-free medium. Different blockers of potassium, calcium and sodium channels had at least partially inhibitory effect on the measured flux. Presented findings indicate that potassium conductance through Ca2+-gated K÷channels is impaired in T cells of old mice. Ca2+-dependent efflux of K ÷ in murine T cells is apparently conducted by a specific class of membrane channels, possibly consisting of two types of channels with different activation kinetics and pharmacological sensitivities (expressed in different subpopulations of T cells?). Impaired potassium conductance in old T cells is discussed as one of possible causes of age-related dysfunction of the immune system.

Key words." Potassium efflux; Calcium-dependent; T cells; Mice; Aging

*Present address: Institute of Gerontology, University of Michigan, 300N lngalls Building, Room 1074, Ann Arbor, MI 48109-2007, USA. 0047-6374/93/$06.00 © 1993 Elsevier Scientific Publishers Ireland Ltd. All rights reserved. SSD1 0047-6374(93)01369-J

112

J.M. Witkowski/Mech. Ageing Dev. 71 (1993) 111-129

1. Introduction

Cation fluxes across the T lymphocyte membrane are known to occur during mitogen- and antigen-induced stimulation of these cells. The transient increase of Ca 2÷ concentration in the cytoplasm of stimulated lymphocytes is an essential, well documented signal for further activation steps [1,2]. Another set of early signals, although not well characterized so far, consists of transient changes in lymphocyte membrane permeability for sodium and potassium cations and changes in the cell transmembrane electrical potential (TMP) [3-6]. Calcium ionophores A23187 and ionomycin were found to be mitogenic for murine and human T lymphocytes [7-10]. We showed recently that A23187 in micromolar concentrations selectively hyperpolarized T lymphocytes from mouse spleen (i.e. increased their TMP) [6], thus confirming the earlier, similar observation suggesting the existence of Ca2+-dependent potassium channels in murine T cells [11]. Early changes in the TMP of murine lymphocytes stimulated with T cell mitogens are also known to include transient hyperpolarization [3,6,12]. However, we have also demonstrated that the A23187-induced hyperpolarization of T cells of old mice was delayed and protracted in time when compared to that of young mice [6]. Furthermore, we found that the pattern of TMP changes in Con A-stimulated young and old T cells was different, with a depolarization as a predominant feature in young cells which is absent (in fact, replaced with strong hyperpolarization) in the case of old ones [6]. These findings lead us to the suggestion of changed properties of Ca2+-dependent K + conductance in T cells of old mice. Thus, early changes in Ca2÷-dependent potassium effiux from T lymphocytes of young and old mice are quantitatively examined and compared here. In this work, we attempted to answer the following questions: (i) is the Ca 2+dependent K ÷ effiux from T cells of young and aged mice quantitatively and/or qualitatively different?, (ii) does it differ in both age groups with respect to the stimulus used?, and (iii) what are the pharmacological characteristics of this effiux? 2. Materials and methods

2.1. Cells Subcutaneous lymph nodes (cervical, axillary and inguinal) from young (2-3 months) and old (20-24 month) BALB/C mice of both sexes were removed after the animals were killed with ethyl ether. The nodes were gently dissociated in Trischoline medium (TC) containing (mM): choline chloride, 145; TRIS-HCI buffer, 20; CaC12, 1; MgCI2, 1; glucose, 1; pH 7.4. The resulting crude cell suspension was left standing at room temperature for 10 min to allow cell clumps to settle, then filtered through nylon mesh and washed four times in the medium. Cells were then suspended at 1 × 107/ml of the same medium. Alternatively, another medium (marked TN) was used, where choline chloride had been replaced with an isoosmotic concentration of NaCI. In pilot experiments, TC medium was found to yield higher signal than TN medium in response to the same concentration of K ÷, while not hampering the viability of tested cells. Both solutions were made using glass-distilled, deionized

J.M. Witkowski/Mech. Ageing Dev. 71 (1993) 111-129

113

water (specific resistance greater than 18 Mfl × cm -1) and were nominally 'potassium-free'. Cells of young and old mice are referred to in the text as young and old cells, respectively. 2.2. Purification of T and B lymphocytes

In order to avoid direct activation of T cells which, in theory, might affect the measured flux, Sephadex G10 and nylon wool adhesion methods using active separation of B lymphocytes were adopted. Purity of the thus obtained T lymphocyte population was checked with FITC-coniugated anti-Thy-l.2 and anti-mouse Ig antibodies and was routinely better than 95%, with no more than 3-5% of contaminating B cells. B cells recovered from the nylon wool column were further depleted of remaining T lymphocytes by antibody-dependent complement lysis, using either anti-Thy-l.2 or anti-CD4 plus anti-CD8 monoclonal antibodies (kind gift from Dr. H.S. Micklem, University of Edinburgh). The final purity of the B cell suspension was around 95%. Purified T and B lymphocytes were suspended in the fresh portion of TC or TN medium at 1 x 107/ml and further processed as unseparated cells. Viability of cells assessed with the trypan blue exclusion test was always above 97% immediately after the separation and more than 93% at the end of every experimental session. 2.3. Analysis of the K + efflux

The measurements were done in a 2-ml Perspex chamber fitted with heating jacket and thermostated at 37°C. The chamber was loaded with 5 x 10 6 cells in 2 ml of medium for each measurement. Settling of the cells in the chamber was prevented by slow stirring, with a magnetic bar contacting the suspension shaped so as to minimize damage to the cells. A K+-selective electrode (sensor: valinomycin/PCV membrane) and a reference electrode with a double electrolytic bridge (1M KCI internal, 1M (NH4)2SO4 external) were both purchased from the Scientific Marketing House DHN, Warsaw). Before every experimental session, the electrodes were conditioned in 10 mM KC1 for 30 min. Between measurements, the chamber was extensively washed with prewarmed medium until isoelectric 'null' readouts were obtained. The electrodes were connected to a digital pH/mV meter N 5121 (MeraTronik, Warsaw) and the latter coupled to a Y-t recorder TZ4200 (Laborni Pristroje Praha, Czechoslovakia) through a 'bucking voltage device' assembled according to Madeira [13] in order to offset too large a signal obtained at the pH/mV meter output. For every cell sample, two measurements were taken. The first served for the determination of resting efflux of potassium (RE) from undisturbed cells and the second, performed immediately thereafter, recorded the total efflux of K ÷ occuring in the presence of ionophore or mitogen. Increase of RE during an experimental session could indicate loss of viability of the cells in a tested sample; we found it to parallel the results of the trypan blue exclusion test. Usually, we did not see significant change in the RE between 3 consecutive measurements of the same sample, each lasting 5 min (not shown). However, if a difference between RE taken at each mea-

J.M. Witkowski/Mech. Ageing Dev. 71 (1993) 111-129

114

surement and at the very end of the experimental session exceeded 15% of the initial value, the sample was considered to be of poor viability and the results excluded. We found earlier that A231287 used at concentrations inducing hyperpolarization of T lymphocytes did not impair the viability of these cells [6]. Also, work by Miller et al. shows that murine T cells treated with hyperpolarizing concentrations of ionomycin (similar to that used in this work) remain viable and functional, i.e., able to proliferate, manufacture IL-2 and generate precursors of cytotoxic T cells [14]. Therefore, the decrease of viability observed in some (infrequent) samples cannot, in our opinion, be related to the toxicity of ionophore. At the beginning and at the end of every experimental session, electrode calibration was performed with standard solution of 10 mM KC1 in both TC and TN media, spanning the range 0-100 mM and thus covering more than the entire range of concentrations observed during experiments. Ten values of electrode potential proportional to increasing concentrations of K ÷ were obtained at each calibration and served for calculation of standard curve for every experiment. At the low concentrations of potassium observed in our experiments the characteristics of the ionselective electrode (dose-response curve) were described best with a parabolic equation. Thus, a best-fit calibration curve was generated for every experimental session and its equation introduced to a computer program developed for the purpose of calculation of actual rates of potassium release. Raw data from the measuring device were then 'digitized' and values corresponding to 10-s intervals fed into the program. At least 25 time points were estimated for every experimental curve. The program calculated first the actual amount of K ÷ released in the sample due to stimulant action utilizing the standard curve and the experimental data on resting and total efflux, and then the treatment-induced rate of K + release to the medium from a single, average cell. The induced rate of efflux of K + for a single stimulated cell (induced efflux rate, IER) at any time point from the addition of stimulant was calculated as:

IER(t) =

[ K ÷ l r e l ( t ) - [ K + l r e l ( t - 1) cell no. x [t - ( t - 1)]

where IER(t) is the effiux rate at time point t from the addition of stimulant, [K+]rel(t) is the amount of potassium released at time point t, in equivalents, [K+]rel(t - 1) is the amount of potassium released at preceding time point t - 1, in equivalents. 2.4. Induction of K + efflux

The calcium ionophore ionomycin (kind gift from the Calbiochem AG, Lucerne, Switzerland) was dissolved in absolute ethanol as 1 mM stock solution. Working solutions in TC or TN media were prepared extempore. Ten microliters of solution of appropriate concentration were added to the chamber filled with cell suspension alter the baseline effiux was recorded, in order to obtain effective concentrations of ionophore in the range 0.5-5 #M. A23187 (Calbiochem, Switzerland) had been

J.M. Witkowski/Meeh. Ageing Dev. 71 (1993) 111-129

115

prepared in a similar way and used at the same working concentrations. We did not test the effect of ionophores at (presumably acutely toxic) concentrations above 5 #M. Concanavalin A (Sigma, St. Louis) was used at 10 tzg/ml of the cell suspension. Maximal effiux of K + across relatively unchanged plasma membrane was estimated in ceils treated with I tzM valinomycin (Sigma, St. Louis). Any difference in the amount or rate of Ca2+-dependent K + effiux from young and old T cells could, in principle, be due to different availability of free K + in the cytoplasm of these cells. Therefore, it was of interest to check the actual contents of releasable K + in the samples. For that purpose, 50-/~1 samples of lymphocyte suspensions from each experiment underwent three cycles of consecutive freezing and thawing and subsequent centrifugation at 15 000 x g for 30 min at +4°C. The concentration of K ÷ in the supernatants was estimated with the K÷-selective electrode and calculated to yield total releasable K + content per cell.

2.5. Pharmacological modification of K + efflux Tetraethylammonium chloride (TEA, Sigma), 4-aminopiridine (4-AP, Reanal), verapamil (VPL, Sigma), and saxitoxin (STX Calbiochem) were used. Of these, TEA and 4-AP are known as blockers of voltage-gated potassium channels in lymphocytes [15,16], VPL is a blocker of calcium channels in various tissues, but in T cells - - lacking typical calcium channels in their membranes - - it presumably acts also on some potassium channels [1,15]. Finally, STX is a known blocker of sodium channels in nerve cells [17,18]. The reagents were dissolved as stock solutions in dimethylsulphoxide (DMSO) (excluding STX, which is supplied as solution in diluted acetic acid). Dilutions in the medium were prepared ex tempore and 10 #1 of the dilution added to the cell suspension for 1 min before the addition of ionophore, at concentrations indicated in the figures. In separate experiments, DMSO and acetic acid were proven to have no effect on K ÷ effiux at concentrations used. 3. Results

3.1. Ca2+-dependent efflux of K +from young T cells is faster and more abundant than from old ones after treatment with .423187 and ionomycin The rates of ionomycin- and A23187-induced effiux of K ÷ from unseparated lymph node lymphocytes of young mice were significantly higher (P < 0.01) than those from old ones in the whole range of ionophore concentrations used (Fig. 1). For both age groups, the maximal rates increased linearly with the dose of ionophore rising from 1.25 to 5 /zM. For each age group, the maximal rates obtained for A23187 treatment did not differ from those resulting from treatment with equal concentrations of ionomycin (Fig. 1), neither did the time course of the rate change (not shown). For clarity, further results are shown for the effects of ionomycin only. We have previously demonstrated that only T cells hyperpolarized when treated with A23187 [6]. To check the assumption that only murine T and not B cells have Ca2+-dependent K ÷ channels, we now purified T and B cell populations from the lymph nodes of young and old mice and analyzed their ability to expel K ÷ upon

J.M. Witkowski/Mech. Ageing Dev. 71 (1993) 111-129

116

max. rate []

YOUNG A23187 OLD A23187

60

0

:;7

40

20

1.25 uM

2.5uM

5 uM

concentration (uM) Fig. 1. Changes in maximal rates of efflux of K ÷ from young and old lymph node lymphocytes induced with 1.25, 2.5 and 5 t~M ionomycin or A23187. Each point represents the mean of five results. Standard deviations were below 10% of the mean for each data point and are omitted from the graph for the sake of clarity. Ordinate - - maximal rate ('max. rate', expressed in attomols (10 -18 M) of K+/cell per s); abscissa - - ionophore concentration.

stimulation with ionomycin. We found that purified T cells did, while B cells did not release substantial amounts of K ÷ upon ionophore stimulation. Maximal efflux rates from old T cells were significantly smaller than from young ones, while the rates of efflux of K ÷ from ionomycin-stimulated, purified B cells of both young and old animals were negligible (Figs. 2,3). The maximal rate of stimulated potassium release (IER) from T cells was reduced in comparison to that obtained for unseparated lymph node cells in the appropriate age group, especially in the young cells (Fig. 3). This reduction reached in some cases 20-25% of the maximal rate obtained from unseparated cells but even for young cells it was of weak statistical significance (P = 0.52). It is worth noting that we observed some bimodality in the change in efflux rates with time with peaks at 120 and 200-220 s, both in young and old cells (Fig. 2), possibly indicating the existence of two separated populations of Ca2+-dependent K ÷ channels with different properties (see Discussion). Usually, the efflux from ionophore-treated cells tended to attain equilibrium with (probably) active reuptake of K ÷ after about 7-10 min of treatment, which lead to a stepwise decrease of the value of net efflux rate to zero (not shown). The time of cessation of net Ca2+-dependent efflux of K ÷ induced with calcium ionophores was similar for young and old T cells.

J.M. Witkowski/Mech. Ageing Dev. 71 (1993) 111-129

5O

117

rate (amol/cell/sec) •

40

0

YOUNG T

t

YOUNG B

)K

OLD T

. . ~ - ' ~ /

60

120

\

180

240

time (sec) Fig. 2. Comparison of the time-dependent changes in the rates of 5 #M ionomycin-induced efflux of K + from purified T and B cells of young and old mice. Each point represents the mean of four results. Standard deviations were below 10% of the mean for each data point (not shown). Ordinate - - rate (expressed in attomols of K+/cell per s); abscissa - - time after ionophore addition.

The main difference between unsorted cells and purified T cells was the different time course of the change in IER. Purified T lymphocytes expressed a lag period between addition of the ionophore and the appearance of measurable IER. This lag period was practically absent in the unseparated population, for which (in the case of young cells) the detectable efflux of K ÷ started within 10 s (Fig. 4). There was no difference in the length of the lag period between young and old T cells. As expected, the lag period tended to be shorter for higher ionophore concentrations (not shown). The IER from unseparated lymph node cells was elevated in comparison to purified T lymphocytes mostly during the first 2 min of ionomycin action, while later on the two rates converged, both for young and old cells. 3.2. Concanavalin A-stimulated T cells o f old mice release less K ÷ than young ones and with different dynamics

The maximal rate of Con A-induced efflux of K ÷ from old T cells was twice as small as that of young ones, and the difference was statistically significant (P < 0.02) (Fig. 5). In the absolute values, the IER obtained upon mitogen stimulation of young

118

J.M. Witkowski/Mech. Ageing Dev. 71 (1993) 111-129

max. rate 60

50

40

30

20

10

young

old

Fig. 3. Comparison of maximal K + efflux rates (max. rate, expressed in attomols of K+/cell per s) from unseparated lymph node lymphocytes, and purified T and B cells of young and old mice, stimulated with 5 #M ionomycin. Bars represent means of four results each, with standard deviations shown as thin, capped vertical lines.

and old cells was, on average, three times lower than that resulting from stimulation of respective cells with a calcium ionophore at 5 #M. Analysis of the time course of the change in IER, while not revealing the bimodality seen in the case of ionophore stimulation, had nevertheless shown a characteristic difference between young and old cells stimulated with Con A. T cells of young animals expressed a lag period of about 60 s between addition of mitogen and onset of measurable increase in IER; this lag was reduced to less than 10 s in case of old mice (Fig. 5).

3.3. Valinomycin is less effective in old than in young T cells There was practically no detectable lag in the onset of the release of K ÷ from young and old T cells treated with a potassium ionophore valinomycin (Fig. 6). Although the maximal rate of the valinomycin-induced efflux of K ÷ from young lymphocytes was higher than that from old cells they did not differ significantly. However, the rate dropped significantly faster in the case of old cells (Fig. 6). This difference could suggest a lower amount of K ÷ available for release in the cytoplasm of old cells. In order to check this possibility, we lysed the cell samples by

J.M. Witkowski/Mech.Ageing Dev. 71 (1993) 111-129

119

rate (amol/cell/sec) 60

50

~

YOUNG T CELLS

~--

YOUNG UNSEPARATED

-~

OLD T CELLS

q-

.d-

40 30

20

10

0~1~

0

~

-1",

60 time

L

I

1

120

180

240

(sec)

Fig. 4. Comparison of the time-dependent changes in the rates of 5 #M ionomycin-induced effiux of K ÷ from unseparated lymphocytesand purified T cells of young and old mice. Each point represents the mean of four results. Ordinate - - rate (expressed in attomols of K+/cell per s); abscissa - - time after ionophore addition.

three cycles o f freezing and thawing and determined the a m o u n t o f K + released this way with the K+-selective electrode. We did not find a significant difference between the a m o u n t o f K + released from the same number o f y o u n g and old cells (P = 0.82).

3.4. The rates of ionophore- and mitogen-induced K + efflux from lymphocytes only partially depend on the presence of external Ca 2+, and are modified by external Na ÷ Both calcium ionophores as well as C o n A required the presence o f Ca 2+ in the external medium in order to induce the release of K + from the cells (Fig. 7a,b). However, when Ca 2+ was omitted from the m e d i u m and E G T A added to ascertain the very low actual concentration of external Ca 2÷, still some - - both reduced and delayed - - effiux o f K ÷ could be recorded, both from y o u n g as well as from the old cells (Fig. 7). The delay was similar for C o n A- and ionomycin-stimulated cells, andthe rates reached m a x i m u m at approximately the same time. There were no differences between y o u n g and old cells treated with the same agent. However, in the Cafree medium, the maximal rate of K ÷ effiux from ionomycin-treated cells was more than twice that from C o n A-stimulated T lymphocytes. The rate o f K ÷ effiux from

J.M. Witkowski/Mech. Ageing Dev. 71 (1993) 111-129

120

rate (amol/cell/sec) YOUNG

20

OLD

15

10

[]

(~

o

60

120

180

240

time (sec) Fig. 5. Comparison of the time-dependent changes in the rates of Concanavalin A-induced efflux of K ÷ from purified T cells of young and old mice. Each point represents the mean of five results. Standard deviations were below 10% of the mean for each data point. Ordinate - - rate (expressed in attomols of K+/cell per s); abscissa - - time after addition of 10 #g/ml Con A to the cell suspension.

the ionophore-treated old cells was slightly decreased in comparison to young ones, but the difference was not significant. The rates obtained for Con A-induced K ÷ efflux from young and old T cells in a Ca2÷-free medium did not differ (not shown). All results described above were obtained for cells suspended in nominally sodium-free TC medium. However, if the cells were suspended in TN medium, containing 140 mM Na ÷, the maximal, ionomycin-induced rate of efflux of K + from young lymphocytes decreased while, surprisingly, the release of K + from old T cells quickened significantly, especially during the first 2 min of the test (Fig. 8). Thus, the difference between IER for ionomycin-induced cells of young and old mice was much smaller in the medium containing sodium than without Na ÷ and significant during the first 3 min of stimulation only. 3.5. A23187-induced efflux of K÷ from T cells was blocked by TEA, VPL, and STX, but not by 4-AP The peculiar behavior of the calcium ionophore-induced K ÷ efflux in Na ÷containing versus Na+-free medium might indicate specific properties (apparently, significant sodium conductance) of the Ca-dependent membrane channels conduc-

J.M. Witkowski/Mech. Ageing Dev. 71 (1993) 111-129

121

rate (amol/cell/sec) []

25

[]

YOUNG

X

OLD

0

20

15

10

5

I

OI 0

n

60

120

m

•

180

240

time (sec) Fig. 6. Comparison of the changes in the rates of 1 t~M valinomycin-induced effiux of K + from purified T ceils of young and old mice. Each point represents the mean of three results. Ordinate - - rate (expressed in attomols of K+/cell per s); abscissa - - time after ionophore addition.

rate (amol/cell/sec) I~

rate (amol/cell/sec)

A 20

CONTROL

~

GONTRO¢. C=4RE|

B Q

3oi

,0[

10

10-

0

+

60

120

time (see)

180

240

0

eO

120

180

24O

time (sec)

Fig. 7. Calcium dependency of efflux of K ÷ from young T cells induced with 5 #M ionomycin (A) or 10 p,g/ml Con A (B). Each point represents the mean of three results. 'Ca-free' indicates that no Ca 2÷ was added to the medium and 0.1 m M EGTA was present. Control medium contained 1 m M Ca 2÷. Ordinate and abscissa described as in previous figures.

122

J.M. Witkowski / Mech. Ageing Dev. 71 (1993) 111-129

rate (amol/cell/sec) 80

)~ _~

ON

,,

0

~...

TC (YOUNG)

[]

TN (YOUNG)

><

TC (OLD)

,

I

J

60

190

"180

.........

• 240

time (sec) Fig. 8. Sodium dependency of efflux of K + from young and old lymph node lymphocytes induced with 5 #M ionomycin. TC - - TRIS-choline, TN - - TRIS-NaCI (see Materials and Methods for detailed composition of both media).

ting K ÷ in murine T cells. In order to characterize partial pharmacological properties of these channels, we analyzed the I E R in ionomycin-induced young and old T cells treated with 10/~M verapamil, 25 mM tetraethylammonium, 2 mM 4-aminopiridine, and 50-250 nM saxitoxin. We have found that the rate of ionophoreinduced effiux of K ÷ from T cells of young mice was significantly decreased by treatment with both VPL and TEA, although neither of these blockers was able to abrogate the efflux at concentrations used (Fig. 9). Verapamil seemed to act mostly during the first stage of effiux, while the effect of TEA could be seen throughout the test (Fig. 9). 4-AP had no effect on the maximal rate of efflux of K ÷ nor on its time course.

On the other hand, the sodium channel blocker saxitoxin strongly inhibited the effiux in a dose-dependent manner, affecting mostly its first phase (Fig. 9). The effect of all of the blockers on the ionomycin-induced efflux of K + from T cells of old mice was identical to that observed in young cells (not shown). 4. Discussion Our results show that calcium-dependent release of K + from both unseparated lymphocytes and T cells of old mice stimulated with micromolar doses of calcium

ZM, Witkowski/ Mech. Ageing Dev. 71 (1993) 111-129

123

rate (amol/cell/sec) 60

:

@TX

t

TEA

~)'

VPL

[]

[]

4AP X

40

0

CONTROL

60

120

180

240

time (sec) Fig. 9. Effect of channel blockers on the rates of ionomycin-induced efflux of K ÷ from young T cells. STX - - 125 nM saxitoxin, TEA - - 25 mM tetraethylammonium, VPL - - 10/zM verapamil, 4AP - 2 mM 4-aminopiridine.

ionophores ionomycin and A23187 was significantly reduced in comparison to that from young animals. This reduction was seen as a decrease in the maximal rate of Ca2÷-dependent efflux, while the duration of the time of reaction to stimulus remained unchanged (Figs. 1-4). These results cannot be interpreted as related to differences in viability due to handling and/or ionophore treatment of young and old T cells, as such differences were not seen in the present work. Also, we showed earlier that A23187 does not impair the viability of murine T cells during short-term incubation [6], while Miller et al. demonstrated that T cells treated with 2/~M ionomycin retain their proliferative, helper and cytotoxic functions [14]. A similar decrease .in the maximal rate of efflux of K ÷ from old T lymphocytes was also demonstrated when the cells were stimulated with Con A (Fig. 5). The maximal IER due to Con A stimulation was only about 1/3 of that obtained for ionophore-induced cells. The concentrations of both types of stimulators were chosen so as to ascertain rapid generation of the calcium signal, and were slightly above the optimal mitogenic concentrations, while still in the stimulatory range. So, the difference between maximal rates of K + efflux may be explained plausibly as due to different proportions o f T cells responding to calcium ionophores and to Con A. It is known that concanavalin A stimulates mostly the 'naive' T cells that develop their calcium signal upon activation with the mitogen, leaving the 'memory' sub-

124

ZM. Witkowski/Mech. Ageing Dev. 71 (1993) 111-129

population relatively unchanged [19,20]. It was also demonstrated that aging leads to the increase in the proportion of memory T cells with a simultaneous reduction of the number of naive T lymphocytes [19,21]. Thus, the Con A-responsive population makes up for more than 70% of young T cells, but only for about 30% of old ones [19-21]. Assuming different properties of potassium conductances in naive and memory cells (which would need a separate confirmation), this difference could be the sole explanation for the reduced average maximal IER from Con A-stimulated old T cells shown in this work. However, analysis of the change of the IER with time reveals not only a quantitative difference, but also a qualitative change, namely significantly earlier onset of effiux of K ÷ (no 'lag period') from mitogen-activated old T cells (Fig. 5). A plausible explanation of this phenomenon is that other potassium conductance(s), which is(are) not Ca2÷-dependent, is(are) activated early in Con A-treated old T cells, but not in the young ones. Without further electrophysiologic studies it cannot be excluded here that the voltage-gated K ÷ channel repertoire is different in naive and memory T cells. This suggestion agrees with our recent demonstration of early, deep hyperpolarization of Con A-stimulated old T cells [6] and with reports of the existence of several classes of voltage-gated K ÷ channels in murine T lymphocytes [1,2,15,16]. We have found also that similarly activated T cells of young mice depolarized almost immediately after Con A addition; this depolarization started to drop after 1 min of activation, clearly indicating activation of some mechanism temporarily returning the transmembrane potential to the resting values [6]. Calcium-dependent K ÷ channels, opening in young T lymphocytes - - as we demonstrate here - - only after 1 min of cell-mitogen contact, are a good candidate for this depolarization-reverting mechanism. Similarly, an approximately 1-min-long lag period was seen in the purified T cells between ionophore addition and the beginning of the measurable efflux of K ÷. This lag period was dependent on the ionophore concentration (not shown), but identical for T cells of both young and old mice, indicating the lack of age-dependent change of the effectiveness of the ionophore action. In both cases, the lag can be explained as the amount of time necessary for the build-up of the cytoplasmic concentration of Ca 2÷ to the level required by the putative Ca2+-dependent K ÷ channels. The proportion o f T cells responding to ionomycin has been demonstrated to drop with advanced age [21,22]. The ionophore-resistant population is speculated to have an increased activity of a calcium-extruding membrane pump, and thus not to be able to raise Ca 2÷ concentration to a level necessary for adequate stimulation, and to belong mostly to the memory T cell subpopulation [21]. This again suggests that decreased effiux of K ÷ from old T cells might, at least in part, depend on the naiveto-memory shift. To confirm this suggestion, we now plan a study in which T cells will be separated according to their naive or memory phenotype, and then analyzed for their ability to elicit mitogen- or ionophore-induced efflux of K ÷, expecting memory T cells to have at least substantially reduced Ca2÷-dependent K ÷ conductance. However, without direct electrophysiological studies we cannot exclude at this time the possibility that the number of Ca2+-gated K ÷ channels is uniformly reduced in all old T cells or their conductance uniformly diminished. The effiux of K ÷ from old T cells treated with a potassium ionophore valinomycin - - also differed from that of young cells; its maximal rate was lower and

J.M. Witkowski/Mech. Ageing Dev. 71 (1993) 111-129

125

it ended significantly earlier (Fig. 6). The virtually instantaneous start of K ÷ extrusion from valinomycin-treated cells confirms the direct action of the ionophore as K ÷ carrier, as opposed to the indirect effects of both calcium ionophores and Con A. We did not find a significant difference in the amount of releasable K ÷ in young and old T cells. Thus, the difference in valinomycin action on young and old cells is most probably due to altered interaction of the ionophore and its immediate environment, i.e., membrane lipids, known to change their composition and physical properties in old lymphocytes [23]. Changes in the membrane lipid composition may also influence the properties of ion channels in old T cells, and such a possibility will be examined in future study. As described in Materials and methods, we performed our analyses on either an unseparated preparation of murine lymph node lymphocytes, or on purified T or B cells. The latter had shown no Ca2÷-dependent efflux of K ÷ when stimulated with either A23187, ionomycin or Con A. This observation confirms previous speculations about the existence of Ca2÷-dependent K ÷ channels in T, but not B cells [6,11]. We found that unseparated lymphocyte suspensions gave higher maximal efflux rates than T cells purified from the same material and that efflux from the T cells started significantly after contact with calcium ionophore than that from unseparated cells. Excluding B cells, the only cell type present in substantial numbers in our 'crude' cell suspensions was macrophages. These cells were mostly removed by passage of the cell suspension through the G10 column. It has been demonstrated that murine cells of monocyte/macrophage origin express Ca2÷-dependent, as well as voltage-gated K ÷ channels [24]; thus, their presence in the initial, 'crude' cell suspension could add to Ca2÷-dependent efflux of K ÷ from T cells, effectively increasing its rate. Calcium-gated K ÷ channels of non-lymphocytic cells might also be responsible for the early onset of efflux of K ÷ from the crude cell suspension; to validate this suggestion, further comparative study of the kinetics of Ca2÷-gated K ÷ channels in T lymphocytes and macrophages is required. Our observations of significantly delayed and reduced, but not absent efflux of K ÷ from the cells stimulated with either ionomycin or Con A in a Ca2+-free media (Fig. 7) indicate that apart from their action on Ca2+-dependent K ÷ channels in the cell membrane, utilizing exogenous Ca 2÷ for the channels' activation, both types of agents activate another mechanism which then promotes efflux of K ÷ from the stimulated cells. In the case of calcium ionophores, the mechanism might be simply an insertion of the ionophore molecules into cytoplasmic membranes including endoplasmic reticulum, leading to the release of Ca 2÷ from this 'calcium store' and, in consequence, the activation of Ca2÷-gated channels. Release of Ca 2÷ from the internal stores, but involving prior hydrolysis of phosphatidylinositol and formation of IP3 triggering calcium channels in the reticulum is probably also responsible for the efflux of K ÷ from Con A-stimulated T cells in the Ca2+-free medium. It is worth mentioning here that our results show similar dynamics of change of the efflux rate for ionophore- and mitogen-stimulated cells in the absence of external Ca 2÷ (Fig. 7), suggesting a similar underlying mechanism. The ability of T cells of old mice to hydrolyze phosphatidylinositol and to form IP3 and other metabolites upon stimulation is reported to remain unchanged [25]. Accordingly, we did not see a decrease in the rate of efflux of K ÷ from old T cells in a Ca2+-free medium (not shown).

126

J.M. Witkowski / Mech. Ageing Dev. 71 (1993) 111-129

Existence of Ca2+-gated K + channels in murine and human T cells was long postulated on the basis of circumstantial evidence, like Ca2+-dependent hyperpolarization [6,11,26], but until recently eluded the electrophysiological approach. In the last couple of years, however, different groups of researchers reported, using the patch clamp technique, the detection of Ca2+-dependent potassium conductance in T cells of mice, rats and humans, varying in actual conductance from as low as 6-8 to more than 90 picosiemens [27-30]. That big a difference might indicate that in fact not one, but at least two populations of Ca2+-dependent K + channels do exist on murine T cells, possibly delineating their different subpopulations. Analysis of the time course of change of the efflux rate of ionomycin-treated cells shows some bimodality, indicating the possibility of two populations of channels with different kinetics (compare Figs. 2, 4 and 8). It is tempting to speculate that memory T cells are expressing channels of low conductance, while naive T lymphocytes express mostly high conductance channels. Such a hypothesis, which would fit with our interpretation of lower Ca2+-dependent K ÷ efflux from old T cells due to naive-tomemory shift, would again require purification of both naive and memory T lymphocytes and patch clamp analysis of purified cells. The ionophore-induced efflux of K+ was decreased in the case of young, and strongly increased in old cells suspended in sodium-containing medium, as compared to a sodium-free one (Fig. 8). These experiments were done on unseparated cells; however, murine and human macrophages apparently do not have typical, voltagegated sodium channels [24,31] and their participation in the results of these experiments should be considered negligible. The result seen in young cells could suggest some competition between potassium and sodium ions trying to pass through the same channel in opposite directions, leading to lower net rate of K + efflux. Voltagegated sodium channels are only rarely found in T cells [1]. This makes channels observed by us a possible passive path for early sodium entrance into the cytoplasm of an activated, young T cell. Sodium ions, entering through the Ca2+-gated channels, would depolarize young (mostly naive) T cells [6], but not old (mostly memory) cells, in which Ca2+-dependent conductance is reduced. The stimulatory effect of Na + on the ionomycin-induced efflux of K+ from old T cells is more difficult to explain. It can be speculated, for instance, that the sodium-containing environment is more likely than the Na+-free one to increase the probability of early opening of voltage-gated K ÷ channels in old T cells. Thus, under physiological conditions, the decreased effectiveness of Ca2+-dependent conductance of K + in old T cells might well be at least partially masked by its voltage-gated counterpart. Analysis of the effect of different channel blockers on the Ca2+-dependent efflux of K ÷ from young and old T cells did not show any age-associated difference (not shown). The time course of change of A23187-induced IER in the presence of different blockers did suggest the existence of two populations of channels with different sensitivity to the inhibitors. A typical blocker of voltage-gated K ÷ channels - - TEA - - reduced the effiux both in the first and (predominantly) second half of the observation period, while verapamil was effective only during the first 2 min of efflux, leaving its rate during the later part unchanged. We have recently also observed that other calcium antagonists - - nifedipine and diltiazem - - had the same effect (not shown). Also 125 nM STX fully blocked effiux during first 2 min, but only

J.M. Witkowski / Mech. Ageing Dev. 71 (1993) 111-129

127

partially that occuring later (Fig. 9). It is worth mentioning here that in our previous observations 25 mM TEA inhibited the development of early depolarization of Con A-stimulated young T cells, while it only partially reduced the hyperpolarization of old ones [61. At this moment it is difficult to speculate on the consequences of decreased Ca 2+dependent K + conductance for the immune functions of T cells. If the decrease is, as we believe, a feature of the (old) memory T cells, it changes the potassium homeostasis in these cells only and, in principle, might be at least partially responsible for their impaired reactivity. It is currently not known what the significance of changes of cytoplasmic K + concentration is for cellular functions, although it was linked with certain cellular phenomena potentially important for lymphocyte functions like actin aggregation [32], intra-lysosomal pH [33], and certain enzymatic activities [34]. Earlier we showed the reduction in the activity of acid phosphatase in T cells of old people [35] and a related drop in the cytotoxic activity [36] which, in theory, might be related to impaired K + homeostasis in old lymphocytes. Changes in the transmembrane potential of activated cells, occurring as a consequence of quantitatively and qualitatively modified output of K + from old T cells can influence the development and progression of the calcium signal [27,37], which is known to be reduced in old T lymphocytes [38]. It was recently shown that depolarization of human T cells by blocking of potassium channels with charybdotoxin or by incubation in the medium with elevated concentration of K + decreased the production of IL2 by these cells [39]. In principle, transmembrane potential changes can also affect functions of charged membrane molecules (receptors, receptor-linked enzymes, carriers, etc.). It is worth noting here that changes in the membrane potential, resulting from the observed Ca2+-dependent K ÷ efflux only, are in the range of 20-75 mV, depending on the initial conditions and, especially, the age of animal. This strong hyperpolarization, as we have shown earlier, lasts in both young and old T cells for more than 5 min [6] and thus might have an effect on other activation events occuring in the same time frame. Such a possibility, in the light of our findings of resting depolarization of T cells of old mice [5,40] and modified pattern of their 'action potential' [6] requires further, thorough study.

5. Acknowledgment The author would like to thank Professor Andrzej Mysliwski for his help and stimulating discussions during most phases of preparation of this paper.

6. References l S. Grinsteinand J. Dixon, Ion transport, membranepotentialand cytoplasmicpH in lymphocytes: changes during activation. Physiol. Rev., 69 0989) 417-481. 2 B. Seligmann,Monovalention transport and membranepotential changesduring leukocyteactivation: lymphocytes.In Current Topics in Membranes and Transport, Vol. 35, Academic Press Inc., 1991, pp. 103-125. 3 H. Kiefer,A.J. Blumeand H.R. Kaback,Membranepotentialchangesduringmitogenicstimulation of mouse spleen lymphocytes.Proc. Natl. Acad Sci. USA., 77 0980) 2200. 4 H. Kiefer and R. Schulze, The membrane potential of lymphocyteschanges only in response to specific stimulation. Biosci. Rep., 2 (1982) 583.

128 5

6 7 8

9 l0 II 12 13 14

15 16 17 18 19 20 21 22 23 24 25 26

27 28 29 30

J.M. Witkowski / Mech. Ageing Dev. 71 (1993) 111-129 J.M. Witkowski and H.S. Micklem, Transmembrane electrical, potential of lymphocytes in aging mice. Flow cytometric studies with oxonol, cyanine and rhodamine-123 dyes. Aging." Immunol. Infect. Dis., 2 (1990) 287-294. J.M. Witkowski and H.S. Micklem, Transmembrane electrical potential of lymphocytes in ageing mice. Flow cytometric analysis of mitogen-stimulated cells. Mech. Ageing Dev., 62 (1992) 167-179. K.E. Akerman and L.C. Andersson, Direct mitogenic effect ofionophore A23187 on isolated human T helper lymphocytes. Eur. J. lmmunoL, 14 (1984) 286-288. M. Burger, M.W. Hess and H. Cottier, Stimulation of proliferation of murine lymphocytes by the calcium ionophore A23187 in the absence of serum: the requirement for thiols. ImmunoL Lett., 12 (1986) 121-125. R.L. Geller, G. Gromo, L. lnverardi, E. Ferraro, and F.H. Bach, Stepwise activation ofT cells. Role of the calcium ionophore A23187. J. ImmunoL. 139 (1987) 3930-3934. M. Nagarkatti and P.S.Nagarkatti, Calcium ionophores at concentrations mitogenic to normal murine T cells inhibit the proliferation of tumor cells in vitro. Cancer Commun., I (1989) 329-334. H.A. Wilson and T.M. Chused, Lymphocyte membrane potential and Ca-sensitive potassium channels described by oxonol dye fluorescence measurements. J. Cell Physiol.. 125 (1985) 72-81. R.Y. Tsien, T. Pozzan and T.J. Rink, T-cell mitogens cause early changes in cytoplasmic free Ca 2+ and membrane potential in lymphocytes. Nature, 295 (1982) 68-71. V.M. Madeira, A rapid and ultrasensitive method to measure Ca ++ movements across biological membranes. Biochem. Biophys. Res. Commun., 64 (1975) 870-876. R.A. Miller, K. Flurkey, M. Molloy, T. Luby and M.J. Stadecker, Differential sensitivity of virgin and memory T lymphocytes to calcium ionophores suggests a buoyant density separation method and a model for memory cell hyporesponsiveness to Con A. J. ImmunoL, 147 (1991) 3080-3086. K.G. Chandy, T.E. DeCoursey, M.D. Cahalan and S. Gupta, Ion channels in lymphocytes. J. Clin. ImmunoL, 5 (1985) 1-6. T.E. DeCoursey, K.G. Chandy, S. Gupta and M.D. Cahalan, Voltage-dependent ion channels in T lymphocytes. J. NeuroimmunoL, 10 (1985) 71-95. W.A. Caterall, Neurotoxins that act on voltage-sensitive sodium channels in excitable membranes. Annu. Rev. PharmacoL Toxicol., 20 (1980) 15-43. G. Strichartz, T. Rando and G.K. Wang, An integrated view of the molecular toxicology of sodium channel gating in excitable cells. Annu. Rev. Neurosci., 10 (1987) 237-267. A. Lerner, T. Yamada and R.A. Miller, Pgp-lhi T lymphocytes accumulate with age in mice and respond poorly to concanavalin A. Fur. J. lmmunol., 19 (1989) 977-982. B. Philosophe and R.A. Miller, Diminished calcium signal generation in subsets of T lymphocytes that predominate in old mice. £ GerontoL: BioL Sci., 45 (1990) B87-B93. R.A. Miller, Accumulation of hyporesponsive, calcium extruding memory T cells as a key feature of age-dependent immune dysfunction. Clin. ImmunoL ImmunopathoL, 58 (1991)305-317. R.A. Miller, B. Philosophe, I. Ginis, G. Well, and B. Jacobson, Defective control of cytoplasmic calcium in T lymphocytes from old mice. J. Cell. Physiol.. 138 (1989) 175-182. B. Rivnay, A. Globerson and M. Schinitzky, Viscocity of lymphocyte plasma membrane in aging mice and its possible relation to serum cholesterol. Mech. Ageing Dev., 10 (1979) 71-79. C. Randriamampita and A. Trautmann, Ionic channels in murine macrophages. J. Cell BioL, 105 (1987) 761-769. A. Lerner, B. Philosophe and R.A. Miller, Defective calcium influx and preserved inositol phosphate generation in T cells from old mice. Aging: lmmunol. Infect. Dis.. 1 (1988) 149-157. Y. Ishida and T.M. Chused, Heterogeneity of lymphocyte calcium metabolism is caused by T cellspecific calcium-sensitive potassium channel and sensitivity of the calcium ATPase pump to membrane potential. J. Exp. Med., 168 (1988) 839-852. M.P. Mahaut-Smith, and L.C. Schlichter, Ca2+-activated K + channels in lymphocytes. Pflugers Arch., 414 Suppl. 1 (1989) S164-S165. P. Gardner, Patch clamp studies of lymphocyte activation. Annu. Rev. lmmunoL. 8 (1990) 231-252. F.V. McCann, D.C. McCarthy, and R.J. Noelle, Patch clamp profile of ion channels in resting murine B lymphocytes. J. Membr. Biol., 114 (1990) 175-188. M.P. Mahaut-Smith, and M.J. Mason, Ca2+-activated K + channels in rat thymic lymphocytes: activation by concanavalin A. J. PhysioL (London) 439 (1991) 513-528.

J.M. Witkowski/Mech. Ageing Dev. 71 (1993) 111-129 31 32 33 34

35 36

37 38 39 40

129

H.J. Woehlck, and F.V. McCann, Action potentials in human macrophages are calcium spikes. Cell Biol. Int. Rep., 10 (1986) 517-525. K. Takagi, K. Zenita and Y. Ichikawa, Characterization of 'abp38' in cytoskeleton of mouse myeloid leukemia cells and distribution of abp38 in various tissues. J. Biochem., 101 (1987) 63-71. Y. Moriyama, Potassium ion-dependent proton effiux and depolarization from spleen lysosomes. Biochem. Biophys. Res. Commun., 156 (1988)211-216. G.F. Fuhrmann, W. Schwarz, R. Kersten and H. Sdun, Effects ofvanadate, menadione and menadione analogs on the Ca2+-activated K + channels in human red cells. Possible relations to membranebound oxidoreductase activity. Bioehim. Biophys. Acta 820 (1985) 223-234. J. Mysliwska, A. Mysliwski, K. Grzegorzewski and J. Witkowski, Age-related changes of activity of acid phosphatase in PHA-stimulated lymphocytes. Exp. Gerontol., 20 (1985) 7-14. J. Mysliwska, A. Mysliwski and J. Witkowski, Age-dependent decline of natural killer and antibodydependent cell mediated cytotoxicity activity of human lymphocytes is connected with decrease of their acid phosphatase activity. Mech. Ageing Dev., 31 (1985) 1-11. E.W. Gelfand, R.K. Cheung and S. Grinstein, Role of membrane potential in the regulation of lectin-induced calcium uptake. J. Cell. Physiol., 121 (1984) 533. R.A. Miller, B. Jacobson, G. Weil and E. Simons, Diminished calcium influx in lectin-stimulated T cells from old mice. J. Cell. Physiol., 132 (1987) 337-342. B.D. Freedman, M.A. Price, and C. Deutsch, Evidence for voltage modulation of IL-2 production in mitogen-stimulated human peripheral blood lymphocytes. J. Immunol., 149 (1992) 3784-3794. J. Witkowski and H.S. Micklem, Decreased membrane potential of T lymphocytes in ageing mice: flow cytometric studies with a carbocyanine dye. Immunology, 56 (1985) 307-313.