Intracellular free calcium levels in mononuclear cells of patients with cystic fibrosis and normal controls

Intracellular free calcium levels in mononuclear cells of patients with cystic fibrosis and normal controls


786KB Sizes 0 Downloads 2 Views

Cetl Calcium (1987) 8.5343 0 Longman UK Ltd 1987





M.H. Schtlni,F. Schllni-Affolter,0. Jeffery and S.Katz Faculty of Pharmaceutical Sciences, University of British Columbia, 2146 East Mall, Vancouver, B.C., V6T lW5, Canada. (Reprint requests to SK) ABSTRACT The time course of resting free intracellular calcium concentrations in isolated mononuclear blood cells following a one hour incubation period with the fluorescent dye quin2 was evaluated. Under equal experimental conditions, a slaw time-dependent increase of intracellular free calciun in patients with cystic fibrosis and normal healthy controls was noted. Using regression analysis, cystic fibrosis patients were seen to exhibit significantly higher free intracellular calcium concentrations than the controls over the time span covered. At an arbitrarily selected time (60 minutes) the free calcium level was 143.7 f 4.3 nM (SEM) in the patients, and 125.5 f 2.6 nM in contr 1s. ')i From these data it is concluded that neglecting the time-dependent (Ca!! changes following quin2 incubation leads to over- and/or underestimation of the unstimulated resting, basic free calcium levels and prevents the detection of differences between normals and cystic fibrosis patients. INTRODUCTION It has been suggested that in cystic fibrosis (CF), the autosomal recessive inherited disease, calciun metabolism might be impaired at the cellular level (1). Abnormalities in intracellular calciun regulation have been reported to occur in red blood cells (2), fibroblasts (3), buccal epithelial cells (4) and parotid acinar cells (5) of CF patients. Free ionized intracellular calciun plays a key role in cellular calcium metabolism. The availability of a new technique for direct measurements of free intracellular calcium in living cells has enhanced the insight into the suggested putative calcium disturbance in CF. Using the fluorescent dye quin2, intracellular free calciun levels have been determined in a variety of cells from animals (6, 7, 8) and humans (9, 10) including lymphocytes (11, 12, 131, lymphoblasts (11) and polynuclear granulocytes (13, 14) of CF patients. Either normal (11, 12, 13) or elevated (13) calciun levels have been found in CF blood cells, and


some evidence has been presented that cellular calcium metabolism might be impaired. Despite these studies, repetitive measurements of unstfmulated resting free intracellular calcium levels fCa2+)i in human blood cells have not been done. In addition, measurement of such levels over time following incubation with quin2 have so far not been reported. In this study we collected data that show that the time factor in such experiments may play a crucial role in elucidating minor abnormalities between patients and controls. MATERIALS AND METHODS After informed consent, venous blood from 10 CF patients (age 15 to 26 years, disease confirmed by elevated sweat chloride and sodium levels together with chronic pulmonary disease and pancreatic insufficiency) and 14 agematched healthy controls was drawn either in precooled heparinized or EDTA tubes and kept on ice. Within 30 minutes, mononuclear cells were isolated by means of Ficoll-Hepaque gradient centrifugation (400 xg, 30 minutes, 1O'C) and then washed 3 times in Hepes (4-(P-hydroxyethylj-1-peperazineethylsulphonic acid) buffered RPM1 (lymphocyte incubation medium from Gibco Laboratories) 1640 (30 mM Hepes, pi 7.4). The isolated cells were counted before and after the incubation with the fluorescent dye and at intervals in the procedure; cell viability was assessed by trypan blue exclusion. From a 10 mM quin2tetraacetoxymethylester (quin2/AM, Calbiochem, La Jolla, CA) stock solution in dry DMSO a 25 uM quin2/AM loading solution for cell incubation was prepared and the mononuclear cells incubated at different cell densities ranging from 5 - 20 x 106 cells in a final volume of 2 ml. After 30 minutes of incubation at 37-C the cells were diluted 1:5 with prewarmed fresh RPM1 and a further 30 minute incubation performed. The loading was stopped by centrifugation at 200 xg for 10 minutes, the cells were resuspended and washed three times in a Hepes-buffer containing 1 mM CaC12 (NaCl 123 mM, KC1 5 mM, Na2HPO4 1 mM, MgS04 0.5 mM, Glucose 5 mM, Hepes 25 mM, @i 7.4 at 37'C) and kept at room temperature until fluorescent reading. Cell counts and viability determinations were done before transferring cells to the cuvette, to assure that the same cell amount was present in every determination. A short ( 1 min) high speed centrifugation step (Eppendorf microcentrifuge) was done before transferring the cells into the cuvette, to remove external quin2 and cell debris. The cells were then resuspended in the above buffer for measurements. The quin2 fluorescence was recorded on a Perkin Elmer 650-10s fluorimeter (excitation 339 nm, slit 4; emission 492 nm, slit 10) with temperature control (37'C) and slight stirring (cuvette stirrer, Medtronics, Philadelphia). After at least 3 minutes of temperature equilibration, the fluorescence trace was recorded for a further 3 minutes to assess the stable fluorescent signal F (expressed in relative fluorescent intensity units). The cells were then disrupted by the addition of Triton X-100 (final cont. O.l%, v/v) and Fmax was determined; Fmin was achieved by adding EGTA (final cont. 4 mM) and Tris base (final cont. 20 mM) in excess (Fmax and Fmin representing fluorescence at high and very low calcium concentration in the medium) (6). In all studies, values were corrected for cell autofluorescence over the entire experimental period; autofluorescence was found to be very weak (5 R.F .I.). In some experiments, the ionophore ionomycin (final cont. 0.5 uM, Calbiochem) was used to establish Fmax in the cells. Occasionally dye leaking was checked and corrected by the reverse procedure described by Arslan et al. (15) adding EGTA and Tris before Triton X-100 treatment. The time span (in minutes) from the end of the quin2/ AM incubation period until the fluorescent reading was carefully recorded and 54

at least two and in general three to five repetitive determinations from the cells of the same batch and the same subject were done. (Ca*+)i was determined using the formula given by Tsien (Cat* = 115 x (F-Fminf/ Fmax-F)) assuming a Kd of 115 nM (6). Intracellular quin2 (free acid, tetrapotassium salt) concentrations (Q)i were calculated assuming a similar cell volume in CF and controls of 0.28 u1/106 cells (16) using standard curves from determinations of quin2 fluorescence in solutions containing 1 mM C&12 and disrupted non-loaded cells at the same cell density. Cell counts in the cuvette were routinely performed and equal cell densities for the s me subject were used. In general, cell densities ranging from 1.5 to 4.5 x 108 cells, depending on the subject's yield, were used. Loading efficiency was determined and hydrolysis of quin2/AM was checked by emission scanning giving similar results as indicated by Tsien (6). The results of the single determinations were plotted against time after incubation (time 0 = end of incubation period). The regression lines of the calcium values versus time for each patient and control (n = 10, n = 14, respectively) were comparable to the regression line of the pooled single values versus time in each group (n = 27, n= 41, respectively). Subsequently, therefore, data of the single determinations were used. Regression analysis and Anova was used to calculate time dependence and differences between CF and controls (computer program Statpro from Wadsworth Electronic Publishing Company, Boston, USA, run on an extended Apple II+). The level of significance was set at P












1.62 67.5






49 24

52 24.5 46 22

48 23.5


0 0





3~10~ calls/ml [llUill]i : 2.06 mM release/h: 0.11 mM


- 5.3 K

0 0







TIME hid

FIG. 1. Representative experiment with mononuclear cells of a healthy control. For abbreviations, see text. Figure 1 shows the time course of (Ca2+)i in mononuclear cells of a normal healthy control subject over five hours after incubation together with the corresponding alterations in F, Fmax, Fmin and (Q)i; the same amount of cells--corrected for the observed cell death--were used in each experiment. Great care was taken to always collect the same amount of cells from the stock solution and to meticulously clean the cuvette and stirrer after each determination to avoid carry over of EGTA from the preceeding experiment. As can be seen in this typical experiment, a linear increase in the resting free calcium levels occurs which is paralleled by a slow decrease of Fmax, F and Fmin. The initial intracellular quin2 concentration fell from 2.08 mM to 1.62 mM resulting in a calculated decrease of 5.3% over one hour. This value is cornarable to those reported in the literature (6). After six hours the (Cak+ )i rose from 110 nM to 155 nM corresponding to a calcium-saturation increase Sf 9% and thus representing a 40% change from the initially determined (Ca +)i. Figure 2 shows the pooled data and regression lines of F measured as a function of time in each individual subject. Identical slopes of the regression lines as well as comparable variations for CF and normals are indicated. Similar results were obtained by plotting Fmax (CF: ~379.7 -0.086x ; normals: y=85.3 - 0.106x) and Fmin (CF: y-25.4 - 0.0192x; normals; y=26.7 - 0.0219x) against time (graphs not shown). However, the regression lines for all these parameters were distinctly different from zero (~~0.01). Computing the dif56

0 0

N: y m 55.0- 0.&21x6-d CF: ym 55.9-o.@!i17x t-1


FIG. 2. Time course of F following the end of the incubation period (= time 0) with regression lines for patients (n=29) and controls (n=41). ference Fmax-F (denominator of the calculation formula) for every value and plotting the results versus time (Fig. 3) indicated a statistically significant difference between the elevation (intercept with the y-axis) of the regression lines for the CF patients versus the controls (pCO.01). On the other hand, the elevation and the slopes of the regression lines of the nofflinator (F-Fmin) were equal (not shown). Consequently, as shown in Figure 4 on the basis of Tsien's formula the calculated (Ca2+)i values differed significantly between CF and controls. The mean change in the calcium concentration over time was 6.36 nM/hour for CF and 12.18 nM/hour for the controls. This increase as indicated by the slope of the regression line was significantly different from zero for the controls (p






o N:

yz 29.5-0.0540x6-d

l IX:

yr 23.7-0.0340x &)



TIME (min)

FI6. 3. Time course and regression lines for the calculated difference Fmax-F for CF patients and controls (same n as in Figure 2). Plots of (Ca2+)i against (Q)i showed no differences between CF and controls and the regression lines were not different from zero revealing no influence of the quin2 concentrations on the intracellular calcium levels (not shown). Stepwise correlations for arbitrarily chosen intracellular quin2 concentrations (0.5-1.5 mM, 1.51-2.5 mM and 2.51-3.5 mM) versus (Ca+2)i indicated no significant differences; the corresponding regression lines were not different from zero. Furthermore, partial regression lines calculated for every hour were not different from the overall regression line for the whole time span, indicating that there was no line weighting due to values being determined over a relatively long period. Taking into account the observed time dependence of fCa2+)i in CF and normals and the above-mentioned similarity in the slppes .forthe regression lines, two normalized regression equations with equal slop+ were computed (CF: y = 133.6 + 0.168x; normals: y = 115.4 + 0.168~). Based on these equations, an arbitrarily selected time (60 minutes following the end of the incubation period was chosen to calculate the 6O(Ca1'11 values. The CF patients exhibited a value of 143.7 f 4.3 nM and the normal controls, 125.5 f 2.6 nM; this difference was statistically significant (p < 0.001, Students t-test). To rule out several possible interfering factors a nunber of changes in the experimental design were carried out: cells were oxygenated after incubation tg5% O2 + 5% CO21 and kept in the dark to avoid possible bleaching by 58

o N:

y~106.6*0.203x c--)

l CF: y=145.7+0.106x c-1

[CA+*]i (nM)

ml * *

0 0





50 'i 50



350 TlME(min)

Time course and regression lines of intracellular calciun levels FIG. 4. in CF patients and controls (same n as in Figure 2). common light. Both interventions did not influence the time increase in (Ca*+)i. Altering the procedure by keeping the cells in fresh RPM1 after incubation and changing to buffer in the last step did.not neutralize the effect. Nor did Fmax generation by ionomycin eliminate the time dependence noted. Using the reverse procedure of adding EGTA and Tris before lysing the cells was not successful in preventing the time dependent change in (Ca+*)i. DISCUSSION Despite the broad application of the quin2 methodology for determining intracellular free calcium in living cells, reports on repetitive measurements fran the same batch of cells in humans have not been published. Investigation of the reproducibility of such determinations has generated little attention. This may be due to the specific aim of most of the studies. In general, changes in intracellular calcium over a short period in response to extra- or intracellular stimuli are studied (17). In such experiments, the procedure described herein is only used to calibrate the initial free calcium level. Subsequent changes.are then determined as an increase or decrease from this initial value. Only a few studies thus far published attempted to compare resting intracellular calciun levels in patients and control samples (11, 12, 13, 14, 18). Such determinations were mainly done in cystic fibrosis patients, with conflicting results (11, 12, 13, 14). As can be seen by our 59

[n]i hd

o N:




l CF:











FI6. 5. I ntracellular quin2 concentrations with regression lines plotted inst time following the end of the incubation period (same n as in Figure

2). data, the protocol and timing of (Ca2+)i measurements may have a profound influence on the interpretation of such results. Neglecting the time dependence of the measurements leads to over- or underestimation of the basic, unstimulated free calcium levels and prevents the detection of differences between CF and controls. This especially occurs when measurements are made at different times foliowing the incubation period. Comparison of data fran two different groups (diseased and healthy) under these conditions is inaccurate. It might be argued that no determinations have been done before 30 minutes following the end of the incubation period. This is due to the standard procedure described by Tsien (61, and used by others where incubation is stopped by centrifugation and followed by washing steps either in fresh RPM1 or calciun buffer. Following this procedure it is hardly possible to start fluorescent reading before 20 to 30 minutes after incubation since it takes this amount of time to process the cells. Using repetitive manual cell counting prolongs this time period. Since the generation of F, Fmax and Fmin after temperature equilibration in the cuvette takes another lo-15 minutes, it is obvious that repetitive determinations from the same cell batch should be carefully time-monitored. The present findings illustrate that if the time factor is neglected, the application of Tsien s formula ml ht lead to inconsistant results. Furthermore, for a given l%change of F 9representing 1 relative fluorescent unit) at constant Fmax and Fmin a 15% change in calculated free calciun is obtained. A 7% change results when either Fmin or Fmax are 60

altered by one unit and F is constant. The observed instability of F, Fmin or Fmax over time is hard to overcome, since several factors contribute to these fluorescent properties. As long as these changes run in parallel, the final calculated calciun concentration is not affected. Instrunentation stability, cell amount and consequently cell death, ATP depletion, buffering capacity, carry-over of quin2 and bleaching as well as energy loss in the cell due to toxic product generation (formaldehyde, protons, acetate) may all contribute and hamper the final results (19). Concerns that these results can be explained on the basis of somewhat higher intracellular quin2 concentrations must be considered. It has been shown that this factor must be taken into account (9). However, we believe that our calculated higher concentrations were mainly due to the incubation of smaller amounts of cells in combination with a 1:5 dilution after the first 30 minutes of incubation. Furthermore, the use of standard curves obtained in the measurement buffer, which is not adjusted for intracellular ionic content, results in higher calculated quin2 concentrations. However, comparison of the loading efficiency, the time for quin2/AM hydrolysis, quin2 leakage and the independence of (Ca+2)i to the measured intracellular quin2 concentration, showed similar results to published reports and therefore do not support the suggestion that the quin2 concentrations used were crucial to the results obtained. Comparing cells from CF patients and normals under these same conditions resulted in the observation of significant differences. From these data it seems quite reasonable to deduce Ia calcium abnormality in CF cells. This is in contrast to the results of Grinstein et al. Ill), which have been confined by Waller et al. (12). No time control is mentioned in these reports. However, Waller et al. (12) indicated that the total calciun concentration (free and bound) in CF lymphocytes was considerably higher than in controls. Furthermore, conflicting results of intracellular calciun levels in other blood cells (neutrophilsl of CF patients have been reported by Suter et al. (14) and Cabrini et al. (13). Our current data suggest that free calcfun levels in mononuclear cells of CF patients are altered; taking into account the report of Waller et al. (121, one can speculate that the transient movement of calciun into overfilled intracel ular stores might be disturbed resulting in temporarily higher free (Cah'Ii. Further work is required to elucidate the putative calcitm alteration in CF cells. ACKNOWLEDGEMENTS M.Sch. was supported by grants of the Swiss National Foundation for Scientific Research, the Swiss CF Association and the Swiss Pneunology Society. The help of Drs. A.G.F. Davidson and L.T.K. Wong is gratefully acknowledged. This study was supported by the Canadian Cystic Fibrosis Foundation. REFERENCES 1.

Katz S, Schbni MH, Bridges MA (1984). The calciun hypothesis of cystic fibrosis. Cell calcium 5421-440



Katz S (1978). Calcium and sodium transport recesses in patients with cystic fibrosis. A specific decrease in (Mg+B + Ca+2)-ATPase activity in erythrocyte membranes from cystic fibrosis patients. Pediatric Research 12:1033-1038


Katz S, Ansah T-A (1980). (Mg+2 + Cat2)-ATPase activity in plasma membrane preparation of cultured fibroblasts from cystic fibrosis patients and normal volunteers. Clin Chim Acta 100:245-252


Mangos JA, Donnelly WH (1981). Isolated parotid acinar cells from patients with cystic fibrosis: morphology and composition. J Dent Res 60:19-25


Mangos JA (1983). Cystolic Cat2 measurements in epithelial cells from cystic fibrosis patients and controls. Abstract Booklet, 12th Annual Meeting of the European Working Group for Cystic Fibrosis, p 21, Athens


Tsien RY, Pozzan T, Rink TJ (1982). Calcium homeostasis in intact lymphocytes: Cytoplasm free calciun monitored with a new intracellularly trapped fluorescent indicator. J Cell Biol 94:325-334.


Blinks JR, Wier WG, Hess P, Prendergast G (1982). Measurement of Ca2+ concentrations in living cells. Prog Biophys Molec Biol 4&l-114


Hesketh TR, Smith GA, Moore JP, Taylor MY, Metcalfe JC (1983). Free cytoplasmic calciun concentration and the mitogenic stimulation of lymphocytesr J 8iol Chem 258:4876-4882


Lew PD, Wollheim CB, Waldvogel FA Pozzan, T (1984). Modulation of cytosolic-free calcimn transients by changes in intracellular calcium-buffering capacity: correlation with exocytosis and 02-production in human neutrophils. J Cell Biol 99: 1212-1220


Rink TJ, Smith W, Tsien RY (1982). Cytoplasmic free Cat2 in human platelets: Cat2 thresholds and Ca-independent activation for shapechange and secretion. FEBS letters 148: 21-26


Grinstein S, Elder B, Clarke CA, Buchwald M (1984). Is cytoplasmic Ca2+ elevated in cystic fibrosis Biochem Biophys Acta 769:270-274


Waller RL, Brattin WJ, Dearborn DG (1984). Cytosolic free calcium concentration and intracellular calcimn distribution in lymphocytes from cystic fibrosis patients. Life Sci 35: 775-781


Cabrini G, De Togni P (1985). Increased cytosolic calcium in cystic fibrosis neutrophils; effect on stimulus-secretion coupling. Life Sci 36:1561-1567


Suter S, Lew PD, Ballaman J, Waldvogel FA (19851. Intracellular calcitnn handling in cystic fibrosis: normal cytosolic calciu?,and intracellular calcium stores in neutrophils. Pediatr Res 19:346-349



Arsldn P2 Di Virgilio F, Beltrame M, Tsien RY, Pozzan T (1985). Cytosolic Ca + homeostasis in Ehrlich and Yoshida carcinomas. J Biol Chem 260:2719-2727


Wissenschaftliche Tabellen Geigy (1979). Teilband Haematologie, Humangenetik, 8. Auflage, Sasel, Switzerland, p 188


Tsien RY, Pozzan T, Rink TJ (1984). Measuring and manipulating cytosolic Ca+2 with trapped indicators. Trends in Biochem Sci 9:263-266


Erne P, Bolli P, BOrgisser E, BllhlerFR (1984). Correlation of platelet calcium with blood pressure. Effect of antihypertensive theraPY* N Engl J Med 310:1084-1088


Rink TJ, Pozzan T (1985). IJsingQuin2 in cell suspensions. Cell Calcium 6:133-144

received revised

28.5.86 version


and accepted