Free Cytoplasmic Calcium Concentration and the Mitogenic Stimulation of Lymphocytes*

The effedts of the lectins concanavalin A, succiny- lated concanavalin A, and wheat germ agglutinin on the free cytoplasmic Ca2+ concentration in mouse thy- mocytes were measured using the fluorescent Ca2+ indicator “quin 2” (Tsien, R. Y. (1980) Biochemistry 19, 2396-2404) and compared with the metabolic and mi- togenic effects of the lectins on the cells. Within l min of adding each ligand, there is a dose-dependent in- crease in the free cytoplasmic Ca2+ concentration reported by quin 2. This response is selective for Ca2+, but it does not coincide closely with the subsequent mitogenic stimulation at 48 h by concanavalin A or succinyl concanavalin A. The nonmitogenic lectin wheat germ agglutinin also causes an increase in free cytoplasmic Ca2+ concentration and early metabolic stimulation of the cells, but stimulation is self-aborted before DNA synthesis occurs. At the intracellular concentrations of quin 2 required for measurement of the free Ca2+ concentration, the chelator causes early met- abolic stimulation of the cells very similar to that produced by concanavalin A and the mitogenic Ca2+ iono- phore A23187. Thus, phosphatidylinositol metabolism and lactate production are stimulated in mouse thy- mocytes and pig lymphocytes within 1 h of loading with quin 2 and significant increases in RNA synthesis occur after 8 h. Quin 2 causes mitogenic stimulation of pig lymphocytes measured as

3 Supported by grants from the Science Research Council. cated that mitogenic lectins such as concanavalin A can cause substantial increases in 45Ca2+ associated with the cells, but the magnitude and duration of the effect and its relevance to mitogenic stimulation are highly controversial (5)(6)(7). We have proposed as a working hypothesis that an increase in the free cytoplasmic Ca2+ concentration ([Cali) within the range to M constitutes the primary mitogenic signal for the transition out of the resting state (Go) into the cell cycle and that this signal must persist for about 20 h to commit the maximum proportion of cells to DNA synthesis (1). If [Cali is lowered back into the Go range (for example by removing the mitogen from the cells) before commitment to DNA synthesis has occurred, then the cells will return to the resting (Go) state. If [Cali is raised above about M, however, mitogenic stimulation is reversibly inhibited in the short term and toxic effects may occur if [Cali is maintained at high concentrations for long periods.
Recently, it has been demonstrated that new Caz+ chelators derived from EGTA' can be introduced into a variety of cells, using the acetoxy methyl ester derivatives of the chelators (8). The nonpolar ester derivatives cross the plasma membrane and are hydrolyzed intracellular to the parent tetracarboxylate anions which are membrane-impermeant. One of the chelators developed by Tsien (9) was recently used as a fluorescent indicator to detect increases in [Cali in lymphocytes within 1 min of the addition of ConA or phytohemagglutinin (10). The increase in [Cali indicated by quin 2 from about 1.2 X 10" M in control cells to about 2.4 X M after addition of mitogenic concentrations of the lectins is consistent with the "calcium hypothesis" for the primary mitogenic signal summarized above.
To assess further the relationship between the early [Cali changes and subsequent mitogenic stimulation, we have compared the two responses as a function of the concentration of mitogenic and nonmitogenic lectins. However, we found that intracellular quin 2 at the concentrations required to measure [Cali itself caused immediate metabolic and subsequent mitogenic stimulations of lymphocytes which were very similar to those produced by the mitogens ConA and A23187. These data imply that all of themeasurements of [Cali using quin 2 are made in cells already stimulated by the presence of the chelator. We have therefore examined the effect of the concentration of intracellular quin 2 and of extracellular Ca*+ on the indicated [Cali, on the assumption that UCa]i would remain constant if the cells were not perturbed by quin 2.

RESULTS
[Ca], in Thymocytes Loaded with quin 2-The estimation of the per cent saturation of quin 2 by Ca2' is illustrated in Fig. 1. The cells were treated with 50 nM ionomycin which causes the intracellular quin 2 to be saturated with Ca2' from the medium (1.0 m). The background signal from the cells (autofluorescence and scatterin) was determined by adding 0.5 mM MnC12. The affinity of Mn2' for quin 2 is 102.7 greater than the affinity of Ca2+ (Table IS in Miniprint) and therefore 0.5 mM Mn2+ quenches quin 2 fluorescence by >99% in the presence of 1.0 mM Ca2+. Addition of 50 nM ionomycin and 0.5 mM MnClz to cells without quin 2 did not affect the background signal. The complete quenching of intracellular quin 2 fluorescence by Mn2+ ions therefore demonstrates that all of the quin 2 becomes accessible to extracellular Mn2+ in the presence of ionomycin. The per cent saturation of quin 2 by Ca2+ was calculated from the expression given in the legend to Fig. 1 Fig. 2 A . The increases in fluorescence induced by ConA were observed with internal quin 2 concentrations ranging from about 0.1 to 2.0 m, with similar delays and time courses for the fluorescence increases at all quin 2 concentrations. Control experiments in which 0.1 mM Mn2' (or 0.01 mM Ce3') was added before ConA showed that the fluorescence increase is similar to that in the absence of quenching M"' ions ( Fig. 2B). This indicates that the fluorescence increase is not due to a significant leakage of quin 2 out of the cells when ConA binds to the cell surface. Further control experiments confirmed that [3H]quin 2 leakage caused by the addition of ConA accounts for only a minor part (<20%) of the fluorescence increase in response to the ligand. The observation that the quin 2 fluorescence increase in response to ConA is obtained in the presence of external Mn2' or CeB+ shows that the increase in intracellular cation concentration is highly selective for Ca2+ compared with the fluorescence quenching ions (see Table IS in Miniprint).
The ion selectivity of the effect of ConA is consistent with an increase in [Cali as a specific response of thymocytes to the ligand, rather than a general increase in ion permeability of the plasma membrane as a result of the cross-linking of proteins by ConA. This prompts the question of whether the increase in [Ca], reported by quin 2 is correlated with mitogenic stimulation by the lectin. In Fig. 3A, the increase in per cent saturation of quin 2 by Ca2+ is shown as a function of the amount of ConA added. Serial additions of 0.05 p g / d of ConA x 100 where Z and AI are the fluorescence int,ensities as illustrated and the fluorescence intensity of free quin 2 or the Mg2+-quin 2 complex at pH 7.1 is taken as 0.16A1 (see Table IS in Miniprint). The correction to Z for leakage of quin 2 into the medium was less than 5% in all experiments (see Miniprint).
give proegressive increases in fluorescence, with the maximal fluorescence increase remaining constant between 0.5 and 10 pg/ml of ConA. This profile can be compared with the mitogenic stimulation of the same cells by ConA (Fig. 3B), which shows ( 4 0 % of maximal stimulation below 0.3 pg/ml or above 2 p g / d of ConA. The concentration of ConA required for maximal mitogenic stimulation therefore coincides quite closely with the lowest ConA concentration required to produce the maximum increase in quin 2 fluorescence. On the other hand, the, self-inhibition of mitogenic stimulation by supraoptimal concentrations of ConA cannot be attributed t.o an excessive early incrase in [Cali. The results for ConA may be compared with the corresponding data for succinyl-ConA (Fig. 3, A and B ) . Succinyl-ConA causes maximal mitogenic stimulation over a wider range of concentrations than ConA and at 10 pg/ml caused a similar increase in [Cali to optimal ConA. It should be noted, however, that the dose-response curves for the increase in [Ca], and mitogenic stimulation by succinyl-ConA do not coincide closely (compare A and B of Fig. 3 ) . In marked contrast, the threshold concentrations of both ConA and succinyl-ConA for the stimulation of PI metabolism and mitogenic stimulation are closely correlated (compare B and C of Fig. 3 ) .
Effect of Wheat Germ Agglutinin-The response to WGA is of interest in that it causes an increase in [Ca], with a concentration dependence intermediate between ConA and succinyl-ConA, although it is not mitogenic (Fig. 3, A and B ) .
WGA does not cause a further increase in quin 2 fluorescence after the maximum quin 2 response to ConA has been elicited and concentrations of WGA which cause maximal or submaximal increases in [Cali completely block the mitogenic action of optimal concentrations of ConA (Fig. 3 B ) . Since WGA is Metabolic and Mitogenic Responses to Quin 2-In the course of the above experiments it was found that quin 2 itself caused early metabolic stimulation of mouse thymocytes and pig lymphocytes. In pig lymphocytes, these early responses ultimately lead to increased DNA synthesis at 48 h. Since the mechanism of mitogenic stimulation by quin 2 is of considerable interest, the progressive responses of the cells to quin 2 were characterized.
-2Phosphatidylinositol Metabolism-A well established early response in T cells to mitogens is an increase in the turnover of PI (see above), usually assayed as an increase in the incorporation of 32Pi into phospholipid and detectable within 10 min of the addition of mitogen (11). The effect on the incorporation of [3H]inositol into PI of incubating mouse thymocytes and pig lymphocytes for 1 h with increasing concentrations of quin 2 AM ester is shown in Fig. 4. It can be seen that at concentrations above about 0.1 p~ quin 2 AM ester there is a substantial increase in [3H]PI obtained from both cell types. It is clear that loading the cells with quin 2 using concentrations of the quin 2 AM ester in the range 0.1-1 PM causes increases in r3H]PI similar to those observed with mitogenic concentrations of Con A and A23187 (compare with Fig. 4

s in Miniprint).
Lactate Production-Lymphocytes respond to ConA and other polyvalent lectins at optimal mitogenic concentrations with an increase in lactate output which follows the time course of increased DNA synthesis (12).   Procedures." Lactate production is expressed relative to control samples without C3H]quin 2 (100% = 9-12 fmol/cell/h for pig lymphocytes and 2-3 fmol/cell/h for mouse thymocytes).
which have no detectable effect on the ATP level (see Fig. 1s in Miniprint). However, it should be emphasized that although both quin 2 and other mitogens induce early changes in lactate production, many factors can influence glycolysis. For example, inhibition of mitochondrial ATP synthesis by quin 2 might stimulate glycolysis to cause the large increases in lactate production shown in Fig. 5 for thymocytes, and the decrease in lactate production above 0.4 mM quin 2 may result from the progressive loss in cell viability at high internal quin 2 concentrations, as evidenced by the decrease in cellular ATP levels ( Fig. 1s in Miniprint). In contrast to the mouse thymocytes, there was only a small stimulation of lactate output from pig lymphocytes (up to 25%), but a significant inhibition of output to below the level of unloaded control cells occurred at above about 1 mM internal quin 2.
Stimulation of Uridine Uptake by quin 2-While the relationship between the early metabolic changes which occur within an hour of the addition of mitogens and the subsequent stimulation of DNA synthesis remains obscure (l), the prior stimulation of RNA synthesis is apparently essential for mi-I *. togenic stimulation. The specific inhibition of RNA synthesis by a-amanitin causes complete and reversible block of the subsequent stimulation of DNA synthesis by mitogens (13). Fig. 6 shows that quin 2 caused a large increase in uptake of [3H]uridine into pig lymphocytes measured at 18-19 h after loading the cells. Quin 2 caused small but significant increases in [3H]uridine uptake into both pig and mouse lymphocytes measured at 8 to 9 h. No effect was observed in mouse cells at the later time due to toxic effects of quin 2 on these cells. Stimulation of Thymidine Uptake by quine 2-Mitogenic stimulation of pig lymphocytes measured as r3H]thymidine uptake 42 to 48 h after the addition of quin 2 AM ester is shown in Fig. 7. Significant stimulation was observed in most experiments ( n = 27), varying from 15 to 100% of the mitogenic stimulation by A23187 of the same pig lymphocyte prepara-tions in the absence of quin 2 (Fig. 7). Both A23187 and quin 2 AM ester show maximal mitogenic activity at about 0.1

Correlation of ICaJ Changes with Mitogenic Stimulation-Concentrations
of ConA and succinyl-ConA which cause optimal mitogenic stimulation also cause the maximal increase in [Cali as indicated by quin 2. The two responses do not, however, coincide closely for either ligand. Submitogenic concentrations of ConA caused detectable increases in [Cali, whereas no increase in [Cali was observed at suboptimal succinyl-ConA concentrations which caused substantial mitogenic stimulation. In contrast, the dose-response curves for PI metabolism are closely correlated with mitogenic stimulation by ConA and succinyl-ConA in the suboptimal concentration range. Until it can be established whether quin 2 itself affects [Cali and its response to these ligand, the correlation between changes in [Cali and mitogenic stimulation remains uncertain. It is significant, however, that the increase in [Ca] i in response to ConA is a specific effect and not attributable to a general increase in permeability to M"+ ions.
At supramitogenic concentrations of ConA and succinyl-Con A, all of the early responses associated with mitogenic stimulation are observed (i.e. increases in [Cali, PI metabolism, and lactate production). The increase in [Cali is the same at mitogenic and supramitogenic concentrations of ConA, suggesting that self-inhibition of mitogenic stimulation cannot be attributed to an excessive early increase in [Ca],, although a subsequent slow increase in [Cali a t supramitogenic concentrations cannot be excluded.
We have suggested previously (14) that self-inhibition of mitogenic stimulation by supraoptimal ConA concentrations may result from premature removal of the ligand and its receptors from the cell surface by rapid cap formation. This terminates the primary mitogenic signal and the cells return to Go before commitment to DNA synthesis can occur. We have shown here that WGA stimulates both early [Cali increases and PI metabolism, but is not mitogenic. Since WGA has been reported previously to cap rapidly on lymphocytes (15), we compared the rates of cap formation by fluoresceinlabeled WGA and ConA at the minimum concentration of each ligand which produced the maximum increase in [Cali (10 pg/ml and 0.8 pg/ml, respectively). The WGA capped with a half-time of 30 min compared with 9 h for ConA. We have shown previously that supramitogenic concentrations of ConA (>3 pg/ml) cap with a half-time of less than 4 h and we therefore suggest that WGA is not mitogenic because it selfaborts stimulation by capping prematurely at concentrations which produce the required increase in [Cali or other primary mitogenic signals.
Metabolic and Mitogenic Stimulation by quin 2"Ambiguity in correlating [Cali changes with mitogenic stimulation may arise because all of the changes in [Cali in response to mitogens are measured in cells already stimulated metabolically by quin 2. The correlation of metabolic stimulation by intracellular chelators with their affinities for Ca2+ suggests that quin 2 may stimulate the cells by affecting the intracellular concentration of Ca2+ or other M n + ions (see Miniprint).
The large Ca2+ fluxes across the plasma membrane of thymocytes, equivalent to -0.2 mM/cell/min (see Miniprint) imply that extracellular Ca2+ equilibrate rapidly with intracellular quin 2. Under these conditions, [Cali is determined by the kinetic parameters of the Ca2+ fluxes across the plasma membrane and it is possible in principle to load the cells with increasing concentrations of quin 2 without affecting [Cali a t steady state. Thus, quin 2 would act purely as an indicator without affecting [Ca],. However, the apparent fractional sat-uration of quin 2 by Ca2+ increased as the intracellular concentration of quin 2 was increased from 0.1 to 0.5 mM. This may indicate either that quin 2 causes an increase in [Cali or that there are impermenat M"+ ions in the cells which quench a significant proportion of the quin 2 fluorescence at low quin 2 concentrations. Either mechanism implies a perturbation of intracellular cation concentrations by quin 2. It should be possible to distinguish these mechanisms from each other and from any systematic experimental errors in the use of quin 2 a t low intracellular concentrations (see Miniprint) by using more sensitive analogues of quin 2.
Perturbation of [Ca], homeostasis by quin 2 may account for the increase in [Cali indicated by quin 2 in response to increases in Ca'+ concentration in the medium (see Miniprint). The increase in [Cali when the external Ca2+ concentration is raised from 0.1 to 1 mM is comparable to that produced by mitogenic concentrations of lectins. However, the same change in external Ca2+ concentration does not cause stimulation of PI metabolism, lactate production, or mitogenesis in cells without quin 2. The key question is whether significant [Cali changes occur in cells without quin 2 in response to changes in external Ca'+ concentration or whether strict [Cali homeostasis is maintained in unstimulated cells as would be expected if the primary mitogenic signal involves the small increase in [Cali indicated by quin 2.
The mitogenic action of quin 2 on pig lymphocytes does not bear directly on the Ca2+ hypothesis, since any changes in free cytoplasmic Ca2+ concentration induced by quin 2 remain to be established using more sensitive indicators. The pattern of metabolic responses of the cells to quin 2 is very similar to that of well-characterized mitogens and the experiments described in the miniprint section show that it is the hydrolyzed form of quin 2 AM ester inside the cells which causes stimulation. The upper limit to the concentration of intracellular quin 2 required for optimal mitogenic stimulation of pig lymphocytes is about 0.1 mM, assuming that all of the quin 2 AM ester is taken up into the cells and hydrolyzed to quin 2. This concentration of intracellular quin 2 causes early metabolic concentrations of ConA and A23187.
We conclude that while the ionic specificity and pattern of responses of [Cali indicated by quin 2 in response to mitogens may be consistent with an increase in [Cali as a primary mitogenic response, a critical assessment of this hypothesis requires a more sensitive indicator of [Ca], than quin 2 that does not perturb the resting cells.