Stimulation of Dictyostelium discoideum with cAMP evokes an elevation of the cytosolic free Ca2+ concentration ([Ca2+]i). The [Ca2+]i-change is composed of liberation of stored Ca2+ and extracellular Ca2+-entry. The significance of the [Ca2+]i-transient for chemotaxis is under debate. Abolition of chemotactic orientation and migration by Ca2+-buffers in the cytosol indicates that a [Ca2+]i-increase is required for chemotaxis. Yet, the iplA- mutant disrupted in a gene bearing similarity to IP3-receptors of higher eukaryotes aggregates despite the absence of a cAMP-induced [Ca2+]i-transient which favours the view that [Ca2+]i-changes are insignificant for chemotaxis.
We investigated Ca2+-fluxes and the effect of their disturbance on chemotaxis and development of iplA- cells. Differentiation was altered as compared to wild type amoebae and sensitive towards manipulation of the level of stored Ca2+. Chemotaxis was impaired when [Ca2+]i-transients were suppressed by the presence of a Ca2+-chelator in the cytosol of the cells. Analysis of ion fluxes revealed that capacitative Ca2+-entry was fully operative in the mutant. In suspensions of intact and permeabilized cells cAMP elicited extracellular Ca2+-influx and liberation of stored Ca2+, respectively, yet to a lesser extent than in wild type. In suspensions of partially purified storage vesicles ATP-induced Ca2+-uptake and Ca2+-release activated by fatty acids or Ca2+-ATPase inhibitors were similar to wild type. Mn2+-quenching of fura2 fluorescence allows to study Ca2+-influx indirectly and revealed that the responsiveness of mutant cells was shifted to higher concentrations: roughly 100 times more Mn2+ was necessary to observe agonist-induced Mn2+-influx. cAMP evoked a [Ca2+]i-elevation when stores were strongly loaded with Ca2+, again with a similar shift in sensitivity in the mutant. In addition, basal [Ca2+]i was significantly lower in iplA- than in wild type amoebae.
These results support the view that [Ca2+]i-transients are essential for chemotaxis and differentiation. Moreover, capacitative and agonist-activated ion fluxes are regulated by separate pathways that are mediated either by two types of channels in the plasma membrane or by distinct mechanisms coupling Ca2+-release from stores to Ca2+-entry in Dictyostelium. The iplA- strain retains the capacitative Ca2+-entry pathway and an impaired agonist-activated pathway that operates with reduced efficiency or at higher ionic pressure.
Aggregation of Dictyostelium discoideum proceeds by an oriented migration of the amoebae towards a source of the attractant cAMP which is synthesized and released periodically by cells in the center of the aggregate. Stimulation with cAMP activates liberation of stored Ca2+ and extracellular Ca2+-entry  leading to a [Ca2+]i-transient [2-4]. Chemotaxis proceeds in the presence of extracellular EGTA but not in the presence of intracellular Ca2+ buffers, so a [Ca2+]i-elevation is necessary and release of stored Ca2+ is sufficient for oriented migration . On the other hand, the view that a [Ca2+]i-increase is essential for chemotaxis was called into question by analysis of a cell line where the iplA gene was disrupted by homologous recombination . The iplA gene is the only gene known in the Dicyostelium genome so far that shares homology with IP3-receptors of higher eukaryotes. However, whether its protein product indeed constitutes a functional IP3-receptor and its cellular localization are not known. The iplA- mutant was found to aggregate and to form fruiting bodies although neither cAMP-activated 45Ca2+-entry nor a [Ca2+]i-elevation were detected . From these results the authors concluded that the iplA gene product is central to the regulation of [Ca2+]i and that its presence and thus the presence of an agonist-activated [Ca2+]i-increase is not necessary for proper chemotaxis and development. However, agents that interfere with IP3-receptor mediated signaling such as XestosponginC  were found to influence not only cAMP-induced Ca2+-fluxes but also the chemotactic response and aggregation of Dictyostelium . In this study we aimed to clarify these conflicting findings and analyzed both, capacitative and chemoattractant-induced Ca2+-fluxes and the effect of their disturbance on chemotaxis and differentiation of the iplA- mutant. Mn2+-influx was used to determine the rates of ion fluxes into cells with filled and emptied stores and related to Ca2+-electrode recordings in cell suspensions. We found that ion fluxes, chemotaxis and differentiation were sensitive towards alteration of the Ca2+-homeostasis. Capacitative Ca2+-entry was normal in the mutant and upon stimulation with agonist Ca2+- and Mn2+-fluxes occurred, yet to a considerably reduced extent. Spontaneous motility and chemotactic performance of mutant amoebae was strongly impaired by the intracellular presence of a Ca2+-chelator.
Extracellular [Ca2+] affects development and chemotaxis of wild type and iplA-
As iplA- cells formed fruiting bodies, albeit somewhat smaller in size, it was concluded that chemotactic aggregation and differentiation was normal . We analyzed development of the mutant in parallel with wild type at various conditions. When cells differentiated on H5-agar plates (control situation) we consistently found a delay in the onset of aggregation by 1–2 h in the mutant; the smaller size of fruiting bodies was due to breaking of aggregation strands yielding smaller mounds (Fig. 1). Next we asked whether the absence or presence of Ca2+ affects development. Differentiation on EGTA-containing agar plates and thus the steady reduction of internal Ca2+-levels dose dependently resulted in a delay of aggregation and a decrease in the size of aggregates and fruiting bodies in both strains. Doses of 5–10 mM EGTA in the agar did not significantly alter the time point of aggregation which is in accordance with previous data  showing requirement of additional multiple washing of amoebae with EGTA in order to affect aggregation. At concentrations of 15–20 mM EGTA however, aggregation occurred at later time points (Fig. 2), on average at 13 ± 3 h in wild type and at 19 ± 2 h in iplA- cells (mean ± s.e.m. from 5 experiments); despite the daily variations in aggregation timing the mutant strain was delayed as compared to the wild type in each of the experiments performed. On the other hand, the presence of Ca2+ in the agar and therefore the continuous loading of cells with Ca2+  resulted in stronger impairment of aggregation in the wild type. The delayed aggregation of wild type cells in the presence of Ca2+ was not due to inhibition of chemotaxis (see below). Now the formation of aggregates was observed consistently at earlier time points in the iplA- strain than in wild type (Fig. 3; on average 7 ± 0.5 h vs. 15 ± 5 h until aggregate formation in iplA- and wild type amoebae, respectively, at 20 mM CaCl2 in 6 independent experiments); indeed, under this condition differentiation of iplA- cells resembled that of wild type observed in the control situation.
Figure 1. iplA- cells have an altered pattern of development. Differentiation of the mutant and the wild type strain was assayed in parallel on agar plates. Cells at different time points of development on H5-agar are shown. Wild type amoebae aggregated at t7, whereas aggregation of the mutant strain was delayed and aggregation strands broke (t10); therefore, smaller fruiting bodies were formed as compared to the wild type. The full width of the image corresponds to 12.5 mm.
Figure 2. Development of iplA- cells is impaired by depletion of internal Ca2+-stores due to EGTA-treatment. Differentiation of the wild type and the mutant on plates containing 20 mM EGTA is shown. Aggregation was delayed in both strains till t15 and t18 in wild type and iplA- cells, respectively. The size of the aggregates and the fruiting bodies were much smaller than under control conditions. The full width of the image corresponds to 12.5 mm.
Figure 3. In the presence of external Ca2 aggregation is accelerated in iplA- cells. Differentiation of the mutant and the wild type strain was assayed in parallel on agar plates supplemented with 20 mM CaCl2. Aggregate formation occurred earlier in iplA- cells (at t7) than in wild type (starting at t19) in the presence of Ca2+. The full width of the image corresponds to 12.5 mm.
Then we analyzed the effect of treatment with either EGTA or Ca2+ on basal cell motility. We found that under control conditions the general morphology of the cells as well as extension of pseudopods was practically identical in both strains (Fig. 4 A, B, [see 1]). Preincubation with 10 mM EGTA for 60 min led to strong rounding of wild type and mutant amoebae (Fig. 4 C, D) with reduced extension of small pseudopods [see 2]. By contrast, pretreatment with 10 mM CaCl2 did not affect the morphology (Fig. 4 E, F) or the extension of pseudopods [see 3] in both strains. Next we tested chemotaxis of amoebae towards a cAMP-filled glass capillary. Under control conditions cells of both strains oriented and migrated towards the tip of the capillary (Fig. 5 A, B); the average chemotactic speed was not different between wild type and the mutant strain (10.7 ± 2.1 vs. 11.0 ± 0.7 μm/min; mean ± s.e.m. of 20 wild type and 43 mutant amoebae analyzed in 3 and 4 independent experiments, respectively). Incubation in 10 mM EGTA for 60 min abolished chemotaxis in most of the wild type and the iplA- amoebae: small pseudopods were extended randomly and the cells did not approach the capillary tip (Fig. 5 C, D). Only rarely, cells of both strains exhibited an oriented but highly reduced migration towards the cAMP source (3% and 5% of 33 wild type and 19 mutant cells analyzed in 3 independent experiments, respectively). Thus the loss of Ca2+ from stores impairs both, orientation and migration also in the absence of the iplA gene product. By contrast, when amoebae were incubated in 10 mM CaCl2 for 60 min to load stores, they oriented and migrated towards the cAMP capillary (Fig. 5 E, F). The chemotactic speed of wild type cells (9.9 ± 1.2 μm/min; mean ± s.e.m. of 17 cells tested in 3 independent experiments) was comparable to that under control conditions whereas mutant amoebae chemotaxed significantly faster (13.6 ± 1.7 μm/min; mean ± s.e.m. of 15 mutant amoebae analyzed in 3 independent experiments) than wild type cells in the presence of 10 mM CaCl2 (Mann-Whitney rank sum test, p = 0.041).
Figure 4. General morphology of wild type and iplA- cells under control conditions (A, B), in the presence of 10 mM EGTA for 60 min (C, D) or in the presence of 10 mM CaCl2 for 80 min (E, F). In H5-buffer or in the presence of 10 mM CaCl2 the morphology was not significantly different between wild type and mutant amoebae. However, in the presence of EGTA the cells of both strains were rounded. Photographs were taken at t5. Basal motility under these conditions can be viewed in the accompanying movies.
Additional File 2. Basal cell motility after preincubation of wild type and mutant amoebae with 10 mM EGTA for 60 min (t4–t5). Images of cells at t5 were captured every 15 sec for 20 min in the continued presence of EGTA. Cells were rounded and extended smaller pseudopods than under control conditions.
Additional File 3. Cell motility of wild type and mutant cells after preincubation with 10 mM CaCl2 for 80 min (t4–t5.3) is shown. Images of cells at t5.3 were captured every 15 sec for 20 min in the continued presence of CaCl2. The behaviour of treated cells was not different from control amoebae.
Figure 5. Chemotaxis of wild type and mutant amoebae at different experimental conditions. The tracks of individual cells (in red) migrating during chemotactic stimulation (position of the tip of the cAMP-filled capillary: green star) are shown. In H5-buffer (A, B) both cell types migrated in an oriented manner towards the capillary tip, albeit not always in a straight line. After preincubation with 10 mM EGTA for 60 min and in its continued presence during the chemotaxis assay (C, D) the cells remained stationary with random pseudopod extension. Preincubation of amoebae with 10 mM CaCl2 (E, F) did not impair chemotaxis; rather, the cells of both strains migrated towards the capillary tip. Chemotaxis experiments were done at t6.
Buffering of intracellular [Ca2+] impairs chemotaxis
The observation that aggregation occurred in the mutant cell line although a cAMP-activated increase in [Ca2+]i was not detectable resulted in the conclusion that [Ca2+]i-changes were not necessary to accomplish chemotaxis . We used the mobile buffer approach originally described by Speksnijder et al.  which allows to analyze the requirement of a [Ca2+]i-gradient for a given response. If in Dictyostelium a [Ca2+]i-increase was necessary for chemotaxis, the presence of a Ca2+-chelator in the cytosol should impair orientation and/or migration. In a previous study, we had introduced the Ca2+-chelator BAPTA and its derivatives into the cytosol of wild type amoebae which indeed had inhibited chemotactic migration and reduced chemotactic orientation . Here we used the Ca2+-indicator Fura2-dextran to clamp [Ca2+]i and loaded the indicator into wild type and mutant cells in the absence of external CaCl2. The treatment affected chemotactic performance of wild type as well as iplA- amoebae. Lack of extracellular CaCl2 during the loading process induced strong rounding of the amoebae and loss of migration. Figure 6 shows that the capacity to orient chemotactically and to extend pseudopods towards the capillary tip was reduced by 58% in wild type (93 cells tested in 4 independent experiments). Inhibition was also evident in iplA- cells (Fig. 6) showing sensitivity of the mutant towards buffering of intracellular Ca2+-levels and eradication of [Ca2+]i-changes: the fraction of pseudopods extended in direction of the cAMP-source was reduced by 75% (74 cells tested in 3 independent experiments). These results show that not only in wild type but also in the iplA- cell line the ability to orient and to migrate in fact depends on an agonist-activated [Ca2+]i-elevation.
Figure 6. Chemotaxis of iplA- cells is impaired in the intracellular presence of a Ca2+-buffer. Wild type and mutant amoebae were loaded with Fura2-dextran and their ability to protrude pseudopods towards a cAMP-filled glass capillary was compared to that of untreated cells. In both strains the presence of the chelator in the cytosol led to a decrease in the fraction of cells extending pseudopods and migrating towards the cAMP source.
Analysis of Ca2+-fluxes
Our findings that differentiation was sensitive towards depletion of Ca2+ or loading of the cells with Ca2+ and that chemotaxis was blocked by intracellular Ca2+-buffers led us to investigate Ca2+-fluxes in the mutant cell line. We used a Ca2+-sensitive electrode in cell suspensions to measure Ca2+-fluxes, an approach different from that of Traynor et al.  who had studied ion fluxes by 45Ca2+-measurements. First we tested whether the coupling of stores to the plasma membrane, i.e. capacitative Ca2+-fluxes without prior stimulation with agonists were altered. In Dictyostelium induction of capacitative Ca2+-influx requires active intracellular Ca2+-pumps. Their inhibition by either thapsigargin or 2,5-di-(t-butyl)-1,4-hydroquinone (BHQ) does not evoke influx; rather, stores have to be emptied by treatment with EGTA . Capacitative Ca2+-fluxes were studied early during differentiation (t2–t4). At this time Ca2+-influx and Ca2+-efflux are at an equilibrium which held true for both, wild type and mutant cells (not shown); in suspensions of cells at later stages of development influx strongly exceeds efflux . In iplA- and in wild type cells emptying of storage compartments via preincubation of amoebae with EGTA induced capacitative Ca2+-entry (Fig. 7 A) which was blocked by addition of 1 mM NaN3 (Fig. 7 B). The characteristics of influx were comparable in wild type and mutant cells. These data show that capacitative Ca2+-influx does not depend on the product of the iplA gene.
Figure 7. Recordings of Ca2+-fluxes in iplA- and wild type amoebae. [Ca2+]e was measured in cell suspensions with a Ca2+-sensitive electrode. (A) Treatment of amoebae with 5 mM EGTA for 30 min activated capacitative Ca2+-influx (one out of 12/6 determinations in 4/3 independent experiments is shown for iplA- and wild type, respectively). (B) Capacitative influx was blocked by the addition of 1 mM NaN3 (one out of 5/4 determinations in 3/3 independent experiments). Measurements were done at t3.
On the other hand, agonist-activated 45Ca2+-entry had been reported to be absent in the mutant strain; in their study, Traynor et al. had stimulated cells with cAMP in the presence of 0.1 mM CaCl2 . The use of a Ca2+-sensitive electrode allows to measure much lower levels of extracellular Ca2+ to analyze Ca2+-fluxes, in the range of approximately 1 μM Ca2+. Indeed, we found that under this condition reversible Ca2+-entry occurred after addition of 1 μM cAMP (Fig. 8 A) that amounted to 10.2 ± 4.3 pmol Ca2+/107 cells (mean ± s.d. from 9 experiments). The level of influx represented roughly 5% of wild type influx (Fig. 8 B and [13,14]). Ca2+-influx was delayed in the mutant and the time to reach the maximum was longer than in wild type cells [13,14]. In addition, challenge with arachidonic acid (AA) induced influx (Fig. 9 A). Again, the mutant was less sensitive and higher concentrations were required than those reported to evoke Ca2+-entry in wild type cells (Fig. 8 B, 9 C and ). Neither 10 nor 20 μM AA were effective; in the wild type 10 μM AA activates influx of 190 ± 58 pmol Ca2+/107 cells . At 60 μM AA entry occurred in the mutant strain which amounted to an average of 107 ± 21 pmol Ca2+/107 cells (mean ± s.d. from 7 experiments). Preincubation of cells with the SERCA-type Ca2+-ATPase inhibitor BHQ reduced AA-induced influx by 82 % (Fig. 9 B) to an average of 21 ± 18 pmol Ca2+/107 cells (mean ± s.d. from 3 experiments). These data show that Ca2+-fluxes across the plasma membrane do occur in iplA- cells as well but at a reduced level.
Figure 8. Agonist-activated Ca2+-fluxes in suspensions of iplA- and wild type amoebae. cAMP elicited reversible Ca2+-influx in the mutant (A) and in wild type ((B) and see ); measurements were done at t7–t7.5. Note the different doses of CaCl2 added for calibration. The time points of cAMP-addition (1 μM) and of AA-addition (6 μM) in the wild type are indicated by arrows.
Figure 9. Fatty acids activate Ca2+-fluxes in iplA- amoebae. (A) 60 μM AA evoked a transient decrease in [Ca2+]e representing Ca2+-influx; measurement was done at t6 (B) After preincubating amoebae with the SERCA-type Ca2+-ATPase blocker BHQ (100 μM) for 20 min to inhibit uptake of Ca2+ into internal storage compartments, the AA-activated response was absent; measurement was done at t7.5. Results of measurements with wild type (C, D) stimulated with 6 μM AA are shown for comparison.
Next we tested whether the mutant strain was able to release stored Ca2+ when stimulated with cAMP or AA. Fluxes were measured in suspensions of cells with permeabilized plasma membranes; any change in [Ca2+]e thus reflects efflux of Ca2+ from storage compartments. Both, cAMP and arachidonic acid activated release of stored Ca2+ (Fig. 10 A). On average, addition of 1 μM cAMP released 7.3 ± 3.4 pmol Ca2+/107 cells and 16.3 ± 7.2 pmol Ca2+/107 cells were liberated after stimulation with 3 μM AA (mean ± s.d. from 10 and 3 experiments, respectively). The amount of Ca2+-efflux from stores after cAMP stimulation was 61% of that found in wild type cells (Fig. 10 B and ) whereas release upon AA-challenge was in the range of 5–10% of wild type (see Fig. 2 in ). In addition to the Ca2+-electrode recordings, we studied Ca2+-fluxes in suspensions of partially purified storage compartments fluorimetrically. ATP induced Ca2+-sequestration (Table 1) was of similar magnitude and rate as in wild type stores. This result indicates that the decreased release of Ca2+ from the stores measured with the Ca2+-electrode is not due to a lack of storage capacity. Moreover, the addition of AA evoked release of Ca2+ from stores as did inhibition of Ca2+-pump(s) by thapsigargin. XestosponginC that inhibits Ca2+-uptake and activates Ca2+-release in wild type  and the ionophore ionomycin also resulted in substantial Ca2+ release in the mutant cell line (Table 1). All values were in the same range as those for wild type.
Table 1. Determination of Ca2+-fluxes in partially purified storage compartments of the iplA- mutant and of wild type. Ca2+-sequestering vesicles were prepared as outlined in Methods. Measurements were performed with the pellet and supernatant fraction. ATP-induced uptake and release activated by different agents is given as nmol Ca2+-uptake/min and mg of protein and pmol Ca2+-release/tube, respectively (mean ± s.d.). In release experiments 60–75 μl of pellet and 120–140 μl of supernatant fraction were used per tube. Numbers in brackets give number of experiments; n.d.: not determined.
Figure 10. cAMP and arachidonic acid elicit Ca2+-release from internal stores. (A) [Ca2+]e was recorded at t7 in iplA- cells with permeabilized plasma membranes. Amoebae were challenged with 1 μM cAMP and 3 μM AA, respectively. (B) The response of permeabilized wild type stimulated with 1 μM cAMP at t6 is shown for comparison.
The [Ca2+]e-recordings in suspensions of cells as described above detect the sum of Ca2+-influx and efflux. Therefore, a complementary approach which monitors influx only was pursued, by using the Mn2+-quenching technique in single intact amoebae. This method is based on the fact that many Ca2+-channels are permeable to Mn2+  and that the Ca2+-indicator Fura2 binds Mn2+ with high affinity. Fluorescence of the indicator is quenched upon binding . We compared quenching of Fura2-dextran fluorescence activated by addition of Mn2+ alone or in combination with 1 μM cAMP in wild type and mutant cells. Higher concentrations of MnCl2 were required to quench Fura2-dextran fluorescence in iplA- amoebae (Fig. 11, Table 2). Reduction of fluorescence occurred at 1 μM Mn2+ and 1 μM Mn2+/cAMP in wild type (Fig. 11 A, B; see also ) but not in the mutant (Fig. 11 C, D) where addition of 100 μM Mn2+/cAMP was necessary; 100 μM Mn2+ alone was not effective (Fig. 11 C, D) but started at 200 μM Mn2+ (not shown). These data show that the reduction in ion fluxes in the mutant were indeed due to an alteration in entry mechanisms.
Table 2. Rate of basal and cAMP-induced Mn2+-influx. Amoebae were challenged with 1 or 100 μM Mn2+ either with or without 1 μM cAMP. Cells were preincubated with 0.1 mM EGTA as outlined in Methods. Mn2+ quenching of fura-2-dextran fluorescence was tested in H5-buffer and is expressed as decrease in fluorescence units/sec (mean ± s.e.m.). Numbers in brackets give number of cells tested and number of experiments.
Figure 11. Basal and cAMP-induced Mn2+-influx. Influx was assayed by quenching of Fura2-dextran fluorescence. (A, B) The response of wild type amoebae is shown for comparison; 1 μM Mn2+ ± 1 μM cAMP was added. iplA- cells in nominally Ca2+-free buffer were challenged with 100 μM Mn2+ ± 1 μM cAMP at t7 (closed symbols); when 1 μM Mn2+ was added (open symbols) no influx was detected (C, D). After preincubation with EGTA influx was observed at 1–2 μM Mn2+ ± 1 μM cAMP (E, F). Fluorescence intensity at 360 nm excitation is shown as mean ± s.e.m
In principle, the iplA gene product could form a channel in the plasma membrane or in membranes of internal stores. The lack of the iplA gene product in the stores might impair their coupling to the plasma membrane. As we had observed capacitative Ca2+-entry in the mutant we asked whether manipulation of the filling state of the stores altered ion fluxes. First we tested the effect of emptying of stores on Mn2+-influx. When cells were preincubated with EGTA, the requirement for high doses of Mn2+ to quench fluorescence was abrogated. Now capacitative and also agonist-activated Mn2+-influx occurred at concentrations of MnCl2 comparable to those used under control conditions in wild type, in the range of 1–2 μM (Fig. 11 E, F). This result renders the possibility that the plasma membrane is altered in the mutant unlikely. Yet, the rate of Mn2+-influx observed in EGTA-treated mutant amoebae was still less than in wild type cells with respect to both basal and cAMP-activated fluxes (53 and 58% of wild type , respectively).
In wild type cells treatment with EGTA augments responsiveness and cAMP-elicited [Ca2+]i-transients are detected at low extracellular [Ca2+] . However, an agonist-induced [Ca2+]i elevation was not observed in iplA- cells under these conditions. On the other hand, when stores were loaded in the continued presence of CaCl2, we observed that differentiation of the mutant resembled the wild type as described above. This led us to compare the [Ca2+]i-response of wild type and mutant amoebae after pretreatment with CaCl2. In wild type cells preincubation with 1 mM CaCl2 for 10–15 min and its continued presence during the [Ca2+]i-imaging experiment is required to activate a [Ca2+]i-transient after challenge with cAMP (Fig. 12 A and see ); in nominally Ca2+-free medium or at very low extracellular [Ca2+] a cAMP-activated [Ca2+]i-elevation is not observed . In accordance with the data of Traynor et al.  this condition (Fig. 12 B) and even increasing the concentration of CaCl2 to 20 mM resulted in no detectable [Ca2+]i-increase in the mutant strain (not shown). However, we found that the basal [Ca2+]i level in iplA- amoebae was significantly lower than in wild type and amounted to an average of 36 ± 3 nM (mean ± s.e.m. of 8 determinations in 3 independent experiments) as compared to 50 ± 2 nM in wild type (mean ± s.e.m. of 29 determinations in 9 independent experiments; Mann-Whitney rank sum test, p = 0.002).
Figure 12. [Ca2+]i-recordings in cells preincubated with CaCl2 in order to load stores. (A) The response of wild type upon stimulation with 1 μM cAMP (arrow) at standard conditions, i.e. after preincubation with 1 mM CaCl2 for 10–15 min and stimulated in the presence of 1 mM CaCl2 is shown. Values give mean ± s.e.m. of 7 cells. (B) When iplA- cells were stimulated with 1 μM cAMP at standard conditions (as outlined in (A)), no response was observed. Mean ± s.e.m. of 6 cells is shown. (C) Wild type was preincubated with 1 mM CaCl2 for 4–5 h; after washing cells were incubated in H5-buffer supplemented with 1 μM CaCl2 and challenged with 1 μM cAMP (arrow). Values give mean ± s.e.m. of 16 cells. (D) iplA- incubated for 3 h with 20 mM CaCl2 were washed and subsequently [Ca2+]i-imaging was done in buffer containing 1 mM CaCl2. Arrow indicates the time point when 1 μM cAMP was added. Mean ± s.e.m. of 10 cells is shown.
For stronger loading of Ca2+-stores, we preincubated cells with 1 mM CaCl2 for 4 h. After this treatment a [Ca2+]i-elevation upon addition of cAMP was detected in 60% of wild type amoebae even at low extracellular [Ca2+] levels, i.e. when the buffer used to wash the cells had been supplemented with only 1 μM CaCl2 (Fig. 12 C). Starting from a basal level of 48 ± 4 nM, the height of the [Ca2+]i-transient amounted to 23 ± 2 nM (mean ± s.e.m. of 12 determinations in 8 independent experiments). Again, these conditions were not effective in iplA- cells; rather, the sensitivity of the mutant was shifted to higher Ca2+ concentrations as had been found with Mn2+-quenching experiments. We preincubated iplA- amoebae with 20 mM CaCl2 for 3 h; after washing thoroughly, a cAMP-induced [Ca2+]i-elevation in the presence of 1 mM CaCl2 was observed (Fig. 12 D) in 28% of the cells; its average height amounted to 67 ± 11 nM (mean ± s.e.m. of 15 determinations in 4 independent experiments) starting from a basal level of 39 ± 2 nM. In the course of these experiments we once observed a response also under standard conditions, i.e. at 1 mM [Ca2+]e without prior incubation in 20 mM CaCl2; the height of the increase amounted to 54 ± 6 nM (mean ± s.e.m.). Yet, this was a rare event (once in 31 determinations).
The role of cAMP-activated [Ca2+]i-changes for chemotaxis has been questioned by Traynor et al.  who reported results obtained with the iplA- mutant cell line favouring insignificance of [Ca2+]i for the chemotactic response. The authors had observed formation of fruiting bodies even though neither 45Ca2+-fluxes nor an agonist induced [Ca2+]i-elevation were detectable. The discrepancy between our view that a [Ca2+]i-elevation is necessary for a proper chemotactic response  and the conclusion of Traynor et al. prompted us to analyze chemotaxis, differentiation and the [Ca2+]-regulation of the iplA- mutant in detail. In particular, we tested not only basal and cAMP-activated ion fluxes but also capacitative Ca2+-entry which is induced by emptying internal Ca2+-stores via preincubation of amoebae with EGTA .
Aggregation and development of wild type and mutant cells on agar plates was sensitive towards continuous emptying or loading of Ca2+-stores. These effects are not necessarily caused by altering chemotactic migration. It is conceivable that other Ca2+-dependent processes were affected, e.g. that the timing or pattern of gene expression or the establishment of cell contacts was altered. Although incubation of mutant amoebae for 2 h with 20 mM CaCl2 or of wild type with 1 mM CaCl2 for 4–5 h or with 1 mM EGTA for 1 h  did not significantly increase or lower basal [Ca2+]i it is possible that the continued presence of 10 mM EGTA or CaCl2 for many hours affects basal levels of [Ca2+]i which in turn might mediate effects on gene expression as was shown for prolonged incubation of cells with BHQ . During development Dictyostelium cells form Ca2+-dependent and EDTA/EGTA-sensitive cell-cell contacts that are mediated by gp24 and DdCAD-1 ([19,20]; for review see ). Therefore, chelation of extracellular Ca2+ might also inhibit cell adhesion. However, mutant cells whose gene encoding DdCAD-1 had been disrupted show normal chemotaxis and cell streams. Furthermore, mound formation was accelerated and only culmination was delayed by about 6 hours . If only Ca2+-dependent cell adhesion was affected in our development assay in the presence of EGTA we would expect a similar phenotype. However, aggregation was clearly delayed. This argues for additional Ca2+-dependent processes during aggregation.
When we tested the influence of the presence of EGTA or CaCl2 on spontaneous motility and chemotaxis we found that in both strains motility in general was strongly impaired and that chemotaxis of the amoebae towards the cAMP-filled glass capillary was virtually abolished upon depletion of internal stores by the extracellular presence of EGTA. This effect is time dependent; after 30 min of incubation with 10 mM EGTA the behaviour of wild type amoebae was found to be unaltered and only after treatment for 0.5–1 h rounding and reduction of pseudopod elongation towards the capillary tip occurred . Prolonged incubation for more than 1 h as carried out in this study completely inhibited the chemotactic response. These results strengthen the view that Ca2+ has a necessary role in chemotaxis in wild type and in the mutant as well. When the cellular Ca2+ content falls below a critical value Ca2+-dependent cytoskeletal rearrangements [22,23] that are necessary for both, random pseudopod extension during spontaneous motility and oriented pseudopod formation after chemotactic stimulation no longer take place correctly. On the other hand, the presence of 10 mM CaCl2 induced no alteration of basal cell motility in wild type or mutant amoebae. Yet, during chemotactic stimulation the average speed of migration towards the capillary tip was higher in mutant than in wild type cells. In this respect it is of importance that the basal level of [Ca2+]i was significantly lower in the former. At standard conditions the reduced basal [Ca2+]i does not impair the capacity of the mutant to chemotax. Therefore, this particular mutant strain represents the "minimal solution" with respect to the concentration of cytosolic Ca2+ necessary to accomplish cytoskeletal rearrangements and extrusion of a pseudopod correctly. However, in the presence of 10 mM extracellular Ca2+ during cAMP-stimulation Ca2+-fluxes are enhanced allowing more efficient formation of pseudopods. We had shown previously that a small global elevation of [Ca2+]i activates the extension of pseudopods all over the cell's circumference whereas a larger increase induces contraction of the amoebae . In our view, the strongest evidence that a [Ca2+]i-transient is necessary for the extension of pseudopods rests upon the experiment where a Ca2+-chelator was introduced into the cytosol of the amoebae. This treatment led to rounding of the amoebae and a general reduction of pseudopod formation (see also ). Upon stimulation with a cAMP-filled capillary, the extension of oriented pseudopods was greatly reduced and migration towards the capillary tip was abolished. As Speksnijder et al.  had pointed out the fact that the presence of a chelator has an effect shows that a [Ca2+]i-gradient is essential for a given response. In summary, these data support the notion that an elevation of [Ca2+]i is required to extend pseudopods; suppression of the [Ca2+]i-elevation inhibits motility in general. Upon chemotactic challenge with cAMP this [Ca2+]i-gradient has to be established in a locally restricted fashion in order to allow local, oriented pseudopod formation (see [2,25,26]); otherwise pseudopods would be extended in all directions (see above, ). Our results imply that in iplA- cells such a [Ca2+]i-gradient occurs as well, either nonrestricted allowing extension of pseudopods at random sites during spontaneous motility or restricted locally after chemotactic stimulation leading to oriented pseudopod formation. The fact that in the mutant cell line cAMP-activated [Ca2+]i-changes were practically undetectable under our standard condition argue for a [Ca2+]i-increase that is either smaller and/or more restricted to distinct domains within the cell than in wild type amoebae. Indeed, in only one out of roughly 30 determinations did we observe a cAMP-activated [Ca2+]i-transient under standard conditions. These results imply a crucial role but not an absolute necessity of the iplA gene product for the regulation of cAMP-induced [Ca2+]i-changes.
By using a Ca2+-sensitive electrode in cell suspensions, we analyzed which aspects of [Ca2+]i are controlled by the iplA gene product. Besides studying agonist-induced Ca2+-fluxes we also investigated capacitative Ca2+-entry and found that this type of influx was similar in mutant and wild type cell suspensions. We obtained equivalent results by testing Mn2+-quenching of Fura2-dextran fluorescence which showed that capacitative entry is independent of the iplA gene product.
On the other hand, using the Ca2+-sensitive electrode, we found that in the iplA- mutant the agonist cAMP and also AA did activate Ca2+-entry into intact cells. The difference between the data published by Traynor et al.  and our results is most likely due to the experimental conditions: the magnitude of the Ca2+-fluxes that we observed was considerably lower than in wild type cells and detectable at low extracellular [Ca2+] only. The 45Ca2+-flux studies had been performed at 100 μM external CaCl2; so the fraction of 45Ca2+ entering the cells was presumably too low to be detected reliably. Moreover, we found cAMP- and AA-induced Ca2+-release from stores in cells with permeabilized plasma membranes. These data show that cAMP-induced Ca2+-release from stores in iplA- cells is functional. However, much like the Ca2+-influx, agonist-activated liberation from stores was smaller than in wild type amoebae. In line with these results are the findings using Mn2+-quenching to assay ion fluxes in intact single cells: higher doses of Mn2+ were necessary to detect influx.
There are several interpretations for the results above. (i) There are two types of channels responsible for Ca2+-influx: one type being activated by emptying of the stores and sustaining capacitative Ca2+-entry which is unaffected in iplA- cells and the other one mediating agonist-induced Ca2+-fluxes, the latter being under the control of the iplA gene product. The view that there are two strictly separated ion channels seems unlikely as under conditions of emptied stores cAMP-activated Mn2+-quenching occurred in the mutant as well. (ii) The same channel(s) mediate capacitative and agonist-activated fluxes but upon stimulation with agonists it cannot be addressed properly when iplA is disrupted. This implies a role of the protein in the liberation of Ca2+ from the stores which is a prerequisite for the triggering of Ca2+-entry . In the mutant this cannot proceed normally so subsequent activation of Ca2+-influx and the generation of a full [Ca2+]i-increase is impaired. The results of the experiments where stores were strongly loaded with Ca2+ prior to stimulation support this notion. In this situation release from stores should be augmented. Indeed, in both, wild type and mutant cells, cAMP-activated [Ca2+]i-elevations occurred at an extracellular [Ca2+] (see Fig. 12) where without pretreatment no increase was observed. Presumably, release of Ca2+ from the filled stores contributed to the observed [Ca2+]i-increase to a greater extent than under standard conditions. The requirement for 20 fold higher concentrations of CaCl2 during preincubation to elicit an agonist-induced [Ca2+]i-elevation in iplA- cells are most likely due to the reduction in Ca2+-entry which necessitates a higher concentration gradient across the plasma membrane to fill the stores efficiently.
An as yet unresolved issue is the mechanism that induces Ca2+-entry upon liberation of Ca2+ from the stores. From our data we conclude that in Dictyostelium these signals are different when the stores are emptied by EGTA or by agonist-activated signaling cascades. Otherwise one cannot explain normal capacitative Ca2+- and Mn2+-influx induced by EGTA-treatment and a requirement for 100 fold higher ion concentrations to induce Mn2+-entry by cAMP. If indeed the iplA gene product constitutes an IP3-receptor like channel that is located on membranes of stores the physical coupling of the receptor to channels in the plasma membrane as a mechanism to activate extracellular Ca2+-entry  should be missing in the mutant. On the other hand, emptying of stores by EGTA-treatment influences not only the IP3-sensitive store but also other stores and thus exerts a much more general effect on the cells. Studies using microarrays should reveal whether the expression of other genes is affected by the absence of iplA and thus might give a clue how [Ca2+]i is regulated in the mutant although one type of Ca2+-store is malfunctional.
Our results show that Ca2+ fluxes and regulation of Ca2+ homeostasis take place in the iplA- mutant and that chemotaxis and development of the mutant are sensitive to disturbance of the Ca2+ homeostasis. In wild type cells and in cells lacking the iplA gene changes in [Ca2+]i are necessary to orient and to migrate chemotactically; their abolition causes loss of chemotaxis towards a cAMP source. The iplA gene product exerts a crucial role in the control of basal [Ca2+]i and of agonist induced Ca2+-fluxes. It is not required to activate capacitative Ca2+-influx. Thus the mechanisms responsible for capacitative and agonist-activated Ca2+-fluxes are different.
Fura2-dextran and Fura2 were purchased from MobiTec; cAMP was from Boehringer.
D. discoideum wild type strain Ax2 and the iplA- cell lines HM1049 and HM1038 (kindly provided by Dr. D. Traynor) were cultured as described  in the absence or presence of 10 μg/ml Blasticidin S, respectively. There was no difference between the two mutant strains with respect to the assays performed; therefore, results of measurements with either HM1038 or HM1049 are shown. Cells were washed by repeated centrifugation and resuspension in cold Sørensen phosphate buffer (17 mM Na+/K+-phosphate, pH 6.0). Amoebae were shaken at 2 × 107 cells/ml, 150 rpm and 23°C until use. The time, in hours, after induction of development is designated tx.
[Ca2+]e in cell suspensions was recorded as described elsewhere . Cells at t5–t8 were washed by repeated centrifugation and resupended at 5 × 107 cells/ml in 5 mM Tricine, 5 mM KCl, pH 7.0. Permeabilization was done by addition of filipin (15 μg/ml) to cell suspensions exactly as outlined in . Capacitative Ca2+-influx was analyzed in cells with emptied storage compartments : amoebae at t2–t4 were incubated with 5 mM EGTA for 30 min before washing in the above buffer.
[Ca2+]i-determination and Mn2+-quenching experiments
Cells were loaded with Fura2-dextran (5 mg/ml + 1 mM CaCl2) at t4–t5 as described . Aliquots (2–5 μl) of washed cells in H5-buffer (5 mM Hepes, 5 mM KCl, pH 7.0) were placed on glass coverslips and incubated in a humid chamber. 10–15 min prior to the experiment, 85–88 μl of H5-buffer + 1 mM CaCl2 were added. In a series of experiments to load stores, wild type and iplA- cells were incubated with 1 mM CaCl2 for 4–5 h and with 20 mM CaCl2 for 2–3 h, respectively. Then they were thoroughly washed exactly as described previously  and incubated either in H5-buffer supplemented with 1 μM CaCl2 (wild type; free [Ca2+] in the solution was measured to be 2–2.5 μM, see also ) or in H5-buffer +1 mM CaCl2 (iplA-); final volume was 90 μl. Single cell [Ca2+]i-imaging was performed at t7–t8 as described ; stimulation was done by adding 10 μl of cAMP (10 μM). For Mn2+-quenching assays, washed cells were incubated in H5-buffer and challenged with Mn2+ or Mn2+/cAMP. In order to study fluxes in cells with partially emptied internal storage compartments cells were preincubated with EGTA (10 μl of H5-buffer plus 0.1 mM EGTA for 1–2 h). 10–15 min prior to the experiment this solution was carefully removed and 100 μl of H5-buffer was added. This was repeated three times; final volume was 90 μl. Fluorescence quenching was measured at 360 nm excitation; influx rates are given as decrease of fluorescence units/sec.
Measurement of Ca2+-fluxes in partially purified storage compartments
Analysis of vesicular Ca2+-fluxes was done as described . In brief, 3 ml of cells at t1–t6 (2 × 108 cells/ml) in 20 mM Hepes, pH 7.2, were lysed by passage through Nuclepore filters. A final concentration of 3 % sucrose, 50 mM KCl, 1 mM MgCl2, 20 μg/ml leupeptin, 1 μg/μl aprotinin, 2.5 mM dithiothreitol and 1 μM microcystin were added; unbroken cells were removed by centrifugation at 3000 g for 5 min. The supernatant was centrifuged again at 12000 g for 20 min. The sediment (P) was resuspended in 1 ml of the above buffer. The rate of uptake and release was determined in the pellet and supernatant fraction by measuring the extravesicular [Ca2+] with Fura2.
Cells were analyzed for chemotaxis towards a capillary filled with 0.1 mM cAMP . 250 μl of 1 × 105 cells/ml in H5-buffer were pipetted onto a glass coverslip and allowed to settle for 60 min. Chemotaxis was recorded on a video recorder for 30–45 min. Chemotaxis was also assayed in the presence of EGTA or CaCl2; then amoebae were incubated in the respective agents for 60 min before they were challenged with cAMP. Images were digitized and the behaviour of the cells was analyzed using a computer program written for this purpose. For determination of cell velocity, a square area of interest (AI) of variable size (usually roughly 1/3 of the area of the cell) was placed at the perimeter of the cell in the first image digitized at the beginning of the assay. In the next image (images were digitized at a 2–4 sec time interval) the program analyzed an area larger than the AI (this area was defined by adding a given number of pixels on each side of the AI) for a pattern that resembled that of the AI; when such a pattern was found then the AI was placed on this new spot. The difference between the position of the AI in the first image to that in the second image was expressed as a vector of a given length. The changes in cell shape during migration were compensated by updating the pattern within the AI for every consecutive image analyzed. Calibration of the system allowed to convert the sum of the vector lengths to the distance in μm that the cells had migrated at the end of the experiment and to calculate the velocity of the amoebae. To test the effect of the intracellular presence of a Ca2+-buffer on chemotaxis, cells were loaded with Fura2-dextran (5 mg/ml in the loading solution) by electroporation in the absence of added external Ca2+. The amount of indicator present in the cytosol is in the range of 2–5% of the concentration present during electroporation . 20 min after loading, cells were stimulated for 3–4 min by placing the cAMP-filled glass capillary at a distance of 10–20 μm of the cells and the number of cells that extended oriented pseudopods and thus elongated towards the capillary tip within this time period was counted. We had shown previously that loading of amoebae with FITC-dextran as a control does not alter chemotaxis as compared to untreated cells .
Analysis of differentiation
Time lapse recordings of the development of Ax2 and iplA- cells on 1.5 % agar in H5-buffer (H5-agar) were done by placing 4 × 106 cells each on one half of a petri dish (∅ 35 mm) at t1. The two populations were separated from each other by a thin plastic disc that had been inserted in the melted agar during cooling. Only after removal of fluid and slight drying of the plate the disc was removed which resulted in a thin rim separating the strains. Differentiation was recorded by capturing an image of the plate every 30 min using a stereo microscope (Stemi 2000, Zeiss) equipped with a CCD camera (AVT Horn) under the control of the AxioVision software package (Zeiss). In addition, development was assessed at various levels of extracellular CaCl2. Then H5-agar contained either 5–20 mM EGTA or 5–20 mM CaCl2.
List of abbreviations
Cytosolic free Ca2+ concentration: [Ca2+]i
Arachidonic acid: AA
Area of interest: AI
RS recorded extracellular [Ca2+] in cell suspensions and participated in the design of the study. DFL participated in the recordings of extracellular [Ca2+] and the design of the study. KBR analyzed chemotaxis and differentiation of wild type and mutant cells. KH performed [Ca2+]i measurements. DM analyzed fluxes in partially purified storage compartments and was involved in the design of the study. CS participated in [Ca2+]i measurements, did Mn2+-flux studies, participated in the design of the study and wrote the manuscript.
The authors wish to thank Dr. D. Traynor and Prof. R. Kay for generously providing the iplA- strain and Rupert Mutzel and Gerd Knoll for many helpful discussions and critical reading of the manuscript. We are indebted to Georg Heine and Williams Pascual from the Wissenschaftliche Werkstätten of the University of Konstanz for computer programming and to Katja Drews, Cyrus Nassiri and Frank Zucchetti for performing initial Ca2+-electrode measurements. This work was supported by the Deutsche Forschungsgemeinschaft.
Miyamoto S, Izumi M, Hori M, Kobayashi M, Ozaki H, Karaki H: Xestospongin C, a selective and membrane-permeable inhibitor of IP3 receptor, attenuates the positive inotropic effect of alpha-adrenergic stimulation in guinea-pig papillary muscle.
FEMS Microbiol Lett 1984, 21:21-25. Publisher Full Text
Schlatterer C, Buravkov S, Zierold K, Knoll G: Calcium-sequestering organelles of Dictyostelium discoideum: changes in element content during early development as measured by electron probe X-ray microanalysis.
Flaadt H, Jaworski E, Schlatterer C, Malchow D: Cyclic AMP- and Ins(1,4,5)P3- induced Ca2+-fluxes in permeabilised cells of Dictyostelium discoideum: cGMP regulates Ca2+ entry across the plasma membrane.
EMBO J 1996, 15:5177-5183. PubMed Abstract
Wong E, Yang C, Wang J, Fuller D, Loomis WF, Siu CH: Disruption of the gene encoding the cell adhesion molecule DdCAD-1 leads to aberrant cell sorting and cell-type proportioning during Dictyostelium development.
Adv Exp Med Biol 1990, 269:99-102. PubMed Abstract
Schlatterer C, Schaloske R: Calmidazolium leads to an increase in the cytosolic Ca2+ concentration in Dictyostelium discoideum by induction of Ca2+ release from intracellular stores and influx of extracellular Ca2+.
Biophys Struct Mech 1982, 9:131-136. Publisher Full Text
Schlatterer C, Gollnick F, Schmidt E, Meyer R, Knoll G: Challenge with high concentrations of cyclic AMP induces transient changes in the cytosolic free calcium concentration in Dictyostelium discoideum.
Science STKE 2004, 2004:pe33. Publisher Full Text
Cell Motil Cytoskel 1993, 25:298-307. Publisher Full Text