BMC Cell Biology BioMed Central

Background A number of transgenic mice carrying different deletions in the Locus Control Region (LCR) of the mouse tyrosinase (Tyr) gene have been developed and analysed in our laboratory. We require melanocytes from these mice, to further study, at the cellular level, the effect of these deletions on the expression of the Tyr transgene, without potential interference with or from the endogenous Tyr alleles. It has been previously reported that it is possible to obtain and immortalise melanocyte cell cultures from postnatal mouse skin. Results Here, we describe the efforts towards obtaining melanocyte cultures from our Tyr transgenic mice. We have bred our Tyr transgenic mice into Tyr c-32DSD mutant background, lacking the endogenous Tyr locus. In these conditions, we failed to obtain immortalised melanocytes. We decided to include the inactivation of the Ink4a-Arf locus to promote melanocyte immortalisation. For this purpose, we report the segregation of the Ink4a-Arf null allele from the brown (Tyrp1b) mutation in mice. Finally, we found that Ink4a-Arf +/- and Ink4a-Arf -/- melanocytes had undistinguishable tyrosine hydroxylase activities, although the latter showed reduced cellular pigmentation content. Conclusion The simultaneous presence of precise genomic deletions that include the tyrosinase locus, such as the Tyr c-32DSD allele, the Tyr transgene itself and the inactivated Ink4a-Arf locus in Tyrp1B genetic background appear as the crucial combination to perform forthcoming experiments. We cannot exclude that Ink4a-Arf mutations could affect the melanin biosynthetic pathway. Therefore, subsequent experiments with melanocytes will have to be performed in a normalized genetic background regarding the Ink4a-Arf locus.


Background
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 Ca 2+ and extracellular Ca 2+ -entry [1] leading to a [Ca 2+ ] i -transient [2][3][4]. Chemotaxis proceeds in the presence of extracellular EGTA but not in the presence of intracellular Ca 2+ buffers, so a [Ca 2+ ] i -elevation is necessary and release of stored Ca 2+ is sufficient for oriented migration [5]. On the other hand, the view that a [Ca 2+ ] 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 [6]. The iplA gene is the only gene known in the Dicyostelium genome so far that shares homology with IP 3receptors of higher eukaryotes. However, whether its protein product indeed constitutes a functional IP 3 -receptor and its cellular localization are not known. The iplAmutant was found to aggregate and to form fruiting bodies although neither cAMP-activated 45 Ca 2+ -entry nor a [Ca 2+ ] i -elevation were detected [6]. From these results the authors concluded that the iplA gene product is central to the regulation of [Ca 2+ ] i and that its presence and thus the presence of an agonist-activated [Ca 2+ ] i -increase is not necessary for proper chemotaxis and development. However, agents that interfere with IP 3 -receptor mediated signaling such as XestosponginC [7] were found to influence not only cAMP-induced Ca 2+ -fluxes but also the chemotactic response and aggregation of Dictyostelium [8]. In this study we aimed to clarify these conflicting findings and analyzed both, capacitative and chemoattractant-induced Ca 2+ -fluxes and the effect of their disturbance on chemotaxis and differentiation of the iplAmutant. Mn 2+ -influx was used to determine the rates of ion fluxes into cells with filled and emptied stores and related to Ca 2+ -electrode recordings in cell suspensions. We found that ion fluxes, chemotaxis and differentiation were sensitive towards alteration of the Ca 2+ -homeostasis. Capacitative Ca 2+ -entry was normal in the mutant and upon stimulation with agonist Ca 2+ -and Mn 2+ -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 Ca 2+chelator.

Extracellular [Ca 2+ ] affects development and chemotaxis of wild type and iplA -
As iplAcells formed fruiting bodies, albeit somewhat smaller in size, it was concluded that chemotactic aggregation and differentiation was normal [6]. 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 Ca 2+ affects development. Differentiation on EGTA-containing agar plates and thus the steady reduction of internal Ca 2+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 [9] 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 iplAcells (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 Ca 2+ in the agar and therefore the continuous loading of cells with Ca 2+ [10] resulted in stronger impairment of aggregation in the wild type. The delayed aggregation of wild type cells in the presence of Ca 2+ was not due to inhibition of chemotaxis (see below). Now the formation of aggregates was observed consistently at earlier time points in the iplAstrain than in wild type ( Fig. 3; on average 7 ± 0.5 h vs. 15 ± 5 h until aggregate formation in iplAand wild type amoebae, respectively, at 20 mM CaCl 2 in 6 independent experiments); indeed, under this condition differentiation of iplAcells resembled that of wild type observed in the control situation.
Then we analyzed the effect of treatment with either EGTA or Ca 2+ 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 Additional file 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 Additional file 2]. By contrast, pretreatment with 10 mM CaCl 2 did not affect the morphology (Fig. 4 E, F) or the extension of pseudopods [see Additional file 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 iplAamoebae: 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 Ca 2+ from stores impairs both, orientation and migration also in the absence of the iplA gene prod-iplAcells have an altered pattern of development Figure 1 iplAcells 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 t 7 , whereas aggregation of the mutant strain was delayed and aggregation strands broke (t 10 ); therefore, smaller fruiting bodies were formed as compared to the wild type. The full width of the image corresponds to 12.5 mm.
Development of iplAcells is impaired by depletion of internal Ca 2+ -stores due to EGTA-treatment Figure 2 Development of iplAcells is impaired by depletion of internal Ca 2+ -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 t 15 and t 18 in wild type and iplAcells, 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.
uct. By contrast, when amoebae were incubated in 10 mM CaCl 2 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 CaCl 2 (Mann-Whitney rank sum test, p = 0.041).

Buffering of intracellular [Ca 2+ ] impairs chemotaxis
The observation that aggregation occurred in the mutant cell line although a cAMP-activated increase in [Ca 2+ ] i was not detectable resulted in the conclusion that [Ca 2+ ] ichanges were not necessary to accomplish chemotaxis [6]. We used the mobile buffer approach originally described by Speksnijder et al. [11] which allows to analyze the requirement of a [Ca 2+ ] i -gradient for a given response. If in Dictyostelium a [Ca 2+ ] i -increase was necessary for chemotaxis, the presence of a Ca 2+ -chelator in the cytosol should impair orientation and/or migration. In a previous study, we had introduced the Ca 2+ -chelator BAPTA and its derivatives into the cytosol of wild type amoebae which indeed had inhibited chemotactic migration and reduced chemotactic orientation [5]. Here we used the Ca 2+ -indicator Fura2-dextran to clamp [Ca 2+ ] i and loaded the indicator into wild type and mutant cells in the absence of external CaCl 2 . The treatment affected chemotactic performance of wild type as well as iplAamoebae. Lack of extracellular CaCl 2 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 iplAcells ( Fig. 6) showing sensitivity of the mutant towards buffering of intracellular Ca 2+ -levels and eradication of [Ca 2+ ] 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 iplAcell line the ability to orient and to migrate in fact depends on an agonist-activated [Ca 2+ ] i -elevation.

Analysis of Ca 2+ -fluxes
Our findings that differentiation was sensitive towards depletion of Ca 2+ or loading of the cells with Ca 2+ and that chemotaxis was blocked by intracellular Ca 2+ -buffers led us to investigate Ca 2+ -fluxes in the mutant cell line. We used a Ca 2+ -sensitive electrode in cell suspensions to measure Ca 2+ -fluxes, an approach different from that of Traynor et al. [6] who had studied ion fluxes by 45 Ca 2+measurements. First we tested whether the coupling of stores to the plasma membrane, i.e. capacitative Ca 2+fluxes without prior stimulation with agonists were altered. In Dictyostelium induction of capacitative Ca 2+influx requires active intracellular Ca 2+ -pumps. Their In the presence of external Ca 2 aggregation is accelerated in iplAcells Figure 3 In the presence of external Ca 2 aggregation is accelerated in iplAcells. Differentiation of the mutant and the wild type strain was assayed in parallel on agar plates supplemented with 20 mM CaCl 2 . Aggregate formation occurred earlier in iplAcells (at t 7 ) than in wild type (starting at t 19 ) in the presence of Ca 2+ . The full width of the image corresponds to 12. inhibition by either thapsigargin or 2,5-di-(t-butyl)-1,4hydroquinone (BHQ) does not evoke influx; rather, stores have to be emptied by treatment with EGTA [12]. Capacitative Ca 2+ -fluxes were studied early during differentiation (t 2 -t 4 ). At this time Ca 2+ -influx and Ca 2+ -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 [1]. In iplAand in wild type cells emptying of storage compartments via preincubation of amoebae with EGTA induced capacitative Ca 2+ -entry ( Fig. 7 A) which was blocked by addition of 1 mM NaN 3 (Fig. 7 B). The characteristics of influx were comparable in wild type and mutant cells. These data show that capacitative Ca 2+influx does not depend on the product of the iplA gene.
On the other hand, agonist-activated 45 Ca 2+ -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 CaCl 2 [6]. The use of a Ca 2+ -sensitive electrode allows to measure much lower levels of extracellular Ca 2+ to analyze Ca 2+ -fluxes, in the range of approximately 1 µM Ca 2+ . Indeed, we found that under this condition reversible Ca 2+ -entry occurred after addition of 1 µM cAMP ( Fig. 8 A) that amounted to 10.2 ± 4.3 pmol Ca 2+ /10 7 cells (mean ± s.d. from 9 experiments). The level of influx represented roughly 5% of wild type influx ( Fig. 8 B and [13,14]). Ca 2+ -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 Ca 2+ -entry in wild type cells (Fig. 8 B, 9 C and [13]). Neither 10 nor 20 µM AA were effective; in the wild type 10 µM AA activates influx of 190 ± 58 pmol Ca 2+ /10 7 cells [14]. At 60 µM AA entry occurred in the mutant strain which amounted to an average of 107 ± 21 pmol Ca 2+ /10 7 cells (mean ± s.d. from 7 experiments). Preincubation of cells with the SERCA-type Ca 2+ -ATPase inhibitor BHQ reduced AA-induced influx by 82 % (Fig. 9 B) to an average of 21 ± 18 pmol Ca 2+ /10 7 cells (mean ± s.d. from 3 experiments). These data show that Ca 2+ -fluxes across the plasma membrane do occur in iplAcells as well but at a reduced level.
Next we tested whether the mutant strain was able to release stored Ca 2+ when stimulated with cAMP or AA. Fluxes were measured in suspensions of cells with permeabilized plasma membranes; any change in [Ca 2+ ] e thus reflects efflux of Ca 2+ from storage compartments. Both, cAMP and arachidonic acid activated release of stored Ca 2+ (Fig. 10 A). On average, addition of 1 µM cAMP released 7.3 ± 3.4 pmol Ca 2+ /10 7 cells and 16.3 ± 7.2 pmol Ca 2+ /10 7 cells were liberated after stimulation with 3 µM AA (mean ± s.d. from 10 and 3 experiments, respectively). The amount of Ca 2+ -efflux from stores after cAMP stimulation was 61% of that found in wild type cells (Fig. 10 B and [15]) whereas release upon AA-challenge was in the range of 5-10% of wild type (see Fig. 2 in [14]). In addition to the Ca 2+ -electrode recordings, we studied Ca 2+fluxes in suspensions of partially purified storage compartments fluorimetrically. ATP induced Ca 2+ -sequestration (Table 1) was of similar magnitude and rate as in wild type stores. This result indicates that the decreased release of Ca 2+ from the stores measured with the Ca 2+ -electrode is not due to a lack of storage capacity. Moreover, the addition of AA evoked release of Ca 2+ from stores as did inhibition of Ca 2+ -pump(s) by thapsigargin. XestosponginC that inhibits Ca 2+ -uptake and activates Ca 2+ -release in wild type [8] and the ionophore ionomycin also resulted in substantial Ca 2+ release in the mutant cell line ( Table  1). All values were in the same range as those for wild type.

Mn 2+ -quenching experiments
The [Ca 2+ ] e -recordings in suspensions of cells as described above detect the sum of Ca 2+ -influx and efflux. Therefore, a complementary approach which monitors influx only was pursued, by using the Mn 2+ -quenching technique in single intact amoebae. This method is based on the fact that many Ca 2+ -channels are permeable to Mn 2+ [16] and Chemotaxis of iplAcells is impaired in the intracellular pres-ence of a Ca 2+ -buffer Figure 6 Chemotaxis of iplAcells is impaired in the intracellular presence of a Ca 2+ -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.
that the Ca 2+ -indicator Fura2 binds Mn 2+ with high affinity. Fluorescence of the indicator is quenched upon binding [17]. We compared quenching of Fura2-dextran fluorescence activated by addition of Mn 2+ alone or in combination with 1 µM cAMP in wild type and mutant cells. Higher concentrations of MnCl 2 were required to quench Fura2-dextran fluorescence in iplAamoebae (Fig.  11, Table 2). Reduction of fluorescence occurred at 1 µM Mn 2+ and 1 µM Mn 2+ /cAMP in wild type (Fig. 11 A, B; see also [12]) but not in the mutant (Fig. 11 C, D) where addition of 100 µM Mn 2+ /cAMP was necessary; 100 µM Mn 2+ alone was not effective (Fig. 11 C, D) but started at 200 µM Mn 2+ (not shown). These data show that the reduction in ion fluxes in the mutant were indeed due to an alteration in entry mechanisms.
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 Ca 2+ -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 Mn 2+ -influx. When cells were preincubated with EGTA, the requirement for high doses of Mn 2+ to quench fluorescence was abrogated. Now capacitative and also agonist-activated Mn 2+ -influx occurred at concentrations of MnCl 2 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 Mn 2+ -influx observed in EGTAtreated 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 [12], respectively).

[Ca 2+ ] i -determination
In wild type cells treatment with EGTA augments responsiveness and cAMP-elicited [Ca 2+ ] i -transients are detected at low extracellular [Ca 2+ ] [12]. However, an agonistinduced [Ca 2+ ] i elevation was not observed in iplAcells under these conditions. On the other hand, when stores were loaded in the continued presence of CaCl 2 , we observed that differentiation of the mutant resembled the Recordings of Ca 2+ -fluxes in iplAand wild type amoebae Agonist-activated Ca 2+ -fluxes in suspensions of iplAand wild type amoebae Figure 8 Agonist-activated Ca 2+ -fluxes in suspensions of iplAand wild type amoebae. cAMP elicited reversible Ca 2+ -influx in the mutant (A) and in wild type ((B) and see [14]); measurements were done at t 7 -t 7.5 . Note the different doses of CaCl 2 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.
wild type as described above. This led us to compare the [Ca 2+ ] i -response of wild type and mutant amoebae after pretreatment with CaCl 2 . In wild type cells preincubation with 1 mM CaCl 2 for 10-15 min and its continued presence during the [Ca 2+ ] i -imaging experiment is required to activate a [Ca 2+ ] i -transient after challenge with cAMP ( Fig. 12 A and see [12]); in nominally Ca 2+ -free medium or at very low extracellular [Ca 2+ ] a cAMP-activated [Ca 2+ ] i -elevation is not observed [12]. In accordance with the data of Traynor et al. [6] this condition (Fig. 12  B) and even increasing the concentration of CaCl 2 to 20 mM resulted in no detectable [Ca 2+ ] i -increase in the mutant strain (not shown). However, we found that the basal [Ca 2+ ] i level in iplAamoebae 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).
For stronger loading of Ca 2+ -stores, we preincubated cells with 1 mM CaCl 2 for 4 h. After this treatment a [Ca 2+ ] i -elevation upon addition of cAMP was detected in 60% of wild type amoebae even at low extracellular [Ca 2+ ] levels, i.e. when the buffer used to wash the cells had been supplemented with only 1 µM CaCl 2 (Fig. 12 C). Starting from a basal level of 48 ± 4 nM, the height of the [Ca 2+ ] itransient amounted to 23 ± 2 nM (mean ± s.e.m. of 12 determinations in 8 independent experiments). Again, these conditions were not effective in iplAcells; rather, the sensitivity of the mutant was shifted to higher Ca 2+ concentrations as had been found with Mn 2+ -quenching experiments. We preincubated iplAamoebae with 20 mM CaCl 2 for 3 h; after washing thoroughly, a cAMP-induced [Ca 2+ ] i -elevation in the presence of 1 mM CaCl 2 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 [Ca 2+ ] e without prior incubation in 20 mM CaCl 2 ; the height of the increase amounted to 54 ± 6 nM (mean ± s.e.m.). Yet, this was a rare event (once in 31 determinations).

Discussion
The role of cAMP-activated [Ca 2+ ] i -changes for chemotaxis has been questioned by Traynor et al. [6] who reported results obtained with the iplAmutant cell line favouring insignificance of [Ca 2+ ] i for the chemotactic response. The authors had observed formation of fruiting bodies even though neither 45 Ca 2+ -fluxes nor an agonist induced [Ca 2+ ] i -elevation were detectable. The discrepancy between our view that a [Ca 2+ ] i -elevation is necessary for a proper chemotactic response [5] and the conclusion of Traynor et al. prompted us to analyze chemotaxis, differentiation and the [Ca 2+ ]-regulation of the iplAmutant in detail. In particular, we tested not only basal and cAMP-activated ion fluxes but also capacitative Ca 2+entry which is induced by emptying internal Ca 2+ -stores via preincubation of amoebae with EGTA [12].
Aggregation and development of wild type and mutant cells on agar plates was sensitive towards continuous emptying or loading of Ca 2+ -stores. These effects are not necessarily caused by altering chemotactic migration. It is conceivable that other Ca 2+ -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 CaCl 2 or of wild type with 1 mM CaCl 2 for 4-5 h or with 1 mM EGTA for 1 h [12] did not significantly increase or lower basal [Ca 2+ ] i it is possible that the continued presence of 10 mM EGTA or CaCl 2 for many hours affects basal levels of [Ca 2+ ] i which in turn might mediate effects on gene expression as was shown for prolonged incubation of cells with BHQ [18]. During development Dictyostelium cells form Ca 2+ -dependent and EDTA/EGTA-sensitive cell-cell contacts that are mediated by gp24 and DdCAD-1 Table 1   ( [19,20]; for review see [21]). Therefore, chelation of extracellular Ca 2+ 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 [20]. If only Ca 2+ -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 Ca 2+ -dependent processes during aggregation.

: Determination of Ca 2+ -fluxes in partially purified storage compartments of the iplAmutant and of wild type. Ca 2+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 Ca 2+ -uptake/min and mg of protein and pmol Ca 2+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
When we tested the influence of the presence of EGTA or CaCl 2 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 [5]. Prolonged incubation for more than 1 h as carried out in this study completely inhibited the chemotactic response. These results strengthen the view that Ca 2+ has a necessary role in chemotaxis in wild type and in the mutant as well.
When the cellular Ca 2+ content falls below a critical value Ca 2+ -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 CaCl 2 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 [Ca 2+ ] i was significantly lower in the former. At standard conditions the reduced basal [Ca 2+ ] 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 Ca 2+ necessary to accomplish cytoskeletal rearrangements and extrusion of a pseudopod correctly. However, in the presence of 10 mM extracellular Ca 2+ during cAMP-stimulation Ca 2+fluxes are enhanced allowing more efficient formation of pseudopods. We had shown previously that a small global elevation of [Ca 2+ ] i activates the extension of pseudopods all over the cell's circumference whereas a larger increase induces contraction of the amoebae [24]. In our view, the strongest evidence that a [Ca 2+ ] i -transient is necessary for the extension of pseudopods rests upon the experiment where a Ca 2+ -chelator was introduced into the cytosol of the amoebae. This treatment led to rounding of the amoe-bae and a general reduction of pseudopod formation (see also [5]). 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. [11] had pointed out the fact that the presence of a chelator has an effect shows that a [Ca 2+ ] igradient is essential for a given response. In summary, these data support the notion that an elevation of [Ca 2+ ] i is required to extend pseudopods; suppression of the [Ca 2+ ] i -elevation inhibits motility in general. Upon chemotactic challenge with cAMP this [Ca 2+ ] 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, [24]). Our results imply that in iplAcells such a [Ca 2+ ] 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 [Ca 2+ ] i -changes were practically undetectable under our standard condition argue for a [Ca 2+ ] 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 [Ca 2+ ] 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 [Ca 2+ ] i -changes.
By using a Ca 2+ -sensitive electrode in cell suspensions, we analyzed which aspects of [Ca 2+ ] i are controlled by the iplA gene product. Besides studying agonist-induced Ca 2+fluxes we also investigated capacitative Ca 2+ -entry and found that this type of influx was similar in mutant and wild type cell suspensions. We obtained equivalent results by testing Mn 2+ -quenching of Fura2-dextran fluorescence which showed that capacitative entry is independent of the iplA gene product.
On the other hand, using the Ca 2+ -sensitive electrode, we found that in the iplAmutant the agonist cAMP and also AA did activate Ca 2+ -entry into intact cells. The difference between the data published by Traynor et al. [6] and our results is most likely due to the experimental conditions: the magnitude of the Ca 2+ -fluxes that we observed was considerably lower than in wild type cells and detectable at low extracellular [Ca 2+ ] only. The 45 Ca 2+ -flux studies had been performed at 100 µM external CaCl 2 ; so the fraction of 45 Ca 2+ entering the cells was presumably too low to be detected reliably. Moreover, we found cAMPand AA-induced Ca 2+ -release from stores in cells with permeabilized plasma membranes. These data show that cAMP-induced Ca 2+ -release from stores in iplAcells is functional. However, much like the Ca 2+ -influx, agonist-activated liberation from stores was smaller than in wild type amoebae. In line with these results are the findings using Mn 2+ -quenching to assay ion fluxes in intact single cells: higher doses of Mn 2+ were necessary to detect influx.
There are several interpretations for the results above. (i) There are two types of channels responsible for Ca 2+influx: one type being activated by emptying of the stores and sustaining capacitative Ca 2+ -entry which is unaffected in iplAcells and the other one mediating agonist-induced Ca 2+ -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 Mn 2+ -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 Ca 2+ from the stores which is a prerequisite for the triggering of Ca 2+ -entry [12]. In the mutant this cannot proceed normally so subsequent activation of Ca 2+ -influx and the generation of a full [Ca 2+ ] i -increase is impaired. The results of the experiments where stores were strongly loaded with Ca 2+ 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 [Ca 2+ ] i -elevations occurred at an extracellular [Ca 2+ ] (see Fig. 12) where without pretreatment no increase was observed. Presumably, release of Ca 2+ from the filled stores contributed to the observed [Ca 2+ ] iincrease to a greater extent than under standard conditions. The requirement for 20 fold higher concentrations of CaCl 2 during preincubation to elicit an agonist-induced [Ca 2+ ] i -elevation in iplAcells are most likely due to the reduction in Ca 2+ -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 Ca 2+ -entry upon liberation of Ca 2+ 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 Ca 2+ -and Mn 2+ -influx induced by EGTA-treatment and a requirement for 100 fold higher ion concentrations to induce Mn 2+ -entry by cAMP. If indeed the iplA gene product constitutes an IP 3receptor 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 Ca 2+ -entry [27] should be missing in the mutant. On the other hand, emptying of stores by EGTAtreatment influences not only the IP 3 -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 [Ca 2+ ] i is regulated in the mutant although one type of Ca 2+ -store is malfunctional.

Conclusion
Our

Materials
Fura2-dextran and Fura2 were purchased from MobiTec; cAMP was from Boehringer.
Cell culture D. discoideum wild type strain Ax2 and the iplAcell lines HM1049 and HM1038 (kindly provided by Dr. D. Traynor) were cultured as described [14] 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 × 10 7 cells/ml, 150 rpm and 23°C until use. The time, in hours, after induction of development is designated t x .
[Ca 2+ ] e -electrode recordings [Ca 2+ ] e in cell suspensions was recorded as described elsewhere [14]. Cells at t 5 -t 8 were washed by repeated centrifugation and resupended at 5 × 10 7 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 [15]. Capacitative Ca 2+ -influx was analyzed in cells with emptied storage compartments [12]: amoebae at t 2 -t 4 were incubated with 5 mM EGTA for 30 min before washing in the above buffer.

[Ca 2+ ] i -determination and Mn 2+ -quenching experiments
Cells were loaded with Fura2-dextran (5 mg/ml + 1 mM CaCl 2 ) at t 4 -t 5 as described [12]. 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 CaCl 2 were added. In a series of experiments to load stores, wild type and iplAcells were incubated with 1 mM CaCl 2 for 4-5 h and with 20 mM CaCl 2 for 2-3 h, respectively. Then they were thoroughly washed exactly as described previously [12] and incubated either in H5-buffer supplemented with 1 µM CaCl 2 (wild type; free [Ca 2+ ] in the solution was measured to be 2-2.5 µM, see also [12]) or in H5-buffer +1 mM CaCl 2 (iplA -); final volume was 90 µl. Single cell [Ca 2+ ] i -imaging was performed at t 7 -t 8 as described [14]; stimulation was done by adding 10 µl of cAMP (10 µM). For Mn 2+ -quenching assays, washed cells were incubated in H5-buffer and challenged with Mn 2+ or Mn 2+ /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 Ca 2+ -fluxes in partially purified storage compartments
Analysis of vesicular Ca 2+ -fluxes was done as described [8]. In brief, 3 ml of cells at t 1 -t 6 (2 × 10 8 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 MgCl 2 , 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 [Ca 2+ ] with Fura2.

Chemotaxis assays
Cells were analyzed for chemotaxis towards a capillary filled with 0.1 mM cAMP [28]. 250 µl of 1 × 10 5 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 CaCl 2 ; 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 pat-tern 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 Ca 2+ -buffer on chemotaxis, cells were loaded with Fura2-dextran (5 mg/ml in the loading solution) by electroporation in the absence of added external Ca 2+ . The amount of indicator present in the cytosol is in the range of 2-5% of the concentration present during electroporation [29]. 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 [5].

Analysis of differentiation
Time lapse recordings of the development of Ax2 and iplAcells on 1.5 % agar in H5-buffer (H5-agar) were done by placing 4 × 10 6 cells each on one half of a petri dish (∅ 35 mm) at t 1 . 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 CaCl 2 . Then H5-agar contained either 5-20 mM EGTA or 5-20 mM CaCl 2 . Additional material