Changes in Intracellular Ca2+ Levels Induced by Cytokines and P2 Agonists Differentially Modulate Proliferation or Commitment with Macrophage Differentiation in Murine Hematopoietic Cells*

The role of intracellular Ca2+ (Ca2+i) on hematopoiesis was investigated in long term bone marrow cultures using cytokines and agonists of P2 receptors. Cytokines interleukin 3 and granulocyte/macrophage colony stimulator factor promoted a modest increase in Ca2+i concentration ([Ca2+]i) with activation of phospholipase Cγ, MEK1/2, and Ca2+/calmodulin kinase II. Involvement of protein kinase C was restricted to stimulation with interleukin 3. In addition, these cytokines promoted proliferation (20 times) and an increase in the Gr-1-Mac-1+ population with participation of gap junctions (GJ). Nevertheless ATP, ADP, and UTP promoted a large increase in [Ca2+]i, moderate proliferation (6 times), a reduction in the primitive Gr-1-Mac-1-c-Kit+ population, and differentiation into macrophages without participation of GJ. It is likely that Ca2+i participates as a regulator of hematopoietic signaling: moderate increases in [Ca2+]i would be related to cytokine-dependent proliferation with participation of GJ, whereas high increases in [Ca2+]i would be related to macrophage differentiation without maintenance of the primitive population.


Intracellular Ca 2ϩ (Ca 2ϩ
i ) 2 has important roles in many intracellular signaling pathways that participate in distinct cell functions. Time, intensity, and localization of Ca 2ϩ i events control its effects. Several intracellular proteins and ionic channels are sensitive to changes in [Ca 2ϩ ] i , translating these changes into cellular physiological effects. The major family of kinases associated with Ca 2ϩ i signaling are the classical Ca 2ϩ -dependent protein kinase C (PKC) and calmodulins, which activate the calmodulin-dependent kinases (CaMKs).
Intracellular Ca 2ϩ stores such as those in endoplasmic and nuclear reticula and mitochondria have an important role in Ca 2ϩ i signaling. Many types of agonists promote Ca 2ϩ i release from cellular stores, e.g. agonists that bind to G q/11 -coupled receptors, which activate phospholipase C␤ (PLC␤), or agonists that operate receptors with tyrosine phosphorylation, which can be recognized by PLC␥. Both PLC␤ and PLC␥ catalyze the hydrolysis of phosphatidylinositol 4,5-bisphosphate to produce inositol 1,4,5-trisphosphate (InsP 3 ) and diacylglycerol (1,2). Diacylglycerol and InsP 3 act as second messengers; diacylglycerol activates PKC on the cellular membrane, whereas InsP 3 is spread in the cytoplasm releasing Ca 2ϩ i from the endoplasmic and nuclear reticula.
Cytokines are the most important regulators in hematopoiesis. They activate distinct cytokine receptors with intrinsic kinase activity such as receptor of interleukin 6 (IL-6) and stem cell factor (SCF), which bind to c-Kit receptor; moreover cytokines also activate receptors without intrinsic kinase activity that are Janus kinase-dependent receptors. Janus kinases catalyze tyrosine phosphorylation on a great variety of cytokine receptors such as receptor of IL-3, granulocyte/macrophage colony stimulator factor (GM-CSF), granulocyte colony stimulator factor (G-CSF), macrophage colony stimulator factor (M-CSF), erythropoietin (EPO), etc. Some studies have described that some cytokines such as IL-1, IL-2, and SCF are able to promote Ca 2ϩ i increase in chondrocytes, neutrophils, and CD34 ϩ cells (3)(4)(5); however, direct participation of Ca 2ϩ i in hematopoiesis has not been described.
In recent years, the involvement of P2 receptors in hematopoiesis has been investigated. These receptors play a role in megakaryocytic formation (6), hematopoietic stem cell (HSC) migration (7), and CD34 ϩ cell proliferation (8). P2 receptors are divided into ionic channel P2X receptors and G-protein-coupled P2Y receptors; they are activated by ATP and its analogs, and in most cases they increase [Ca 2ϩ ] i to promote the effects of P2 receptors.
Thus, cytokine and P2 receptor activation may promote Ca 2ϩ i release through PLC␥ or PLC␤ activation, respectively, or by Ca 2ϩ i influx through P2X receptors. Therefore, Ca 2ϩ i is likely to participate in hematopoiesis. In this study, we have shown participation of Ca 2ϩ i in proliferation and differentiation of hematopoietic cells in long term bone marrow cultures (LTBMC), which support myelopoiesis. A modest increase in Ca 2ϩ concentration by IL-3 and GM-CSF may act mainly on proliferation with participation of GJ, whereas large [Ca 2ϩ ] i increases by ATP and analogs promote weak proliferation and induce differentiation of hematopoietic cells into macrophages without participation of GJ.

EXPERIMENTAL PROCEDURES
Extraction of Bone Marrow and LTBMC-To establish the stroma layer, femur bones were excised from mice (C57Bl/6), and the medullary cavities were aseptically flushed with Iscove's modified Dulbecco's medium (IMDM; Invitrogen). The cells so obtained were seeded and incubated (37°C in 5% CO 2 ) in a humidified incubator. Half of the medium in each flask was replaced weekly with an equal amount of fresh medium. IMDM was supplemented with 5% fetal bovine serum (FBS; Cultilab), 20% horse serum (StemCell Technologies), and 10 Ϫ6 M hydrocortisone (Sigma). At the end of the 8th week, after stroma formation, the remaining hematopoietic cells were removed. New bone marrows from other mice were collected in supplemented IMDM (10 ml) and cultured (2 h) in tissue culture flasks (75 cm 2 ). Non-adherent cells were collected by removing the medium, and 10 6 cells/well (12-well plates) were added to precultured stroma.
After 1-week co-culture, the cells were further cultured in 0.5% FBS, IMDM for 24 h. Subsequently LTBMC were stimulated with IL-3 (Sigma) or GM-CSF (Sigma) in 12-well plates in 0.5% FBS, IMDM (1 ml) to evaluate the proliferation and cell populations. Inhibitor was added 1 h before stimulation with cytokine. The number of cells present in the supernatant was evaluated every 24 h by using a Neubauer chamber. Cell counts were normalized by the number of cells present before stimulation. The experiments were approved by the Ethics Committee of the Federal University of São Paulo (1464/03).
Calcium Measurements in LTBMC-Bone marrow cells were seeded on cover glass slides (25 mm) in 6-well plates. For [Ca 2ϩ ] i measurements, the cells were incubated (40 min at room temperature) with fluo-4/AM (10 M) and pluronic acid (0.01%) and washed with Tyrode's solution. Images were captured in two Z planes with a microscope (Axiovert 100 M, Zeiss, Heidelberg, Germany) equipped with a laser scanner (LSM 510 META, Zeiss) and using an objective (Plan-Neofluor, 63ϫ, 1.4 numerical aperture) under oil immersion. The fluo-4 probe was excited with argon laser ( Ex ϭ 488 nm), and light emission was detected by using a bypass filter ( Em ϭ 500 -550 nm). The pinhole device was not used for [Ca 2ϩ ] i measurements. Images were collected at ϳ4.5-s intervals for about 2 min. Fluorescence intensity was normalized with reference to the basal fluorescence by using Examiner 3.2 (Zeiss) and Spectralyzer softwares (Spectralyzer, Philadelphia, PA). Max refers to the value of maximal intensity per pixel after stimulation by the agonists, and Basal refers to the maximal value per pixel obtained before stimulation. The mean intensity by Basal and Max images (8 bits ϭ 256 levels) in supplemental Fig. S1 was obtained using the Image J software.
Statistical Analysis-The fluo-4 fluorescence intensity were normalized with reference to basal intensity and were shown to be representative pseudocolored images according to a fluorescence intensity scale ranging from 0 (black) to 255 (white). Data were expressed as the means Ϯ S.E. Statistical comparisons were performed by using Student's t test or analysis of variance. p values Ͻ0.05 were considered statistically significant.

Hematopoietic Cytokines Promote Increase in [Ca 2ϩ ] i of
Hematopoietic Cells-LTBMC reproduce myelopoiesis, the formation of granulocytic, monocyte/macrophage, and erythroid cells, although erythrocyte formation depends on external erythropoietin (9). In this kind of culture, a visible formation of cobblestones is observed where hematopoietic progenitor cells can be found in close relationship with stromal cells (9,10). LTBMC are composed of stromal cells and hematopoietic cells that grow in the stroma. LTBMC were incubated in IMDM supplemented with 0.5% FBS for 24 h before any stimulus to synchronize the cellular cycle; hematopoietic cells were sensitive to lower concentrations of FBS (data not shown).
To determine participation of Ca 2ϩ i in hematopoiesis, Ca 2ϩ i concentration was monitored in cobblestone areas of LTBMC after addition of agonists. In the first step it was determined which cytokines promoted an increase in

Versatility of Ca 2؉
i Signaling in a Hematopoietic System [Ca 2ϩ ] i in cobblestone areas. Thus, measurements of [Ca 2ϩ ] i were obtained in two planes: from the inferior Z plane (down) where more stromal and some hematopoietic progenitor cells were found and on the superior Z plane (up) where more hematopoietic cells were observed (Fig. 1). Because hematopoietic cells are smaller cells with circular shape and stromal cells are bigger with variable shape, these morphological features were used in their identification.
EC 50 values for cytokines range from 0.05 to 3 ng/ml as informed by Sigma-Aldrich. These values were confirmed by cellular proliferation assays with IL-3 and GM-CSF, and maximum proliferation efficacy for these agonists occurred at about 10 -50 ng/ml (supplemental Fig. S2A). All cytokines tested (SCF, IL-3 (supplemental Video S1), IL-6, GM-CSF, EPO, and IL-7) promoted an increase in [Ca 2ϩ ] i ( Fig. 1) in hematopoietic and stromal cells in the two Z planes

Versatility of Ca 2؉ i Signaling in a Hematopoietic System
evaluated. This cytokine-dependent increase in [Ca 2ϩ ] i was not easily observed and was lower than the ATP-dependent [Ca 2ϩ ] i increase (see supplemental Fig. S1). Interestingly cytokines that activated hematopoietic progenitor cells such as SCF or specific cytokines such as IL-7 and EPO were able to produce an increase in [Ca 2ϩ ] i in stromal and hematopoietic cells.

Ca 2ϩ i Signaling Inhibitors Decrease IL-3-and GM-CSF-dependent Proliferation-Because all cytokine tested promote an increase in [Ca 2ϩ ] i it is likely that Ca 2ϩ
i participates in cell proliferation and differentiation. To evaluate the role of Ca 2ϩ i in these processes we used two important hematopoietic cytokines, IL-3 and GM-CSF, which promote proliferation and differentiation in hematopoietic cells. IL-3 and GM-CSF pro-moted proliferation in LTBMC that was about 20 times the initial cell number after 3 days (Fig. 2, A and B). Proliferation was followed for 3 days (supplemental Fig. S2B); in addition, cell cycle analysis also revealed a proliferation state in the primitive population Fig. S2C). All stimuli by agonists were performed in IMDM, 0.5% FBS.
Lymphocytes, erythrocytes, and remaining cells of bone marrow coculture can be found in the Gr-1 Ϫ Mac-1 Ϫ population, and the primitive population c-Kit ϩ corresponds to 1% of the total cells of the culture. The c-Kit ϩ population present in LTBMC after stimulus with cytokines was still able to form either colony-forming unit granulocyte-macrophage or colony-forming unit granulocyte-erythroblastmacrophage-megakaryocyte colonies (supplemental Fig. S3).
Participation of Ca 2ϩ i in proliferation was evaluated by using Ca 2ϩ i signaling inhibitors. LTBMC were incubated with the signaling inhibitors, and the Ca 2ϩ chelator BAPTA-AM for 1 h before stimulation with IL-3 and GM-CSF. Because all inhibitors were diluted in dimethyl sulfoxide (DMSO), stimulus by IL-3 and GM-CSF in DMSO were used as controls in statistical analysis. Supernatant cells in LTBMC stimulated by IL-3 and GM-CSF were counted for 3 days in the presence or absence of signaling inhibitors. InsP 3 receptor (2APB), PLC (U73122), CaMKII (KN-62), and MEK (PD98059) inhibitors and BAPTA significantly decreased IL-3-and GM-CSF-dependent cell proliferation ( Fig. 2A). PKC inhibitors (GF109203 and chelerythrine) partly blocked IL-3-dependent cell proliferation ( Fig.  2A) but did not promote a decrease in GM-CSF-dependent cell proliferation (Fig. 2B). The concentration of inhibitors is shown in supplemental Fig. S2B. The signaling inhibitor was also able to decrease the percent fraction of cells in phase S/G 2 /M (supplemental Fig. S4, A and B). These results confirm the role of PKC in the effect of IL-3 described previously in hematopoietic lineages (11,12). Because the fraction of apoptotic cells in LTBMC was not altered by the presence of inhibitors, their specific effect in intracellular signaling is inferred (supplemental Fig. S4, C and D).

Versatility of Ca 2؉ i Signaling in a Hematopoietic System
Quantification of cell populations in LTBMC was assessed by flow cytometry after the 3rd day of stimulation with cytokines. As expected, LTBMC stimulated by IL-3 and GM-CSF produced myeloid cells. IL-3 and GM-CSF produced an increase in the Gr-1 Ϫ Mac-1 ϩ population as compared with its control (0.5% FBS). Proliferation and flow cytometry analyses were performed by using the same samples. When LTBMC were stimulated by IL-3, a significant increase in the Gr-1 Ϫ Mac-1 ϩ population occurred in the presence of 2APB and KN-62 (Fig. 2C). When cells were stimulated by GM-CSF in the presence of the CaMKII inhibitor KN-62 a significant difference was observed with an increase in the Gr-1 Ϫ Mac-1 ϩ population (Fig. 2D). PD98059, an MEK inhibitor, promoted an increase in the Gr-1 ϩ Mac-1 ϩ population when cells were stimulated with GM-CSF. The statistic analysis is described in detail in supplemental Table S1. The inhibitors used in this study can inhibit cell proliferation and act at distinct moments of myeloid differentiation. The fraction of progenitor Gr-1 Ϫ Mac-1 Ϫ c-Kit ϩ cells in the presence of cell signaling inhibitors was not altered (Fig.  2, E and F).
IL-3 and GM-CSF Activate Intracellular Ca 2ϩ -dependent Kinases-Because Ca 2ϩ signaling inhibitors were able to inhibit IL-3-and GM-CSF-dependent cell proliferation, antibodies to active forms related to Ca 2ϩ i signaling such as PLC␥, PKC, and CaMKII were tested. Anti-P-MEK antibody, a protein that acts on cytokine signaling by activating MAPKs, was used as a positive control. The cells were stimulated by IL-3 and GM-CSF in Tyrode's solution (37°C for 5 min). Under this condition, an increase in the expression of P-MEK and P-PLC␥ was observed (Fig. 3, A and  B). Participation of PKC in the IL-3 response was shown previously by using PKC inhibitors ( Fig. 2A), and Fig. 3C confirms participation of PKC in IL-3-dependent proliferation. Activation of CaMKII may be associated mainly with GM-CSF (Fig. 3D); however, CaMKII does not act exclusively in GM-CSF signaling because IL-3 partially activated P-CaMKII (Fig. 3D), and IL-3dependent proliferation was also decreased by KN-62 ( Fig. 2A). Therefore, cytokines IL-3 and GM-CSF promote high cell proliferation with participation of Ca 2ϩ i and activation of kinases without a decrease in the primitive population. As a negative control, a secondary antibody, rabbit or goat anti-IgG-Alexa Fluor 488 conjugate, was used (supplemental Fig. S5). To know the ability of P2 receptors to promote proliferation in hematopoietic cells, LTBMC were stimulated with ATP and analogs. Maximal cell proliferation in LTBMC was obtained with 1 mM ATP and its analogs (supplemental Fig. S6A). P2 agonists induced a significant proliferation on the 1st day (Fig.  4B); however, such proliferation was transient, and its efficacy was lower than that of the cytokine-dependent proliferation (Fig. 2, A and B). Of note, daily stimulation with 1 mM ATP did not cause a higher proliferative effect compared with that induced by a single stimulation (data not shown). However, the possibility that the transient effect of P2 agonists on hematopoietic cell proliferation and differentiation is because of their faster degradation deserves further investigation.

Activation of P2 Receptors Produces Both a Higher Increase in [Ca 2ϩ ] i and Macrophage Differentiation of Hematopoietic Cells in LTBMC-[Ca
Interestingly evaluation of populations of LTBMC showed that stimulation with ATP, ADP, and UTP induced an increase in the Gr-1 ϩ Mac-1 ϩ population (Fig. 4C) and a decrease in the  NOVEMBER 14, 2008 • VOLUME 283 • NUMBER 46

JOURNAL OF BIOLOGICAL CHEMISTRY 31913
Gr-1 Ϫ Mac-1 Ϫ c-Kit ϩ immature population (Fig. 4D). Stimulus of LTBMC by ATP and analogs produced mainly macrophages as revealed by Giemsa/May-Grünwald stain (Table 1). Reduction in the Gr-1 Ϫ Mac-1 Ϫ c-Kit ϩ population by ATP and its analogs can account for the transient proliferation observed. In addition, BAPTA blocked the P2 agonist-dependent effects (Fig. 4, C and D). Another compound used to promote Ca 2ϩ increase was the Ca 2ϩ ionophore A23187, which was able to induce an increase in the Gr-1 ϩ Mac-1 ϩ population (supplemental Fig. S6B).

Participation of P2Y Receptors in Maturation of
Hematopoietic Cells-In hematopoietic cells, P2Y receptors are activated mainly by ADP and UTP. The presence of P2Y receptors seems to be a characteristic on hematopoietic cells (7,(13)(14)(15). However, the presence of P2X receptors has also been described in macrophages, platelets, and granulocytic cells (14, 16, 17).  IL-3, GM-CSF, UTP, and ADP promoted a decrease in the number of apoptotic cells (Fig. 5A). Only ATP, the agonist that promotes less proliferation, was unable to promote a significant decrease in apoptosis.
Among P2Y receptors, ADP-activated P2Y 1 receptor is expressed in the human CD34 ϩ progenitor cell population (8,18) and murine HSCs (Lin Ϫ c-Kit ϩ Sca1 ϩ ) and is expressed in lesser amounts in mature hematopoietic cells (data not shown). Thus, it is likely that this receptor is involved in hematopoiesis. For these reasons, participation of P2Y 1 receptor in the ADPdependent response was evaluated by using its specific inhibitor, MRS2179, which was able to block both the increase of Gr-1 ϩ Mac-1 ϩ cells (Fig. 5B) and the decrease of immature Gr-1 Ϫ Mac-1 Ϫ c-Kit ϩ cells (Fig. 5C). MRS2365 (10 Ϫ5 M), a specific P2Y 1 agonist, also induced [Ca 2ϩ ] i increase (supplemental Fig. S7D). The primitive Gr-1 Ϫ Mac-1 Ϫ c-Kit ϩ population expressed the P2Y 1 , P2Y 2 , P2Y 4 , P2Y 6 , P2Y 12 , and P2Y 14 receptors (Fig. 5D) similarly to the short term HSC (c-Kit ϩ Sca-1 ϩ Lin low ) population (data not shown). Participation of P2X receptors in proliferation and differentiation of hematopoietic cells was not tested; however, ␣,␤-methylene ATP and benzoyl ATP promoted a higher [Ca 2ϩ ] i increase in stromal and hematopoietic cells (supplemental Fig. S7). ADP and UTP induced similar effects, although they bind to different receptors.
Intercellular Communication of Cytokine-dependent Ca 2ϩ Signaling Occurs through Connexins-When LTBMC were stimulated with cytokines, Ca 2ϩ i events such as Ca 2ϩ i waves ( Fig. 6A and supplemental Video S3), frequently associated with the presence of GJ, oscillating [Ca 2ϩ ] i increases, and cell mobility were observed (Fig. 6B). In Fig. 6B GJ are among distinctive structures related to intercellular communication. GJ comprise two hemichannels, the connexons, each of them composed of six connexin molecules that are transmembrane proteins. Previously the presence of Cx43 has been described in hematopoietic tissue (19). In Fig. 7A, the presence of Cx43 in a cobblestone area (i) and in detail (ii) between two hematopoietic cells is shown.
Participation of GJ in Ca 2ϩ i signaling was tested by the presence of carbenoxolone (100 M), an inhibitor of GJ. Cells were incubated for 2 h with carbenoxolone and stimulated with IL-3 or GM-CSF, as described above, followed by addi-

Macrophages Monocytoid Neutrophils
Immature forms Blasts NOVEMBER 14, 2008 • VOLUME 283 • NUMBER 46 tion of ATP, and the rate of cells responsive to agonists was calculated. As shown in Fig. 7B, the IL-3-dependent [Ca 2ϩ ] i increase was lower than the ATP-induced increase observed; moreover the [Ca 2ϩ ] i increase was higher in hematopoietic than in stromal cells. Carbenoxolone was able to decrease the number of cells responsive to IL-3 and GM-CSF; however, alteration of the ATP response was not observed (Fig.  7C). Cells with increases in F t /F 0 higher than 0.2 above the resting level after stimulus were considered responsive. Carbenoxolone was also able to block IL-3-and GM-CSF-dependent proliferation showing the important role of connexins in hematopoietic proliferation (Fig. 7D). However, carbenoxolone was not able to significantly block ADP-dependent proliferation (Fig. 7D).

DISCUSSION
Cytokines are the main controllers in the hematopoietic system; these agonists bind to specific receptors (cytokine receptors) that are classified in many families. The intracellular pathways Ras-Raf-MEK-MAPK and Janus kinases/signal transducers and activators of transcription are the two main pathways triggered by cytokines (20,21). However, recent studies have shown participation of other intracellular pathways that control proliferation, differentiation, and the cell death process in hematopoietic stem cells. For example, participation of the Wnt and Notch pathways in HSC maintenance and differentiation was reported recently (22,23). In this report, we describe the role of Ca 2ϩ i pathway; Ca 2ϩ i is an important second messenger associated with many intracellular processes in murine myelopoiesis.
We demonstrate that low increases in [Ca 2ϩ ] i are related to high IL-3-and GM-CSF-dependent proliferation because the cytokine effects were sensitive to the PLC (U73122) and InsP 3 receptor (2APB) inhibitors, which are active components in the Ca 2ϩ i release process (Fig. 2, A and B). In addition, activation of PLC␥, which recognizes phosphorylated tyrosine in cytokine receptors, was demonstrated by confocal microscopy (Fig. 3B). Release of Ca 2ϩ i induced activation of Ca 2ϩdependent proteins (PKC and CaMKII) that translate their signals into physiological responses (Fig. 3, C and D).
Participation of PKC in hematopoietic proliferation was partial and restricted to responses to IL-3 ( Fig. 2A). Participation of PKC was described previously in the differentiation process of hematopoietic precursors. High concentrations of phorbol 12-myristate 13-acetate induced myelomonocyte differentiation, whereas low concentrations induced eosinophil differentiation (24). Whetton et al. (11) also showed that IL-3 and phorbol 12-myristate 13-acetate are able to promote proliferation, an increase in [Ca 2ϩ ] i , and PKC activation in factor-dependent cell established at the Paterson

Versatility of Ca 2؉ i Signaling in a Hematopoietic System
Institute with mixed differentiation potential (FDCP-Mix), an IL-3-dependent stem cell line. CaMKII, another Ca 2ϩ -dependent kinase expressed in most cells, apparently acts on IL-3 and GMCSF responses because KN-62 partly inhibited IL-3-and GM-CSF-dependent proliferation and induced an increase in the Gr-1 Ϫ Mac-1 ϩ population. Other CaMKIIs such as CaMKIV, related to quiescence of murine HSCs, also can control hematopoiesis by cyclic AMP response element-binding protein activation (25).
Ca 2ϩ i ion is a very versatile messenger, and it is not neces-sary that the intensity of Ca 2ϩ i release be directly proportional to the observed effect. For example, in many processes, Ca 2ϩ evokes a biphasic response like that in InsP 3 and ryanodine receptors. An initial increase in [Ca 2ϩ ] i induces a positive feedback and opens Ca 2ϩ channels. Inversely high Ca 2ϩ concentrations induce a negative feedback and block Ca 2ϩ channels (26). Ca 2ϩ sparks released between the cell plasma membrane and the endoplasmic reticulum can produce relaxation in vascular smooth muscle by opening of Ca 2ϩ -dependent K ϩ channels, whereas a global [Ca 2ϩ ] i increase in the same cell can induce contraction (27). This versatility of Ca 2ϩ i signaling allows these ions to act on hematopoiesis in distinct ways depending on stimulus duration, amount of Ca 2ϩ released, and type of protein activated. In the hematopoietic system, cytokines and P2 receptor agonists altered the proportion of Gr-1, Mac-1, and c-Kit populations by different [Ca 2ϩ ] i increases. IL-3 and GM-CSF induced a low increase in [Ca 2ϩ ] i , high proliferation, and an increase in the Gr-1 Ϫ Mac-1 ϩ population without a change in the percent value for the primitive Gr-1 Ϫ Mac-1 Ϫ c-Kit ϩ cells and with participation of GJ. However, ATP, ADP, and UTP induced a large [Ca 2ϩ ] i increase and low proliferation and differentiation of hematopoietic cells (macrophage differentiation and a decrease of the primitive Gr-1 Ϫ Mac-1 Ϫ c-Kit ϩ populations). Thus, low [Ca 2ϩ ] i release may induce proliferation with normal differentiation, whereas high [Ca 2ϩ ] i release can induce preferential rapid cell differentiation of the primitive population. In human CD34 ϩ , monocyte, and HL-60 cells, a large Ca 2ϩ i increase by Ca 2ϩ ionophores induces differentiation in dendrocytes (28 -30).
Participation of GJ is another distinctive characteristic of cytokines and P2 agonist response. A high [Ca 2ϩ ] i increase by ATP and its analogs was independent of GJ, whereas the IL-3and GM-CSF-dependent [Ca 2ϩ ] i increase was partly inhibited by carbenoxolone (Fig. 7C). The function of GJ in hematopoiesis is not clear, but it is known that Cx43 is highly expressed in bone marrow of neonatal animals and decreases after birth (31).

Versatility of Ca 2؉ i Signaling in a Hematopoietic System
However, expression of Cx43 can be increased in regeneration processes (32). Montecino-Rodriguez et al. (33) showed that Cx43 Ϫ/ϩ heterozygote animals have problems with myeloid and lymphoid regeneration. GJ also act on the maintenance of primitive human CD34 ϩ cells co-cultivated with L87/4 stromal fibroblasts (34). In addition, the stroma regulates quiescence of primitive cells through GJ by keeping more CD34 ϩ cells quiescent in the stromal cell line S17, whereas higher proliferation occurs in the stroma of leukemia patients, who have fewer GJ (35). Proliferation inhibition by carbenoxolone shows the important role of GJ in proliferation (Fig. 7D). Participation of GJ in Ca 2ϩ i release by cytokines explains how stromal cells are also responsive to specific cytokines such as SCF, IL-7, and EPO. Consequently primitive cells may conduct Ca 2ϩ i signals to stromal cells that trigger others effects, like cytokine release, modification of extracellular matrix, and cellular mobilization. These effects have not yet been investigated.
In the supplemental Video S3, a Ca 2ϩ wave is clearly observed in the stromal cell showing the functional presence of intercellular communication. Ca 2ϩ waves without stimulation were seldom observed; however, movements between hematopoietic and stromal cells are often observed and would occur by pseudopodia, which have been described previously in murine hematopoietic primitive cells (36).
The action of purinergic receptors is not clear. Herein we propose that these receptors promote differentiation in the primitive population. In another study, action of ATP was shown to be synergic with cytokines, increasing the number of progenitor cells (8). The source of ATP in the hematopoietic system is another intriguing fact. The release of ATP in the bone marrow microenvironment may occur with other neurotransmitters such as adrenergic transmitters that regulate the attraction of stem cells to their niches (37) or by either release of endothelial cells or release of cytoplasm in bone fracture (38 -40).
These results allowed us both to show an important role for Ca 2ϩ i in hematopoiesis and to evidence how Ca 2ϩ signaling contributes to this process. Temporal changes in Ca 2ϩ signals and intercellular communication should be further investigated to better understand the role of Ca 2ϩ in the hematopoietic system. The low cytoplasm volume and the small size of hematopoietic cells raise difficulties in better evaluating subcellular aspects associated to Ca 2ϩ i signaling. In this study, redundancy in Ca 2ϩ signaling was evidenced for all agonists used. Although distinct effects were induced, Ca 2ϩ was utilized as the coordinator of their functions. Participation of different subtypes of Ras, MAPK, PLC␥, PLC␤, PKC, and CaMK proteins have to be investigated. A cross-talk between Ca 2ϩ i signaling and Ras-Raf-MEK-MAPK pathways are under investigation in murine and human HSCs.
It may be stated that Ca 2ϩ i participates in hematopoiesis depending on signal intensity, type of kinase activated, and participation of GJ. Herein we report results of an ongoing study aimed at the understanding of Ca 2ϩ signal translators and other factors that control the development of murine hematopoiesis.