,
,
* Departments of Medicine and
Biochemistry, McGill University, and
Endocrine Laboratory and
Division of Hematology, McGill University Health Centre, Montreal, Canada
Correspondence: L. F. Congote, McGill University Health Centre, Endocrine Laboratory, 687 Avenue des Pins, Ouest, Montreal, Canada H3A 1A1. E-mail: luis.f.congote{at}muhc.mcgill.ca
 TOP ABSTRACT INTRODUCTIONMATERIALS AND METHODSRESULTSDISCUSSIONREFERENCES |
|---|
Key Words: Stat5 • CD34
|
TOPABSTRACTINTRODUCTIONMATERIALS AND METHODSRESULTSDISCUSSIONREFERENCES |
|---|
We have found that insulin-like growth factor II (IGF-II) can be produced in insect cells as a properly folded and secreted peptide as a chimera [BOMIGF; NQPMVHTY-hIGF-II(9–67)] composed of the signal peptide and the first nine amino acids of the insect IGF bombyxin, followed by amino acids 9–67 of human IGF-II [2 ]. The cDNA of human interleukin (IL)-3 was inserted at the C-terminal end of BOMIGF with a flexible linker to construct the chimera BOMIGF–IL-3 [3 ]. This bivalent HGF fusion protein had a definite, proliferative advantage on hematopoietic cells in comparison with the combined administration of its individual components. In addition to an enhanced proliferation effect, BOMIGF–IL-3 appeared to promote the survival of TF-1 cells at extremely low concentrations (pico- to femtomolar range).
Although great effort has been focused over the last few decades toward documenting and characterizing the effects of IGF-I and IGF-II on the development and growth of cells of different tissue origins, it is only recently becoming evident that these factors also play a prominent role in promoting cell survival. IGF-I has been shown to prevent apoptosis ensuing from cytokine withdrawal in IL-3-dependent hematopoietic cells and murine factor-dependent cell progenitors [4 ]. The addition of IGF-I to cultures of burst-forming units–erythroid cells significantly reduces the amount of DNA fragmentation characteristically observed during apoptotic cell death [5 ]. The hematopoietic cell-survival effects exerted by IL-3 are also widely recognized. The addition of IL-3 to suspension cultures enriched with committed hematopoietic stem cells effectively prevents the loss of progenitors giving rise to hematopoietic cells of the erythroid, granulocyte, and macrophage pathways [6 ]. Removal of IL-3 from cultures of 32D cells results in cessation of proliferation and the induction of programmed cell death, but these effects can be prevented by the overexpression of the antiapoptotic protein bcl-2 [7 ]. In this paper, we investigated the possibility that the bifunctional protein chimera BOMIGF–IL-3 has a selective advantage on cell survival of the CD34-positive hematopoietic cell line TF-1 as compared with the action of the mixture of equimolar amounts of the single components BOMIGF and IL-3. We analyzed the effects of the chimera on different aspects of cell death, apoptosis, and proliferation, followed with a study on the role of the fusion protein on specific downstream signal-transduction pathways that have been implicated in the antiapoptotic and proliferative actions of IGFs and IL-3, namely the expression of the antiapoptotic protein BclxL and the stimulation of phosphatidylinositol-3 kinase (PI-3K) activity [8 9 10 ].
|
TOPABSTRACTINTRODUCTIONMATERIALS AND METHODSRESULTSDISCUSSIONREFERENCES |
|---|
Cell culture
TF-1 cells [American Type Culture Collection (ATCC), Manassas, VA] were maintained in culture in RPMI-1640 medium supplemented with 10% (v/v) fetal bovine serum (FBS; Invitrogen, Burlington, Ontario, Canada, or Biomedia, Drummondville, Quebec, Canada). This cytokine-dependent hematopoietic cell line required the addition of granulocyte macrophage-colony stimulating factor (GM-CSF) for growth (0.1–1 ng/ml culture medium). Once large amounts of baculovirus-produced recombinant proteins were available, GM-CSF was replaced with BOMIGF–IL-3 as a growth supplement. All incubations with cytokines were done in a serum-free medium consisting of RPMI 1640 containing 300 µg/ml (w/v) fatty acid-free, tissue culture-tested bovine serum albumin (Sigma Chemical Co., Mississauga, Ontario, Canada) and 30 µg/ml (w/v) bovine transferrin (ICN Biochemicals, Costa Mesa, CA). Cells were starved overnight in the serum-free medium supplemented with 0.5% FBS (v/v) before studies on the effect of cytokines on growth or apoptosis. HL-60 cells (ATCC) were cultured as indicated above, without cytokine supplements. Human umbilical cord endothelial cells were obtained from Clonetics (Cambrex, Walkesville, MD) and cultured as indicated for bone marrow endothelial cells [11
] but under normal (nonhypoxic) oxygen tension. Umbilical cord cells are more sensitive to starvation than bone marrow cells. Therefore, the starvation period was reduced to 12 h and was done in confluent cell cultures. Cell proliferation and cyotoxicity were measured using the Alamar Blue technique (Biosource) as described previously [11
].
Cell-cycle analysis
TF-1 cells, previously starved overnight as indicated above, were washed and resuspended in serum-free medium. Aliquots of 106 cells were dispensed into wells (24-well microplates, Costar, Fisher Scientific, Whitby, Ontario, Canada) containing the different growth factors in a final volume of 2 ml incubation medium and a factor concentration of 1 nM. After incubation periods of 12, 24, and 36 h, cells were collected, washed with ice-cold phosphate-buffered saline (PBS) solution, fixed in 70% (v/v) ethanol in water, and kept at 4°C until staining with cell-cycle staining solution consisting of 50 µg/ml propidium iodide (PI; Sigma Chemical Co.) and 20 µg/ml RNase (Amersham-Pharmacia, Baie d’Urfe, Quebec, Canada) in PBS. Flow cytometry analysis of 10,000 events was performed on a FACScan instrument equipped with an argon laser tuned to 488 nm.
BclxL Western blot
TF-1 cells were incubated in starvation medium overnight, washed, and incubated in the serum-free medium at a cell density of 106 cells/well in 24-well plates in a final volume of 2 ml in the presence or absence of the cytokines (1 nM). Cells were collected at selected time points, washed three times with PBS, and conserved frozen in liquid nitrogen until lysis. Cells were incubated for 30 min with lysis buffer consisting of 50 mM HEPES, pH 7.5, 150 mM NaCl, 100 mM NaF, 10 sodium pyrophosphate, 5 mM EDTA, 10% (v/v) glycerol, 0.5% (v/v) Triton X-100, 5 µg/ml (w/v) aprotinin, 2 µg/ml (w/v) leupeptin, and 1 mM phenylmethylsulfonyl fluoride. Cell lysate was cleared of insoluble material by centrifugation for 15 min at 12,000 RPM. Total cell lysate protein content for each sample was quantitated by the Bradford method using the Bio-Rad protein kit (Mississauga, Ontario, Canada). Total cell lysate protein (50 µg) was resolved on a 12% sodium dodecyl sulfate-polyacrylamide gel electrophoresis (SDS-PAGE) and subsequently electroblotted onto a nitrocellulose membrane (Hybond, Amersham-Pharmacia). Membranes were blocked in Tris-buffered saline/Tween 20 buffer (50 mM Tris-HCl, pH 7.6, 100 mM NaCl, 0.05% Tween 20) containing 0.25% (w/v) gelatin and incubated with a rabbit polyclonal anti-BclxL antibody (Transduction Laboratories, Lexington, KY) followed by incubation with a secondary goat–anti-rabbit horseradish peroxidase (HRP)-conjugated antibody (Bio-Rad). Detection was performed with the enhanced chemiluminescence (ECL) system (Amersham-Pharmacia). Amounts of BclxL were quantified using the Bandleader program of Ma’ayan Aharoni (TechKnowledge, Tel-Aviv). The uniformity of protein loadings was confirmed in some experiments after stripping the nitrocellulose membranes and repeating the immunodetection using antiactin antibodies (Sigma Chemical Co.).
IL-3 receptor, Janus kinase (Jak)-2, and signal transducer and activator of transcription (Stat)5 phosphorylation
TF-1 cells were starved as indicated above and incubated for different time periods in serum-free medium containing the cytokines to be tested at a concentration of 25 nM. The cells were collected and lysed as indicated above. About 1 mg protein from total lysate was used for immunoprecitation with a rabbit polyclonal antibody against the ßc subunit of human IL-3 receptor or rabbit anti-Jak-2 antiserum mixed with Protein A agarose beads at 4°C overnight. Beads were washed and resuspended in loading buffer before applying onto 10% SDS-PAGE gels. After semi-dry electrotransfer of proteins onto a nitrocellulose membrane, immunodetection was done with antiphosphotyrosine (4G10, Upstate Biotechnologies, Lake Placid, NY) as a primary antibody followed by a goat–anti-rabbit secondary antibody linked with HRP. Detection was done with ECL as indicated above. The uniformity of blotting was checked by reprobing the membranes with IL-3 receptor and Jak-2 rabbit antibodies. Stat5 phosphorylation was measured directly without immunoprecipitation using a monoclonal antibody to the phosphorylated form of Stat5a/b Y694/Y699 (Zymed Laboratories, San Francisco, CA) as described previously [12
]. In some experiments the chemiluminiscent detection method was replaced with a colorimetric method using a goat–anti-mouse alkaline phosphatase-congugated antibody to facilitate quantification [13
].
PI-3K assay
This assay was performed according to the method of Royal and Park [14
]. About 107 TF-1 cells/sample were incubated in starvation medium overnight, washed, and incubated with the cytokines (25 nM) in serum-free medium. Cells were collected at selected time points, washed with PBS, and treated with the ice-cold lysis buffer as indicated above. About 1 mg protein from total cell lysate was used for immunoprecipitation with 2 µg mouse monoclonal antiphosphotyrosine antibody 4G10 and 50 µl of a 50% suspension of Protein A agarose (Invitrogen) at 4°C overnight. The agarose immunocomplexes were washed four times in 50 mM HEPES, pH 7.5, 150 mM NaCl, 0.5 mM EGTA, 0.2 mM Na3VO4, and 10% (v/v) glycerol and were resuspended in 50 Fl kinase buffer (20 mM Tris-HCl, pH 7.5, 100 mM NaCl, 5 mM MgC2, 2 mM MnCl2, 0.5 mM EGTA) containing a mixture of phosphatidylinositol and phosphatidylserine at 100 µg/ml each and 0.2 mM adenosine. The kinase reaction was performed in the presence of 20 µCi [32P]adenosine 5'-triphosphate (ATP; ICN Biochemicals) and 0.02 mM ATP for 30 min at room temperature. The reaction was terminated with the addition of 100 µl 1 M HCl. The supernatants were transferred to fresh tubes, and phospholipids were extracted with 200 µl 1:1 chloroform:methanol. The upper organic phase was washed with 100 Fl 1:1 methanol:1 M HCl, dried down, resuspended in 20 Fl 95:5 chloroform:methanol, spotted on a silica 60 thin-layer chromatography plate (Merck, Fisher Scientific, Ontario, Canada), and resolved in chloroform:methanol:28% ammonium hydroxide:water (86:76:10:14). The phosphorylated products were visualized by autoradiography and quantified using the Bandleader program of Ma’ayan Aharoni.
|
TOPABSTRACTINTRODUCTIONMATERIALS AND METHODSRESULTSDISCUSSIONREFERENCES |
|---|
Fluorescein-activated cell sorter (FACS) analysis of cell cultures stained with PI can be used to determine whether the cytokines increase cell numbers by increasing the number of cells entering the S-phase of the cell cycle or/and by reducing the number of dying TF-1 cells. Previously starved TF-1 cells were cultured in serum-free medium or medium supplemented with 1 nM of the chimera, the individual factors, or a mixture of the two factors for a period of 12, 24, and 36 h. Cells were subsequently stained with PI and analyzed by FACS. Figure 1 shows the distribution of TF-1 cells as a function of their staining intensity, an indication of their DNA content. Distribution markers were assigned to the regions of the histogram that are normally associated with each specific cell-cycle state. Cells that are dead or dying have lower DNA content and therefore show less staining capacity and are grouped under the first marker at the left end of the histogram (A or M1), whereas cells between the G1 and G2 peaks of the cell cycle represent those cells in the DNA synthetic phase (S or M3). Cultures containing BOMIGF–IL-3 had the lowest number of cells with subdiploid DNA content (A). The DNA distribution of cells cultured with BOMIGF alone was very similar to that observed in control cell cultures, whereas incubation in the presence of IL-3 or the mixture of IL-3 and BOMIGF had a higher amount of subdiploid DNA-containing cells than that observed with the chimera. The results of FACS experiments similar to those shown in Figure 1 were expressed as percent distribution, and the averages of the different experimental groups are shown in Figure 2 . Figure 2 , upper, confirms the observations made in Figure 1 concerning the number of cells with subdiploid DNA. The chimera had a significantly lower amount of apoptotic cells than the equimolar mixture of BOMIGF and IL-3 after a 24-h incubation (P<0.04). Although when added alone, BOMIGF protects TF-1 cells from cell death marginally, this factor appears to cooperatively decrease TF-1 cell death when applied simultaneously with IL-3. The BOMIGF–IL-3 chimeras significantly increased the proportion of cells at the S-phase of the cell cycle as compared with the equimolar mixture of both cytokines after a 24- and 36-h incubation (P<0.05; Fig. 2 , lower). It can be concluded that the improved survival of TF-1 cells mediated by the chimera can be explained by a combination of decreased cell death and increased proportion of cells at the DNA synthetic phase of the cell cycle.
 View larger version (54K): [in a new window] |
Figure 1. Cell-cycle analysis of TF-1 cells. Cells were starved overnight as indicated. Cells (106) were cultured for 12, 24, and 36 h in 2 ml medium containing 1 nM of the indicated cytokines. The cells were fixed, stained with PI, and analyzed by flow cytometry as indicated in Materials and Methods. The profile of a 24-h incubation is shown. A, Cells with low DNA content (apoptotic); S, cells in the synthetic phase of the cell cycle, situated between the G1 and G2 peaks. This figure is representative of three experiments.
|
 View larger version (44K): [in a new window] |
Figure 2. Effect of different cytokines on the proportion of cells in the S-phase of the cell cycle or undergoing apoptosis. The percentage of TF-1 cells present in the M1 (Apoptotic) fraction (A fraction of Fig. 1
) and the M3 (S-phase of the cell cycle; S of Fig. 1
) were quantitated and expressed as mean ± SEM (n=3). BOMIGF, 1 nM; IL-3, 1 nM; BOMIGF–IL-3 fusion protein (Mixture), 1 nM BOMIGF + 1 nM IL-3; Chimera, 1 nM. ANOVA and Student-Newman-Keuls test indicated that the chimera significantly increased the proportion of cells in the S-phase after 24 h and 36 h incubation as compared with the equimolar mixture of BOMIGF and IL-3 (P<0.05). The proportion of apoptotic cells was significantly lower in cultures treated with the chimera as compared with the mixture of BOMIGF and IL-3 after a 24-h incubation (P<0.04).
|
 View larger version (45K): [in a new window] |
Figure 3. Effect of the cytokine chimera on the expression of BclxL. TF-1 cells, previously starved as indicated in Figure 1
, were washed with serum-free medium and transferred in aliquots of 106 cells to 24-well plates with 2 ml serum-free medium alone (lanes 1), supplemented with 1 nM BOMIGF (lanes 2), 1 nM IL-3 (lanes 3), 1 nM BOMIGF–IL-3 chimera (lanes 4), or a mixture of 1 nM BOMIGF and 1 nM IL-3 (lanes 5). After 12, 24, and 36 h incubation, the cells were washed and lysed, and 50 µg protein of the cell lysates was separated by SDS-PAGE. Western blots of the separated proteins were tested for the presence of BclxL or the control protein actin using specific antibodies as indicated in Materials and Methods. Upper, Typical pattern of BclxL expression and the corresponding Western of Actin as control. Lower, Amounts of BclxL expressed as % of control cell cultures. The values are mean ± SEM of seven to eight experiments. The amounts of BclxL present in lysates of cells treated with the chimera were significantly higher than those measured in cells treated with the equimolar mixture of BOMIGF and IL-3 after a 36-h incubation (P<0.05, nonparametric Wilcoxon-test).
|
 View larger version (27K): [in a new window] |
Figure 4. Effect of cytokines on the phosphorylation (PY) of the IL-3 receptor (IL-3R), Jak-2, and Stat5. TF-1 cells were starved as indicated in the previous figures and incubated as indicated in Materials and Methods in control, serum-free medium (lanes 1) or in medium containing BOMIGF (lanes 2), IL-3 (lanes 3), BOMIGF–IL-3 chimera (lanes 4), or an equimolar mixture of BOMIGF and IL-3 (lanes 5). The cells were washed and lysed. Cell lysates were separated on PAGE, transferred to nitrocellulose membranes, and tested as indicated by Materials and Methods for the presence of tyrosine-phosphorylated IL-3 receptor (A), Jak-2 phosphorylation (B), or Stat5 phopshorylation (C). Blots tested with anti-IL-3R, anti-Jak-2, and anti-Stat5 show the homogeneous loading of the gels. (A and B) The incubation time was 10 min.
|
 View larger version (19K): [in a new window] |
Figure 5. Time course of Stat5 phosphorylation in the presence of the cytokine chimera. TF-1 cells were cultured as indicated in Figure 7C
for 5, 40, and 80 min, and the degree of Stat5 phosphorylation was measured as indicated in Materials and Methods in cells incubated in the presence of the chimera ( ) or the corresponding equimolar mixture of BOMIGF and IL-3 ( ). Mean ± SEM (n=4, 5 and 80 min; n=5, 40 min). Although there was a wide variation of basal phophorylation from one experiment to the other, the paired t-test indicated that the stimulation of Stat5 phosphorylation after 5 and 40 min was significantly higher than that observed in the presence of BOMIGF and IL-3 added simultaneously (P<0.04 and P<0.03, respectively).
|
 View larger version (25K): [in a new window] |
Figure 6. PI-3K assay of TF-1 cells. TF-1 cells (107) were starved as indicated in Figure 1
and then treated for 15, 45, and 90 min with 25 nM BOMIGF–IL-3 chimera or an equimolar mixture of 25 nM BOMIGF and 25 nM IL-3 (A). The cells were washed, lysed, and immunoprecipitated with an antiphosphotyrosine antibody. The PI-3K activity was measured by adding phosphatidylinositol with [32P]ATP and separating the products by thin-layer chromatography as indicated in Materials and Methods. (B) A typical profile of the separation after a 90-min incubation. The top spot of the chromatogram corresponds to the marker phosphatidylinositol-3 phosphate (PIP3). The total amount of the products was scanned and quantitated. The amounts of PIP3 were significantly higher in lysates of cells cultured for 90 min with the chimera than in cells treated with the equimolar mixture of BOMIGF and IL-3 (n=7, Wilcoxon test, P<0.04). (C) Effect of increasing concentrations of LY294002 on TF-1 cell numbers in the presence of 10 nM chimera () or 10 nM each BOMIGF and IL-3 (). The cells were counted 1 day after addition of the inhibitor. The results are expressed as percent of cells remaining in culture as compared with the total number of cells in the absence of inhibitor (100%). Mean ± SEM (n=6). The higher sensitivity of TF-1 cells toward the PI-3K inhibitor in the presence of the chimera as compared with cultures in the presence of the equimolar mixture was significant at the concentrations of 2 and 20 µM (P<0.001, Student-Newman-Keuls multiple comparisons test).
|
 View larger version (17K): [in a new window] |
Figure 7. Cell survival in the presence of a PI-3K inhibitor. Effect of increasing concentrations of LY294002 on cell numbers in the presence of 10 nM chimera () or 10 nM each BOMIGF and IL-3 (). (A) Human umbilical endothelial cells were starved for 12 h and then incubated in the presence of the cytokines for 48 h. Alamar blue was added, and the cells remaining in the monolayers were measured spectrophotometrically [11
]. The results are expressed as percent of cells remaining in culture as compared with the total number of cells in the absence of inhibitor (100%). Mean ± SEM (n=5). The higher sensitivity of endothelial cells toward the PI-3K inhibitor in the presence of the chimera as compared with cultures in the presence of the equimolar mixture was significant at the concentrations of 2 and 20 µM (P<0.05 and P<0.001, respectively, Dunn’s multiple comparisons test). (B) Human HL-60 cells were starved for 24 h and then incubated in the presence of the cytokines for 24 h. Alamar blue (10%) was added, and the cells numbers remaining in the cultures measured spectrophotometrically [17
]. The results are expressed as percent of cells as compared with the total number of cells in the absence of inhibitor (100%). Mean ± SEM of triplicate determinations.
|
|
TOPABSTRACTINTRODUCTIONMATERIALS AND METHODSRESULTSDISCUSSIONREFERENCES |
|---|
As it is the case with IGFs, IL-3-mediated receptor induction of cell survival is associated with the activation of multiple signaling pathways. IL-3-dependent stimulation of hematopoietic cells has been shown to induce the activation of the PI-3K-Akt pathway [28 ]. The activation of this pathway is likely to mediate suppression of apoptosis by leading to the phosphorylation of Bcl2/BclxL-associated death promoter (BAD) and its sequestration by 14.3.3 proteins [29 , 30 ]. Alternatively, studies performed with constitutively active mutant components from the Ras signal-transduction pathway [31 ] and the Jak-2–Stat5 pathway [32 , 33 ] have also demonstrated the importance of both of these pathways in providing protection against apoptosis resulting from IL-3 withdrawal in IL-3-dependent cell lines. Although the PI-3K/Akt pathway is believed to mediate cell survival by inhibiting the activity of a proapoptotic protein component (i.e., BAD), the protective effects of the Ras/MAPK and Jak–Stat pathways may be related to their capacity to affect the production of the antiapoptotic protein BclxL [8 ].
BclxL is a member of the Bcl2 family of proteins that are known to inhibit (e.g., Bcl2, BclxL) or promote (e.g., Bax, Bak, BclxS) programmed cell death. The increased expression of the antiapoptotic protein BclxL has been shown to benefit cell survival in a variety of cells of hematopoietic origin [34 35 36 ]. Stat5 binds to the BclxL promoter, and Stat5 knockout mice have severe anemia, indicating the importance of the Jak-2–Stat5 pathway in the regulation of apoptosis of another critical hematopoietic cytokine, erythropoietin [37 ].
The hematopoietic cell line TF-1 can be kept in culture in the presence of IL-3. IGF-I, on the contrary, is not a very active growth factor for these cells. There is only a very minor increase in cell proliferation at doses higher than 100 nM. However, it can have a potent, synergistic effect with other cytokines [15 ]. We also observed a synergism between IL-3 and the IGF-II analog BOMIGF on cell proliferation in TF-1 cells, and this stimulation was particularly striking in the presence of the chimeric protein BOMIGF–IL-3 [3 ]. The results of the present study clearly indicate that the chemical linkage of an IGF analog with IL-3 results in a cytokine that is more effective than the equimolar mixture of the single components, as evidenced by the effects of the chimera on several cellular events associated with survival, such as lower proportion of apoptotic cells (Figs. 1 and 2) and increased amounts of the antiapoptotic protein BclxL (Fig. 3) . The lower proportion of apoptotic cells was associated with an increased number of cells in the S-phase (Fig. 2) , which corresponds with the higher thymidine incorporation observed with the chimera in our previous studies [3 ]. It is noteworthy to mention that the IGF-II analog BOMIGF alone had only a marginal effect on many of these processes (Figs. 1 2 3) . It is likely that the number of IGF-I receptors as compared with the number of IL-3 receptors is too low for IGFs, given individually, to make a notable difference in protecting TF-1 cells from apoptosis. Wang et al. [15 ] also found that although IGF-I evoked DNA synthesis in the TF-1 cell line and to a greater extent, in the murine IL-3-dependent hematopoietic cell line follicular dendritic cell (FDC)-P1, it was impossible to maintain the cells with IGF-I alone. To maintain FDC-P1 cells with IGF-I in the absence of IL-3, McCubrey et al. [38 ] had to intruduce the human IGF-I receptor with a retroviral construct, which effectively increased the number of IGF-I receptors, 100–400 times more than in uninfected cells. Therefore, the discrepancy between the relative numbers of IL-3 and IGF-I receptors in TF-1 cells may explain our failure to detect any difference at the level of phosphorylation of the ßc subunit of the IL-3 receptor or of the closely associated Jak-2 kinase after incubation with IL-3, the chimera, or the equimolar mixture of BOMIGF and IL-3 (Fig. 4A and 4B) . It is known that the IL-3–G-CSF chimera Leridistim can bind simultaneously with the IL-3 and G-CSF receptors in AML-193.1.3 cells, which proliferate in response to activation of either receptor. Intriguingly, despite this double-receptor interaction, the levels of phosphorylation of the ßc subunit of the IL-3 receptor and Jak-2 kinase were equivalent with those observed with the single components [24 ]. Therefore, it seems that a potentiation of phosphorylation by simultaneous stimulation of two receptors by bifunctional chimeras may be technically undetectable at the receptor level, even in the presence of an equivalent number of receptors for both ligands. Regardless of this limitation, an infrequent, simultaneous ligand contact may still be sufficient to trigger measurable differences in some downstream signal-transduction pathways. The early stimulation of Stat5 phosphorylation corroborates this assumption (Figs. 4C and 5) . Stat5 phosphorylation is a reasonable target for a possible interaction of IGFs and IL-3, as it has been previously reported that IGF-I enhances the erythropoietin-induced Stat5 phosphorylation in the hematopoietic cell line F-36P [16 ]. Moreover, Stat5 binds to the BclxL promoter [37 ].
As indicated in the introduction, the stimulation of the PI-3K is a critical step for the action of IGFs and IL-3. Therefore, it was reasonable to assume that it may play a critical role in the synergism observed between BOMIGF and IL-3 and that it may likewise participate in mediating the superior performance of the BOMIGF–IL-3 chimera. The results shown in Figure 6A and 6B , indicate that the BOMIGF–IL-3 chimera significantly increased the activity of this enzyme. Obviously, the PI-3K activation plays a key role in the synergism observed when IGF-II and IL-3 are linked together as a bifunctional cytokine. The addition of the PI-3K inhibitor LY294002 to the cell cultures specifically eliminated the selective advantage of the chimeric molecule on cell survival, not only in the transformed cell line TF-1 (Fig. 6C) but also in normal human umbilical endothelial cells (Fig. 7A) . We have also found that this inhibitor specifically eliminated the selective advantage of the chimera in the stimulation of TF-1 cell migration, whereas the chimera and the equimolar mixture were equally potent in the presence of Y-27632, an inhibitor of the Rho kinase, which is another important migration-related signal pathway [39 ]. It is interesting to point out that even in hematopoietic cell lines with large number of IGF-I receptors, such as the myeloma cell line RPMI 8226, a maximal activation of the PI-3K takes place at concentrations of IGF-I, which are 10 times lower than those required to observe a maximal phosphorylation of the IGF-I receptor [40 ]. Therefore, the fact that cells have low IGF-I receptor numbers may have contributed to better observe the synergism of the fusion protein of BOMIGF and IL-3. We are trying to study in more detail the contribution of the IGF-I receptor in hematopoietic cell lines that are more suitable than TF-1 cells to IGF-I receptor transfection. However, nontransfected cells are likely to reflect more closely the normal physiological response in vivo, whereas overexpression of IGF-I receptors may lead to transformation [38 ].
In all experiments described in this paper and in our previous publications, the factor-dependent TF-1 cell line has to be depleted from the cytokines and growth factors used for its routine maintenance by washing in a serum-free medium followed by a period of starvation. This period seemed to be appropriate for most investigations on cell proliferation [3 ], migration [39 ], or apoptosis. However, we noticed that it is sometimes difficult to completely eliminate the basal phosphorylated state of Stat5 (Fig. 4C) and that the immunoprecipitated PI-3K activity remains high, whereas it is known to return to basal levels very shortly after IL-3 stimulation in other hematopoietic cell lines [41 ]. Preliminary experiments indicate that basal Stat5 phosphorylation levels may depend on the cell density used during the period of starvation. Furthemore, the PI-3K activity was measured in vitro after immunoprecipitation using antiphosphotyrosine antibodies. This method of detection does not always correlate with the direct measurement of the PI-3K products in the cells [41 ]. The experiments using the PI-3K inhibitor clearly indicate the strong correlation between PI-3K and the preferential action of the chimera on apoptosis and cell migration, but it remains to be seen whether these effects are mediated by the products of PI-3K or are a result of the association of the phosphorylated enzyme with other proteins.
Bifunctional cytokines such as the IGF-II–IL-3 chimera described here could be used as tools to study cross-talk between different cytokines or signal transduction pathways and may eventually result in more effective adjuvants for ex vivo expansion of hematopoietic cells or as promising therapeutic agents. Another example of a synergistic effect of IGF-containing protein chimeras on cell proliferation has been obtained with the linkage of BOMIGF and the 
1-proteinase inhibitor [17
]. We are working on a new vector for the large-scale production of IGF-I–cytokine chimeras using insect cells. An additional advantage of cytokine chimeras containing an IGF moiety is the possible use of its binding with IGF-binding proteins, which play an important role in delaying the clearance and modifying the action of IGFs. This possibility remains to be explored.
Received August 14, 2002; revised November 4, 2002; accepted November 13, 2002.
|
TOPABSTRACTINTRODUCTIONMATERIALS AND METHODSRESULTSDISCUSSIONREFERENCES |
|---|
| ||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||
