The tyrosine kinase inhibitor dasatinib induces a marked adipogenic differentiation of human multipotent mesenchymal stromal cells

Abstract

Background: The introduction of specific BCR-ABL inhibitors in chronic myelogenous leukemia therapy has entirely mutated the prognosis of this hematologic cancer from being a fatal disorder to becoming a chronic disease. Due to the probable long lasting treatment with tyrosine-kinase inhibitors (TKIs), the knowledge of their effects on normal cells is of pivotal importance.

Design and methods: We investigated the effects of dasatinib treatment on human bone marrow-derived mesenchymal stromal cells (MSCs).

Results: Our findings demonstrate, for the first time, that dasatinib induces MSCs adipocytic differentiation. Particularly, when the TKI is added to the medium inducing osteogenic differentiation, a high MSCs percentage acquires adipocytic morphology and overexpresses adipocytic specific genes, including PPARγ, CEBPα, LPL and SREBP1c. Dasatinib also inhibits the activity of alkaline phosphatase, an osteogenic marker, and remarkably reduces matrix mineralization. The increase of PPARγ is also confirmed at protein level. The component of osteogenic medium required for dasatinib-induced adipogenesis is dexamethasone. Intriguingly, the increase of adipocytic markers is also observed in MSCs treated with dasatinib alone. The TKI effect is phenotype-specific, since fibroblasts do not undergo adipocytic differentiation or PPARγ increase.

Conclusions: Our data demonstrate that dasatinib treatment affects bone marrow MSCs commitment and suggest that TKIs therapy might modify normal phenotypes with potential significant negative consequences.

Figure 1. Effect of dasatinib on MSCs phenotype.

Panel a) MSCs were seeded at a density of 10⁴ cells/cm². After one day, 25 nM dasatinib (Das) was added. After one week, the osteogenic medium replaced the growth medium. The osteogenic medium containing the TKI was changed every 3 days. Finally, after additional 3 weeks, the images of MSCs were captured as described in Materials and Methods. Panel b) The experiments were performed exactly as reported in panel a, except that the dasatinib was employed at 25 and 50 nM. The figure is representative of 3 independent experiments.

Figure 2. Characterization of dasatinib effect on MSCs osteogenesis.

Panel a) MSCs cells were treated for one week with 25 nM or 50 nM dasatinib (Das). Then, the activity of alkaline phosphatase was evaluated as described in Materials and Methods and expressed as percentage of control. Panel b) MSCs cells were treated with 25 nM or 50 nM dasatinib (Das) following the scheme reported in Figure 1. At the end of the treatment, the calcium content was evaluated by and o-cresophtalein complexone method, and expressed as percentage of control. Panel c) MSCs cells were treated with 25 nM or 50 nM dasatinib (Das) following the scheme reported in Figure 1. At the end of the treatment, the calcium content was evaluated by alizarin red staining and expressed as percentage of control. Each panel is representative of 5 independent experiments.

Figure 3. Characterization of dasatinib-dependent MSCs adipogenesis.

Panel a) MSCs were treated as in Figure 1. At the end of experiments the image of dasatinib-treated cells was captured (A) and, then, the cells were stained with the Sudan Red staining method (B). Panels b–d) MSCs were seeded at a density of 10⁴ cells/cm². After one day, 25 nM dasatinib was added plus β-glycerophosphate (Das/β-GP, Panel c) or plus 100 nM dexamethasone (Das/Dex, Panel d). Control cells were shown in Panel b. Finally, after additional 3 weeks, the images were captured. The panels are representative of 4 independent experiments.

Figure 4. Effect of dasatinib on PPARγ expression in MSCs.

Panel a) MSCs were seeded at a density of 10⁴ cells/cm². After one day, 25 nM dasatinib (Das) were added. After one week, 100 nM dexamethasone (Dex) was added in some samples as reported in the panel. After further 30 days, total RNA was prepared and the content of PPARγ transcript evaluated by semiquantitative RT-PCR. MW, molecular weight standards. Glyceraldheyde 3-phosphate dehydrogenase (GAPDH) was used as internal standard. Further details of RT-PCR are reported under Materials and Methods Section. Panel b) The experiment is similar to that in Panel 4a except that the content of PPARγ transcript was determined by quantitative RT-PCR. Further details on this analysis are reported under Materials and Methods Section. The results are expressed as percentage of control. The panels are representative of 5 independent experiments.

Figure 5. Effect of dasatinib or bosutinib with or without dexamethasone on the expression of PPARγ, CEBPα, LPL and SREBP1c in MSCs.

Panel a) MSCs were seeded at a density of 10⁴ cells/cm². After one day, the cells were treated with: 100 nM dexamethasone (Dex); 25 nM dasatinib/100nM dexamethasone (Das/Dex); 100 nM bosutinib/100nM dexamethasone (Bos/Dex); 1 µM dexamethasone (Dex); 25 nM dasatinib alone (Das). After 10 days, total RNA was prepared and the content of PPARγ, CEBPα, LPL and SREBP1c was evaluated by semiquantitative RT-PCR. MW, molecular weight standards. Glyceraldheyde 3-phosphate dehydrogenase (GAPDH) was used as internal standard. Furthermore, in the analysis of SREBP1c transcript, a K562 erythroid cell total RNA was employed, since it has been reported that these cells are rich of SREBP1c transcript. Additional details of RT-PCR are reported under Materials and Methods Section. Panel b) MSCs were seeded at a density of 10⁴ cells/cm². After one day, the cells were treated respectively with: 25 nM dasatinib/100nM dexamethasone (Das/Dex); 25 nM dasatinib alone (Das), and 100 nM bosutinib (Bos) alone. After 20 days, total RNA was prepared and the content of PPARγ, CEBPα and LPL was evaluated by semiquantitative RT-PCR. MW, molecular weight standards. Glyceraldheyde 3-phosphate dehydrogenase (GAPDH) was used as internal standard. In this experiment, we have increased the number of PCR cycles from 30 to 35 in order to identify the occurrence of low amount of transcript. Additional details of RT-PCR are reported under Materials and Methods Section. The panels are representative of 5 independent experiments.

Figure 6. Effect of TKIs with (or without) dexamethasone on the content of PPARγ and p27^Kip1 proteins in MSCs and primary human fibroblasts.

The experiment was performed essentially as reported in Figure 5. Panel a) In this case, MSCs total cell extracts were prepared and investigated by immunoblotting for the content of PPARγ and p27^Kip1 (p27) protein. Actin was used for confirming the equal loading. Two different film expositions (1 and 5 minutes) were performed to obtain a better evaluation of PPARγ protein content. Additional details are reported under Materials and Methods Section. Panel b) MSCs were seeded at a density of 10⁴ cells/cm². After one day, MSCs were treated with 25 nM dasatinib (Das) or cultured in adipogenic or osteogenic media. Subsequently, MSCs were cultured for 25 days and then collected. Finally, cell extracts were prepared and investigated by immunoblotting for the content of PPARγ and p27^Kip1 (p27) protein. Panel c) MSCs and human primary skin fibroblasts were seeded at a density of 10⁴ cells/cm². After 24 hours, 25 nM dasatinib (Das) or 100 nM bosutinib (Bos) were added to the cells, After 40 days, the cells were collected and nuclear and cytosolic fractions prepared by Ne-Per as reported in Materials and Methods. The purity of nuclear and cytosolic compartments was evaluated by immunoblotting with anti-HDAC1 (a nuclear protein) and anti-cofillin (a cytolic protein) antibodies. The results confirmed the purity of fractions (data not reported). Finally, the content of PPARγ and p27^Kip1 (p27) proteins was determined by immunoblotting. Panels d and e) MSCs were seeded at a density of of 10⁴ cells/cm². After one day, the medium was removed and substituted with osteogenic medium or osteogenic medium plus dasatinib. At the reported days, MSCs were recovered and employed to evaluate (Panel d) phosphoErk1/2 and Erk1/2 levels, or (Panel e) the level of phospho-tyrosine protein levels (pY). The panels are representative of 3 independent experiments.