Enhancement of growth and survival and alterations in Bcl-family proteins in beta-thalassemic erythroid progenitors by novel short-chain fatty acid derivatives

Abstract

Accelerated apoptosis of erythroid progenitors is a characteristic of beta-thalassemia which presents a significant barrier to definitive therapeutic approaches utilizing induction of endogenous fetal globin gene expression. gamma-globin gene expression may not be inducible in, or may not be able to rescue, erythroid cells in which programmed cell death is initiated early in erythroblast development. In this report, short-chain fatty acid derivatives (SCFADs) which induce fetal globin gene expression were tested for their ability to promote proliferation and survival of erythroid progenitors cultured from beta-thalassemic subjects, and of cytokine-dependent erythroid cell lines. Certain SCFADs promoted thalassemic Bfu-e growth and cytokine-independent growth and survival of erythroid cell lines. A 40-80% increase in erythroid Bfu-e colony number was observed in cultures established with any of five mitogenic SCFADs, compared to control or butyrate-treated cultures from the same subjects. Immunoblot analysis demonstrated that these same SCFADs also regulated the expression of specific protein inhibitors of apoptosis. Anti-apoptotic ratios of the proteins Bcl-xL/Bcl-xS in thalassemic Bfu-e were increased by 30-120% with exposure to the SCFDs, compared to the ratios in the same cells cultured under control conditions. Similar anti-apoptotic increases in Mcl-1L/Mcl-1S ratios were induced by the SCFADs. These findings suggest that select fetal globin-inducing SCFADs which enhance proliferation of beta-thalassemia progenitors may enhance survival of these progenitors by altering levels of Bcl-family protein members. This combination of effects should enhance erythroid cell survival in the beta-thalassemia syndromes, allowing fetal globin gene expression to be induced more effectively than currently available, growth-suppressing, fetal globin-inducing agents, such as the butyrates or chemotherapeutic agents.

Fig. 1

(A) Proliferation curves of TF-1 cells treated with SCFADs. TF-1 cell cultures were depleted of GM-CSF as described in Material and methods, and cultured with or without SCFADs, in the absence of any or severely depleted cytokines, for 36 h and then enumerated. AB (Arginine Butyrate), ST7 (sodium α-methylhydrocinnamate), ST20 (sodium 2,2-dimethylbutyrate), ST79 [sodium 3,4-(Methylenedioxy)cinnamate], RB3 [2-(quinazolin-4-ylamino)butanoic acid], and RB4 ({4-[(trifluoromethyl)sulfanyl]aniline} acetic acid). Results from three independent experiments are shown, with bars indicating SE. ANOVA results for SCFAD-treated cultures compared to controls were as follows: for AB, ST20, ST79, and RB4, P < 0.05; for ST7 and RB3, P < 0.004. (B) Survival curves of TF-1 cells treated with SCFADs. TF-1 cell cultures were depleted of GM-CSF, and washed to remove any remaining GM-CSF, as described in Material and methods, and cultured with or without SCFADs, in the absence of any cytokines, for 36 h and then enumerated. Results from three independent experiments are shown, with bars indicating SE. ANOVA results for SCFAD-treated cultures compared to controls were as follows: for AB P < 0.05; for RB4 and ST 20, P < 0.015; for ST7, ST79 and RB3, P < 0.008. (C) Effect of selected SCFADs on Bfu-e growth in cultures derived from β-thalassemia patients. CD34+ enriched cells were collected and cultured as described in Materials and methods, in a cytokine mixture designed for optimal Bfu-e growth, with or without SCFADs. Bfu-e were enumerated after 12 days. The number of Bfu-e colonies is expressed as the mean relative to cells from the same patient cultured without SCFADs (control [C]), for nine β-thalassemia patient cultures (±SE). Comparison between SCFAD-treated and control values was performed by the Wilcoxon signed-rank test. *P < 0.05; **P < 0.01; ***P < 0.0001.

Fig. 2

(A) Bcl-2 family protein expression in 32D cells cultured with SCFADs. 32D cells were cultured under low rIL-3 conditions (0.5 μg/ ml) as described in Materials and methods, in the presence or absence of the indicated SCFADs (B is Arginine Butyrate; 20 is ST20; 7 is ST7; 79 is ST79), or without SCFADs (C) for 36 h, then lysed and analyzed by immunoblotting for Bcl-x or Bcl-2. The Bcl-2 immunoblot was stripped and reprobed for β-actin as a loading control. The sample in lane “IL-3” was derived from a culture treated with high concentrations of rIL-3 (25 U/ml) for 1 h before harvesting. (B) Bcl-xL/xS ratios in 32D cells cultured with SCFADs. Ratios are derived from three separate experiments, carried out as described in the legend to A, and quantitated by densitometric scanning.

Fig. 3

A. Effect of selected SCFADs on Bcl-x_L/Bcl-x_S protein ratios in Bfu-e cultures from β-thalassemia patients. CD34+ enriched cells were collected and cultured as described in Materials and methods, in a cytokine mixture designed for optimal Bfu-e growth, with or without the SCFADs. Bfu-e were harvested on day 12, cells were incubated overnight in the presence or absence of the SCFADs, and the cell lysates were analyzed for levels of Bcl-x proteins. Bars represent the mean of Bcl-x_L/Bcl-x_S ratios (±SE) detected by immunoblotting in Bfu-e cultures from four β-thalassemia patients. The Bcl-x_L/Bcl-x_S ratios of the cultures not treated with SCFADs (control [C]) were normalized to 1, and the ratios from cultures treated with specific SCFADs are expressed as fold change of the ratio in the control culture conditions. Comparison between compounds and control values was performed by means of the Wilcoxon signed-rank test. *P < 0.05; **P < 0.01; ***P < 0.001. The insert is a typical immunoblot of Bfu-e-derived Bcl-X protein levels. (B) Effect of selected SCFADs on Bcl-x_L/ Bcl-x_S protein ratios in TF-1 cells. TF-1 cell cultures were depleted of GM-CSF as described in Materials and methods, and cultured in the absence of any cytokines, with or without (control [C]) the indicated SCFADs, for 36 h and then harvested and lysed for Bcl-x protein analysis by immunoblotting. Protein ratios were determined by densitometric scanning of the film. (C) Effect of selected SCFADs on Mcl-1_L/Mcl-1_S protein ratios in TF-1 cells. TF-1 cell cultures were depleted of GM-CSF, cultured in the absence of any cytokines, with or without (control [C]) the indicated SCFADs, as described in the legend to B, and Mcl-1 protein levels were assayed by immunoblotting and quantitated. The lane marked EPO contains lysates from TF-1 cells cultured in the presence of erythropoietin (3 U/ml) for 36 h. The insert depicts an immunoblot of Mcl-1 proteins from a typical experiment, with the position of the Mcl-1_L and Mcl-1_S species indicated. D. Effect of selected SCFADs on Mcl-1 protein levels in Bfu-e cultures from thalassemia subjects. CD34+ enriched cells were collected and treated as described in the legend to B described in Materials and methods, and lysates were immunoblotted for Mcl-1 protein. The level in the untreated control lane (C) was assigned an arbitrary value of 1.0, and the relative levels of Mcl-1 protein in the other lanes are indicated, after normalization to β-actin as a control for loading.

Fig. 4

(A) Effect of selected SCFADs on percentage of F-cells determined in cultures from β-thalassemia patients. CD34+ enriched cells were collected and cultured as described in Materials and methods, in a cytokine mixture designed for optimal Bfu-e growth, with or without the SCFADs. Bfu-e were harvested on day 12, cells were incubated overnight in the presence or absence of the test compounds, and the cells were analyzed for % F-cells. The percentage of F-cells (%F-cells) in cultures exposed to the test compounds is expressed relative to untreated (control) cultures (C) for seven β-thalassemia patient cultures (±SE). Comparison between compounds and control values was performed by the Wilcoxon signed-rank test. *P < 0.05; **P < 0.01. (B) Effect of selected SCFADs on γ-globin mRNA production in TF-1 cells. TF-1 cells were cultured in the absence of GM-CSF, with or without SCFADs, as described in Materials and methods, for 36 h and then harvested for globin mRNA determination using an RNAse protection assay kit. A typical autoradiogram is shown here. 18s RNA serves as a loading control. (C) Effect of selected SCFADs on γ-globin mRNA production in TF-1 cells. γ- and β-globin transcript levels were determined in TF-1 cells treated as described in the legend to B and the ratios plotted. Bar labeled “GM-CSF” indicates RNA isolated from TF-1 cells cultured in recombinant human GM-CSF at 4 ng/ml.