Dexamethasone and lenalidomide have distinct functional effects on erythropoiesis

Abstract

Corticosteroids and lenalidomide decrease red blood cell transfusion dependence in patients with Diamond-Blackfan anemia (DBA) and myelodysplastic syndrome (MDS), respectively. We explored the effects of dexamethasone and lenalidomide, individually and in combination, on the differentiation of primary human bone marrow progenitor cells in vitro. Both agents promote erythropoiesis, increasing the absolute number of erythroid cells produced from normal CD34(+) cells and from CD34(+) cells with the types of ribosome dysfunction found in DBA and del(5q) MDS. However, the drugs had distinct effects on the production of erythroid progenitor colonies; dexamethasone selectively increased the number of burst-forming units-erythroid (BFU-E), whereas lenalidomide specifically increased colony-forming unit-erythroid (CFU-E). Use of the drugs in combination demonstrated that their effects are not redundant. In addition, dexamethasone and lenalidomide induced distinct gene-expression profiles. In coculture experiments, we examined the role of the microenvironment in response to both drugs and found that the presence of macrophages, the central cells in erythroblastic islands, accentuated the effects of both agents. Our findings indicate that dexamethasone and lenalidomide promote different stages of erythropoiesis and support the potential clinical utility of combination therapy for patients with bone marrow failure.

Figures

Figure 1

Dexamethasone increases BFU-E colony formation, whereas lenalidomide increases CFU-E colony formation, and the combination of drugs is additive. After 3 days of treatment with the compounds in liquid culture, cells were plated in methylcellulose without drugs. Data are presented as the number of colonies per 3200 cells plated. The effects of dexamethasone (Dex) are shown in panel A and the response to lenalidomide (Len) is shown in panel B. The effects of combinations of Dex and Len on CFU-E are shown in panel C and the effects on BFU-E in panel D. The expected number of colonies was calculated as follows: expected number of colonies = (Dex-alone colonies − control) + (Len-alone colonies − control) + control. The experiments were performed in triplicate and repeated with similar results. In panels A and B, a 2-tailed Student t test was used. **P ≤ .01; *P ≤ .05. In panels C and D, ANOVA was used and the statistical findings are discussed in the text.

Figure 2

Dexamethasone and lenalidomide increase the production of erythroid cells from CD34+ cells expressing RPS19 or RPS14 shRNAs without increasing apoptosis. In panels A and D, a Western blot shows the decreased level of protein with RPS19 knockdown (56.7% knockdown by qRT-PCR) and with RPS14 knockdown (61.9% knockdown by qRT-PCR). The absolute number of erythroid cells after 10 days in liquid culture was determined by multiplying the number of cells counted per well by the percentage of those cells expressing any erythroid markers (CD71, glycophorin A) as assayed by flow cytometry. Data are presented as relative numbers to eliminate differences caused by cell number and infection rate. The effects of dexamethasone (Dex), lenalidomide (Len), and control on cells infected with a RPS19 shRNA are shown in panel B. The effects of Dex, Len, and control on cells infected with a RPS14 shRNA are shown in panel E. No increase in annexin staining was noted in CD34+ cells expressing RPS19 or RPS14 shRNAs and treated with either Dex or Len compared with vehicle control, as shown in panels C and F. The experiments were performed in triplicate and the entire experiment was repeated with similar results with an independent shRNA (supplemental Figure 5). A 2-tailed Student t test was used. **P ≤ .01; *P ≤ .05.

Figure 3

The combination of dexamethasone and lenalidomide has an additive effect on increasing the production of erythroid cells from CD34+ cells expressing RPS19 or RPS14 shRNAs. The effects of combination treatment with dexamethasone (Dex) and lenalidomide (Len) and control on cells infected with a RPS19 shRNA are shown in panel A and with a RPS14 shRNA in panel B. The absolute number of erythroid cells after 10 days in liquid culture was determined by multiplying the number of cells counted per well by the percentage of cells expressing any erythroid markers (CD71, glycophorin A) as assayed by flow cytometry. Data are presented as relative number to eliminate differences caused by cell number and infection rate. The expected number of erythroid cells was calculated as follows: expected number of cells = (absolute number of erythroid cells with Dex treatment alone − control) + (absolute number of erythroid cells with Len treatment alone − control) + control. The experiments were performed in triplicate and the entire experiment was repeated with similar results with an independent shRNA (supplemental Figure 6). ANOVA was used and the statistical findings are discussed in the text.

Figure 4

Dexamethasone and lenalidomide induce distinct but overlapping gene-expression profiles. The heat map (panel A) illustrates the top 25 genes induced by dexamethasone (Dex), lenalidomide (Len), and the combination of both drugs. Genes induced by one drug and the combination are labeled as genes induced by the combination. The genes were arranged by hierarchical clustering. High gene expression is portrayed in red and low expression is blue. Expression of RPS19 and RPS14 in response to drug treatment is shown in panels B and C, respectively.

Figure 5

Stromal cells promote colony formation without altering the effects of dexamethasone and lenalidomide. After 3 days of treatment with the compounds in liquid culture, hematopoietic cells were separated from GFP+ OP9 cells by flow sorting, and the hematopoietic cells were plated in methylcellulose without drugs. Data are presented as the number of colonies per 3200 cells plated. The effects of dexamethasone (Dex) and lenalidomide (Len) on CFU-E colony formation are shown in panel A. The effects of Dex and Len on BFU-E colony formation are shown in panel B. The experiments were performed in triplicate and the entire experiment was repeated on HS5 stromal cells with similar results. ANOVA was used and the statistical findings are discussed in the text.

Figure 6

The effects of dexamethasone and lenalidomide on erythropoiesis are influenced by macrophage cocultures. Macrophages were confirmed by presence CD68 by flow cytometry (A) and by immunofluorescence (B), with the green stain representing CD68 and the blue stain representing DAPI. After 3 days of treatment with the compounds in liquid culture, the amount of TNF-α in the supernatant was quantified by a chemiluminescent immunoassay (C). The experiment was performed in triplicate and repeated with similar results. After 3 days of treatment with the compounds in liquid culture, cells were plated in methylcellulose without drugs. Data are presented as the number of colonies per 3200 cells plated. The effects of dexamethasone (Dex) and lenalidomide (Len) on CFU-E colony formation are shown in panel D. The effects of Dex and Len on BFU-E colony formation are shown in panel E. The experiments were performed in triplicate and repeated with similar results. In panel C, a 2-tailed Student t test was used. **P ≤ .01; *P ≤ .05. In panels D and E, ANOVA was used and the statistical findings are discussed in the text.