Glucocorticoid-induced apoptosis and glucocorticoid resistance: molecular mechanisms and clinical relevance

The ability of glucocorticoids (GC) to efficiently kill lymphoid cells has led to their inclusion in essentially all chemotherapy protocols for lymphoid malignancies. This review summarizes recent findings related to the molecular basis of GC-induced apoptosis and GC resistance, and discusses their potential clinical implications. Accumulating evidence suggests that GC may induce cell death via different pathways resulting in apoptotic or necrotic morphologies, depending on the availability/responsiveness of the apoptotic machinery. The former might result from regulation of typical apoptosis genes such as members of the Bcl-2 family, the latter from detrimental GC effects on essential cellular functions possibly perpetuated by GC receptor (GR) autoinduction. Although other possibilities exist, GC resistance might frequently result from defective GR expression, perhaps the most efficient means to target multiple antileukemic GC effects. Numerous novel drug combinations are currently being tested to prevent resistance and improve GC efficacy in the therapy of lymphoid malignancies.     

A conceptual view on glucocorticoid-lnduced apoptosis, cell cycle arrest and glucocorticoid resistance in lymphoblastic leukemia

Glucocorticoids (GC) control cell cycle progression and induce apoptosis in cells of the lymphoid lineage. Physiologically, these phenomena have been implicated in regulating immune functions and repertoire generation. Clinically, they form the basis of inclusion of GC in essentially all chemotherapy protocols for lymphoid malignancies. In spite of their significance, the molecular mechanisms underlying the anti-leukemic GC effects and the clinically important phenomenon of GC resistance are still unknown. This review summarizes recent findings related to GC-induced apoptosis, cell cycle arrest, and GC resistance with particular emphasis on acute lymphoblastic leukemia (ALL). We hypothesize that under conditions of physiological Bcl-2 expression, GC might induce classical programmed cell death by directly perturbing the Bcl-2 rheostat. In the presence of anti-apoptotic Bcl-2 proteins, cell death might result from accumulating catabolic and/or other detrimental GC effects driven by, and critically dependent on, GC receptor (GR) autoinduction. Although still controversial, there is increasing evidence for release of apoptogenic factors through pores in the outer mitochondrial membrane, rather than deltapsiloss-dependent membrane rupture, with maintenance of mitochondrial function at least in the early phase of the death response. GC-induced cell cycle arrest in ALL cells appears to be independent of apoptosis induction and vice versa, and critically depends on repression of both cyclin-D3 and c-myc followed by increased expression of the cyclin-dependent kinase inhibitor, p27Kip1. Since development of GC-resistant clones requires both cell cycle progression and survival, GC resistance might frequently result from structural or regulatory defects in GR expression, perhaps the most efficient means to target both pathways concurrently.     

Cyclin D3 and c-MYC control glucocorticoid-induced cell cycle arrest but not apoptosis in lymphoblastic leukemia cells

Glucocorticoids (GC) induce cell cycle arrest and apoptosis in lymphoblastic leukemia cells. To investigate cell cycle effects of GC in the absence of obscuring apoptotic events, we used human CCRF-CEM leukemia cells protected from cell death by transgenic bcl-2. GC treatment arrested these cells in the G1 phase of the cell cycle due to repression of cyclin D3 and c-myc. Cyclin E and Cdk2 protein levels remained high, but the kinase complex was inactive due to increased levels of bound p27(Kip1). Conditional expression of cyclin D3 and/or c-myc was sufficient to prevent GC-induced G1 arrest and p27(Kip1) accumulation but, importantly, did not interfere with the induction of apoptosis. The combined data suggest that repression of both, c-myc and cyclin D3, is necessary to arrest human leukemia cells in the G1 phase of the cell division cycle, but that neither one is required for GC-induced apoptosis.     

Expression of the glucocorticoid receptor from the 1A promoter correlates with T lymphocyte sensitivity to glucocorticoid-induced cell death

Glucocorticoid (GC) hormones cause pronounced T cell apoptosis, particularly in immature thymic T cells. This is possibly due to tissue-specific regulation of the glucocorticoid receptor (GR) gene. In mice the GR gene is transcribed from five separate promoters designated: 1A, 1B, 1C, 1D, and 1E. Nearly all cells express GR from promoters 1B-1E, but the activity of the 1A promoter has only been reported in the whole thymus or lymphocyte cell lines. To directly assess the role of GR promoter use in sensitivity to glucocorticoid-induced cell death, we have compared the activity of the GR 1A promoter with GC sensitivity in different mouse lymphocyte populations. We report that GR 1A promoter activity is restricted to thymocyte and peripheral lymphocyte populations and the cortex of the brain. The relative level of expression of the 1A promoter to the 1B-1E promoters within a lymphocyte population was found to directly correlate with susceptibility to GC-induced cell death, with the extremely GC-sensitive CD4+CD8+ thymocytes having the highest levels of GR 1A promoter activity, and the relatively GC-resistant alphabetaTCR+CD24(int/low) thymocytes and peripheral T cells having the lowest levels. DNA sequencing of the mouse GR 1A promoter revealed a putative glucocorticoid-response element. Furthermore, GR 1A promoter use and GR protein levels were increased by GC treatment in thymocytes, but not in splenocytes. These data suggest that tissue-specific differences in GR promoter use determine T cell sensitivity to glucocorticoid-induced cell death.     

Fas-mediated apoptosis of proliferating, transiently growth-arrested, and senescent normal human fibroblasts

Previous studies suggest that apoptotic signaling may require proteins that are critical to cellular proliferation and cell cycle regulation. To further examine this question, proliferating, transiently growth-arrested, and senescent normal human fibroblasts were induced to undergo apoptosis in response to two distinct mediators of apoptosis-Fas (APO-1/CD95) death receptor and staurosporine. Ligation of the Fas receptor in the presence of cycloheximide or actinomycin D resulted in apoptosis of proliferating cells, cells transiently growth arrested by gamma-irradiation or serum starvation (i.e., G(0) arrest), and permanently growth-arrested senescent fibroblasts. Proliferating and G(0)-arrested cells were also susceptible to staurosporine-mediated apoptosis. Surprisingly, gamma-irradiated cells did not undergo staurosporine-mediated apoptosis, and remained viable for a prolonged time. Fas-mediated apoptosis of senescent fibroblasts was evidenced by chromosome condensation and by activation of caspase-8 and -3, proteases crucial for the execution of the Fas apoptosis pathway. In addition, ligation of the Fas receptor in G(0)-arrested cells did not result in the activation of p34(cdc2) kinase, arguing that activation of this kinase is not essential in this apoptotic process. From these studies we conclude that proliferating, transiently growth-arrested, and senescent normal human fibroblasts are susceptible to apoptotic signals and that apoptosis is not necessarily dependent upon cell cycle or proliferative state of the cell.     

P element excision and repair by non-homologous end joining occurs in both G1 and G2 of the cell cycle

P element excision generates a DNA double-strand break at the transposon donor site. Genetic studies have demonstrated a strong bias toward repair of P element-induced DNA breaks by homologous recombination with the sister chromatid, suggesting that P element excision occurs after DNA replication, in G2 of the cell cycle. We developed methods to arrest Drosophila tissue culture cells and assay P element excision in either G1- or G2-arrested cells. Dacapo or tribbles transgene expression arrests cells in either G2 or G2, respectively. RNA-mediated gene interference (RNAi) directed against cyclin E or cyclin A arrests cells in G1 or G2, respectively. P element excision occurs efficiently in both G1- and G2-arrested cells, suggesting that cell cycle regulation of P element transposase does not occur in our somatic cell system. DNA double-strand break repair occurs by two predominant mechanisms: homologous recombination (HR) and non-homologous end joining (NHEJ). HR is thought to be restricted to the post-replicative, G2, phase of the cell cycle, while NHEJ may occur throughout the cell cycle. Our results indicate that NHEJ repair of an extrachromasomal plasmid substrate occurs at least as efficiently in G2-arrested cells as in asynchronous cells or in G1-arrested cells.     

Glucocorticoids engage different signal transduction pathways to induce apoptosis in thymocytes and mature T cells

Glucocorticoids (GC) induce apoptosis in a variety of cells, but their exact mode of action is controversial. Although initiation relies on the GC receptor (GR) and de novo gene expression, the effector phase differs among cell types. Proteasomal degradation as well as caspase-3, - 8, and -9 activity are essential for GC-induced apoptosis in murine thymocytes, but the same enzymes are dispensable in splenic T cells. Live imaging by confocal microscopy revealed that lysosomal cathepsin B, an unrecognized component of this pathway to date, becomes rapidly activated in thymocytes after GC exposure. This is followed by leakage of cathepsin B into the cytosol, nuclear condensation, and processing of caspase-8 and -3. According to our model, activation of caspase-3 by caspase-9 in thymocytes occurs both directly as well as indirectly via a lysosomal amplification loop. Interestingly, acute T lymphoblastic leukemia cells depend on caspase activity to undergo GC-induced cell death similar to thymocytes. Collectively, the apoptotic program induced by GCs comprises cell type-specific as well as common features.     

Dual regulation of glucocorticoid-induced leucine zipper (GILZ) by the glucocorticoid receptor and the PI3-kinase/AKT pathways in multiple myeloma

Glucocorticoids (GCs) are effective therapeutics commonly used in multiple myeloma (MM) treatment. Clarifying the pathway of GC-induced apoptosis is crucial to understanding the process of drug resistance and to the development of new targets for MM treatment. We have previously published results of a micro-array identifying glucocorticoid-induced leucine zipper (GILZ) as GC-regulated gene in MM.1S cells. Consistent with those results, GCs increased GILZ in MM cell lines and patient samples. Reducing the levels of GILZ with siRNA decreased GC-induced cell death suggesting GILZ may mediate GC-killing. We conducted a screen to identify other pathways that affect GILZ regulation and report that inhibitors of PI3-kinase/AKT enhanced GILZ expression in MM cell lines and clinical samples. The combination of dexamethasone (Dex) and LY294002, wortmannin, triciribine, or AKT inhibitor VIII dramatically up regulated GILZ levels and enhanced apoptosis. Addition of interleukin-6 (IL-6) or insulin-like growth factor (IGF1), both which activate the PI3-kinase/AKT pathway and inhibit GC killing, blocked up regulation of GILZ by GC and PI3-kinase/AKT inhibitors. In summary, these results identify GILZ as a mediator of GC killing, indicate a role of PI3-kinase/AKT in controlling GILZ regulation and suggest that the combination of PI3-kinase/AKT inhibitors and GCs may be a beneficial MM treatment.     

Mechanisms for growth factor-induced pituitary tumor transforming gene-1 expression in pituitary folliculostellate TtT/GF cells

PTTG1, a securin protein, also behaves as a transforming gene and is overexpressed in pituitary tumors. Because pituitary folliculostellate (FS) cells regulate pituitary tumor growth factors by paracrine mechanisms, epidermal growth factor (EGF) receptor (EGFR)-mediated PTTG1 expression and cell proliferation was tested in pituitary FS TtT/GF cells. EGFR ligands caused up to 3-fold induction of Pttg1 mRNA expression, enhanced proliferating cell nuclear antigen, and increased entry of G0/1-arrested cells into S-phase. PTTG binding factor mRNA expression was not altered. EGF-induced Pttg1 expression and cell proliferation was abolished by preincubation of TtT/GF cells with EGFR inhibitors AG1478 and gefitinib. Phosphatidylinositol 3 kinase, protein kinase C, and MAPK, but not c-Jun N-terminal kinase and Janus activating kinase signaling regulated EGF-induced Pttg1, as well as proliferating cell nuclear antigen mRNA expression and entry into S-phase. EGF-induced EGFR and ERK1/2 phosphorylation was followed by rapid MAPK kinase/ERK kinase-dependent activation of Elk-1 and c-Fos. EGF-induced Pttg1 expression peaked at the S-G2 transition and declined thereafter. Pttg1 cell cycle dependency was confirmed by suppression of EGF-induced Pttg1 mRNA by blockade of cells in early S-phase. The results show that PTTG1 and its binding protein PTTG binding factor are expressed in pituitary FS TtT/GF cells. EGFR ligands induce PTTG1 and regulate S-phase, mediated by phosphatidylinositol 3 kinase, protein kinase C, and MAPK pathways. PTTG1 is therefore a target for EGFR-mediated paracrine regulation of pituitary cell growth.