Infrequent alterations of the RAR alpha gene in acute myelogenous leukemias, retinoic acid-resistant acute promyelocytic leukemias, myelodysplastic syndromes, and cell lines

Retinoids are important regulators of cell growth and differentiation in vitro and in vivo and they exert their biologic activities by binding to nuclear retinoic acid receptors (RARs; alpha, beta, and gamma) and retinoid X receptors (RXRs; alpha, beta, and gamma). All- trans retinoic acid (RA) induces complete remission in patients with acute promyelocytic leukemia (APL) presumably by binding directly to RAR alpha of APL cells. Leukemic blasts from APL patients initially responsive to RA can become resistant to the agent. HL-60 myeloblasts cultured with RA have developed mutations of the ligand-binding region of RAR alpha and have become resistant to RA. Furthermore, insertion of an RAR alpha with an alteration in the ligand-binding region into normal murine bone marrow cells can result in growth factor-dependent immortalization of the early hematopoietic cells. To determine if alterations of the ligand binding domain of RAR alpha might be involved in several malignant hematologic disorders, the mutational status of this region (exons 7, 8, and 9) was examined in 118 samples that included a variety of cell lines and fresh cells from patients with myelodysplastic syndromes (MDS) and acute myeloid leukemias (AML), including 20 APL patients, 5 of whom were resistant to RA and 1 who was refractory to RA at diagnosis, using polymerase chain reaction-single- strand conformational polymorphism (PCR-SSCP) analysis and DNA sequencing. In addition, 7 of the 20 APLs were studied for alterations of the other coding exons of the gene (exons 2 through 6). No mutations of RAR alpha were detected. Although the sensitivity of PCR-SSCP analysis is less than 100%, these findings suggest that alterations of RAR alpha gene are rare and therefore other mechanisms must be involved in the onset of resistance to retinoids and in the lack of differentiation in disorders of the myeloid lineage.     

Retinoic acid-resistant HL-60 cells exclusively contain mutant retinoic acid receptor-alpha

Sequence analysis of the retinoic acid receptor-alpha (RAR alpha) gene from a subline of HL-60 cells (RA-res) stably resistant to all-trans retinoic acid (RA) disclosed a single-base change in codon number 411, the same C to T transition previously reported in an independently selected HL-60 RA resistant clone by Robertson et al (Blood 80:1885, 1992). This mutation eliminates a FokI restriction endonuclease site. Using primers framing this mutation in exon 9 of the RAR alpha gene, we showed that polymerase chain reaction products amplified from either mRNA or genomic DNA templates from the RA-res subline were completely resistant to FokI digestion whereas those from wild-type (wt) HL-60 cells could be digested to completion. The lack of a normal allele in the RA-res cells was confirmed by mixing experiments and hybridization analyses. Southern blot analysis of DNA from the RA-res and wt cells versus control placental DNA indicated that the RAR alpha gene is not haploid. The independent isolation of the same RAR alpha mutation in different laboratories suggests either that the mutation exits in a small subpopulation in the wt line or that this is a mutational "hot spot." Furthermore, the results indicate that if a dominant negative mode of resistance is involved in the RA-res subline, this must involve interference with the function of heterologous receptor proteins such as the retinoid X receptors. The lack of any normal RAR alpha in this subline may facilitate studies of the mode of action of retinoids.     

Stimulation of tissue-type plasminogen activator expression by retinoic acid in human endothelial cells requires retinoic acid receptor beta 2 induction

We previously showed the involvement of retinoic acid receptor alpha (RAR alpha) in the induction of tissue-type plasminogen activator (t- PA) synthesis by RA in human umbilical vein endothelial cells (HUVECs). However, the rather slow onset of this induction of t-PA synthesis suggested an indirect role of RAR alpha. Here, we show that the protein synthesis inhibitor, cycloheximide completely blocks the induction of t- PA by RA, which points to the need of an intermediary protein in t-PA stimulation. This intermediary protein is likely to be RAR beta 2 on the basis of the following findings: (1) the induction of RAR beta by RA exactly precedes that of t-PA; (2) HUVECs with elevated RAR beta mRNA levels show an undelayed t-PA induction on stimulation with RA, and this response can be almost completely inhibited with an RAR antagonist; and (3) an antisense oligodeoxynucleotide against the translation initiation site of RAR beta 2 mRNA greatly reduces the t-PA induction by RA. Thus, induction of t-PA by RA in HUVECs involves a 2- step mechanism requiring induction of RAR beta 2 via RAR alpha, followed by induction of t-PA synthesis via RAR beta 2. Each of these steps is shown to have a different activation profile with RA and 9 cis RA.     

All-trans-retinol is a ligand for the retinoic acid receptors.

Competition of all-trans-retinol and all-trans-retinaldehyde with 3H-labeled all-trans-retinoic acid (RA) for binding to retinoic acid receptors (RARs) was examined in human neuroblastoma cell nuclear extracts. All-trans-retinol was 35-fold less potent than all-trans-RA, whereas all-trans-retinaldehyde was 500-fold less active in binding to the nuclear receptors. To confirm that all-trans-retinol binds to RARs, experiments were carried out with RARs alpha, beta, and gamma expressed as bacterial fusion proteins. All-trans-retinol was only 4- to 7-fold less potent than all-trans-RA in binding to all three RAR subtypes. The all-trans-retinol binding observed was not the result of metabolism of retinol to RA or some other active compound during the binding experiment. Retinyl acetate was virtually inactive in competition binding experiments, while very slight activity was observed with 13-cis-RA and all-trans-retinaldehyde. Significant competition occurred with 4-hydroxy-RA and 4-keto-RA, which were 15- to 40-fold less potent than all-trans-RA. The 9-cis isomer of RA was equipotent with all-trans-retinol in these studies. These results suggest that all-trans-retinol cannot be excluded as a physiologically significant ligand for RAR-mediated gene expression.     

Exposure to an obesity-inducing diet early affects the pattern of expression of peroxisome proliferator, retinoic acid, and triiodothyronine nuclear receptors in the rat

Since evidence has appeared that alpha and gamma isoforms of the peroxisome proliferator receptors (PPARs) are involved in the regulation of triglyceride homeostasis and in the control of the differentiation of adipocytes that is required for the development of obesity, a large number of studies have investigated the physiologic role of nuclear receptors in the control of energy balance. The aim of this study was to determine the early effects of an obesity-inducing diet on the expression of PPAR alpha and gamma and other nuclear receptors such as all-trans retinoic acid receptor (RAR) and triiodothyronine receptor (TR), which all form functional heterodimers with a common partner, the 9-cis retinoic acid receptor (RXR). The experiment used a cafeteria diet where 60% of the energy was supplied as lipids. This diet was offered to young rats for 8 and 28 days and the expression of nuclear receptors was determined at the end of each experimental time period (1) in the liver by assaying the binding properties of RAR and TR and by quantifying mRNA levels of RAR beta, TR alpha(1)beta(1), RXR alpha, and PPAR alpha, and (2) in the white adipose tissue (WAT) by quantifying mRNA levels of RAR alpha, RXR alpha, TR alpha(1)beta(1), and PPAR gamma(2). After 8 days of cafeteria diet a significant decrease of RAR and TR maximal binding capacity (MBC) was observed in the liver (-20.1% and -35.0%, respectively, P <.05) and the level of the mRNA of RAR beta was significantly decreased (-17.4%, P <.05). After 28 days of cafeteria diet, the level of the mRNA of PPAR alpha and acyl-CoA oxidase (ACOX) was significantly increased (+54.5% and +37.8%, P <.01 and P <.05, respectively), whereas the MBC of RAR and TR was significantly decreased (-16.0% and -23.4%, P <.01), as were the mRNA levels of RAR beta and TR alpha(1) beta(1) (-28.5% and -32.0%, P <.05). The level of RXR alpha mRNA was unchanged. In WAT, the mRNA level of PPAR gamma(2) was significantly increased after 28 days of cafeteria diet (+49.5%, P <.05) and the mRNA levels of RAR alpha and TR alpha(1) beta(1) significantly decreased (-22.3% and -31.0%, P <.05). These results as a whole showed that a high-fat diet can induce early modifications in the pattern of expression of nuclear receptors in the liver and the WAT. These modifications could be compatible with an early adaptive phenomenon. Further investigations are necessary to better understanding the link between the modifications of the pattern of expression of these receptors and plasticity of adipose tissue leading to the onset of obesity.     

1 alpha,25-dihydroxyvitamin D3-induced upregulation of calcineurin during leukemic HL-60 cell differentiation

Cyclosporin A and FK506, at concentrations that inhibited phosphatase activity of calcineurin in HL-60 cellular lysates, augmented the proliferation of leukemic HL-60 cells. These immunosuppressants did not affect 1 alpha,25-dihydroxyvitamin D3 [1,25(OH)2D3]-induced monocytic differentiation of HL-60 cells, but did abrogate the 1,25(OH)2D3- induced inhibition of HL-60 cell growth. Treatment with 20 nmol/L 1,25(OH)2D3 led to a progressive increase in calcineurin phosphatase activity in subcellular fractions from HL-60 cell extracts, the increase in this activity appeared to parallel the phenotypic and functional changes of HL-60 cells during monocytic differentiation induced by 1,25(OH)2D3. Immunoblot analysis indicated that increase in calcineurin activity was concordant with the increased expressions of calcineurin catalytic subunit isozymes, calcineurin A alpha (CNA alpha), and calcineurin A beta(CNA beta), and a regulatory calcineurin B subunit (CNB) proteins, which were preceded by a coordinate increase in the levels of CNA alpha, CNA beta and CNB mRNAs. The expression of calmodulin remained unaltered throughout 1,25(OH)2D3-induced monocytic differentiation. These results suggest that calcineurin activation has a net negative effect on HL-60 cell proliferation, and that the increased expression of calcineurin may be involved in 1,25(OH)2D3- induced inhibition of HL-60 cell proliferation.     

PML protein expression in hematopoietic and acute promyelocytic leukemia cells

Acute promyelocytic leukemia (APL) is thought to be caused by the t(15,17) translocation that fuses the PML gene to that of the retinoic acid receptor alpha (RAR alpha) and generates a PML/RAR alpha fusion protein. Yet, paradoxically, APL cells are exquisitely sensitive to retinoic acid (RA), as they terminally differentiate upon RA exposure. In this report, we have examined the expression of PML and PML/RAR alpha in normal and APL cells. By immunofluorescence or immunocytochemistry, we show that PML has a speckled nuclear pattern of expression that contrasts with that of PML/RAR alpha (mostly a micropunctuated nuclear pattern or a cytoplasmic localization). The APL- derived cell line NB4 that expresses both the PML and PML/RAR alpha genes also shows the fine micropunctuated nuclear pattern, suggesting a dominant effect of the fusion protein over the localization of wild- type PML. RA treatment of NB4 cells or clones expressing PML/RAR alpha gradually leads to a PML pattern before apparent morphologic maturation. In 14 untreated APL patients, the PML-reactive proteins were cytoplasmic (by immunocytochemistry) or both cytoplasmic and nuclear with a micropunctuated pattern (by immunofluorescence). Strikingly, in 4 patients, after 1 to 2 weeks of RA therapy, the speckled nuclear PML pattern reappeared concomitant with the onset of differentiation. These results establish that fusion of PML to RAR alpha results in an altered localization of PML that is reverted upon RA treatment. This observation, which highlights the importance of PML, is likely to be a key to unravelling the molecular mechanism of both leukemogenesis and RA-induced differentiation of APL.     

Recombinant human interferon alpha enhancement of retinoic-acid-induced differentiation of HL-60 cells

We have previously demonstrated that a combination of interferon beta and a differentiation agent, dimethyl sulfoxide (DMSO), is cytotoxic for HL-60 cells, a human promyelocytic leukemic cell line. We now report that a combination of recombinant interferon alpha (Intron; Schering) and retinoic acid is synergistically cytostatic for HL-60 cells. Retinoic acid (RA) induced the differentiation of HL-60 cells into granulocytes. Interferon (IFN) alone at 1-1000 IU/ml had no effect on either differentiation or proliferation of HL-60 cells. The addition of 1000 IU/ml of IFN and 10(-7) M RA at the initiation of culture reduced the number of viable cells to 28% of that observed for cells treated with RA alone. The decreased number of cells was a result of decreased cellular proliferation, rather than of a cytotoxic effect of the combination. IFN-RA-treated cells differentiated more rapidly than cells treated with RA alone. In addition, the final percentage of mature cells was increased at day 7 in IFN-RA-treated cultures, as compared with RA-treated cells. Simultaneous treatment of the cells with IFN and RA decreased the concentration of RA needed to induce differentiation or to exert a cytostatic effect. Significant changes in the nuclear structure of RA-treated HL-60 cells after 24 h have been reported. Cells were pulsed with RA for 24 h, washed, and IFN added. At day 7, cell growth was inhibited to the same extent as that of cells continuously exposed to IFN-RA. However, while 70% of the continuously exposed cell differentiated, cells pulsed with RA and subsequently treated with IFN did not differentiate. The results of this investigation further support our findings that combinations of IFN and inducers of differentiation may be of importance in the treatment of leukemia.     

In vitro interaction of recombinant tumor necrosis factor alpha and all- trans-retinoic acid with normal and leukemic hematopoietic cells

Both human recombinant tumor necrosis factor alpha (TNF alpha) and all- trans-retinoic acid (RA) inhibit the in vitro clonal growth of human myeloid leukemic cells. We investigated the in vitro interaction of TNF alpha and RA with normal and a variety of leukemic myeloid cells. With the promyelocytic HL-60 cells, TNF alpha (greater than or equal to 2.5 U/mL) in combination with RA synergistically inhibited clonal growth; TNF alpha at lower concentrations (less than or equal to 1 U/mL) plus RA (10(-9) mol/L) were antagonistic in their inhibition of growth. The ability of RA (10(-8) mol/L) plus TNF alpha (2.5, 5 U/mL) to enhance differentiation of HL-60 cells paralleled their ability to inhibit clonal growth of these cells. In addition, RA (10(-9) to 10(-7) mol/L) increased the number of TNF alpha receptors on HL-60 cells 1.3- to 1.7- fold without changing the affinity for the TNF alpha receptor. With the more immature KG-1 myeloblasts, concentrations of TNF alpha greater than 10 U/mL synergistically interacted with RA to inhibit clonal growth; at lower concentrations of TNF alpha (less than 10 U/mL), RA appeared to inhibit the expected effect of TNF alpha. KG-1 cells were not induced to differentiate with either agent alone or in combination. With four of nine leukemic patients, TNF alpha in combination with RA (10(-7) mol/L) inhibited leukemic clonal growth to a greater extent than each agent alone. No marked effect of the combined treatment was seen in two other patients. The RA reversed the inhibitory action of TNF alpha on normal human granulocyte-macrophage colony forming cells (GM-CFC) and on clonal growth of leukemic cells from three patients. Our study suggests that TNF alpha and RA interact in a complex manner with normal and leukemic hematopoietic cells.