All-trans retinoic acid (ATRA), an active metabolite of vitamin A, regulates gene expression in a number of physiological processes, including embryogenesis, vision, growth, skin and lung maturation, cell differentiation and homeostasis. Furthermore, nowadays it is a cardinal therapy against acute myeloid leukemia (AML), as it is capable of inducing the terminal differentiation of malignant cells which is usually followed by their death. It was believed that this compound last century exerted its effect through the nuclear retinoid receptor RAR-alpha. At the beginning of the 21st century, it became clear that it could bind to other cellular proteins. In this way, the scientists found that retinoic acid would prepare the cellular ground for subsequent genomic responses. One of these cellular proteins is the protein kinase c-Raf1, which is critical for the activation of the mitogenic MAPK pathway. However, depending on the cellular context, MAPK can also drive cellular maturation and commitment to a differentiated state.
The old axioms of cyclic AMP-cell differentiation and MAP kinase-cell proliferation have undergone many research updates over the past 40 years. Traditionally, the treatment of cultured tumor cells with analogues of the second cyclic AMP messenger (dbc-AMP) or activators of the enzyme adenylate cyclase (forskolin), has been shown to induce growth arrest, with the possibility of maturing (differentiation) or posthumous cell death. Instead, the activation of transductive pathways coupled to MAP kinases mediated by growth factors and hormones, has been classically associated with cell duplication and greater malignant expansion. However, over time, anomalies have been recorded, whereby the treatment of some tumor lines with cyclic AMP agonists induced a paradoxical effect of stimulating tumor growth. The first case was noted with thyroid cancer, and the reason was initially linked to the property of the thyroid stimulating hormone (TSH) to stimulate the production of cyclic AMP after binding with its receptor.
But it was evident that thyroid tumor cells must have an alternative cellular target for cyclic AMP, different from the conventional receptor, namely protein kinase A (PKA). The discovery of the Epac proteins solved the puzzle, because it uncovered the bridge between cyclic AMP and MAP kinase-regulated proliferation in cancer cells. An important breakthrough in understanding these mechanisms was achieved using the HL-60 model (promyelocytic leukemia; AML) in which ATRA causes nuclear enrichment of c-Raf which drives differentiation. The mechanism behind this process has been difficult to understand as the c-Raf / MAPK pathway has traditionally been observed to trigger cell proliferation. Besides, original data indicated that despite the absence of cyclic AMP involved, the nuclear activity of retinoic acid bound to ATRA involved the regulatory subunit of PKA (RIIa), which has been speculated either to behave as a transcription factor and/or a scaffold platform for other nuclear proteins or transcriprional regulators.
Scientists speculated that when c-Raf is associated with a different set of protein combinations, it would be able to change its behavior. Furthermore, in this experimental set c-Raf is associated with retinoic acid, which is not supposed to be its regular modulator. Other second messengers, in fact, have been seen to bind and regulate c-Raf such as phosphatidic acid and ceramide biolipids, which can increase its kinase activity or redistribute the protein into cellular subcompartments. Indeed, almost 20 years ago, some publications reported that ceramide was able to induce MAPK kinase acitvity in HL-60 cells, driving them subsequentially toward maturation into monocytes. Another signaling molecule required for the full effect of ATRA on HL-60 cells is the PI-3-kinase, which belong to the PI3K/c-Akt kinase axis involved in cell proliferation and survival. Nuclear effects of either PI3K and c-Akt are well known, though usually associated with cell proliferation but not differentiation.
But again PI3K enzymatic activity allows ATRA to stimulate the expression of a differentiation marker, tissue transglutaminase (TG-ase II), which is consistent with a terminal cell maturation. However, this could change if the intervention of other second messengers is brought up. Beside RAR-alpha and c-Raf1, ATRA could have additional yet unknown cellular targets; in other words, behave like a second messenger itself. One among the most likely is proten kinase C family, since these enzyme are able to bind many kind of hydrophobic molecules. Beginning this century it was demonstrated that, along with Raf kinases, some members of the PKC family may also bind retinol with exceptional high affinity; and that the delta isofrom is able to direcly bind ATRA itself. Furthermore, PKC isoforms (alpha, lambda and iota) are actively involved in the differentiation program triggered by ATRA. It is thought that these kinase may influence gene expression by targeting various transcription factors, and may drive subsequent steps of the RARalpha-driven gene expression as well.
More recent research has shown that a cell cycle inhibitor, roscovitin, enhances the nuclear translocation of some traditionally cytoplasmic transductive molecules and differentiation after cycle arrest. The initial phase probably involves a direct interaction between ATRA and the protein kinase c-Raf1 followed by ATRA-induced cell cycle arrest mediated by loss of phosphorylation of the pRB gene. As a result, pRb would lose its association with c-Raf, which would be free to promote differentiation. Roscovitin is a cyclin-dependent kinase inhibitor (Cdk1, Cdk2, and Cdk4) that also enhanced ATRA-induced nuclear enrichment of other signaling molecules traditionally perceived as cytoplasmic proliferation promoters, but now known to promote differentiation. These included in particular: c-Lyn tyrosine kinases and Src-like c-Fgr; adapter proteins such as c-Cbl and SLP-76; a nucleotide exchange factor, Vav1; and the transcription factor IRF-1.
ATRA causes HL-60 cells to undergo G0 cell cycle arrest and thus myeloid differentiation, which is dependent on a sustained MAPK pathway signal with an unexpected translocation of c-Raf1 within the cell nucleus. This topic has also been analyzed in the past: i.e. transient activation vs. the prolonged activation of MAPK in triggering various outcomes such as cell proliferation. The potential ability of signal intensity and duration to affect growth factor signal outcome has been suggested in different contexts. For example, greater amounts of activated MAPK/cell might target genes whose expression required phosphorylation of more or different transcription factors. Greater signal duration has been implicated in promoting MAPK translocation to the nucleus. Another possibility is that MAPK has targets other than transcription factors that become involved with higher levels of activation.
In the case of RA-enhanced ERK2 activity in HL-60 cells, the duration is relatively long, extending from 4 h to at least 48 h when the onset of differentiation and arrest has already occurred. This is in contrast to the shorter signal durations typical of mitogenic responses to growth factors. These results are consistent with recent findings that intense Ras/Raf signaling can cause cell cycle arrest, whereas lower levels promote proliferation, Usually once a nuclear receptor is bound with a ligans (like the steroid receptors), it become phosphorylated by the simulatenous MAPK signaling; this is not the case with retinoid acid receptor (RAR-alpha). When scientists genetically knocked down c-Lyn, it augmented certain roscovitine enhancements of ATRA effects on nuclear signaling and cell cycle regulatory molecules. This makes sense since this tyrosine kinase is involved in cell proliferation but not differentiation.
Protein tyrosine phosphorilation is classically coupled with a mitogenic MAPK signaling in the absence of a pro-differentiating push. Likely this scenario changes when differentiation is enhanced by molecules like ATRA. In HL-60 leukemia differentiation induce by retinoic acid, Vav1 becomes tyrosine phosphorylated and also interacts with PU.1, a transcription factor that coordinates the expression of the core network of T cell regulatory genes and human stemness during embryogenesis. Other nuclear proteins are involved in the process and many step of chromatin remodeling are in sequence for the complete committment toward cell differentiation. Clinicians are most interested in these aspects of cellular biology to find the perfect target against blood malignancies. The emplyment of ATRA has already changed the prognosis of some bone marrow cancers. Yet ATRA alone may not be enough since its activity seem to become complete woth other drugs.
Most recently, indeed, ATRA has been found to strongly synergize with chimaera Bcr-Abl inhibitors (bosutinib), but c-Raf inbitors (GW5074) and Src-like inhibitors (PP-2) antagonize its activity, especially in ATRA-resistant leukemia forms. Src-family kinases c-Lyn and c-Fgr, c-Raf1 and PI3K display highly correlated signaling changes during ATRA treatment, while activation of traditional downstream targets (Akt, ERK) were poorly correlated with c-Raf or Lyn during differentiation. This suggests that an interrelated kinase module may work in a nontraditional way during retinoic-induced maturation. The acute promyelocytic leukemia (APL) has been treated with ATRA for decades. But while ATRA has largely been ineffective in non-APL AML subtypes, co-treatments with other drugs are currently in clinical trials to unhinge these mechanisms of resistance.
This being the case, researcher must be aware first of how cellular signaling intertwine in order to find a precise target and/or avoid unpleasant surprises. This, unfortunately, does often take some time.
- Edited by Dr. Gianfrancesco Cormaci, PhD, specialist in Clinical Biochemistry.
Rashid A et al. Oncotarget. 2020; 11(12):1017-36.
Wang S et al. Drug Des Devel Ther 2020; 14:297.
Shao X et al. Pharmacol Res. 2020; 151:104545.
Liang C et al. Food Chem Toxicol. 2019; 126:303.
Klobuch S et al. Front Pharmacol. 2018; 9:1380.
Hammerling U. Subcell Biochem. 2016; 81:201-30.
Hoyos B et al. J Exp Med 2000; 192(6):835-45.
Zhang JW et al. J Biosci. 2000 Sep; 25(3):275-84.
Yen A et al. Res. 1998 Jul 15; 58(14):3163-3172.
Rained MA et al. J Biol Chem 1993; 268(20):14572.
Briggs RC et al. J Cell Physiol. 1988; 136(1):198.