All trans-retinoic acid analogs promote cancer cell apoptosis through non-genomic Crabp1 mediating ERK1/2 phosphorylation

www.nature.com/scientificreports OPEN received: 09 November 2015 accepted: 11 February 2016 Published: 03 March 2016 All trans-retinoic acid analogs promote cancer cell apoptosis through non-genomic Crabp1 mediating ERK1/2 phosphorylation Shawna D. Persaud1,*, Sung Wook Park1,*, Mari Ishigami-Yuasa2, Naoko Koyano-Nakagawa3, Hiroyuki Kagechika2 & Li-Na Wei1 All trans retinoic acid (atRA) is one of the most potent therapeutic agents, but extensive toxicity caused by nuclear RA receptors (RARs) limits its clinical application in treating cancer. AtRA also exerts nongenomic activities for which the mechanism remains poorly understood. We determine that cellular retinoic acid binding protein 1 (Crabp1) mediates the non-genomic activity of atRA, and identify two compounds as the ligands of Crabp1 to rapidly and RAR-independently activate extracellular signal regulated kinase 1/2 (ERK1/2). Non-canonically activated ERK activates protein phosphatase 2A (PP2A) and lengthens cell cycle duration in embryonic stem cells (ESC). This is abolished in Crabp1-null ESCs. Re-expressing Crabp1 in Crabp1-negative cancer cells also sensitizes their apoptotic induction by atRA. This study reveals a physiological relevance of the non-genomic action of atRA, mediated by Crabp1, in modulating cell cycle progression and apoptosis induction, and provides a new cancer therapeutic strategy whereby compounds specifically targeting Crabp1 can modulate cell cycle and cancer cell apoptosis in a RAR-independent fashion, thereby avoiding atRA’s toxicity caused by its genomic effects. All-trans retinoic acid (atRA), the active ingredient of Vitamin A, affects diverse biological processes including development, growth, immune function, neural function, reproduction, and vision. Like other steroid hormones, atRA elicits genomic action via its nuclear receptors retinoic acid receptors (RARs) and non-genomic action. The non-genomic mediator of atRA and its mechanism of action remain poorly understood. For estrogens, studies reported membrane-associated estrogen receptor (ER) as the principal endogenous mediator for their non-genomic action1–3. For thyroid hormones, a study reported a yet-to-be identified membrane-localized receptor for its non-genomic action modulating nitrate oxide synthase, guanylyl cyclase and protein kinase GII4. Interestingly, glucocorticoids can involve a membrane-bound glucocorticoid receptor (GR) and a non-classical membrane GR that is a glycoprotein5 for non-genomic activities. Typical genomic action of steroid hormones occurs after a time lag of several hours, and alters target gene expression lasting for an extended period of time6. On the contrary, non-genomic steroid actions such as thyroid hormone occur much faster, ranging from minutes to seconds, and can affect multiple signaling pathways6. However, the physiological implication of these non-genomic actions has been debated. Most critically, a lack of specific ligands that can selectively elicit the “non-genomic” action of these hormones without acting on their nuclear receptors has prevented a validation of these studies and hindered the progress in the field. For atRA, non-genomic action has been reported in multiple studies7, such as activating PKCα 8, binding to membrane RAR to affect neuron spine formation9, and rapidly activating extracellular signal regulated kinase 1/2 (ERK1/2)10,11. ERK activation by atRA has been most widely detected in different experimental model systems. This current study establishes a physiological role for Crabp1 in mediating the non-genomic activities of atRA, and identifies specific ligands for Crabp1 that can elicit atRA’s non-genomic actions without involving RARs. 1 Department of Pharmacology University of Minnesota, Minneapolis, MN 55455, USA. 2Tokyo Medical and Dental University (TMDU), Institute of Biomaterials and Bioengineering, 2-3-10 Kanda-Surugadai, Chiyoda-ku, Tokyo 1010062, JAPAN. 3Department of Medicine, University of Minnesota, MN 55455, USA. *These authors contributed equally to this work. Correspondence and requests for materials should be addressed to L.-N.W. (email: weixx009@ umn.edu) Scientific Reports | 6:22396 | DOI: 10.1038/srep22396 1 www.nature.com/scientificreports/ Figure 1. RAR-independent, rapid activation of ERK1/2 by compounds 3 and 4. (A) Structure of compounds. (B) Activation of ERK1/2 in Cos-1 assessed by western blot analyses under 100 nM for 30 min. Upper band depicts ERK1 (44 kDa) and lower band depicts ERK2 (42 kDa). (C) Compounds 3 and 4 do not activate RAR activity as detected by RAR reporter assay performed in Cos-1 cells treated with compounds at 250 nM for 24 hrs. (D) Pan-RAR antagonist AGN 193109 (100 nM, 1 hr pretreatment) fails to block rapid ERK activation in Cos-1 cells. Data (B–D) are representative of at least 3 independent experiments. (E) In vitro competition assay to displace 3H-atRA from Crabp1 by atRA, C3, and C4. Data is displayed as counts per minute (CPM). Asterisk shows significance: RA (p <  0.001), C3 (p =  0.023) and C4 (p =  0.014) versus control, and mean ±  S.E.M (n =  4). The non-genomic effects of atRA are particularly interesting in a stem cell context such as embryonic stem cells (ESCs) and cancer cells. Maintaining their self-renewal requires a tight control over cell cycle progression that ultimately governs cell proliferation, differentiation, senescence and apoptosis12. It is known that during differentiation protein phosphatase 2A (PP2A) activity gradually increases; and inhibiting PP2A promotes ESC self-renewal12. Interestingly, in this current study we identify PP2A as a target of non-genomic ERK1/2 activation, elicited by holo-Crabp1, in ESC. From a translational point of view, atRA can be a potent therapeutic for various diseases because of its anti-proliferative, anti-oxidative, pro-apoptotic, and differentiation-inducing activities. It has been most successful in treating acute promyelocytic leukemia13. In animal models of skin, oral, lung, breast, bladder, ovarian, and prostate cancers, atRA has also been found to suppress carcinogenesis14–19. In various human clinical trials to prevent oral, head and neck, non-melanoma skin cancers, breast, cervical, and hepatocellular carcinoma, retinoids were a highly promising category of compounds that have been investigated20. However, toxic side effects and retinoic acid syndrome have drastically limited clinical use of these compounds. These toxicities are mainly attributed to retinoids’ canonical, genomic actions through binding to RARs21. Furthermore, at therapeutic doses atRA loses efficacy as RA resistance develops. RA resistance currently is attributed to RAR action and hinders therapeutic potential of retinoid compounds22,23. An important question is, whether it is possible to elicit certain desirable effects of retinoids without activating the wide spectrum RAR-mediated genomic actions. To this end, this study reports two compounds that act specifically (...truncated)