Targeting Nrf2-Mediated Oxidative Stress Response Signaling Pathways as New Therapeutic Strategy for Pituitary Adenomas

Oxidative stress and oxidative damage are the common pathophysiological characteristics in pituitary adenomas (PAs), which have been confirmed with many omics studies in PA tissues and cell/animal experimental studies. Nuclear factor erythroid 2 p45-related factor 2 (Nrf2), the core of oxidative stress response, is an oxidative stress sensor. Nrf2 is synthesized and regulated by multiple factors, including Keap1, ERK1/2, ERK5, JNK1/2, p38 MAPK, PKC, PI3K/AKT, and ER stress, in the cytoplasm. Under the oxidative stress status, Nrf2 quickly translocates from cytoplasm into the nucleus and binds to antioxidant response element /electrophile responsive element to initiate the expressions of antioxidant genes, phases I and II metabolizing enzymes, phase III detoxifying genes, chaperone/stress response genes, and ubiquitination/proteasomal degradation proteins. Many Nrf2 or Keap1 inhibitors have been reported as potential anticancer agents for different cancers. However, Nrf2 inhibitors have not been studied as potential anticancer agents for PAs. We recommend the emphasis on in-depth studies of Nrf2 signaling and potential therapeutic agents targeting Nrf2 signaling pathways as new therapeutic strategies for PAs. Also, the use of Nrf2 inhibitors targeting Nrf2 signaling in combination with ERK inhibitors plus p38 activators or JNK activators targeting MAPK signaling pathways, or drugs targeting mitochondrial dysfunction pathway might produce better anti-tumor effects on PAs. This perspective article reviews the advances in oxidative stress and Nrf2-mediated oxidative stress response signaling pathways in pituitary tumorigenesis, and the potential of targeting Nrf2 signaling pathways as a new therapeutic strategy for PAs.


INTRODUCTION
Pituitary adenoma (PA) is a common intracranial neoplasm that occurs in the central regulatory organ pituitary gland in the hypothalamic-pituitary-target organ axis system, which seriously affects human endocrine system and health. PAs account for 10-25% of all intracranial tumors, and are classified into benign (∼65%), invasive (∼35%), and malignant (carcinoma; only 0.1-0.2%) PAs according to the malignancy level (Stalla et al., 2019). PAs are divided into macroadenomas (≥10 mm) and microadenomas (<10 mm) according to tumor size (Lopes, 2017). They are also divided into clinically functional and nonfunctional PAs (FPAs and NFPAs) according to the level of hormone secretion (Zhan et al., 2016). FPAs are hormonesecreting PAs, which result in hyperpituitarism, including acromegaly derived from growth hormone (GH)-secreting PAs, hyperprolactinemia derived from prolactin (PRL)secreting PAs, and Cushing's syndrome derived from adrenocorticotropin (ACTH)-secreting PAs. NFPAs are nonhormone-secreting PAs . The main clinical symptoms of PAs include inappropriate hormone secretion syndrome, and compression of the neighboring tissues and structures such as headache, visual field defect, and increased intracranial pressure (Reimondo et al., 2019). PA is a multifactor, multi-process, and multi-consequence complex disease, which is involved in a series of molecular alterations at the levels of genome, transcriptome, proteome, peptidome, metabolome, and radiome; and these molecules mutually associate and function in a molecular network system (Zhan and Desiderio, 2010b;Hu et al., 2013;Grech et al., 2015;Cheng and Zhan, 2017;Lu and Zhan, 2018). Thus, one must shift the research and practice strategy from a single-factor model to a multiparameter systematic model for predictive, preventive, and personalized medicine in PAs (Hu et al., 2013;Grech et al., 2015;Cheng and Zhan, 2017). Multiomics is an effective approach to realize this multi-parameter systematic strategy model shift, which can establish signaling pathway systems for in-depth understanding of molecular mechanisms of PAs, identify molecular network-based biomarkers for prediction, diagnosis, and prognostic assessment of PAs, and discover signaling pathway network-based therapeutic targets for effective treatment of PAs (Grech et al., 2015;Cheng and Zhan, 2017;Lu and Zhan, 2018).
A series of omics analyses have been performed in PAs to reach our long-term goals that clarify molecular mechanisms and discover effective biomarkers and therapeutic targets for PAs (Zhan and Desiderio, 2010a;Long et al., 2019;Cheng et al., 2019;, including NFPA quantitative transcriptomics (differentially expressed genes, DEGs) (Moreno et al., 2005;Cheng et al., 2019), NFPA quantitative proteomics (differentially expressed proteins, DEPs) (Moreno et al., 2005), NFPA proteomic mapping (Zhan and Desiderio, 2003;Cheng et al., 2019), NFPA nitroproteomics (Zhan and Desiderio, 2006), invasive NFPA quantitative transcriptomics (Galland et al., 2010;Zhou et al., 2011;, invasive NFPA quantitative proteomics (Zhan et al., 2014b), control pituitary proteomic mapping (Beranova- Giorgianni et al., 2002;Giorgianni et al., 2003;Zhao et al., 2005), pituitary control nitroproteomics (Zhan and Desiderio, 2004;Zhan and Desiderio, 2007), control pituitary phosphoproteomics (Giorgianni et al., 2004;Beranova-Giorgianni et al., 2006), PRL-secreting adenoma proteomics and transcriptomics (Evans et al., 2008), and ACTH-secreting adenoma proteomics and metabolomics (Feng et al., 2018). Integrative analysis of these omics data has revealed some important signaling pathway network alterations in PA pathogenesis, including mitochondrial dysfunction, oxidative stress, cell cycle dysregulation, and mitogen-activated protein kinase (MAPK) signaling pathway alteration (Zhan and Desiderio, 2010a;Long et al., 2019). Mitochondrial dysfunction pathway network and mitochondrial dynamics , and MAPK signaling pathway-based drug therapeutic targets  have been discussed in detailed in PAs. It is well-known that mitochondria are the energy factories of the body, and mitochondrial metabolism is the source of reactive oxygen species (ROS). The imbalance between free radicals reactive oxygen/nitrogen species (ROS/ RNS) and antioxidant system leads to oxidative stress, which plays an important role in diseases. Many studies focus on oxidative stress system as therapeutic strategy; for example, benfotiamine is an efficient antioxidant, which could prevent oxidative stress in the anterior tibialis muscle and heart of mice (Gonçalves et al., 2019). Another research shows that pancreatic oxidative damage in the diabetic state is caused by ROS, and scavenging the various ROS generated in the disease is one of effective ways to treat this disease (Afolabi et al., 2018). Studies have clearly demonstrated that mitochondrial dysfunction and oxidative stress pathway changes operate in PAs (Zhan and Desiderio, 2010a), and nuclear factor erythroid 2 p45-related factor 2 (Nrf2)-mediated oxidative stress response significantly impacts the pathogenesis of PAs and modulates the energy metabolism reprogramming for PAs (Sabatino et al., 2018). It is well-known that PAs can lead to abnormal hormone secretion, which might affect oxidative stress and Nrf2 signaling in PAs; for example, human growth hormone (hGH) can attenuate inflammation and oxidative stress attained by Cisplatin probably through inhibition of Nrf2/heme oxygenase 1 (HO-1) pathway (Mahran, 2020). More studies show that Nrf2 signaling and oxidative stress can be regulated by cortisol (Wu et al., 2019), thyroid hormone (Mishra et al., 2019), follicle-stimulating hormone (FSH) , luteinizing hormone (LH) , GH (Mahran, 2020), ACTH (Benlloch et al., 2016), and PRL (Ebokaiwe et al., 2020). These findings clearly demonstrate the importance of oxidative stress in PAs. This present review article will focus on oxidative stress response signing pathway network in PA pathogenesis.

REDOX HOMEOSTASIS AND NRF2 AS THE HEART OF OXIDATIVE STRESS RESPONSE
Oxidative stress is derived from the imbalance between the upload of free radicals ROS/RNS from in vivo and in vitro environmental approaches and the ability of endogenous antioxidants to detoxify these ROS/RNS (Prasad et al., 2016;Klaunig, 2018;Sajadimajd and Khazaei, 2018). It results in the injuries of multiple biomacromolecules such as DNAs, RNAs, proteins, and membrane lipids to significantly associate with a wide spectrum of diseases including cancers. Many studies demonstrate that the increased ROS/RNS productions promote carcinogenesis development (Kudryavtseva et al., 2016;Kruk and Aboul-Enein, 2017), and oxidative stress-medicated chronic inflammation is the risk factor of tumorigenesis (Reuter et al., 2010;Qian et al., 2019). The oxidative phosphorylation system in mitochondrial respiratory chain is the central machine that generates ROS products such as superoxide radical (O 2 .-). One study shows that ROS levels and signs of oxidative damage are significantly increased in PAs (Sabatino et al., 2018). One of the most important RNS, nitric oxide (NO), is generated by inducible nitric synthase (iNOS) in many pathogenesis conditions, which can rapidly react with superoxide radical (O 2 .-) to generate more toxic peroxynitrite anion (ONOO -) and highly reactive hydroxyl radical (OH . ) to attack DNAs, RNAs, proteins, and membrane lipids. iNOS has been extensively found in rat and human pituitaries (Ceccatelli et al., 1993;Lloyd et al., 1995;Ueta et al., 1998;Kruse et al., 2002;Pawlikowshi et al., 2003) and has the elevated activities in PAs compared to those in controls (Vankelecom et al., 1997;Kruse et al., 2002). Another study shows that NO functions in the hypothalamic-pituitary-adrenocortical axis (Riedel, 2002) by promoting the release of follicle-stimulating hormone-releasing hormone (FSHRH) and luteinizing hormonereleasing hormone (LHRH) from hypothalamus (McCann et al., 2001;Pinilla et al., 2001;McCann et al., 2003), and regulating secretion of PRL (Duvilanski et al., 1995) and GH in pituitaries and PAs (Cuttica et al., 1997;Pinilla et al., 1999;Bocca et al., 2000). Peroxynitrite anion (ONOO − ) is a key factor in vivo that causes protein tyrosine nitration and alters protein functions. Nine nitrotyrosine-containing proteins have been identified in NFPA tissues, and tyrosine nitration occurs in important structural and functional domains to change protein functions (Zhan and Desiderio, 2006).
With the generation of ROS/RNS, the in vivo antioxidant detoxification system is correspondingly initiated to adapt against the increased ROS/RNS (Valko et al., 2006;Obrador et al., 2019). The endogenous antioxidant detoxification system is a very complex system, including i) enzymatic antioxidants such as superoxide dismutases (CuZnSOD and MnSOD), glutathione peroxidase, and catalase; ii) non-enzymatic antioxidants such as vitamin E, vitamin C, carotenoid, flavonoid, selenium, thiol antioxidant (thioredoxin, lipoic acid, and glutathione), and others; and iii) multiple regulatory factors [Nrf2, NF-kB (nuclear factor kB), and AP-1 (activator protein-1), etc.] that interact with antioxidants (Valko et al., 2006;Obrador et al., 2019). CuZnSOD exists in most parts of cells, while MnSOD is only found in mitochondrial matrix; and both of them are able to effectively scavenge O 2 .and generate H 2 O 2 (Li et al., 1995;Melov et al., 2001;Elchuri et al., 2005). H 2 O 2 can be scavenged by GPX's (glutathione peroxidases) and peroxiredoxins (thioredoxindependent peroxidases) (Chu et al., 2004;Kang et al., 2005). Studies have found that the levels of CuZnMOD and MnSOD are significantly lower in PAs compared to those of controls (Kurisaka et al., 2004;Yang et al., 2012;Ilhan et al., 2018). The abnormal activities of these antioxidant enzymes and non-enzymatic antioxidants are directly associated with carcinogenesis (Neumann et al., 2003;Chu et al., 2004;Harris et al., 2015). The transcription factor Nrf2 is pivotal to the antioxidant response, which is a sensor of oxidative stress in redox homeostasis, and is mainly located in the cytoplasm under basal conditions (Li and Kong, 2009;Furfaro et al., 2016a). When the upload of free radicals ROS/RNS is increased to cause oxidative stress, Nrf2 quickly translocates from cytoplasm into the nucleus to initiate the antioxidant response, protecting against oxidative/nitrative damages (Dhakshinamoorthy and Porter, 2004;Osburn et al., 2006;Mann et al., 2007;Pi et al., 2008). The Nrf2 signaling regulatory system contains at least four components, including Nrf2, Kelch-like ECH-associated protein 1 (Keap1), small musculoaponeurotic fibrosarcoma (Maf), and antioxidant response element (ARE) or electrophile responsive element (EpRE), which in combination are necessary for the antioxidant response (Kwak and Kensler, 2010;Furfaro et al., 2016;de la Vega et al., 2018). Nrf2 signaling pathways regulate multiple biological processes, including i) the expressions of antioxidant genes, ii) ubiquitin-proteasome system, iii) molecular chaperone/stress-response system, and iv) anti-inflammatory response (Kwak and Kensler, 2010;Furfaro et al., 2016). The accumulated evidence clearly demonstrates that Nrf2 signaling pathways are involved in 12 hallmarks of cancer, including sustained proliferative signaling, insensitivity to antigrowth signals, resistance to apoptosis, limitless replicative potential, sustained angiogenesis, tissue invasion and metastasis, metabolic reprogramming, avoiding immune destruction, tumor-promoting inflammation, genome instability, altered redox homeostasis, and proteotoxic stress (de la Vega et al., 2018). Thereby, any decreased capability of the antioxidant protective system in the redox homeostasis might cause more susceptibility to carcinogen toxicity, tumor inflammatory response, oxidative stress, and carcinogenesis (Yates and Kensler, 2007).

MULTIUOMICS REVEALS OXIDATIVE STRESS-RELATED PATHWAY ALTERATIONS IN PAs
Our multiomics studies in PAs (Zhan and Desiderio, 2010a;Long et al., 2019) clearly demonstrate oxidative stress-related pathway changes in PAs. For example, i) Nrf2-mediated oxidative stress response pathway is significantly changed in NFPAs with evidence of upregulation of key molecules [upregulated DEPs: GST (glutathione S-transferase) or GSTM2 (glutathione S-transferase mu 2), and ERP29 (endoplasmic reticulum protein 29], and downregulation of key molecules [downregulated DEPs: HSP22 (heat shock protein 22), HSP27, and HSP90 or GRP94 (94 kD glucose-regulated protein)] in this pathway. ii) Mitochondrial dysfunction pathway is significantly changed in NFPAs with evidence of upregulation of key molecules [upregulated DEPs: NDUFS8 (NADH ubiquinone oxidoreductase core subunit S8), COX6B (cytochrome c oxidase subunit 6B), CAT (catalase), β-secret2, and ATP5B (ATP synthase, H+ transporting mitochondrial F1 complex, beta subunit)], and downregulation of key molecules [downregulated DEPs: GPX4 (glutathione peroxidase 4), and ATP5A1] in this pathway. Mitochondrial dysfunction can increase ROS production in cancer cells to mediate tumorrelated signaling pathways and activate pro-oncogenic signaling . iii) Oxidative phosphorylation pathway is significantly changed in NFPAs with evidence of upregulation of key molecules (upregulated DEPs: NDUFS8, COX6B, and ATP5B) in this pathway. Mitochondrial oxidative phosphorylation system contains mitochondrial complexes I, II, III, IV, and V, which are the major sites that produce endogenous ROS such as OH . and O − 2 ; among these, complexes I, II, and III play a crucial role in the generation of mitochondrial ROS, because the electrons tend to be leaky at complexes I and III, which results in an incomplete reduction of oxygen and thus generates a free radical such as superoxide radical . iv) Glutathione redox reaction I pathway is significantly changed in NFPAs with evidence of downregulation of key molecule (downregulated DEP: GPX4) in this pathway. GPX's (glutathione peroxidases) are important components in the antioxidant defense system: the downregulation of GPX's can decrease the capability of the antioxidant defense system. v) The superoxide radical degradation pathway is significantly changed in NFPAs with evidence of upregulation of key molecule (upregulated DEP: CAT) in this pathway. vi) Aryl hydrocarbon receptor signaling is significantly changed in NFPAs with evidence of upregulation of key molecules [upregulated DEP: GST; upregulated DEGs: HSPCA (heat shock protein 90 alpha family class A member 1), HSPCB (heat shock protein 90 alpha family class B member 1), ESR1 (estrogen receptor 1), and Bax (BCL2 associated X, apoptosis regulator)], and downregulation of key molecules [downregulated DEPs: HSP27, HSP90 or GRP94, and TGM2 (transglutaminase 2); downregulated DEG: ESR2 (estrogen receptor 2)] in this pathway. vii) Glucocorticoid receptor signaling is significantly changed in NFPAs with evidence of upregulation of key molecule [upregulated DEG: PI3K (phosphatidylinositol 3 kinase)], and downregulation of key molecules [downregulated DEGs: HSP70, c-Fos, CCL2 (C-C motif chemokine ligand 2), BCL2, PRL, and POMC (proopiomelanocortin)] in this pathway. viii) Corticotropinreleasing hormone signaling is significantly changed in NFPAs with evidence of upregulation of key molecules [upregulated DEGs: CALM (calmodulin), and IP3R)], and downregulation of key molecules [downregulated DEGs: ACTH, Nur77 (NR4A1 nuclear receptor subfamily 4 group A member 1), and c-FOS)] in this pathway. ix) Melatonin signaling is significantly changed in NFPAs with evidence of nitration of key molecule (PKA) in this pathway. x) Methylglyoxal degradation III pathway is significantly changed in NFPAs with evidence of upregulation of key molecules [upregulated DEPs: aldose reductase or AKR1B1(aldo-keto reductase family 1 member B)] in this pathway. xi) AMPK signaling is significantly changed in NFPAs with evidence of upregulation of key molecules [upregulated DEGs: PP2C (putative protein phosphatase), and PFK (phosphofructokinase)], and downregulation of key molecules [downregulated DEGs: PI3K, PKA, and PDK1 (pyruvate dehydrogenase kinase 1)] in this pathway. Thereby, these signaling pathway changes clearly demonstrate that the disturbance in redox homeostasis, the imbalance between generation and detoxification of free radicals ROS/RNS, results in oxidative stress and damage in human PAs. Recently, these findings are also confirmed with experiments in cell models and animal models, which demonstrate that increased mitochondrial fusion results in bigger mitochondria, increased ROS levels, and oxidative damage in PAs, and that Nrf2 signaling pathway is activated in PAs as an antioxidant response (Sabatino et al., 2018). Thus, it suggests that Nrf2 is the master regulator of the cellular antioxidant response (de la Vega et al., 2018).
One study shows that Nrf2, phosphorylated Nrf2 (p-Nrf2) protein, and mRNA expressions are increased in PAs, and the Nrf2 downstream effector HO-1 is also increased in PAs (Sabatino et al., 2018). This clearly demonstrates the activation of the Nrf2 signaling pathway, likely causing the extensive surviving capability of pituitary tumor cells. The Nrf2/PTENinduced putative kinase protein 1 (PINK1)/Parkin pathway and mitophagy are activated in T-2 toxin-induced toxicities in rat pituitary GH3 cells (Deyu et al., 2018). Antioxidants N-acetylcysteine (NAC) and vitamin E can decrease the expressions of Nrf2 and HO-1 in rat pituitaries (Prevatto et al., 2017). Genetically induced Nrf2 overexpression in melanoma cells promotes tumor growth and increases antioxidant defense in malignant cells, which can be inhibited by anticancer agent pterostilbene (Pter, a natural dimethoxylated analog of resveratrol) through the downregulation of pituitary production of ACTH, plasma corticosterone, and the glucocorticoid receptor-and Nrf2-dependent antioxidant defense systems in growing melanomas (Benlloch et al., 2016). Irradiation can result in oxidative damage in C57/BL6 mice via activation of Nrf2 and HO-1 expressions, which can be blocked by antioxidant agent pituitary adenylate cyclase-activating polypeptide 38 (PACAP38) through inhibiting Nrf2 expression . Chronic restraint stress (CRS) and acute restraint stress (ARS) can upregulate the mRNA expressions of oxidative stress molecules (gp91phox, iNOS, and Nrf2) and inflammationrelated molecules (IL-1β, IL-6, TNFα, and TLR4) in the mouse hypothalamus, which can be alleviated by Iptakalim (Ipt), an ATP-sensitive potassium (K-ATP) channel opener (Zhao et al., 2017). The loss-of-function mutations of the aryl hydrocarbon receptor-interacting protein gene (AIP) are well-recognized in PAs (Hernández-Ramírez et al., 2018). The aryl hydrocarbon receptor signaling is also revealed by multiomics as an oxidative stress-related signaling pathway in PAs (Long et al., 2019). Further studies show that AIP interacts with antioxidants, chaperone and stress response-related proteins, and cytoskeletal proteins, including HSPA5, HSPA9, HSP90AA1, HSP90AB1, HSPA8, SOD1, TUBB, TUBB2A, and NME1; AIP variants show the impaired interaction of AIP with HSPA8, HSP90AB1, NME1, SOD1, TUBB, and TUBB2A; AIP-mutated PAs show the reduced expression of TUBB2A (Cuadrado et al., 2019). The levels of MnSOD and total antioxidant capability (TAC) are significantly decreased in GH-secreting PAs (Ilhan et al., 2018). The frequencies of micronuclei (MN), nucleoplasmic bridges, nuclear buds, apoptotic and necrotic cells, and plasma 8hydroxy-2′-deoxyguanosine (8-OHdG) levels in peripheral blood lymphocytes are significantly increased in PRL-secreting PAs, which indicates the increased oxidative damage in PRL-secreting PAs (Bitgen et al., 2016). Oxidative stress and mitochondrial dysfunction have been revealed by multiple proteomics and nitroproteomics studies in human PAs (Zhan and Desiderio, 2010a;Zhan et al., 2013;Zhan et al., 2014a;Zhan et al., 2014b;Long et al., 2019). Also, tumor inflammation is an important pathophysiological characteristic in human PAs, which is always tightly associated with oxidative stress and chronic inflammation. The relationship among age-related disease, chronic inflammation, and oxidative stress has also been discussed (Pizza et al., 2011;Liguori et al., 2018;Qian et al., 2019). Oxidative stress is also involved in the processes of anti-proliferative effect and cell death induced by dopamine in the pituitary tumor cells via dopamine D2 receptors through p38 MAPK, and ERK pathways . Therefore, oxidative stress and antioxidative stress response extensively exist in PA pathogenesis. Nrf2, as the core of oxidative stress response, could be the novel target used to develop effective therapeutic agents for human PAs (Kwak and Kensler, 2010;Furfaro et al., 2016;de la Vega et al., 2018).

THERAPEUTIC STATUS TARGETING NRF2 SIGNALING PATHWAYS IN CANCERS
Nrf2 signaling, as the heart of oxidative stress response, is extensively related to cancer pathogenesis, which has attracted tremendous attention as possible anticancer therapeutic target. Nrf2 signaling-based anticancer therapeutic studies have been extensively carried out in multiple cancers, including acute myeloid leukemia, gallbladder cancer, renal carcinoma, pancreatic cancer, melanoma, hepatocellular carcinoma, lung cancer, colon cancer, ovarian cancer, breast cancer, esophageal cancer, and glioblastoma ( Table 1). i) In acute myeloid leukemia, studies found that Nrf2 activators [dimethyl fumarate (DMF), tert-butylhydroquinone, or carnosic acid] and vitamin D derivatives can cooperatively induce acute myeloid leukemia cell differentiation to inhibit leukemia progression in a xenograft mouse model via activating the Nrf2/ARE signaling pathway (Nachliely et al., 2019). Novel pyrazolyl hydroxamic acid derivative (4f) can inhibit Nrf2 activity to induce apoptosis of human acute myeloid leukemia cells . ii) In gallbladder cancer, one study found that atypical protein kinase Cι (aPKCι) can promote gallbladder tumorigenesis and chemoresistance of anticancer agent gemcitabine by competing with Nrf2 for binding to Keap1, implying that inhibiting the aPKC1-Keap1-Nrf2 axis might overcome drug resistance for the treatment of gallbladder cancer . iii) In renal carcinoma, one study found that the natural product chitosan oligosaccharide (COS) can inhibit human renal carcinoma cell proliferation in vitro and in vivo by promoting the expressions of Nrf2 and Nrf2 target genes such as HO-1, the modifier subunit of    Frontiers in Pharmacology | www.frontiersin.org March 2021 | Volume 12 | Article 565748 8 glutamate cysteine ligase, solute carrier family 7 member 11, glucose-regulated protein 78, protein RNA-like endoplasmic reticulum kinase, and cytochrome C. (Zhai et al., 2019). iv) In pancreatic cancer, one study found that anticancer agent resveratrol enhances the sensitivity of pancreatic cancer cells to gemcitabine via suppressing NAF-1 (nutrient-deprivation autophagy factor-1) expression, inducing ROS accumulation, and activating Nrf2 signaling pathways . v) In melanoma, the co-treatment of Nrf2 inhibitor (brusatol, BR) and UVA irradiation can effectively inhibit melanoma growth by regulating AKT-Nrf2 pathway . vi) In hepatocellular carcinoma, one study found that potential Nrf2 inhibitors can sensitize chemotherapy drugs in hepatocellular carcinoma (Tian et al., 2018). Cordycepin (CA) can activate the Nrf2/HO-1/NF-κB pathway for its anti-hepatocarcinoma effect in N-nitrosodiethylamine (NDEA)-induced mouse hepatocellular carcinomas (Zeng et al., 2017). The novel indazolo[3,2-b]quinazolinone (IQ) derivatives, IQ-7 and IQ-12, can induce apoptosis of human hepatoma cells Hep3B and inhibit the Nrf2/ARE signaling pathway in Hep3B cells, and IQ-7 is suggested as a degree of specificity against cancer cells (Zhang et al., 2016). Also, dibenzoylmethane (DBM) can protect against carbon tetrachloride (CCl4)-induced liver injury by activating Nrf2 signaling via JNK, AMPK, and calcium signaling . vii) In lung cancer, one study found that the potent anticancer agent isodeoxyelephantopin can induce protective autophagy in lung cancer cells via the Nrf2-p62-keap1 pathway . The Nrf2 activators, DMF and the synthetic oleanane triterpenoids, activate the Nrf2 pathway as well as regulate different subsets of Nrf2 target genes and Nrf2independent genes in lung cancer (Chian et al., 2014;To et al., 2015). viii) In colon cancer, one study found that anticancer agent sulforaphane (SFN) can activate Nrf2 signaling to suppress human colon cancer (Johnson et al., 2017). Also, taxifolin (TAX) can induce antioxidant response pathway and enhance level of Nrf2 protein, and act as effective chemopreventive agent capable of modulating inflammation in colon cancer (Manigandan et al., 2015). ix) In ovarian cancer, one study found that Nrf2 can mediate the response of cancer cells to the anti-HER2 drugs, trastuzumab and pertuzumab, in ovarian cancer cells (Khalil et al., 2016). Also, activation of Nrf2 pathway in ovarian cancer seems to be related to Keap1 mutations within highly conserved domains of Keap1 gene and that Nrf2 may serve as an important therapeutic target for novel drugs capable of preventing or reversing resistance to chemotherapy in ovarian cancer (Konstantinopoulos et al., 2011). x) In breast cancer, Nrf2 serves as a key regulator in chemotherapeutic resistance under hypoxia through ROS-Nrf2-GCLC-GSH pathway and can be a potential treatment for hypoxia induced drug resistance in breast cancer cells Syu et al., 2016). xi) In esophageal cancer, C-28 methyl ester of 2-cyano-3,12-dioxoolean-1,9-dien-28-oic acid (CDDO-Me) can protect the cells against oxidative stress via inhibition of ROS generation, while CDDO-Me at low micromolar concentrations induces apoptosis by increasing ROS and decreasing intracellular glutathione levels in esophageal squamous cancer cells . xii) In glioblastoma, there are many potent anti-cancer agents targeting Nrf2 signaling for chemotherapy and chemoresistance in glioblastoma (Zhu et al., 2014). xiii) In osteosarcoma, the bioengineered Nrf2-siRNA can effectively interfere with the Nrf2 signaling pathway to improve chemosensitivity of human cancer cells . Moreover, the PIM (proviral integration site for moloney murine leukemia virus) kinase inhibitors can reduce Nrf2 signaling and increase ROS to kill hypoxic tumor cells such as prostate cancer cells (PC4-LN4), colon cancer cells (HCT-116), and breast cancer cells (MB-MDA-231 and MB-MDA-231-ARE-Luc) (Warfel et al., 2016). One study shows that proteasome biogenesis is dependent on the Nrf2 transcriptional factor, thus proteasome inhibitors have been actively developed as potential anticancer drugs (Albornoz et al., 2019). Gallic acid (GA), Z-ligustilide (LIG), and senkyunolide A (SA) can individually or cooperatively target Nrf2/ARE pathway to prevent cancer . Therefore, it can be said that Keap1-Nrf2 signaling pathways have different roles at different stages of cancer (Leinonen et al., 2014;Furfaro et al., 2016;de la Vega et al., 2018). Multiple Nrf2 or Keap1 inhibitors have been reported; and some of them are in the stages of pre-and clinical trial towards the Nrf2 signaling for cancers. For example, sulforaphane can target Nrf2 and the Nrf2 target genes NQO1 and GCLC to prevent oral cancer, and a preclinical trail has been performed to study its chemopreventive activity for oral cancer (Bauman et al., 2016). A single centre, single arm prospective phase II clinical trial has been performed for phytosome complex of curcumin targeting Nrf2 signaling as a the complementary therapy of gemcitabine on pancreatic cancer (Pastorelli et al., 2018). However, none of these Nrf2 or Keap1 inhibitors have currently entered into real clinical applications, which suggests that the sole inhibition of Nrf2 might not be sufficient for anticancer. A rational combination of Nrf2 inhibitors with other chemical agents would be a better strategy to treat cancers .

POTENTIAL OF TARGETING NRF2 SIGNALING AS NEW THERAPEUTIC STRATEGY FOR PAs
As described above, many omics studies in human PA tissues and experimental studies in PA cells and animal models demonstrate that oxidative stress and oxidative damage is the important hallmark of PA pathogenesis. Nrf2-mediated oxidative stress response signaling pathways are at the heart of oxidative stress response, and many chemical agents targeting Nrf2 signaling pathways have been developed and tested as potential anticancer drugs for different cancers. This clearly demonstrates the potential of targeting Nrf2 signaling pathways as new therapeutic strategies for PAs. However, the use of Nrf2 signaling as a therapeutic target for PAs has not been studied. We strongly believe that the Nrf2-mediated oxidative stress response signaling pathways are the promising targets for novel therapeutic strategies for PAs. Furthermore, MAPK signaling pathways including ERK, JNK, and p38 MAPK clearly regulate Nrf2 signaling (Figure 1). Moreover, MAPK signaling pathways have been recognized as potential therapeutic targets for PAs . The combined use of Nrf2 inhibitors targeting Nrf2 signaling and ERK inhibitors [e.g., somatostatin analogs pasireotide (SOM230) and octreotide (OCT), or dopamine] plus p38 activators (e.g., cabergoline, bromocriptine, and fulvestrant) or JNK activators (e.g., ursolic acid, UA) targeting MAPK signaling pathways  might produce better anti-tumor effects on PAs. In addition, oxidative stress is tightly associated with mitochondrial dysfunctions, both operate in PAs (Zhan and Desiderio, 2010a;Li and Zhan, 2019;Long et al., 2019). Some drugs targeting mitochondria are also recognized as a therapeutic strategy for PAs, including pyrimethamine, temozolomide, melatonin, melatonin inhibitors, gossypol acetate, 18 beta-glycyrrhetinic acid, T-2 toxin, Yougui pill, cyclosporine A, grifolic acid, paeoniflorin, and dopamine agonists . Therefore, the combined use of Nrf2 inhibitors targeting Nrf2 signaling and drugs targeting metochondria could be another way to generate better anti-tumor effects on PAs.

CONCLUSION
Pituitary adenoma (PA) is a common and important disease that occurs in the hypothalamic-pituitary-target organ axis system and seriously affects human endocrine system and health. The imbalance between oxidative stress and the antioxidant defense system is an important pathophysiological characteristic in PAs, which has been evidenced by many omics analysis in PA tissues and experimental studies in PA cells and animal models. Nrf2 signaling is at the heart of oxidative stress response signaling pathways. Multiple anticancer agents targeting Nrf2-mediated oxidative stress response pathways have been developed and tested as potential therapeutic drugs for different cancers. However, Nrf2 signaling and targeting Nrf2 signaling as a therapeutic strategy has not yet been extensively studied in PAs. We strongly recommend the emphasis on in-depth studies of Nrf2 signaling and potential therapeutic agents targeting Nrf2 signaling pathways in PAs. Furthermore, the combined use of Nrf2 inhibitors targeting Nrf2 signaling and ERK inhibitors plus p38 activators or JNK activators targeting MAPK signaling pathways, or drugs targeting mitochondria dysfunction pathway might produce better anti-tumor effects on PAs.

AUTHOR CONTRIBUTIONS
XZ conceived the concept, collected and analyzed literature, designed, coordinated, wrote and revised manuscript, and was responsible for its financial supports and the corresponding works. JL and TZ participated in literature collection and analysis, and prepared figures. All authors approved the final manuscript.

FUNDING
The financial supports from the Shandong First Medical University Talent Introduction Funds (to XZ), and the Hunan Provincial Hundred Talent Plan (to XZ).