Ferroptosis as a potential molecular mechanism of bipolar disorder

Abstract

The unclear pathogenesis of bipolar disorder (BD) poses a challenge, especially with the striking rates of comorbid medical and psychiatric disorders, treatment resistance, and premature mortality in the absence of a specific diagnostic marker. We put forward the hypothesis of ferroptosis, a recently identified iron-dependent cell death, as a potential underlying mechanism of BD. We aimed to portray the possibility of ferroptosis involvement in BD pathogenesis as a doorway to encourage both animal and clinical studies on the topic. Ferroptosis is associated with multiple psychiatric disorders, including major depressive disorder, stress-induced anxiety, post-traumatic stress disorder, autism spectrum disorder, and alcohol use disorder. In addition, ferroptosis-related genes have been identified in schizophrenia, which shares genetic liabilities with BD. One of the top five most significant genes in BD in a recent genome-wide association study, FADS 2, is involved in ferroptosis. The three hallmarks of ferroptosis intersect with the pathogenesis of BD, including iron dysregulation, lipid peroxidation, and the failure of antioxidant systems. Other pieces of BD pathogenesis, including inflammation, mitochondrial dysfunction, calcium dysregulation, neurotransmission disturbance, and affection of synaptic plasticity and myelination, are either a preface or an aftermath of iron dysregulation. Additionally, circadian rhythm abnormalities and hypothalamic-pituitary-adrenal axis disturbances in BD could be another point where ferroptosis and BD intersect. Moreover, some BD treatments, such as lithium, haloperidol, olanzapine, clozapine, valproic acid, and electroconvulsive therapy, show anti-ferroptosis action in other contexts. These observations present a strong case for ferroptosis as a potential underlying mechanism of BD. Therefore, we call for studies that address iron accumulation in the brain in BD patients, postmortem tissues, and BD animal models. We call for genetic studies to look for the genetic signature of ferroptosis in BD patients. In addition, we call for studies on different BD models to assess the expression of ferroptosis markers. Our hypothesis has substantial implications if validated, including the use of ferroptosis-related genes and ferroptosis markers as a prognostic marker for BD and a potential therapeutic target based on ferroptosis inhibitors.

Fig. 1. Ferroptosis machinery.

Ferroptosis is an iron-dependent cell death which is executed by lipid peroxidation. Iron dysregulation, antioxidant failure, and lipid peroxidation are the three hallmarks of ferroptosis. 1- Iron dysregulation: the labile iron pool (LIP) or free redox iron increases during ferroptosis through three ways. First, imported iron through the upregulated divalent metal transporter 1 (DMT1) and transferrin receptors-1 (TFR1) adds up to the LIP. Second, iron exportation is limited by degradation of the only iron exporter, ferroportin 1 (FPN1), increasing the LIP. Third, iron is released from its ferritin stores by the nuclear receptor co-activator 4 (NCOA4)-mediated ferritinophagy or ferritin degradation. The highly available Ferrous iron (Fe+2) invokes lipid peroxidation and therefore ferroptosis through a Fenton-like reaction. 2-Failure of antioxidant defence: accumulated reactive oxygen species (ROS) overwhelms antioxidant defence system. During ferroptosis, the Solute carrier family 7 member 11 (SLC7A11) is downregulated by the decrease of nuclear factor erythroid 2-related factor 2 (NrF2) and the increase of P53, which prevents cystine entry into the cell and therefore decrease the formation of glutathione (GSH). 3- Lipid peroxidation: acyl-coenzyme A synthetase long-chain family member 4 (ACSL4) increases during ferroptosis and incorporates more polyunsaturated fatty acids (PUFA) into membrane phospholipid by converting them to PUFA-CoAs. ROS initiates a chain reaction of lipid peroxidation forming lipid hydroperoxides (PUFAOOH^•), which is amplified by the increased LIP, promoting ferroptosis. Fe⁺³ ferric iron, TF transferrin.

Fig. 2. Overview of intersection points between bipolar disorder and ferroptosis.

Our hypothesis of ferroptosis being implicated in the pathogenesis of bipolar disorder (BD) is supported by 1-Ferroptosis being involved in the pathogenesis of some of BD comorbidities, including Schizophrenia (SCZ), autism spectrum disorder (ASD), major depressive disorder (MDD), post-traumatic stress disorder (PTSD), and alcohol use disorder (AUD). 2-The existence of ferroptosis hallmarks in BD. 3- Hypothalamic-adrenal axis disturbance in BD pathogenesis, which could induce ferroptosis. 4- Circadian rhythm disturbance in BD, which is echoed by the ferroptosis-induced degradation of circadian rhythm transcriptional factors (clockophagy). 5- Drugs used in BD treatment have antiferroptosis effect in other contexts.

Fig. 3. Hypothetical mechanism of ferroptosis in bipolar disorder pathogenesis.

BD shows high glutamate levels, which can increase the LIP by increasing expression of DMT1 through the NF-κB/PKC. The increased LIP, which could happen through higher iron importation by TFR1 and DMT1 or less exportation by FPN1, can promote the cytotoxicity of dopamine byproducts which is already increased in BD. The increased LIP can promote lipid radical formation (PUFAOOH^•) through a Fenton-like reaction, which ends up in lipid peroxidation and ferroptosis. Mitochondrial dysfunction in BD can lead to overwhelming the antioxidant defence system with ROS production. The high dopamine levels in BD can also participate in the increased ROS levels. The antioxidant failure in BD can cause disturbance in GSH, GPX4, and NrF2 which leads eventually to high ROS, lipid peroxidation, and ferroptosis. BD shows decreased sensitivity of glucocorticoid receptors along with high basal levels of corticosterone, while dexamethasone can induce ferroptosis by activation of p53, which decreases the expression of SLC7A11 and, therefore, the entry of cystine and production of GSH. Epigenetic modulation of circadian transcriptional factors (BMAL1 and Clock) is a prominent event in BD and their degradation (clockophagy) induces ferroptosis. FADS2 is strongly implicated in both BD pathogenesis and ferroptosis induction. Multiple BD treatments inhibit ferroptosis. ECT inhibits ferroptosis by activating NrF2, which increases the expression of SLC7A11 and loads Fe⁺² into ferritin, decreasing the LIP. ECT also activates GPX4 and inhibits ACSL4 and lipid peroxidation. Lithium inhibits ferroptosis by inhibiting the GSK3-β that inhibits NrF2, increasing the expression of SLC7A11. Lithium inhibits lipid peroxidation and ferroptosis by inhibiting both COX-2 and GSK3- β. Carbamazepine inhibits lipid peroxidation and ferroptosis by inhibiting COX-2. Clozapine N oxide, a metabolite of clozapine inhibits ferroptosis through inhibition of NCOA4-mediated ferritinophagy, which decreases ferritin degradation and build-up of the LIP. Haloperidol can inhibit iron dysregulation and thereby ferroptosis. Valproic acid inhibits HDAC, which in turn inhibits the epigenetic modulation of ferroptosis genes and ferroptosis. N -acetylcysteine inhibits ferroptosis by increasing both SLC7A11 and GSH. Both olanzapine and haloperidol inhibit ferroptosis through ROS scavenging action. Lamotrigine decreases the availability of PUFA. BD bipolar disorder, LIP labile iron pool, Fe⁺² ferrous iron, Fe⁺³ ferric iron, DMT1 divalent metallic transporter 1, TFR1 transferrin receptor 1, TF transferrin, FPN1 ferroportin 1, NF-κB nuclear factor kappa B, PKC protein kinase C, NCOA4 nuclear receptor co-activator 4, ECT electroconvulsive therapy, NrF2 nuclear factor erythroid 2-related factor 2, HDAC histone deacetylase, BMAL1 brain and muscle ARNT-like 1 or aryl hydrocarbon receptor nuclear translocator-like protein 1, CLOCK circadian locomotor output cycles kaput, ROS reactive oxygen species, GSH glutathione, GPX4 glutathione peroxidase 4, SLC7A11 solute carrier family 7 member 11, ACSL4 acyl-coenzyme A synthetase long-chain family member 4, GSK3- β glycogen synthase kinase-3 beta, PUFA polyunsaturated fatty acid, FADS1/2 fatty acid desaturase1/2, COX-2 cyclooxygenase-2.