FKBP5 blockade may provide a new horizon for the treatment of stress‐associated disorders: An in‐silico study

Abstract

Objective

We searched for, from the FDA (Food and Drug Administration‐USA)‐approved drugs, inhibitors of FKBP5 with tolerable adverse effect profiles (eg, mild headache, sedation, etc.) and with the ability to cross the blood brain barrier (BBB), using bio‐informatics tools (in‐silico). This may pave the road for designing clinical trials of such drugs in patients with functional seizures (FS) and other stress‐associated disorders.

Methods

Several databases were used to find all the approved drugs that potentially have interactions with FKBP51 protein [ie, CTD gene‐chemical interaction section of FKBP51 protein of Harmonizome of Mayaanlab, DrugCenteral database, PDID (Protein Drug Interaction Database), DGIdb (the Drug Gene Interaction database)]. Other databases were also searched [eg, clinicaltrials.gov; DRUGBANK (the FASTA format of the FKBP51 protein was imported to the target sequencing section of the database to find the associated drugs), and the STITCH database (to find the related chemical interaction molecules)].

Results

After a comprehensive search of the designated databases, 28 unique and approved drugs were identified. Fluticasone propionate and Mifepristone and Ponatinib, Mirtazapine, Clozapine, Enzalutamide, Sertraline, Prednisolone, Fluoxetine, Dexamethasone, Clomipramine, Duloxetine, Citalopram, Chlorpromazine, Nefazodone, and Escitalopram are inhibitors of FKBP5 and have BBB permeability.

Significance

While the current in‐silico repurposing study could identify potential drugs (that are already approved and are widely available) for designing clinical trials in patients with stress‐associated disorders (eg, FS), any future clinical trial should consider the pharmacological profile of the desired drug and also the characteristics and comorbidities of the patients in order to foster a success.

Key Points.

• Several databases were used to identify drugs with FKBP5 blockade property.

• Sertraline, prednisolone, fluoxetine, and dexamethasone were acceptable candidates.

• This may pave the road to design trials in patients with functional seizures.

1. INTRODUCTION

Functional seizures (FS), also known as psychogenic non‐epileptic seizures (PNES) or dissociative seizures (DS), are common occurrences in epilepsy centers. ¹ , ² These seizures consist of sudden and paroxysmal changes in responsiveness, movements, or behavior that may superficially look like epileptic seizures, but are not associated with electrophysiological epileptic changes; they are associated with psychological problems and are often considered as a stress‐associated disorder. ³ , ⁴ , ⁵ Functional seizures have neurobiological bases that are different from those in epileptic seizures. ⁶ , ⁷ , ⁸ , ⁹ , ¹⁰ Across nations and cultures, people who suffer from FS continue to report detrimental effects on social aspects of their lives. ¹¹ , ¹² Furthermore, patients with FS have a standardized mortality rate that is 2.5 times higher than the general population; this is similar to those who have drug‐resistant epilepsy. ¹³ In spite of the significance of this condition, the scientific community cannot offer any promising and curative treatment strategy to these patients yet. ¹⁴

FK506‐binding protein 51 (FKBP5/FKBP51) is a gene that by encoding FKBP51 protein contributes to the regulation of glucocorticoid receptor sensitivity. Mutant variants of FKBP5 polymorphisms are associated with reduced glucocorticoid receptor sensitivity, which can lead to diminished hypothalamus‐pituitary–adrenal (HPA) axis negative feedback (ie, slower “off‐switch” for the stress response) and subsequent maintenance of glucocorticoids in the absence of any threats. ¹⁵ A systematic review and meta‐analysis found strong evidence of interactions between FKBP5 genotypes and early‐life stress (eg, sexual abuse), which could pose a significant risk for stress‐associated disorders such as major depression and post‐traumatic stress disorder (PTSD) later in life. ¹⁶ Sustained elevations of glucocorticoids (eg, due to the combination of FKBP5 polymorphisms and life stresses) may interfere with neurotrophic gene expression and protein synthesis across different brain regions, particularly within the hippocampus and prefrontal cortex, resulting in reduced neurogenesis and neuroplasticity, as well as degraded connectivity between these two regions. ¹⁷

Functional seizures are associated with life stresses (eg, sexual abuse or medical comorbidities), at least in many of these patients. ¹⁸ , ¹⁹ In one recent study, the authors showed that the hypothesis of the association between FKBP5 single nucleotide polymorphisms (SNPs) and FS is a plausible hypothesis. ²⁰ In that study, patients with FS had less GG and more AA genotypes in both rs9470080 and rs1360780 SNPs compared with healthy controls. However, they could not show whether FKBP5 polymorphisms were associated with FS in the absence of depression. ²⁰ Another study showed that patients with FS have basal hypercortisolism that is independent of the acute occurrence of seizures. ²¹ These two findings are in line with the studies that have provided evidence on the interactions between FKBP5 genotypes and early‐life stress to pose a risk for stress‐associated disorders (eg, major depression and PTSD). ¹⁶

Consequently and hypothetically, FKBP5 blockade may hold promise as a treatment option for stress‐associated disorders, including FS. Having hypothesized that repurposing approach can be an efficient way to discover the desired drugs for FKBP5 blockade; this method has lower costs and offers a shorter duration for drug development than the traditional methods of drug discovery. ²² In the current study, we aimed to investigate and discover, from the FDA (Food and Drug Administration‐USA)‐approved drugs, inhibitors of FKBP51 protein with tolerable adverse effect profiles (eg, mild headache, sedation, etc.) and with the ability to cross the blood brain barrier (BBB), using bio‐informatics tools (in‐silico). This may pave the road for designing future clinical trials of such drugs in patients with stress‐associated disorders (eg, FS).

2. MATERIALS AND METHODS

2.1. Compound screening strategy

Here, a “Drug” was defined as any chemical substance (other than food) that is used to treat or relieve symptoms of a disease or an abnormal condition. Several databases were used to find all the approved drugs [by the U.S. Food and Drug Administration (FDA)] that potentially have interactions with FKBP51 protein. The studied drugs have been reported to have FKBP5 blocking properties by previous studies. These drugs were collected from searching the CTD gene‐chemical interaction section of FKBP51 protein of Harmonizome of Mayaanlab, ²³ DrugCenteral database, ²⁴ PDID (Protein Drug Interaction Database), ²⁵ DGIdb (the Drug Gene Interaction database). ²⁶ Other databases were also searched [eg, clinicaltrials.gov; DRUGBANK ²⁷ (the FASTA format of the FKBP51 protein was imported to the target sequencing section of the database to find the associated drugs), and the STITCH database (to find the related chemical interaction molecules)]. ²⁸

2.2. Protein and ligand preparation and molecular docking

The structure of the protein was obtained from the RCSB PDB database available at https://www.rcsb.org/structure/5OMP with the best available resolution (1.88 A). Then, preparations for docking were done with UCSF Chimera 1.8.1. UCSF Chimera is a software for interactive visualization and analysis of molecular structures and the related data, including density maps, trajectories, and sequence alignments. The extra molecules, like ligands or other subunits, were deleted from the protein structure by using this software. The prep dock option was used to delete solvent and add hydrogens and charges to the protein. ²⁹ Canonical SMILES of suggested drugs were obtained from PubChem and then their energy was minimized using PyRx 0.8. Pyrx performs energy minimization was used for the selected molecules by OpenBabel application and also a graphical user interface for changing energy minimization parameters. PyRx is a virtual screening software for computational drug discovery that can be used to screen libraries of compounds against potential drug targets. ³⁰ Autodock was done using PyRx Vina Tool that considers different configurations of a single compound and measures binding affinity and lower band (LB) and upper band (UB) of root mean square deviation (RMSD) of each compound–protein interaction. LB and UB of RMSD show the differences in how the atoms are matched in the distance calculation of the docking process. Drugs that had binding affinities less than −5 and LB and UB of RMSD lower than 2.5 were selected as the probable antagonists for the protein.

2.3. In‐silico BBB permeability analysis

Neuroactive drugs should cross the BBB. In this study, the BBB permeability measures of different drugs were gathered from the DrugBank. In DrugBank, the result of the analysis based on Lipinski's rule of five is studied. ²⁷ This database uses the predicted properties of different drugs and small molecules from the admetSAR database. The absorption, distribution, metabolism, excretion, and toxicity (ADMET) structure–activity relationship database, abbreviated as admetSAR, is an open‐source searchable database that collects, curates, and manages available ADMET‐associated properties data from the published literature. The database includes 22 qualitative classification and 5 quantitative regression models with highly predictive accuracy, allowing to estimate ADMET properties for novel chemicals. AdmetSAR is accessible free of charge at http://www.admetexp.org. ³¹

2.4. Drug safety data

Since this study aimed to investigate the previously FDA‐approved drugs for a new purpose, the adverse effects of these drugs are already known. The lists of the adverse effects of the main candidate drugs were collected from the SIDER (side effect resource) database, ³² DRUGBANK, ²⁷ and Drugs.com. ³³

2.5. Standard protocol approvals, registrations, and patient consents

The Shiraz University of Medical Sciences Institutional Review Board approved this study.

3. RESULTS

3.1. Included molecules

After a comprehensive search of the designated databases, 28 unique and approved drugs were identified. The names of these small molecules and the resources from where they were retrieved are available in the Table S1.

3.2. Molecular docking results

The results of the best docking conformations of the identified drugs are available in Table 1. We arbitrarily classified the identified drugs into three groups: (1) excellent candidates (a binding affinity of less than −10 with an RMSD equal to zero): Rifampin, Tacrolimus, Fluticasone propionate, Mifepristone; (2) acceptable candidates (a binding affinity of more than −10 and less than −5, with an RMSD equal to zero): Ponatinib, Mirtazapine, Doxorubicin, Clozapine, Sirolimus, Enzalutamide, Sertraline, Cyclosporine, Prednisolone, Fluoxetine, Ritonavir, Dexamethasone, Clomipramine, Duloxetine, Paclitaxel, Citalopram, Chlorpromazine, Nefazodone, Escitalopram; and, (3) not desirable (a binding affinity of −5 or higher): Venlafaxine, Gemcitabine, Bupropion, Salicylic acid, and Leuprolide acetate.

TABLE 1.

The results of the docking and BBB studies of the potential FKBP5 inhibitors.

Drug name	Binding affinity	RMSD/UB	RMSD/LB	BBB permeability
Fluticasone propionate	−11.9	0	0	+
Mifepristone	−11	0	0	+
Ponatinib	−7.4	0	0	+
Mirtazapine	−6.7	0	0	+
Clozapine	−6.4	0	0	+
Enzalutamide	−6.2	0	0	+
Sertraline	−6.2	0	0	+
Prednisolone	−6	0	0	+
Fluoxetine	−6	0	0	+
Dexamethasone	−5.8	0	0	+
Clomipramine	−5.7	0	0	+
Duloxetine	−5.6	0	0	+
Citalopram	−5.5	0	0	+
Chlorpromazine	−5.5	0	0	+
Nefazodone	−5.4	0	0	+
Escitalopram	−5.4	0	0	+
Venlafaxine	−5	0	0	+
Gemcitabine	−4.8	0	0	+
Bupropion	−4.8	0	0	+
Salicylic acid	−4.7	0	0	+
Leuprolide acetate	−2.9	0	0	−
Rifampin	−14.4	0	0	−
Tacrolimus	−13.9	0	0	−
Doxorubicin	−6.7	0	0	−
Sirolimus	−6.3	0	0	−
Cyclosporine	−6.1	0	0	−
Ritonavir	−5.9	0	0	−
Paclitaxel	−5.5	0	0	−

Abbreviations: BBB, blood brain barrier; LB, lower band; RMSD, root mean square deviation; UB, upper band.

3.3. In‐silico BBB permeability analysis

From the excellent and acceptable drug candidates (during the docking process), 16 drugs passed the BBB permeability prediction (Table 1). Fluticasone propionate and Mifepristone (from the excellent candidates) and Ponatinib, Mirtazapine, Clozapine, Enzalutamide, Sertraline, Prednisolone, Fluoxetine, Dexamethasone, Clomipramine, Duloxetine, Citalopram, Chlorpromazine, Nefazodone, and Escitalopram (from the acceptable candidates) have BBB permeability.

3.4. Drug safety data

The most common adverse effects of the selected drugs are shown in Table 2. Mifepristone is used, together with another medication called misoprostol, to end an early pregnancy. ³⁴ FDA has determined an “Approved Risk Evaluation and Mitigation Strategies” for Mifepristone. Mifepristone has significant adverse effects [especially in women (eg, endometrial hypertrophy, heavy bleeding, pregnancy problems)]. ²⁷ , ³² Therefore, this is not a desirable drug. Fluticasone propionate is a glucocorticoid used to treat asthma, inflammatory pruritic dermatoses, and non‐allergic rhinitis. ³⁵ This drug is available as inhalers, nasal sprays, and topical treatments for various inflammatory indications. Intranasal bioavailability of fluticasone propionate is <2%, and the inhaled bioavailability is about 9%. The main adverse effects of fluticasone propionate are respiratory tract infections. ²⁷ , ³² For the most common adverse effects of other drugs, please refer to Table 2.

TABLE 2.

The most common adverse effects of the selected drugs.

Drug	Common adverse effects	Frequency (up to)
Fluticasone propionate	Candida infection	13%
	Lower respiratory tract infection	8%
	Upper respiratory tract infection	16%
Mifepristone	Fatigue	48%
	Headache	44%
	Endometrial hypertrophy	38%
	Blood potassium decreased	34%
	Arthralgia	30%
	Vomiting	26%
	Peripheral edema	26%
	Hypertension	24%
Ponatinib	Neutropenia	63%
	Hypertension	71%
	Platelet count decreased	47%
	Thrombocytopenia	47%
	Leukopenia	63%
Mirtazapine	Somnolence	54%
	Dry mouth	25%
	Increased appetite	17%
	Constipation	13%
Clozapine	Drowsiness	39%
	Somnolence	46%
	Salivation	31%
	Weight increased	56%
	Insomnia	33%
	Vertigo	27%
Enzalutamide	Asthenia	50%
	Back pain	26%
	Arthralgia	20%
	Diarrhea	21%
	Hot flush	20%
	Peripheral edema	15%
	Musculoskeletal pain	15%
Sertraline	Ejaculation problem (delayed)	14%
	Diarrhea	24%
	Insomnia	28%
	Fatigue	16%
	Dry mouth	16%
	Dizziness	17%
Prednisolone	Weight increased	10%
	Indigestion	10%
	Sweating	10%
	Anxiety and mood changes	10%
Fluoxetine	Rhinitis	23%
	Nausea	29%
	Insomnia	33%
	Fatigue	10%
	Headache	32%
	Asthenia	21%
Dexamethasone	Intraocular pressure increase	35%
	Conjunctival hemorrhage	30%
	Hypertension	12%
Clomipramine	Dry mouth	84%
	Somnolence	54%
	Dizziness	54%
	Fatigue	39%
	Tremor	54%
	Headache	52%
	Constipation	47%
Duloxetine	Nausea	30%
	Headache	20%
	Dry mouth	18%
	Constipation	15%
	Dizziness	17%
	Fatigue	16%
	Insomnia	16%
	Somnolence	21%
	Diarrhea	13%
Citalopram	Headache	26%
	Dry mouth	20%
	Ejaculation disorder	18%
	Insomnia	15%
	Somnolence	18%
Chlorpromazine	Leukopenia	30%
	Hemolytic anemia	10%
	Thrombocytopenia	10%
	Pancytopenia	10%
Nefazodone	Blurred vision	1%
	Clumsiness or unsteadiness	1%
	Lightheadedness and fainting	1%
	Ringing in the ears	1%
	Skin rash and itching	1%
Escitalopram	Dry mouth	10%
	Sweating	10%
	Insomnia	10%
	Somnolence	10%
	Fatigue	10%
	Dizziness	10%
	Feeling anxious or agitated	10%
	Painful urination	10%
	Nausea and constipation	10%

4. DISCUSSION

The results of this in‐silico study should not be viewed as clinically applicable yet; rather, this endeavor can be simply viewed as an approach to visualize and analyze the existing data to build a hypothesis. Most importantly, the relationship between FKBP5 polymorphisms and FS should be established in future studies. When such a relationship is established, then a search to find drugs that may provide FKBP5 blockade would be clinically more meaningful. However, considering the existing robust evidence on the relation between FKBP5 blockade and other stress‐associated disorders (eg, major depression and PTSD), ¹⁶ the current study adds to the literature. In the current in‐silico study, we could identify fluticasone propionate as a good candidate to deliver FKBP5 blockade. Prednisolone and Dexamethasone have also acceptable pharmacological profiles to be used as FKBP5 inhibitors. Some psychiatric drugs (eg, Mirtazapine, Sertraline, Fluoxetine, and Citalopram) also have acceptable pharmacological profiles to be used as FKBP5 inhibitors (Tables 1 and 2).

Currently, the best available therapeutic option for patients with FS is psychotherapy. However, a recent large clinical trial showed that cognitive behavioral therapy (CBT) plus standardized medical care had no statistically significant advantage compared with standardized medical care alone for the reduction of monthly seizures in these patients. ¹⁴ Considering the significant adverse effects and consequences of FS on patients' lives, ¹¹ , ¹² , ¹³ the suboptimal therapeutic effects of psychotherapy in these patients, ¹⁴ and also limited availability of such services (ie, psychotherapy) in many places in the world, ³⁶ the scientific community should try to discover more promising therapeutic options (eg, a drug) for the treatment of FS.

Many of the identified drugs in the current study could potentially be used in well‐designed randomized clinical trials in patients with FS. Fluticasone propionate is seemingly a good candidate, but whether the currently available dosage forms could affect the brain as a target is a matter of interest and question. Prednisolone and Dexamethasone have appropriate dosage forms to be used in disorders of the brain (ie, oral forms) and are already being used for various brain disorders (eg, brain tumor, Hashimoto encephalopathy, etc.). ³⁷ , ³⁸ However, corticosteroids are associated with some significant adverse effects (eg, weight gain, Cushing's syndrome, high blood sugar, mood changes, etc.), particularly if used for a long time. Mirtazapine, Sertraline, Fluoxetine, Citalopram, and other psychiatric drugs could also be used in such clinical trials. Important adverse effects of these drugs should be considered in their selection (eg, weight gain with Mirtazapine, sleep problems with Sertraline and Fluoxetine and Citalopram, loss in sexual ability with Sertraline, joint or muscle pain with Fluoxetine, loss in sexual ability with Citalopram, etc.). Previous studies have shown that many patients with FS also suffer from psychiatric comorbidities ³⁹ , ⁴⁰ ; using these psychiatric drugs in such clinical trials for patients with FS may also help with their psychiatric comorbidities and these psychiatric drugs could potentially provide dual effects of treating the psychiatric problems and also acting as FKBP5 inhibitors. However, a pilot treatment trial for FS with four arms [medication (sertraline) only, cognitive behavioral therapy informed psychotherapy (CBT‐ip) only, CBT‐ip with medication (sertraline), or treatment as usual] concluded that the sertraline‐only arm did not show a reduction in FS. ⁴¹

As mentioned earlier, FKBP5 exerts an inhibitory effect on glucocorticoid receptor (GR) signaling. Disinhibition of FKBP5 has been liked to various stress‐related illnesses. ¹⁵ Multiple molecular interactions have been explored for FKBP5, among which is its binding to heat‐shock protein 90 (Hsp90) and other co‐chaperones of GR complex that are of physiological importance. ⁴² Inhibition of FKBP5 by means of the proposed drugs in our study, may result in GR complex activation. ⁴³ Additionally, activation of GR complexes in the hippocampus, the brain region that suppresses HPA basal activity and reactivity to stress, provides a negative feedback regulation. ⁴⁴

In‐silico and computer‐aided drug design techniques can accelerate drug development processes by reducing the time, costs, and failures. While the current in‐silico repurposing study could identify potential drugs (that are already approved and are widely available) for designing clinical trials in patients with FS, any future clinical trial should consider the pharmacological profile of the desired drug and also the characteristics and comorbidities of the patients in order to foster a success. We should mention that the identified drugs in the current study are not the only candidates in the path towards discovering efficient therapies for FKBP5 blockade; there are other drugs in the pipeline (not approved yet) that provide selective inhibition of FKBP5 (eg, SaFit‐1 and SaFit‐2). ⁴⁵ Finally, we should re‐emphasize that the relationship between FKBP5 polymorphisms and FS should be established in future studies.

AUTHOR CONTRIBUTIONS

Ali A. Asadi‐Pooya, M.D.: Study conceptualization and design, manuscript preparation. Mahdi Malekpour: Study design, analyses, and manuscript preparation. Bardia Zamiri, Negar Firouzabadi, & Mohammad Kashkooli: analyses and manuscript preparation.

FUNDING INFORMATION

This research was supported by a grant from Shiraz University of Medical Sciences. The funding source had no involvement in study design, in the collection, analysis, and interpretation of data, in the writing of the report, and in the decision to submit the article for publication.

CONFLICT OF INTEREST STATEMENT

Ali A. Asadi‐Pooya: Honoraria from Cobel Daruo; Royalty: Oxford University Press (Book publication); Grant from the National Institute for Medical Research Development. Others: no conflict of interest.

Supporting information

Table S1

Asadi‐Pooya AA, Malekpour M, Zamiri B, Kashkooli M, Firouzabadi N. FKBP5 blockade may provide a new horizon for the treatment of stress‐associated disorders: An in‐silico study. Epilepsia Open. 2023;8:633–640. 10.1002/epi4.12749

DATA AVAILABILITY STATEMENT

Research data sharing is not applicable to this work.