Amiodarone Alters Cholesterol Biosynthesis through Tissue-Dependent Inhibition of Emopamil Binding Protein and Dehydrocholesterol Reductase 24

Abstract

Amiodarone is prescribed for the treatment and prevention of irregular heartbeats. Although effective in clinical practice, the long-term use of amiodarone has many unwanted side effects, including cardiac, pulmonary, hepatic, and neurological toxicities. Our objective was to elucidate effects of amiodarone exposure on the cholesterol metabolism in cultured neuronal and non-neuronal cells and in individuals taking amiodarone. We observed that amiodarone increases distinct cholesterol precursors in different cell types in a dose-dependent manner. In liver and kidney cell lines, amiodarone causes increase in desmosterol levels, and in primary cortical neurons and astrocytes, amiodarone increases zymosterol, zymostenol, and 8-dehydrocholesterol (8-DHC). We conclude that amiodarone inhibits two enzymes in the pathway, emopamil binding protein (EBP) and dehydrocholesterol reductase 24 (DHCR24). Cortical neurons and astrocytes are more sensitive to amiodarone than liver and kidney cell lines. We confirmed the inhibition of EBP enzyme by analyzing the sterol intermediates in EBP-deficient Neuro2a cells versus amiodarone-treated control Neuro2a cells. To determine if the cell culture experiments have clinical relevance, we analyzed serum samples from amiodarone users. We found that in patient serum samples containing detectable amount of amiodarone there are elevated levels of the sterol precursors zymosterol, 8-DHC, and desmosterol. This study illustrates the need for close monitoring of blood biochemistry during prolonged amiodarone use to minimize the risk of side effects.

Graphical Abstract

INTRODUCTION

Amiodarone is commonly prescribed for the treatment and prevention of arrhythmia (amiodarone label). The primary action of amiodarone is blocking potassium rectifier currents in the heart, which prolongs the myocardial cell action potential.^1–4 Other actions of amiodarone include effects on β-adrenergic receptors and calcium and sodium channels.^5–7 Amiodarone was the 198th most prescribed medication in the US in 2016 with 2.9 million prescriptions⁸ and is on the World Health Organization list of essential medicines, the most effective and safe medicines needed in a health system.⁹

The starting dose for amiodarone is 800–1600 mg taken orally in a single daily dose for 1–3 weeks. The dosage is then decreased to 100–200 mg per day in a single dose for maintenance.^10,11 The drug has a long elimination half-life at 58 days on average (25–100 days),¹¹ while its active metabolite mono-N-desethylamiodarone (MDEA) has a 36 day half-life, though it may not reach steady state levels for up to a year.^12,13 Both amiodarone and MDEA accumulate in adipose tissue as well as the liver and lungs.¹⁴ Due to the long half-life and this accumulation in multiple tissues, the adverse effects of amiodarone may take several months to abate upon cessation of the drug. In fact, the major risk factor for clinically significant toxic adverse effects identified by a retrospective study of amiodarone users’ medical records was duration of treatment, with 1.6–2.8% of users being affected.^15,16

The most commonly listed side effect is corneal microdeposits that occur in most users,^12,17 followed by cardiac and pulmonary toxicity, hypo- and hyperthyroidism, and liver toxicity.^18,19 Dermatologic effects such as blue skin discoloration and photosensitivity have also been reported (amiodarone label). Amiodarone crosses the blood–brain barrier (BBB) and long-term use leads to an increased incidence in neurological symptoms including cognitive impairment, tremors, gait ataxia, peripheral neuropathy, and in rare cases quadriplegia.^19–22 Although the mechanism of action in the nervous system is currently unknown, several reports implicated altered cholesterol biosynthesis.^23–25

The most affected pathway in rats treated with amiodarone was cholesterol biosynthesis.^23–25 This is an interesting observation considering the importance of cholesterol for maintaining membrane fluidity and composition that may affect the functioning of potassium channels and other membrane proteins.²⁶ In addition, due to the brain’s reliance on de novo cholesterol biosynthesis, perturbation of this process by prescription drugs is necessary to consider when attempting to understand the etiology of off-target effects.^27–29 Analysis of amiodarone’s effects on the accumulation of pathway intermediates may help to elucidate the mechanism whereby it inhibits cholesterol synthesis. In this study, we analyzed the effects of amiodarone on the production of sterol intermediates in several cell lines and in primary neuronal and glial cultures. To validate the clinical significance of our findings, we analyzed the sterol composition of serum from patients taking amiodarone.

RESULTS

Amiodarone Affects Post-lanosterol Cholesterol Biosynthesis in Neuro2a, HepG2 and HEK293 Cells.

Brain and whole body cholesterol homeostasis are distinct, and therefore we tested amiodarone’s effects on cholesterol synthesis in both neuronal and non-neuronal cells. Figure 1 shows a schematic of post-lanosterol cholesterol biosynthesis including intermediates measured in the present study.³⁰ We focus on lanosterol, the first intermediate with sterol ring structure, and several post-lanosterol sterols. Neuro2a (neuronal cell line), HEPG2 (liver), and HEK293 (kidney) cells were treated with amiodarone across a range of doses, and sterols were analyzed by LC-MS/MS (Figure 2). The consistent finding in all three cell lines was the increased level of 8-DHC in response to an increasing dose of amiodarone. The amiodarone treatment increased cellular levels of desmosterol in HEPG2 and HEK293 cells but not in Neuro2a cells. Note that the Neuro2a cells have already high levels of desmosterol compared to HEPG2 and HEK cells. Cellular cholesterol was not significantly altered. Increases in 8-DHC and desmosterol suggest that amiodarone inhibits enzymes emopamil binding protein (EBP) and dehydrocholesterol reductase 24 (DHCR24).

Figure 1.

Post-lanosterol cholesterol biosynthesis pathway. Chemical structures in blue are sterols analyzed by LC-MS/MS, structures in red are sterols analyzed by GC-MS, and the structure shown in black was not analyzed. Note that one enzyme can catalyze more than one enzymatic step. Drawing adapted from Honda et al., 2008.³⁰

Figure 2.

Amiodarone treatment in Neuro2a, HEPG2, and HEK293 alters sterol synthesis. Cholesterol, desmosterol, 7-DHC, and 8-DHC levels were measured by LC-MS/MS in (A) Neuro2a, (B) HEK293, and (C) HEPG2. *P < 0.05; **P < 0.01; ***P < 0.001; ****P < 0.0001.

Amiodarone Affects Post-lanosterol Cholesterol Biosynthesis in Neuronal and Glial Primary Cells.

We treated primary cortical neurons and astrocytes with increasing concentrations of amiodarone and analyzed levels of sterol intermediates by LC-MS/MS (Figure 3) and GC-MS (Figure 4). While cell viability was not affected over the range of concentrations used (data not shown), several sterol intermediates were affected by the treatment. Notably 8-DHC (Figure 3), zymostenol, and zymosterol (Figure 4) levels are significantly increased compared to vehicle (DMSO) treatment. In cortical neurons, cultured for 10 days, there was a significant increase in desmosterol and 7-DHC and greatly decreased levels of cholesterol (Figure 3). Amiodarone had a lesser effect on astrocytes: there was no change in desmosterol and no change in cholesterol. Although there was a significant increase of 7-DHC in both neurons and astrocytes, the baseline concentration of 7-DHC is smaller in cells compared to total amount of other sterols. The increases in zymostenol, zymosterol, and 8-DHC suggest that amiodarone inhibits EBP, the enzyme responsible for the conversion of zymosterol to 24-dehyrolathosterol, zymostenol to lathosterol, and 8-DHC into 7-DHC (Figure 1).

Figure 3.

Cholesterol, desmosterol, 7-DHC and 8-DHC levels were measured by LC-MS/MS in (A) primary cortical neurons (gray bars) and (B) astrocytes (green bars). Amiodarone has a dose-dependent effect on desmosterol, 7-DHC, 8-DHC, and lanosterol in both astrocytes and neurons. Changes in cholesterol are present in neurons but not astrocytes. **P < 0.01; *** P < 0.001; ****P < 0.0001.

Figure 4.

Amiodarone elevates zymosterol and zymostenol. Zymostenol and zymosterol levels were measured by GC-MS in (A) cortical neurons (gray) and (B) cortical astrocytes (green). Amiodarone has a dose-dependent effect on both zymosterol and zymostenol in both astrocytes and neurons, with astrocytes being more sensitive to the treatment. *P < 0.05; *** P < 0.001; ****P < 0.0001.

Amiodarone Inhibits the Sterol Biosynthetic Enzyme EBP.

To confirm that amiodarone treatment produces a similar sterol profile when compared to EBP inhibition, we generated an Ebp-knockdown Neuro2a cell line utilizing shRNA specific to Ebp. The mouse neuroblastoma-derived Neuro2a cells have been used extensively in cholesterol biosynthesis studies^{27–29,31,32} and Dhcr7-deficient Neuro2a cells were key in elucidating the molecular defects associated with faulty cholesterol synthesis.^32,33 Using LC-MS/MS and GC-MS methods to measure sterol intermediate levels, we compared Ebp-deficient Neuro2a cells with amiodarone treated Neuro2a cells (Figure 5). The cholesterol level is not significantly altered in both models; however zymosterol and zymostenol both are significantly increased. These chemical and genetic modes on Neuro2a cell inhibition both resulted in pronounced increases in zymostenol and to a lesser extent zymosterol.

Figure 5.

Comparison of genetic and chemical inhibition of EBP. Cholesterol, zymostenol, and zymosterol levels were measured by GC-MS in (A) Ebp-deficient Neuro2a cultured in defined media for 3 days and (B) Neuro2a treated with 1 μM amiodarone for 3 days. **P < 0.01; ***P < 0.001.

Amiodarone-Treated Patients Have Elevated Serum Levels of Zymosterol.

On the basis of the findings from our cell culture experiments, we aimed to determine if the changes in cholesterol biosynthesis are measurable in the serum of patients taking the medication. We obtained deidentified serum samples from individuals with amiodarone in their medical records, as well as from samples from age and sex matched control individuals. Samples from patients with amiodarone listed in the medical records were analyzed using LC-MS/MS to confirm the presence of the medication. We detected amiodarone and its metabolite, mono-N-desethylamiodarone (MDEA), in 24 samples (Figure 6A–C). Of these 24 samples, 13 were given amiodarone orally (Figure 6D,E, gray boxes), 8 received intravenous amiodarone (Figure 6D,E, red boxes), and the remaining 3 samples had unknown administration route (Figure 6D,E, black boxes). Samples that came from patients receiving amiodarone orally had higher average levels of amiodarone and MDEA. Amiodarone has a narrow therapeutic range (1.0–2.5 μg/mL),³⁴ and toxicity is linked to concentrations above 2.5 μg/mL.^35,36 The amounts we measured in the blood are below 1 μg/mL and yet correspond with abnormal levels of several sterols. Out of five sterols analyzed, desmosterol and 8-DHC were the most affected by amiodarone (Figure 7). Zymostenol was detectable only in 12 of 13 oral amiodarone samples and 7 out of 8 intravenous amiodarone samples. While zymostenol showed a trend toward elevation in response to treatment, zymosterol was significantly increased in the treated patients’ samples. It is important to take into consideration that amiodarone treatment can range from weeks to years.¹⁶ Unfortunately, the donor length of treatment was not available for these samples, but intravenous administration is primarily used for short-term management of serious arrhythmias, so it follows that those samples would have lesser effects reflected in their serum sterol levels.

Figure 6.

Amiodarone and MDEA are present in patient serum. Typical chromatograms and structures for (A) amiodarone, (B) MDEA, and (C) the internal standard d₈-aripiprazole. Amiodarone and MDEA were present in measurable quantities in 24 samples (D, E) that had amiodarone listed in their medical records. Red boxes indicate intravenous amiodarone administration, gray boxes indicate oral administration, and black boxes show unknown administration.

Figure 7.

Sterol measurements in human serum samples. Serum samples that were confirmed to contain amiodarone or MDEA were analyzed by LC-MS/MS for (A) cholesterol, (B) desmosterol, and (E) 8-DHC content and by GC-MS for (C) zymostenol and (D) zymosterol content along with control samples. Red boxes represent intravenous administration of amiodarone, gray boxes indicate oral intake, and black boxes show unknown administration of amiodarone. Zymosterol levels were below the limit of detection for all control samples. *P < 0.05; **P < 0.01.

DISCUSSION

The number of amiodarone prescriptions in the USA for 2006–2016 was 37 806 878 with yearly average of 3 436 989 (median 3 404 961) (https://clincalc.com/DrugStats/Drugs/AmiodaroneHydrochloride). It is effective but has many potential adverse effects. One of these is that amiodarone is considered a category D drug, indicating positive evidence of human fetal risk, and careful consideration is critical for the use of this drug during pregnancy.³⁷ Our results can be summarized as follows: (A) Amiodarone inhibition of EBP and DHCR24 enzymes is concentration and cell type dependent. At 500 nM concentrations, EBP is inhibited in all cells tested, and at 1 μM concentration, DHCR24 is inhibited. (B) In primary neuronal and astrocytic cultures, amiodarone inhibition of EBP enzyme leads to a dose-dependent increase in zymosterol and zymostenol, in addition to changes in desmosterol and cholesterol. (C) Genetic inhibition and chemical inhibition of EBP yield similar results in Neuro2a cells. (D) Patient samples with detectable levels of amiodarone or MDEA have elevated levels of desmosterol, 8-DHC, and zymosterol when compared to nontreated controls. Amiodarone measurements in the clinical population are in agreement with the study in the cell culture system, suggesting an effect on more than one enzyme.

Biochemical findings in cell culture systems and serum samples show that amiodarone inhibits two sterol enzymes, EBP and DHCR24. The strongest effect in neuronal cells is the inhibition of EBP. EBP, also known as sterol Δ8–Δ7 isomerase, catalyzes the shift of the double bond in the Bring of sterols from C_8–9 to C_7–8. Therefore, EBP mediates the conversion of zymosterol to 24-dehydrolathosterol and zymostenol to lathosterol and isomerization of 7-DHC to 8-DHC. Mutations in this enzyme lead to chondrodysplasia punctata, 2, X-linked dominant (CDPX2) https://omim.org/entry/302960. The disorder, which manifests almost exclusively in females because EBP mutations are usually lethal in males, is characterized by calcification of the bones, rhizomelic shortness, patchy alopecia, cataracts, and midface hypoplasia.^38,39 Biochemically it is characterized by increased levels of 8-DHC and zymostenol (8(9)-cholestenol).⁴⁰ Moebius and colleagues^41–44 analyzed the pharmacological properties of the EBP enzyme and found that, in vitro, EBP can bind with high affinity to structurally diverse drugs, including amiodarone.

At the cellular and molecular levels, since amiodarone is lipophilic, it easily accumulates in cells where it can cause mitochondrial damage and structural and functional disturbances in the late endosomes and lysosomes,^45–47 which may lead to lipid traffic jams. In the liver, amiodarone can cause nonalcoholic steatohepatitis.^48–50 Desmosterol levels in liver and serum correlate with the severity of steatosis and inflammation.⁵¹ Careful clinical evaluation and monitoring of the adverse side effects is warranted in long-term amiodarone use.

Numerous studies have tried to answer the question if plasma lipid levels or pharmaceutical manipulations affecting sterol homeostasis have an impact on cognitive functions and brain neurochemistry. While there is no definitive answer from analyzing cholesterol levels in human serum samples, based on our studies in transgenic mouse models of cholesterol disorders, it is clear that serum sterol levels reflect cholesterol biosynthesis in the brain and other tissues.^33,52,53 However, the magnitude of change is much more pronounced in tissues compared to serum, and the brain undergoes a much larger change than other organs. Also, elevated levels of sterol intermediates are more representative of the defect in cholesterol synthesis than the level of cholesterol itself. The significant increase of desmosterol in the serum samples reported in our study and the study by Simonen et al.^54,55 most likely reflects liver cholesterol synthesis.

Amiodarone is not unique in inhibiting more than one cholesterol synthesis enzyme. Our high-throughput screening experiments showed that clomiphene and tamoxifen inhibit both EBP and DHCR24.^27–29 A comparison of the chemical structures of four drugs is shown in Figure 8. It seems that a similar substructure present in amiodarone, clomiphene, and tamoxifen is also present in triparanol. Triparanol is an inhibitor of DHCR24 and was in clinical use in 1960s but was removed from the market due to its significant side effects.⁵⁶

Figure 8.

Chemical structures of triparanol and three drugs currently in clinical use.

Our study provides insight into the consequences of amiodarone’s inhibition of cholesterol synthesis and warrants careful monitoring of blood biochemistry in the users of amiodarone. Our study points toward potential chemical substructures that may inhibit sterol enzymes, and this should be pursued in follow-up studies.

METHODS

Chemicals.

Unless otherwise noted, all chemicals were purchased from Sigma-Aldrich Co (St. Louis, MO). HPLC grade solvents were purchased from Thermo Fisher Scientific Inc. (Waltham, MA). Amiodarone was obtained from Sigma-Aldrich and dissolved in DMSO for the cell culture experiments. All sterol standards, natural and isotopically labeled, used in this study are available from Kerafast, Inc. (Boston, MA).

Study Population.

Serum samples were obtained from the UNMC biobank that is made up of residual samples from patients who consent to donate any left-over after the laboratory testing. Electronic Health Record personnel identified a group of 94 samples from users of amiodarone and age and sex matched control group. Following LC-MS measurements of the amiodarone and MDEA, samples that had combined levels of amiodarone plus MDEA of 0.5 ng/mL or greater were used to analyze the sterol levels by LC-MS and GC-MS. The characteristics of control samples were 7 males and 6 females ages 23–45, and those of amiodarone users were 11 males and 13 females ages 15–45.

Cell Cultures.

Human hepatocellular carcinoma HEPG2 cells, human embryonic kidney HEK293 cells, and mouse neuroblastoma cell line Neuro2a were purchased from ATCC (Rockville, MD). Ebp-deficient Neuro2a cells were generated using pGIPZ mouse Ebp shRNA (Dharmacon) and electroporation with Lonza Nucleofector Technology. In order to generate a stable cell line expressing Ebp shRNA, puromycin was used to kill nontransfected cells using the optimal concentration as determined by checking the puromycin dose response curve. In addition to control Neuro2a cells, two additional control cell lines were generated: Neuro2a cells transfected with pGIPZ mouse nonsilencing shRNA and Neuro2a cells transfected with empty pGIPZ plasmid. The human and mouse cell lines were maintained in EMEM supplemented with L-glutamine, 10% FBS, and puromycin at 37 °C and 5% CO₂. Cells were subcultured once a week, and the culture medium was changed every 2 days. For experimental purposes, the cells were plated in 96-well plates (for cell viability and sterol analysis). To assess the endogenous sterol synthesis, these cultures were grown in defined medium without cholesterol and without lipids using EMEM with N2 supplement, l-glutamine, and puromycin. At the end point of the incubation, Hoechst dye was added to all wells in the 96-well plate, and the total number of cells were counted using an ImageXpress Pico and cell counting algorithm in CellReporterXpress. After removing the medium, wells were rinsed twice with 1× PBS and then stored at −80 °C for lipid analysis. All samples were analyzed within 2 weeks of freezing.

Primary Neuronal Cultures.

Primary cortical neuronal cultures were prepared from E15 and E18 mice as previously described.^57,58 Briefly, the brain was placed in prechilled HBSS solution (without Ca²⁺ or Mg²⁺), and two cortices were dissected, cut with scissors into small chunks of similar sizes, and transferred to Trypsin/EDTA (0.25%) for 25 min at 37 °C. Trypsin was removed and residual trypsin was inactivated by adding Trypsin Inhibitor (Sigma T6522) for 5 min. Solution was removed and small tissue chunks were resuspended in Neurobasal medium (NBM) with B-27 supplement (Gibco no. 17504-044). Samples were then spun at 80g; the pelleted tissue was resuspended in NBM with B-27 supplement and then triturated with a fire-polished Pasteur pipet. The cells were pelleted by centrifugation for 5 min at 80g. The cell pellet was resuspended in NBM with B-27 supplement, and the cells were counted. The cells were plated on poly(d-lysine) coated 96-well plates at 60 000 cells/well. The growth medium was NMB with B-27 supplement, Glutamax, and 3 μM cytosine arabinoside. Cells were incubated at 37 °C in 5% CO₂ for 6–10 days. At the end point of the incubation, Hoechst dye was added to all wells in the 96-well plate, and the total number of cells were counted using an ImageXpress Pico and cell counting algorithm in CellReporterXpress. After removing the medium, wells were rinsed twice with 1× PBS and then stored at −80 °C for lipid analysis. All samples were analyzed within 2 weeks of freezing. For the current study, we prepared three independent preparations of primary cortical neurons and astrocytes (one from E15 and two from E18). The results were concordant for cultures obtained from both embryonic stages and across three independent set of experiments.

Primary Astrocyte Cultures.

After plating the required number of cells for neuronal cultures, left-over cells were plated in 100 mm dishes at density of 10 × 10⁶ per tissue culture plate in DMEM with 10% FBS. Under these conditions, astrocytes adhere, divide, and completely populate the plate within 10–14 days. Once the plates were full, they were rinsed using the cold jet method.⁵⁹ The astrocytes were trypsinized and plated in 96-well plates in DMEM plus 10% FBS at 30 000 cells/well. The following day, the medium was completely changed, and astrocytes were grown in Neurobasal medium with B-27 supplement in the absence of cholesterol (same medium as neuronal cells without cytosine arabinoside). Cells were incubated at 37 °C in 5% CO₂ for 6 days. At the end point of the incubation, cells were counted as described for neuronal cultures and processed in the same way as neuronal cultures for analysis.

GC-MS and LC-MS/MS Methods.

Sterol levels were analyzed in individual wells of 96-well plate and, for most experiments, cellular levels correspond to 8–12 technical replicates. After rinsing plates with 1× PBS (or removing previously frozen plates from −80 °C freezer), 200 μL of MeOH containing the internal standard cocktail was added, as reported previously.³² The plate was placed on an orbital shaker for 30 min at room temperature. An aliquot (100 μL) of the supernatant was transferred to a PTAD-predeposited plate, sealed with Easy Pierce Heat Sealing Foil followed by 30 min agitation at room temperature, and analyzed by LC-MS/MS (as described in a following section). Values were normalized by average cell count and reported as fold change over control (DMSO for amiodarone, nonsilencing shRNA for Ebp-deficient Neuro2a).

For GC-MS analysis, single wells were extracted with 200 μL of MeOH containing an internal standard cocktail as previously described (d₇-cholesterol, d₇-7-DHC).³² 7 An aliquot (190 μL) of the supernatant was transferred to an Eppendorf tube. Three adjacent wells were combined in each tube before drying. Samples were dried in a SpeedVac concentrator. Dried lipid samples were derivatized with 35 μL of N,O-bis(trimethylsilyl)trifluoroacetamide (BSTFA), and 5 μL of sample was injected onto the column (SPB-5, 0.25 μm, 0.32 mm × 30 m) with the following temperature program: 180 °C was held for 1 min, then increased to 250 °C at 20 °C/min, then raised to 300 °C at 4 °C/min and kept for 7 min. Selected sterol intermediates were analyzed by extracted ion chromatogram (EIC) using the following values: cholesterol (m/z 458), zymostenol (m/z 458), zymosterol (m/z 456), and d₇-Chol (m/z 465). Final sterol numbers were calculated using d₇-Chol as the internal standard. Response factors were calculated for each sterol and used to calculate the total amount. Values were normalized by average cell count and reported as fold change over control (DMSO for amiodarone, nonsilencing shRNA for Ebp-deficient Neuro2a).

Chromatograms and mass spectra for pure standards, DMSO treated astocytes, and amiodarone treated astrocytes can be found in Supplemental Figure 1 (zymostenol) and Supplemental Figure 2 (zymosterol).

Serum Sterol Measurements: LC-MS/MS.

To 10 μL of human serum was added 800 μL of Folch solution containing 0.25 mg/mL TPP, 0.005% BHT, and the internal standards d₇-7-DHC (13 ng), d₇-8-DHC (30 ng), ¹³C₃-Des (100 ng), ¹³C₃-Lano (100 ng), and d₇-Chol (34 ng), followed by the addition of 400 μL of 0.9% NaCl. The resulting mixture was vortexed and centrifuged. The lower organic phase was recovered and dried in SpeedVac. Freshly prepared 4-phenyl-1,2,4-triazoline-3,5-dione (PTAD) solution (200 μL of 1 mg/mL in MeOH) was added to the residues of serum extracts, and the solutions were incubated for 30 min at room temperature with occasional shaking, transferred into sample vials, and placed in an Acquity UPLC system equipped with ANSI-compliant well plate holder coupled to a Thermo Scientific TSQ Quantis mass spectrometer equipped with an APCI source. Then 5 μL was injected onto the column (Phenomenex Luna Omega C18, 1.6 μm, 100 Å, 2.1 mm × 50 mm) with 100% MeOH (0.1% v/v acetic acid) mobile phase for 1.0 min run time at a flow rate of 500 μL/min. Natural sterols were analyzed by selective reaction monitoring (SRM) using the following transitions: Chol 369 → 369, 7-DHC 560 → 365, desmosterol 592 → 560, and lanosterol 634 → 602, with retention times of 0.7, 0.4, 0.3, and 0.3 min, respectively. SRMs for the internal standards were set to d₇-Chol 376 → 376, d₇-7-DHC 567 → 372, ¹³C₃-desmosterol 595 → 563, and ¹³C₃-lanosterol 637 → 605. Final sterol numbers are reported as nmol/μL of serum.

Serum Drug Extraction.

To 75 μL of human serum were added 188 μL of H₂O, 10 μL of internal standard (d₈-aripiprazole), and 75 μL of 4.5 M ammonium hydroxide. The samples were then vortexed vigorously for 1 min. Next, 950 μL of methyl tert-butyl ether (MTBE) was added to each sample followed by another minute of vortexing. Samples were then centrifuged at 10 000g for 10 min. The top organic layer was collected in a glass autosampler vial and dried in a SpeedVac. Next, 150 μL of MeOH/NH₃·H₂O (95:5, v/v) was added to each vial, and the vial was briefly vortexed and then transferred to a chromatography vial insert for LC-MS/MS analysis.

Serum Drug Measurements.

Amiodarone and desethylamiodarone levels were acquired in an Acquity UPLC system coupled to a Thermo Scientific TSQ Quantis mass spectrometer using an ESI source in the positive ion mode. Five microliters of each sample was injected onto the column (Phenomenex Luna Omega C18, 1.6 μm, 100 Å, 2.1 mm × 50 mm) using water (0.1% v/v acetic acid) (solvent A) and acetonitrile (0.1% v/v acetic acid) (solvent B) as mobile phase. The gradient was 10% to 40% B for 0.5 min, 40% to 95% B for 0.4 min, 95% B for 1.5 min, 95% to 10% B for 0.1 min, and 10% B for 0.5 min. Amiodarone and its metabolites were analyzed by selective reaction monitoring (SRM) using the following transitions: amiodarone 646 → 86, and mono-N-desethylamiodarone 618 → 72. The SRM for the internal standards (d₈-aripiprazole) was set to 456 → 293, and response factors were determined to accurately determine the drug levels. Final drug levels are reported as ng/μl of serum.

Statistics.

Statistical analyses were performed using GraphPad Prism 8 for Windows. Unpaired two-tailed t tests were performed for individual comparisons between two groups. The Welch’s correction was employed when the variances between the two groups were significantly different. Ordinary one-way ANOVA and multiple comparisons with Tukey corrections were performed for comparison between three or more groups. When all standard deviations were not equal as determined by the Brown–Forsythe test, Welch’s ANOVA was employed. A P value of <0.05 was considered statistically significant. The results are expressed as mean ± SE.