Animals

Male Wistar Hannover [Crl:WI (Han)] rats (Rattus norvegicus L.) were obtained from the animal facility of NOVA Medical School. The cages had corncob bedding and were changed weekly. The animals were maintained under standard laboratory conditions, i.e., artificial 12 h light/dark cycles (lights on at 8 a.m.), at room temperature (22 ± 2.0 °C) and a relative humidity of 60 ± 10%. Rats were given standard maintenance laboratory diet, in form of dried pellets (Special Diet Service-SDS diets RM1) and reverse osmosis water in drinking bottles, both ad libitum.

All applicable institutional and governmental regulations concerning the ethical use of animals in research were followed, according to the NIH Principles of Laboratory Animal Care (NIH Publication 85-23, revised 1985), the European guidelines for the protection of animals used for scientific purposes (European Union Directive 2010/63/EU) and the Portuguese regulation and laws on the protection of animals used for scientific purposes (Law nº 113/2013). All experimental procedures were approved by the Ethical Committee of the NOVA Medical School for the animal care and use in research (protocol nº 15/2017/CEFCM) and by the Portuguese General Directorate for Animal Health (DGAV—Direcção-Geral de Alimentação e Veterinária).

Chronic intermittent hypoxia (CIH) animal model

The paradigm of CIH employed here was previously described and validated by our group (Diogo et al. 2015; Coelho et al. 2020, 2021; Pimpão et al. 2023) and others (Diogo and Monteiro 2014; Arnaud et al. 2023). Briefly, the animals were housed in a eucapnic atmosphere within medium A-chambers (A-60274-P, Biospherix Ltd) and exposed to 10.5 h of intermittent hypoxia (IH) per day. The chambers were equipped with gas injectors and sensors to monitor O2 and CO2 levels, ensuring the accuracy of the CIH cycles. Oxygen concentration inside the chambers was regulated by electronically controlled solenoid switches, which adjusted the flow of 100% N2 and 100% O2 gas through a three-channel gas mixer, gradually reducing the oxygen from 21 to 5% (OxyCycler AT series, Biospherix Ltd.). A CIH cycle involved the infusion of 100% N2 for 3.5 min to rapidly reduce O2 concentration to ~ 5–6%, followed by an infusion of 100% O2 for 7 min to restore O2 to normal levels (~ 21%). Each cycle lasted 10.5 min (5.6 CIH cycles per hour) and occurred during the inactive period of the animals (9:30 AM–8:00 PM). During the remaining hours of the day, the chambers were ventilated with a continuous flow of room air (21% O2).

Study design

Male Wistar rats were exposed to CIH for 2, 3, 5 and 9 weeks (n = 5 per group) to evaluate the temporal impact of CIH in ketamine biotransformation. Animals maintained under normoxic conditions (Nx) (79% N2 and 21% O2) for each time point of the study were used as controls (n = 5 per group).

To evaluate the role of AhR signaling in ketamine biotransformation, animals were subjected to CIH for 3 weeks, starting on the 22nd day, daily administration of CH-223191 (5 mg/kg/day in vegetable oil by oral gavage). This design was selected because in the first 3 weeks, animals present an increase in blood pressure with the duration of CIH exposure, reaching a plateau up to week 5. Rats were maintained under this treatment and in CIH condition for the subsequent 2 weeks. Animals under Nx or Nx treated with CH-223191 were used as controls (n = 3 per group).

Ketamine/medetomidine administration and tissue collection

At the end of the experiments, rats were intraperitoneally injected with a solution of ketamine (75 mg/kg body weight; Imalgene 1000, Boehringer Ingelheim Animal Health, France) and medetomidine (0.5 mg/kg body weight; Domitor, Pfizer Animal Health, New Zealand). After the anesthetics administration, blood was drawn by cardiac puncture without thoracotomy and tissues were collected. Death was confirmed by cervical dislocation. Both liver and kidney cortex tissues were then collected, snap-frozen in liquid nitrogen and all samples were stored at –80 °C until analyzed.

Metabolomics analysisMetabolite extraction

Before metabolite extraction, samples were randomized. Briefly, 200 µL of cold methanol/water (75%) solution was added to each 40 mg of tissue. Samples were vortexed and sonicated using an ultrasound bath at 4 °C for 60 min to enhance metabolite extraction. Then, samples were centrifuged at 13,200 rpm for 10 min at 4 °C and 160 µL of the supernatant was transferred to a new microcentrifuge tube for dry vacuum. Dried samples were reconstituted with 10% acetonitrile in water in a final volume of 600 µL. A volume of 5 µL of each reconstituted sample was pooled together (quality control pool; QCpool).

Liquid chromatography–high-resolution mass spectrometry (LC-HRMS)

Samples were analyzed by liquid chromatography (Elute UHPLC, Bruker, Bremen, Germany) interfaced with a Bruker Impact II quadrupole time-of-flight mass spectrometer equipped with an electrospray source (Bruker Daltoniks, Bremen). Chromatographic separation was performed on a LUNA C18 column (150 mm × 2.0 mm; 3.0 μm, Phenomenex) (Phenomenex) at 45 °C. A flow rate of 150 μL/min was used and the eluent system consisted on formic acid 0.1% (phase A) and acetonitrile (phase B), with the following gradient program: 1.5 min 5% phase B, then in 13 min to 100% phase B, held for 3.5 min at 100% phase B, subsequently in 0.25 min to 5% phase B and held for 1.75 min at 5% phase B. Calibration was performed by high-precision calibration mode (HPC) on the internal standard segment, consisting of sodium formate solution introduced at the beginning of each analysis. The mass spectrometer was operated in positive ionization mode on the full scan mode and data were acquired in the mass range from m/z 50 to 1000 with a spectra rate of 1 Hz. The capillary was set at 4.5 kV, the End Plate offset at 500 V, the Nebulizer gas (N2) at 40 psi and the Dry gas (N2) at 8 L/min at 200 °C. To evaluate the performance of the instrument, QCpools were injected with every 7 samples. QCpools were also acquired in DDA mode on the same instrument, with an isolation window of 0.5, acquisition rate of 3 Hz and a fixed cycle time of 3 s.

Targeted peak detection of ketamine and ketamine metabolites and medetomidine

Ketamine, ketamine metabolites and medetomidine were identified from the expected m/z values of the precursor ions and the product ions published elsewhere (Turfus et al. 2009; Bijlsma et al. 2011; Dinis-Oliveira 2017) and/or available in Human Metabolome Database. LC–MS files were converted to mzXML files using the ProteoWizard MSConvert software (Chambers et al. 2012). A targeted analysis was then performed with the open-source software MZmine3 (Katajamaa et al. 2006; Pluskal et al. 2010), which consisted of target peak detection and peak matching.

Targeted peak detection was performed with the list of the corresponding m/z and retention time values for each precursor compound (Table S1) and the following parameters: shape tolerance = 50%, noise level = 100, m/z tolerance = 0.005 Da or 15 ppm and retention time tolerance = 0.5 min. Peak matching among samples was performed using the Join aligner algorithm with m/z tolerance = 0.005 Da or 15 ppm, retention time tolerance = 0.5 min, weight for m/z and retention time = 1 and required same identification.

Untargeted LC–MS data preprocessing

Kidney and liver LC–MS data were preprocessed separately using XCMS 3.6.0 using XCMS 3.6.0 (Smith et al. 2006; Tautenhahn et al. 2008) in R environment (http://www.r-project.org/, R version 4.2.2). Preprocessing consisted of peak picking, retention time alignment, peak grouping and gap filling. Peak picking was performed with the centwave algorithm (Tautenhahn et al. 2008) and the following parameters: ppm = 20, peak width = 10–50 s, prefilter = 6, 10,000. Retention time alignment was performed against the average of the QCpools using the Obiwarp method with a m/z width = 0.01. Peak grouping was performed with the following peak density parameters: band width = 30, m/z width = 0.01, minimum fraction = 0.5. Gap filling was performed with a fixed retention time deviation = 15 s. Ions were further filtered according to the relative standard deviation of ions in QCpools (≤ 30%). The final dataset contained a total of 399 and 494 features (peaks with specific retention time and m/z values) with their corresponding m/z, retention time and peak area for the kidney and liver, respectively. The signals corresponding to ketamine, medetomidine and related compounds (isotopes, metabolites, fragments, adducts) were excluded from the dataset. To do that, the features corresponding to ketamine and medetomidine were identified and correlated with the rest of the ions present in the dataset. Those highly correlated features (correlation coefficient ≥ 0.8) were excluded from the dataset. Data were normalized by total area and centered and unit-variance scaled before statistical analysis.

Statistical analysis

Results are presented as mean ± standard error of the mean (S.E.M.). The normality of distributions was assessed using the Shapiro–Wilk test and comparison of unpaired data was performed using Student’s t test or Mann–Whitney test, using GraphPad Prism software version 8 (GraphPad Software, San Diego, CA, USA). p values < 0.05 were considered significant. Additionally, Principal Component Analysis (PCA), Partial Least Square Analysis (PLS) and Partial Least Square Discriminant Analysis (PLS-DA) were performed using SIMCA software (MKS Umetrics, Umeå, Sweden). Details on the datasets used to perform each analysis are provided in figure’s captions.