Chemicals and reagents
COS 2 (purity > 95%, Additional file 1: Figure S1a and c) and COS 3 (purity > 95%, Additional file 1: Figure S1b and d) were prepared in our laboratory, and their purities were measured and confirmed with HPLC-ELSD (Ji et al. 2021a, b). Acetonitrile and methanol were purchased from Fisher Scientific (Fair Lawn, NJ, USA). HPLC–MS-grade ammonia solution (25%, v/v) was purchased from Aladin (Shanghai, China). Ultrapure water was obtained from the Milli-Q system (Millipore, Bedford, MA, USA).
Instrumentation and analytical conditions
The samples were analyzed on a LC-30A liquid chromatography system with a high-pressure LC-30AD pump, a SIL-30AC autosampler, and a CTO-30A column oven (Shimadzu Inc., Japan). The column eluates were analyzed using a Q-Exactive Plus mass spectrometer (MS) equipped with an Orbitrap mass analyzer with a heated electrospray ionization source (Thermo Fisher Scientific Inc., USA). The sequence and data analysis were performed using Xcalibur Qual Browser (Thermo Fisher Scientific Inc.).
The separation of the compound was performed using an ACQUITY UPLC Glycan BEH amide column (100 mm × 2.1 mm, 1.7 µm, Waters Corporation, MA, USA) combined with a Van Guard pre-column (5 mm × 2.1 mm, 1.7 µm, Waters Corporation) at 45 °C. The optimized method used a gradient mobile phase consisting of 0.1% (v/v) ammoniacal aqueous solution (solvent A) and 0.1% (v/v) ammonia in acetonitrile (solvent B). A flow rate of 0.2 mL/min was used with 2 µL of injection volume. The gradient conditions were 0.0–10.0 min, 75%–45% B; 10.0–10.1 min, 75% B; and 10.1–20 min 75% B to stabilize the initial conditions. The total run time was 10 min, and the post-delay time for reconditioning the column with 75% B was 10 min.
For the MS condition, the samples were detected in targeted-selected ion monitoring (t-SIM) for quantification with the m/z 341.1555 ([M + H]+) for COS 2 and m/z 502.2243 ([M + H]+) for COS 3, which operated in a positive ionization mode and the parameters settings were as described in the literature (Ardalani et al. 2021; Chitescu et al. 2015). High-purity nitrogen was used as both the ion source and collision gas. The quantitative parameters were as follows: resolution 70,000 FWHM (m/z 200) and automatic gain target 5 × 104. The extracted ion chromatograms were used for quantitation, selecting m/z 341.1555 (mass tolerance < 6 × 10–6) and m/z 502.2243 (mass tolerance < 6 × 10–6) as the quantitative ions of COS 2 and COS 3, respectively.
Preparation of standard solution and quality control sample
Stock solutions (1 mg/mL) of COS 2 and COS 3 for calibration and quality control were prepared in ultrapure water. A series of standard solutions were prepared, and equal amounts of COS 2 and COS 3 standard solutions were taken together to obtain a mixed standard solution.
To obtain the calibration curves, 5 µL samples of the mixed standard solution were mixed with 95 µL of blank serum or one of the tissue homogenate supernatants (liver, kidney, heart, lung, spleen, pancreas, cerebrum, and cerebellum). The COS 2 and COS 3 calibration curves consisted of seven non-zero concentrations in the range of 0.002–2.56 µg/mL and 0.020–5.12 µg/mL in serum and tissue homogenate supernatants. Quality control (QC) samples were independently prepared with blank serum at concentrations of 0.02 µg/mL (low), 0.20 µg/mL (medium), and 2.00 µg/mL (high) for COS 2 and 0.05 µg/mL (low), 0.50 µg/mL (medium), and 5.00 µg/mL (high) for COS 3. The high concentration serum samples were prepared by spiking COS 2 and COS 3 mixed standard solutions with blank serum to obtain concentrations of 2000 µg/mL and 200 µg/mL for COS 2, and 5000 µg/mL and 500 µg/mL for COS 3, which were 1000 times and 100 times higher than the high QC samples (2.00 µg/mL and 5.00 µg/mL for COS 2 and COS 3, respectively).
Sample processing
The initial sample preparation was based on procedures previously described in the literature (Wang et al. 2020), and we simplified the protocol to maintain high throughput. Briefly, 100 µL of rat serum of pharmacokinetic samples (or QC samples) and 300 µL of methanol–acetonitrile (1:1, v/v) were vortexed for 30 s, and then centrifuged at 13,000× g for 10 min to precipitate the proteins. The supernatant was filtered through a 0.22 µm nylon syringe filter, and the liquid was transferred to a sample vial, and a 2-µL aliquot was injected for UPLC–MS analysis. Tissue samples were added to the cooled saline (1:5, w/v) and two grinding beads and subsequently homogenized in a homogenizer at 60 Hz for 120 s. After 30 s of vigorous vortexing, the sample was sonicated in an ice water bath for 10 min, centrifuged at 3000×g for 10 min at 4 °C, and 100 µL of supernatant was collected. The subsequent steps were the same as those for the serum.
Method validation
The analytical method was validated according to U.S. Food and Drug Administration, Guidance for Industry–Bioanalytical Method Validation (Food and Drug Administration 2018).
Selectivity was assessed by comparing the chromatograms of six different batches of blank serum and tissue homogenization with the corresponding spiked serum and pharmacokinetic samples to ensure there were no interfering peaks (Jin et al. 2021).
A standard curve in the form of y = Ax + B was determined by plotting the peak areas of COS 2 and COS 3 against the known standard concentrations of COS 2 and COS 3. The slope, intercept, and coefficient of determination were estimated using the least squares linear regression method with a weighting of 1/x. The acceptance criterion was that the coefficient correlation (r2) must be greater than 0.990 (Zhou et al. 2021). The lower limit of quantification (LLOQ) is the lowest amount that can be quantitatively determined with a signal noise ratio 10:1.
The precision and accuracy of the method were evaluated by analyzing the COS 2 samples at concentrations of 0.02, 0.20, and 2.00 µg/mL and 0.05, 0.50, and 5.00 µg/mL for COS 3 samples. Five replicates on the same day were analyzed for the intra-day evaluation, and five replicates per day for two consecutive days were used for the inter-day evaluation to verify the repeatability of the method. The accuracy was obtained by comparing the measured values and theoretical values of the QC samples expressed as the relative error (RE). Acceptable levels of accuracy were 85%–115%. The relative standard deviation (RSD) was calculated to assess precision, which should be less than 15% (Zhu et al. 2020a, b).
The matrix effect and extraction recovery of COS 2 and COS 3 were determined at three levels in five replicates. The matrix effects were evaluated by comparing the peak area ratio of the post-extracted QC samples (spiked with analytes in the extracted analyte-free blank serum samples) to the peak area ratio of 75% acetonitrile solution. Recoveries were calculated by comparing the mean peak area of spiked QC samples with post-extracted QC samples (Ma and Wang 2021).
The stability of the analytes in rat serum was evaluated by investigating three QC concentrations in sextuplicate serum samples. The different storage and handling conditions were as follows: (a) autosampler tray stability at 10 °C for 24 h, and (b) freeze–thaw stability after three freeze–thaw cycles at − 20 °C (Penchala et al. 2013).
Dilution effects were used to determine the accuracy and precision of high concentration samples after dilution. Simulated high concentration serum samples were diluted 100-fold and 50-fold with post-dilution concentrations of 2 µg/mL for COS 2 and 5 µg/mL for COS 3. Blank serum was used for the dilutions and five replicates were performed. The RE and RSD values between the calculated and theoretical concentrations were measured to evaluate the accuracy and precision after dilution (Verougstraete et al. 2021).
Carry-over was performed in each analytical run by injecting three blank samples after the high QC sample (2 µg/mL for COS 2 and 5 µg/mL for COS 3). The outcomes of the carry-over effect should be less than or equal to 20% of the tested compounds (Ma and Wang 2021).
Application to the pharmacokinetic study
The pharmacokinetic study was approved by the Institutional Animal Care and Use Committees of SHCQ (permit number SHCQ-20200032). Male Wistar rats (weight, 220 ± 10 g) were obtained from Shanghai SLAC Laboratory Animal Co., Ltd. (license NO. SCXK-hu-20170005). Prior to the start of the experiment, all rats were housed in animal rooms (temperature 22 ± 2 °C, humidity 40 ± 10%, 12 h light/dark cycle) with free access to water and food for a week acclimation period. A total of 24 rats were randomly divided into four groups and deprived of food with free access to water for 12 h before the experiments, and 100 mg/kg and 500 mg/kg of COS 2 and of COS 3 in pure water were intragastrically administered to rats. For intravenous experiments, 10 rats were randomly divided into two groups, and COS 2 and COS 3 were administered via the tail vein at a dose of 100 mg/kg separately to non-fasted rats. Blood samples (200 µL) were collected from the tail vein at 0.08, 0.17, 0.25, 0.5, 0.45, 1, 2, 4, 8, 12, and 24 h after intragastric dosing and 0.08, 0.17, 0.25, 0.5, 1, 1.5, 2, 4, and 8 h after intravenous dosing. Blood samples were left overnight at 4 °C and centrifuged at 3000×g at 4 °C for 10 min to obtain the serum.
Tissue distribution study
A total of 60 rats were randomly divided into 12 groups (n = 5) and they were deprived of food with free access to water for 12 h prior to the experiments; six groups were intragastrically administered 500 mg/kg COS 2, and the remaining rats were treated with 500 mg/kg COS 3. Rats were killed at different times (0.25, 1, 2, 4, 8, and 12 h) by neck amputation and tissue samples (heart, liver, kidney, lung, spleen, pancreas, cerebellum, and brain) were collected. The tissue samples were rinsed with ice-cold 0.9% NaCl solution to remove blood or content and blotted on filter paper. Samples were snap-frozen in liquid nitrogen and stored at − 80 °C until subsequent processing.
Statistical analysis
Each evaluation was repeated at least three times, and the results are expressed as mean values with standard deviation (SD). Charts were prepared using GraphPad Prism 7 (GraphPad Software Inc., San Diego, CA, USA). The pharmacokinetic parameters for each rat were estimated using Drug and Statistics software version 2.0. Non-compartmental analysis was used to determine the pharmacokinetic parameters including maximum serum concentration (Cmax), the area under the serum concentration–time curve (AUC), the half-life (t1/2Z), the mean residence time (MRT), clearance (ClZ/F), and apparent volume of distribution (VZ/F) of COS 2 and COS 3.