Sensitive profiling of phenols, bile acids, sterols, and eicosanoids in mammalian urine by large volume direct injection-online solid phase extraction-ultra high performance liquid chromatography-polarity switching tandem mass spectrometry

Abstract

Bile acids, sterols, and eicosanoids serve as important chemical categories in mammalian urine and, sometimes, as primary diagnostic markers for diverse diseases, whereas phenols, usually derived from the diet, are usually observed as annoying interferents in pharmacokinetic and metabolic profiling of phenolic derivatives. Precise analysis of these substances widely suffers from instability and trace distributions, as well as errors and uncertainties resulting from tedious sample preparation procedures. Herein, directly simultaneous determination of phenols, bile acids, sterols, and eicosanoids, in particular those trace ones, in mammalian urinary matrices was attempted using large volume direct injection-online solid phase extraction-ultra high performance liquid chromatography-polarity switching tandem mass spectrometry (LVDI-online SPE-UHPLC-psMS/MS). Large volumes (500 μL) of liquid-state samples were directly loaded onto an online SPE column via ten consecutive injections. The SPE column was responsible for accumulating targeted components at the loading phase (−4.5 to 5.0 min), while transmitting those trapped analytes into a Waters HSS T3 column at the elution phase (5.0–27.0 min). Phase switching was accomplished using an electronic 6-port/2-channel valve. Analyte detection of all analytes, 28 ones in total, was performed using both positive and negative multiple reaction monitoring modes. Various method validation assays demonstrated the developed method to be extremely sensitive (most limits of quantitation lower than 0.3 ng mL⁻¹), precise (all RSDs of intra- and inter-day variations lower than 15%) and accurate (recoveries ramped from 80.4% to 120.0% with RSDs lower than 20%). Significant variations occurred for urine samples from different species, as well as amongst urinary matrices from different individuals within the same group. The findings proved that the developed LVDI-online SPE-UHPLC-psMS/MS method takes advantage of detecting the trace components and preserving those mutable substances from degradation, thus offering a meaningful tool for widely targeted monitoring substances in biofluids.

---

1. Introduction

Direct analysis has been widely recognized as a feasible way to eliminate, to some extent, the errors and uncertainties that are usually caused by tedious sample preparation procedures, such as liquid–liquid extraction and protein precipitation.^1–3 Moreover, a noticeable portion of endogenous substances in biological samples face extensive chemical degradation because of the exposure to light or organic solvents during sample preparation. As a consequence, it is emergent to propose directly analytical approaches to profile the real chemical compositions of biological samples.

Although fewer proteins are distributed in urinary samples in comparison with plasma and serum, it is still important to remove those inorganic salts as well as other non-volatile substances prior to their arrivals at the ion-source of mass spectrometer.^4,5 Combinatory online solid phase extraction (SPE) column, usually embedded with octadecyl silane particles, and column switching technique exhibits unique advantages for direct analysis of liquid-state samples by online capturing those semi-polar together with hydrophobic components at the meanwhile of expelling proteins along with hydrophilic compounds.⁶ As a consequence, this combination could fulfill the demands for the direct measurements of liquid-state biological samples, e.g., urine, plasma, and serum.

A number of endogenous components show trace, even ultra-trace, distributions in biofluids, and thus giving rise to a significant obstacle for their reliable detection and precise determination. It makes sense that increasing injection volume could advance the sensitivity owing to the subjection of enough amounts of analytes into ultra-high performance liquid chromatography coupled with tandem mass spectrometry (UHPLC-MS/MS) platform,^7–11 which is playing as a work-horse for either qualitative or quantitative characterization of biological samples. Fortunately, RP-C₁₈ SPE column is able to efficiently extract hydrophobic substances from aqueous fluids, and then those trapped analytes that usually focus at the front of the SPE column can be comprehensively loaded onto the analytical column by back-flushing SPE column without sacrificing the peak capacity. Therefore, the integration of large volume injection, online SPE, column switching, and LC-MS/MS techniques can be expected as a feasible solution to understand the real chemical compositions, especially those trace and/or instable substances, of biofluids.

Urinary samples are favored biological sources for metabolome characterization because they are non-invasive and more accessible, and also rich in diverse endogenous substances.^12–14 Bile acids (BAs) and sterols, together with eicosanoids usually serve as important chemical categories in the mammalian urine, and they intervene in a wide range of metabolic pathways and various physiological processes, for instance immunity and lipid absorption & metabolism.¹⁵ Moreover, those chemical clusters play key roles for diverse pharmacological activities and immune defense procedures, such as maintaining human secondary sex characteristics, assimilation, anti-anemia, anti-inflammation, antibiosis, etc.¹⁶ Therefore, the demand for the development of direct, robust, and sensitive method is quite urgent to understand the real distributions of BAs, sterols, and eicosanoids in biological samples, which could extensively benefit clinical diagnosis of various diseases. On the other side, phenols, usually derived from the diets, have been observed as primary interferences when pharmacokinetic and metabolic profiling of phenolic derivatives.^17,18 Thus, it is also critical to clarify the distributions of phenols, mainly isoflavonoids and flavonoids, in the biological samples to differentiate drug-derived metabolites from food-related substances.

Therefore, an attempt to propose a new method being qualified for directly and sensitively quantitative analysis of various targets, including phenols, bile acids, sterols, and eicosanoids, in urinary matrices was made in current study by integrating large volume direct injection (LVDI), online SPE, and UHPLC-MS/MS (Fig. 1). We envision that the developed platform could offer a promising tool for largely targeted analysis of substances in urine as well as some other biofluids.

Fig. 1 Connectivity sketch of the 6-port/2-channel switching valve controlling LVDI-online SPE-UHPLC-psMS/MS system. Loading phase: ten specimen aliquots delivered from the autosampler are loaded onto the SPE column using 5% aqueous ACN to capture all targeted analytes and to expel those hydrophilic and non-volatile substances, while the valve is maintained at A-channel; elution phase: the specimen fraction adsorbed onto the SPE column, including phenols, bile acids, sterols, and eicosanoids, is eluted using a gradient program and subsequent to MS/MS detection, and the valve is maintained at B-channel. Details are described at Section 2.4. LVDI-online SPE-UHPLC-psMS/MS.

2. Materials and methods

2.1. Chemicals

Arachidonic acid (AA) as well as its derivatives, including prostaglandin E₂ (PGE₂), prostaglandin D₂ (PGD₂), prostaglandin F_2α (PGF_2α), 20-hydroxy-eicosatetraenoic acid (20-HETE), 15-HETE, 12-HETE, and 5-HETE, were purchased from Cayman Chemicals (Ann Arbor, MI, USA). Taurocholic acid (TCA), chenotaurocholic acid (CTCA), taurocynocholic acid (TCCA), and testosterone, as well as genistein, daidzein, calycosin, kaempferol, isorhamnetin, and cygnocholic acid¹⁹ were supplied by the State Key Laboratory of Natural and Biomimetic Drugs, Peking University (Beijing, China). Cholic acid (CA), hyodeoxycholic acid (HDCA), ursodeoxycholic acid (UDCA), deoxycholic acid (DCA), chenodeoxycholic acid (CDCA), tauroursodeoxycholic acid (TUDCA), estradiol, cortisol, mestanolone, and cortisone were obtained from National Institute for the Control of Pharmaceutical and Biological Products (NICPBP, Beijing, China). Two internal standards (ISs), including ginsenoside Ro (IS1, purity > 98%) for negative mode and oxypeucedanin (IS2, purity > 98%) for positive ionization were also collected from NICPBP. Purities of all authentic compounds, 30 ones in total, were determined to be greater than 98% by UHPLC-MS/MS, whereas the structural confirmation was carried out via ¹H- and ¹³C-NMR spectroscopy.

Formic acid, dimethylsulfoxide (DMSO), methanol, and acetonitrile (ACN) were of HPLC grade and purchased from Merck (Darmstadt, Germany). Deionized water was prepared by Milli-Q plus water purification system (Millipore, Bedford, MA, USA). Phosphate buffer salt solution (PBS, pH 7.4) that was supplied by Thermo-Fisher (Logan, UT, USA) acted as analyte-free matrix.⁵ The other chemicals were of analytical grade and obtained commercially from Beijing Chemical Works (Beijing, China).

2.2. Sample collection

Human urinary samples (FH1–FH6 and MH1–MH6) were collected from six females and six males (aged from 24 to 28 years) in our laboratory. All experiments were performed in compliance with the relevant laws of the People's Republic of China, and the research protocol was approved by the Internal Ethical Committee of Beijing University of Chinese Medicine (Beijing, China). Written informed consents were obtained for all of the volunteers for urine collection. No restrictions about diet were taken into account and all of them were healthy, non-pregnant, and non-smokers. Moreover, six male Sprague-Dawley rats (MR1–MR6) were supplied by Vital River Laboratories (Beijing, China) and acclimated in laboratory for one week at temperature of 23 ± 1 °C with 12 h-light/dark cycle and 50% relative humidity. Afterwards, all animals were housed in separate metabolic cages and free access to tap water, while urinary samples were collected in ice-water bath from each animal over 0–24 h. Aliquots of all urinary samples were stored at −80 °C immediately after collection.

2.3. Sample preparation

Stock solution (100 mmol L⁻¹ for each) of each reference standard was prepared individually with methanol or DMSO depending on the solubility characteristic and stored at −20 °C until usage. Mixed standard stock solution was then obtained by pooling all stock solutions, and the obtained solution was sequentially diluted using DMSO to afford serial mixed standard solutions with desired concentration levels. Afterwards, 100-fold dilution was performed for each mixed standard solution with PBS solution containing IS1 and IS2 (2 μg mL⁻¹ for either) to yield a serial of calibration samples. Three concentration levels of calibration samples, including high, medium, and low levels, were selected as quality control (QC) samples. On the other hand, all urinary samples were also 100-fold diluted with PBS solution fortified with IS1 and IS2 (2 μg mL⁻¹ for either). Each urinary sample was prepared in triplicate. All calibration samples along with diluted urine samples were centrifuged at 13 000 rpm for 10 min at 4 °C, and each supernatant was immediately subjected onto LVDI-online SPE-UHPLC-psMS/MS platform.

Moreover, stock solution of each reference standard was diluted to appropriate concentration (100–500 ng mL⁻¹) with 70% aqueous methanol and directly infused (flow rate, 7 μL mL⁻¹) into the ion source of a hybrid triple quadrupole-linear ion trap mass spectrometer (Qtrap-MS) via a syringe pump for manual mass parameter optimization.

2.4. LVDI-online SPE-UHPLC-psMS/MS analysis

UHPLC was conducted on a Shimadzu UHPLC system (Kyoto, Japan) consisting of two LC-20AD_XR pumps (pumps A and B), a LC-20AD solvent delivery unit (pump C), DGU-20A3R degasser, and a CBM-20A controller. An ABSciex 5500 Qtrap mass spectrometer (Foster City, CA, USA), mounted with an electrospray ionization (ESI) interface as well as an electronic 6-port/2-channel valve, was utilized to connect with the UHPLC system. The generic layout of instrumentation setup proposed previously was introduced in current study with minor modifications,^5,6,20 and the schematic is elucidated in Fig. 1. Analyst Software package (Version 1.6.2, ABSciex) was implemented to control and synchronize the entire system, and also for data acquisition and processing.

A single analytical run was divided into two phases, namely loading and elution phases, by alternating the electronic valve (Fig. 1A and B). At loading phase (−4.5 to 5.0 min, Fig. 1A), auto-sampler was maintained at 4 °C and responsible for injecting 50 μL of sample for ten times (0.5 min for each sampling). The LC-20AD pump was responsible for delivering 5% aqueous ACN to the SPE column (Security Guard™, C₁₈ 3.0 × 4.0 mm, Phenomenex, Torrance, CA, USA) at a flow rate of 1.0 mL min⁻¹ aiming to facilitate those lipophilic constituents being adsorbed onto the SPE column, while flushing the hydrophilic substances into waste. At elution phase (5.0–27.0 min, Fig. 1B), those trapped components were back-flushed from SPE column into the analytical column (HSS T3 2.1 × 100 mm, 1.8 μm, Waters, Milford, CT, USA) using a programmed gradient elution. Within a single measurement, the mobile phase of the analytical column consisted of 0.1% aqueous formic acid (A) and acetonitrile containing 0.1% formic acid (B), and was delivered by the two LC-20AD_XR pumps (pumps A and B, Fig. 1) in gradient at a total flow rate of 0.3 mL min⁻¹ as follows: −4.5 to 5 min, 30% B; 5–10 min, 30–40% B; 10–12 min, 40% B; 12–15 min, 40–45% B; 15–18 min, 45–70% B; 18–20 min, 70–85% B; 20–21 min, 85–100% B; 21–22 min, 100% B; 22–22.01 min, 30% B; and 22.01–27 min, 30% B. The SPE column was maintained at room temperature (25 °C), whereas the analytical column was maintained at 50 °C in the thermal column oven. At the end of each run, the whole system was switched to the initial status and maintained for another two minutes to re-equilibrate both columns and all tubes.

The mass spectrometer was operated in multiple reaction monitoring (MRM) mode. Mass axis was calibrated using standard polypropylene glycol (PPG) dilution solvents. Nitrogen was used as the nebulizer (GS1), curtain (CUR), heater (GS2), and collision gases. The ion source was heated to 500 °C, and the ion-spray voltages were maintained at 5500 V and −4500 V for positive and negative polarities, respectively. GS1, GS2, and CUR were set as 50, 50, and 35 psi, respectively. Both positive and negative polarities were applied according to the results gained from manual parameter optimization. The polarity-switching schedule, the precursor-to-product transition, optimized declustering potential (DP) values, and collision energy (CE) values are elucidated in Table 1, whereas the dwell time, entrance potential (EP) and collision cell exit potential (CXP) values of all ion transitions were fixed at 120 ms, 10 V, and 12 V, respectively.

Table 1 The precursor-to-product ion transitions, declustering potential values (DP), collision energy values (CE), retention times (t_R) of the 30 targeted compounds and the polarity switching schedule

Period	Duration (min)	Analyte	tR (min)	Ion transitions precursor ion > product ion^a	DP (V)	CE (eV)
a Two ion pairs were optimized for each analyte except UDCA, HDCA, DCA, and CDCA, and the ion transitions in bold were adopted for quantitative analysis, while the other one was utilized as qualifier ion pair.
Period 1 (negative)	0–8.76	CTCA	7.88	514.0 > 124.0; 514.0 > 80.0	−100	−65
		Daidzein	8.09	253.0 > 132.0; 253.0 > 208.0	−150	−49
		Calycosin	8.58	283.0 > 239.9; 283.0 > 211.0	−105	−44
Period 2 (positive)	8.76–9.50	Cortisol	8.88	363.2 > 121.1; 363.2 > 327.1	130	36
		Cortisone	8.96	361.0 > 163.0; 361.0 > 121.0	120	30
Period 3 (negative)	9.50–12.66	Genistein	9.70	269.0 > 159.0; 269.0 > 135.0	−150	−34
		Kaempferol	9.91	285.0 > 143.0; 285.0 > 93.0	−90	−40
		TUDCA	9.93	498.0 > 80.0; 498.0 > 124.0	−150	−120
		TCA	10.19	514.0 > 124; 514.0 > 80.0	−100	−65
		TCCA	10.50	514.0 > 124; 514.0 > 80.0	−100	−65
		Isorhamnetin	10.31	315.0 > 300.0; 315.0.0 > 151.0	−120	−29
		IS1	11.02	1001.0 > 955.0; 1001.0 > 1001.0	−90	−32
		PGF2α	11.16	353.0 > 247.0; 353.0 > 193.0	−120	−32
		PGE2	11.60	351.0 > 315.0; 351.0 > 271.0	−120	−16
		PGD2	12.30	351.0 > 315.0; 351.0 > 271.0	−120	−16
Period 4 (positive)	12.66–15.61	Estradiol	12.99	255.0 > 159.0; 255.0 > 133.0	120	20
		Testosterone	13.92	289.0 > 109.0; 289.0 > 97.0	120	33
		IS2	14.52	287.0 > 203.0; 287.0 > 85.0	160	24
Period 5 (negative)	15.61–17.50	CA	15.76	407.2 > 407.3; 407.3 > 343.1	−140	−40
		Cygnocholic acid	16.35	407.0 > 389.0; 407.0 > 271.0	−180	−50
		UDCA	16.33	391.1 > 391.1	−120	−30
		HDCA	16.57	391.1 > 391.1	−120	−30
Period 6 (positive)	17.50–18.37	Mestanolone	18.16	305.0 > 305.0; 305.0 > 227.0	60	20
Period 7 (negative)	18.37–27.00	DCA	18.93	391.1 > 391.1	−120	−30
		CDCA	18.69	391.1 > 391.1	−120	−30
		20-HETE	19.4	319.1 > 301.1; 319.1 > 389.1	−110	−24
		15-HETE	20.00	319.2 > 219.1; 319.2 > 301.3	−110	−17
		12-HETE	20.24	319.2 > 179; 319.2 > 257.1	−110	−18
		5-HETE	20.40	319.1 > 114.9; 319.1 > 301.1	−110	−16
		AA	22.06	303.1 > 205.3; 303.1 > 259.2	−90	−18

---

The Analyst software (ABSciex) quantification module was used to carry out the quantitative process including peak detection, peak integration, and analyte quantification. Except those default parameters for the automated Analyst classic integration algorithm, the smoothing factor was set as 3 and the bunching factor as 1 for all monitored peaks.

2.5. Method validation

Method validation was carried out in terms of linearity, sensitivity, precision, and accuracy by following the protocols presented in our previous report.²⁰ Because all samples were immediately subjected for measurement after dilution, stability wasn't taken into account.

2.5.1. Linearity and sensitivity. The linearity was assayed using external calibration curves with more than seven concentration levels for each analyte. Each calibration curve generation was performed by plotting the peak area ratio of the analyte and its corresponding IS against the theoretical concentration over the calibration concentration range. The acceptance criterion for each calibration curve was a correlation coefficient (r) greater than 0.99. The limit of detection (LOD) was determined with a signal to noise ratio (S/N) > 3. Limit of quantitation (LOQ) corresponded to the lowest concentration of the linear range, and the S/N and relative standard deviation (RSD, %) of each LOQ should be >10 and <20%, respectively.

2.5.2. Precision. Intra- and inter-day assays were selected to assess the precision of proposed method. For intra-day variability evaluation, QC samples (low, medium, and high concentration levels) were analyzed for six replicates within a single day, while regarding inter-day assay, those solutions were examined in triplicate per day for consecutive three days. Variations were expressed by RSDs (%).

2.5.3. Accuracy. The recovery was used to assess the accuracy of the method. Known amounts (low, medium, and high concentration levels) of mixed standard solutions were added into 1.0 mL of a selected urinary sample (FH1 or MR1) after it was quantitated. Following dilution with PBS containing ISs, the combined solution was subjected for LVDI-online SPE-UHPLC-psMS/MS analysis. The recoveries were calculated using the following formula: recovery (%) = (amount found − original amount)/amount spiked × 100%.

3. Results and discussion

3.1. Method development of LVDI-online SPE-UHPLC-psMS/MS

3.1.1. Instrumentation setup. A generic instrumentation setup^5,6,20 was introduced here to connect all units, such as three pumps (pumps A–C), auto-sampler, column oven, columns, and mass spectrometer, etc. (Fig. 1).

Because of their pivotal roles for the performance of LVDI-online SPE-UHPLC-psMS/MS, both SPE and analytical columns were carefully screened. A Phenomenex guard column embedded with RP-C₁₈ particles was found to be superior to some other candidates, e.g., hydrophilic interaction chromatography (HILIC)-type guard column, RP-C₈ guard column, and RP-C₂ guard column, as well as some short RP-C₁₈ columns (50 × 4.6 mm, 5 μm) in sight of the satisfactory adsorption property as well as low back-pressure; hence, the Phenomenex column served as SPE column to online trap those targeted analytes from aqueous biofluids. Several analytical columns were assayed in term of acceptable separation for all analytes, and a versatile Waters HSS T3 column²¹ was finally employed because it is advantageous at resolution, peak shape, and back-pressure in comparison with some other choices with comparable size, i.e. ACE UltraCore 2.5 SuperC18 column (150 × 3.0 mm, 2.5 μm, Advance Chromatography Technologies Ltd., Aberdeen, Scotland), Ascentis® Express F5 (150 × 2.1 mm, 2.7 μm, Sigma-Aldrich, Bellefonte, DE, USA), and BEH Shield RPC₁₈ (100 × 3.0 mm, 1.7 μm, Waters).

3.1.2. Mass spectrometric parameter optimization. Five phenols, including daidzein, calyconsin, genistein, kaempferol along with isorhamnetin were involved as targeted components. Negative polarity could afford better mass responses for all five phenols in comparison with positive ionization mode. Moreover, the mass cracking rules of those flavonoids (kaempferol and isorhamnetin) and isoflavonoids (daidzein, calyconsin, and genistein) have been well proposed in the literature,²² and it is thereby convenient to obtain the appropriate precursor > product ion transitions as well as optimum mass parameters by following recommended guidelines.²³

BAs are synthesized in hepatocytes and can be subsequently biotransferred into secondary and conjugated BAs, also known as tertiary BAs, via sulfation, glycine conjugation, and taurine conjugation in the intestinal lumen and liver. All BAs exhibited better responses with negative mode than positive polarity. Free BAs are relatively stable in the collision chamber of the tandem mass spectrometer; hence, ion pair of deprotonated molecular ion > deprotonated molecular ion ([M − H]⁻ > [M − H]⁻) was utilized to monitor those unconjugated BAs, such as CA, UDCA, HDCA, DCA, and CDCA, whereas ion transition of [M − H]⁻ > [M − H − H₂O]⁻ was defined for cygnocholic acid because of the extensive occurrence of H₂O-cleavage in the collision chamber of Qtrap-MS. Meanwhile, fragment species at m/z 124 (H₂NC₂H₄SO₃⁻) and 80 (SO₃⁻˙) usually act as the primary product ions for [M − H]⁻ ion of taurine conjugated BAs;²⁴ hence, ion pairs of [M − H]⁻ > m/z 80 and [M − H]⁻ > m/z 124 were employed for those tertiary BAs, such as CTCA, TUDCA, TCA, and TCCA.

Regarding those sterols (cortisol, cortisone, estradiol, testosterone and mestanolone) along with eicosanoids (PGD₂, PGE₂, PGF_2α, 5-HETE, 12-HETE, 15-HETE, 20-HETE, and AA), the optimum mass parameters were also obtained by manual tuning via directly infusing pure compounds into mass spectrometer, and the optimized values were consistent with those archived in the literature.^25–27

Two precursor-to-product ion transitions (quantifier along with qualifier ion pairs)²⁸ were scheduled for each analyte (Table 1), except UDCA, HDCA, DCA, and CDCA, and the more sensitive one usually acted as the quantifier ion transition for each analyte. All optimized DP and CE values are illustrated in Table 1.

3.1.3. Elution program. Suitable solvent could facilitate the retention of targeted analytes in SPE column along with the clearance of those hydrophilic interferences, and 5% aqueous ACN (v/v) was finally employed to elute SPE column following several evaluations ramped from 0% to 10% aqueous ACN (step-size, 1% ACN). To guarantee the removal of interference substances (e.g. proteins along with hydrophilic compounds) and the retention of targeted analytes, the flow rate of the solvent for SPE were finally optimized as 1.0 mL min⁻¹ after assays of 0.5, 0.75, 1.0, 1.25, and 1.5 mL min⁻¹. Regarding the analytical column, combinatory water and ACN was employed as elution solvent after careful assessments between water–ACN and water–methanol. Formic acid (0.1%, v/v) was finally introduced as the additive for either solvent because it could afford better peak shapes along with overall MRM responses than ammonia and ammonium formate, and assays were also performed among the ratios of 0.01%, 0.05%, 0.1%, 0.2%, and 0.3%. Afterwards, the gradient elution program was carefully optimized to afford satisfactory separation for all analytes, in particular between adjacent signals with different ionization polarities (Fig. 2). Moreover, a relative high temperature (50 °C) was applied for the analytical column, because this parameter could significantly decrease the back-pressure of the entire system and modify the peak shapes in comparison than those lower temperatures, e.g. 25 and 35 °C, whereas a higher temperature (60 °C) would significantly shorten the column lifespan.

Fig. 2 Representative chromatograms of mixed standard solution (A) and urinary sample (FH1, B). 1: CTCA (9.56 ng mL⁻¹ in A); 2: daidzein (2.38 ng mL⁻¹ in A); 3: calycosin (5.28 ng mL⁻¹ in A); 4: cortisol (3.37 ng mL⁻¹ in A); 5: cortisone (3.36 ng mL⁻¹ in A); 6: genistein (2.56 ng mL⁻¹ in A); 7: kaempferol (2.67 ng mL⁻¹ in A); 8: TUDCA (0.932 ng mL⁻¹ in A); 9: TCA (1.69 ng mL⁻¹ in A); 10: TCCA (0.960 ng mL⁻¹ in A); 11: isorhamnetin (0.584 ng mL⁻¹ in A); 12: IS1; 13: PGF_2α (0.262 ng mL⁻¹ in A); 14: PGE₂ (0.656 ng mL⁻¹ in A); 15: PGD₂ (0.328 ng mL⁻¹ in A); 16: estradiol (2.54 ng mL⁻¹ in A); 17: testosterone (1.07 ng mL⁻¹ in A); 18: IS2; 19: CA (4.40 ng mL⁻¹ in A); 20: cygnocholic acid (1.52 ng mL⁻¹ in A); 21: UDCA (1.46 ng mL⁻¹ in A); 22: HDCA (1.46 ng mL⁻¹ in A); 23: mestanolone (0.568 ng mL⁻¹ in A); 24: DCA (7.32 ng mL⁻¹ in A); 25: CDCA (1.46 ng mL⁻¹ in A); 26: 20-HETE (0.373 ng mL⁻¹ in A); 27: 15-HETE (0.186 ng mL⁻¹ in A); 28: 12-HETE (0.149 ng mL⁻¹ in A); 29: 5-HETE (0.149 ng mL⁻¹ in A); 30: AA (23.6 ng mL⁻¹ in A).

3.1.4. Injection and phase switching programs. Increasing injection volume could definitely advance the sensitivity, nonetheless, challenging the accumulation potential of the SPE column. After analyzing the effluents collected from the outlet of the SPE column corresponding to stepped injection volumes (step-size, 50 μL), comprehensive trapping could not be accomplished when the injection volume was greater than 500 μL. Because the injection loop was 50 μL, an entire injection program was divided into ten times (Fig. 1). In order to completely expel those hydrophilic substances, SPE column elution with 5% aqueous ACN was maintained for another five minutes which was optimized form 3, 5, and 7 minutes, after the ten injections. Moreover, polarity switching was set to simultaneously monitor all analytes, and the switching program was set on the basis of the elution sequence of all analytes. In total, the elution phase was fragmented into seven periods (Periods 1–7, Fig. 1).

Above all, all optimized parameters were described in Section 2.4. LVDI-online SPE-UHPLC-psMS/MS analysis and Table 1.

3.2. Method validation

Method validation was carried out using linearity, LOQ, LOD, intra- and inter-day, and recovery assays, whereas stability assay wasn't involved because all samples were immediately subjected for measurement after dilution.

3.2.1. Linearity and sensitivity. Correlation coefficients (r) of calibration curves in all inter-run cases were higher than 0.99 over the concentration ranges (Table 2), most of which spanned 2–4 orders of magnitude. Except AA, CTCA, and daidzein, LOQs of all analytes were lower than 0.3 ng mL⁻¹ (Table 2), and most LODs were less than 0.04 ng mL⁻¹ (Table 2), indicating an outstanding performance regarding sensitivity for the developed method.

Table 2 Linear regression data, lower limits of quantification (LOQs) and limits of detection (LODs) for all targeted analytes

Analyte	Linear regression data			LOQ (ng mL^−1)	LOD (ng mL^−1)
	Regression equation	r	Linear range (ng mL^−1)
CTCA	y = 0.0134x − 0.0086	0.998	0.954–239	0.954	0.0362
Daidzein	y = 0.191x − 0.11	0.998	0.952–596	0.952	0.00413
Calycosin	y = 0.738x + 0.0255	0.994	0.212–132	0.212	0.00121
Cortisol	y = 0.335x − 0.124	0.998	0.270–84.3	0.270	0.00914
Cortisone	y = 0.208x + 0.0002	0.999	0.0268–83.9	0.0268	0.00672
Genistein	y = 1.25x − 0.000478	0.996	0.101–64.0	0.101	0.00351
Kaempferol	y = 0.336x − 0.0426	0.999	0.107–66.7	0.107	0.00413
TUDCA	y = 1.01x − 0.0328	0.999	0.0928–23.3	0.0928	0.00492
TCA	y = 0.878x − 0.00565	0.999	0.0337–42.3	0.0337	0.000334
TCCA	y = 0.89x − 0.00195	0.999	0.0192–24.0	0.0192	0.00172
Isorhamnetin	y = 22.1x − 30.9	0.997	0.154–14.6	0.154	0.00461
PGF2α	y = 0.15x + 0.000915	0.999	0.0262–6.56	0.0262	0.0103
PGE2	y = 0.513x + 0.00147	0.997	0.0131–16.4	0.0131	0.00341
PGD2	y = 1.1x − 0.0444	0.999	0.0655–8.20	0.0655	0.00133
Estradiol	y = 0.0674x + 0.0125	0.998	0.253–63.4	0.253	0.0273
Testosteron	y = 0.484x − 0.00142	0.998	0.0215–26.8	0.0215	0.00342
CA	y = 0.164x + 0.00577	0.998	0.0612–110	0.0612	0.00991
Cygnocholic acid	y = 0.0188x + 0.0016	0.999	0.0612–38.0	0.0612	0.000924
UDCA	y = 0.35x + 0.00146	0.999	0.146–36.5	0.146	0.0225
HDCA	y = 0.302x + 0.0372	0.999	0.142–36.5	0.142	0.0252
Mestanolone	y = 0.705x + 0.101	0.991	0.113–14.2	0.113	0.0204
DCA	y = 0.687x + 0.0362	0.999	0.0588–183	0.0588	0.00523
CDCA	y = 0.249x + 0.237	0.999	0.0592–36.5	0.0592	0.0158
20-HETE	y = 0.0602x − 0.000311	0.991	0.0372–9.32	0.0372	0.00571
15-HETE	y = 8.5x − 0.152	0.997	0.0186–2.33	0.0186	1.08 × 10^−5
12-HETE	y = 5.47x − 0.0004	0.997	0.00371–0.932	0.00371	1.77 × 10^−5
5-HETE	y = 12.3x + 0.0931	0.995	0.00753–0.932	0.00753	1.67 × 10^−5
AA	y = 0.00523x + 0.033	0.997	1.13–590	1.13	0.400

---

3.2.2. Precision. RSDs of intra- and inter-day precisions were found to be lower than 15% for all analytes (Table 3), indicating that the precision of the newly developed method could meet the demands for simultaneous determination of those 28 analytes.

Table 3 The results of intra- and inter-day, and recovery assays (n = 6)^a

Analyte	Concentration (ng mL^−1)	Intra-day RSD (%)	Inter-day RSD (%)	Recovery (%)		RSD (%)
				FM1	MR1	FM1	MR1
a N.A.: not applicable.
CTCA	Low	7.45	8.10	120.0	113.2	18.35	13.28
	Medium	8.31	13.14	114.0	91.4	13.55	11.14
	High	5.69	12.88	104.0	98.7	11.29	8.57
Daidzein	Low	10.86	4.50	83.6	115.8	14.53	15.23
	Medium	3.61	13.46	90.0	94.1	11.34	10.06
	High	14.17	11.05	94.5	102.3	8.83	7.14
Calycosin	Low	10.03	13.49	95.4	112.7	7.20	10.77
	Medium	12.71	14.91	93.5	106.5	9.84	9.68
	High	13.77	14.59	95.6	102.4	10.28	7.65
Cortisol	Low	2.69	7.69	119.7	86.3	17.94	14.80
	Medium	5.50	8.09	111.9	89.3	12.48	9.93
	High	5.68	5.38	111.1	95.7	13.08	4.67
Cortisone	Low	7.35	3.23	90.1	111.3	10.71	12.71
	Medium	4.88	5.63	96.9	104.6	5.83	6.38
	High	2.58	5.18	100.0	98.9	16.64	4.96
Genistein	Low	9.52	6.82	111.1	88.8	18.86	15.53
	Medium	7.56	14.73	118.7	91.6	19.28	13.74
	High	11.51	13.71	100.2	98.7	10.37	9.55
Kaempferol	Low	2.91	2.81	95.4	110.1	7.92	11.48
	Medium	0.69	7.68	97.0	106.8	8.30	8.98
	High	3.57	0.91	110.0	103.1	4.43	6.56
TUDCA	Low	13.25	13.61	115.2	86.3	3.74	11.27
	Medium	13.35	13.08	114.0	94.1	4.23	10.06
	High	11.49	11.48	80.4	98.8	20.04	3.57
TCA	Low	11.96	8.62	109.6	87.7	14.74	10.94
	Medium	6.29	14.63	109.8	94.8	3.34	8.86
	High	7.43	12.32	110.3	98.1	4.51	6.48
TCCA	Low	10.21	12.44	N.A.	N.A.	N.A.	N.A.
	Medium	12.29	14.39	N.A.	N.A.	N.A.	N.A.
	High	9.24	11.42	N.A.	N.A.	N.A.	N.A.
Isorhamnetin	Low	0.24	0.34	89.1	111.4	3.64	5.85
	Medium	0.11	0.13	90.4	106.2	10.83	5.46
	High	0.15	0.12	104.4	102.3	11.66	3.59
PGF2α	Low	14.53	13.70	110.5	113.2	13.91	15.43
	Medium	13.31	12.26	113.2	108.4	12.39	13.28
	High	8.80	14.38	106.3	103.2	8.93	9.87
PGE2	Low	8.54	8.62	86.6	116.8	7.54	12.37
	Medium	10.24	14.88	96.6	110.0	6.40	9.86
	High	11.69	13.69	86.7	103.4	15.18	6.48
PGD2	Low	6.52	9.68	N.A.	N.A.	N.A.	N.A.
	Medium	8.05	14.55	N.A.	N.A.	N.A.	N.A.
	High	14.22	10.78	N.A.	N.A.	N.A.	N.A.
Estradiol	Low	5.92	12.16	92.1	N.A.	9.83	N.A.
	Medium	6.07	9.08	90.4	N.A.	11.14	N.A.
	High	2.39	14.93	104.4	N.A.	15.08	N.A.
Testosterone	Low	7.76	5.33	113.8	84.8	7.24	14.13
	Medium	2.05	2.25	107.7	90.1	6.40	10.28
	High	1.10	1.75	117.7	93.4	13.83	7.81
CA	Low	7.70	14.99	111.0	86.3	10.71	11.16
	Medium	11.06	14.47	99.5	89.9	3.63	6.50
	High	4.02	12.65	115.1	94.3	11.04	3.75
Cygnocholic acid	Low	14.9	13.25	88.8	N.A.	9.34	N.A.
	Medium	14.25	11.05	117.7	N.A.	5.44	N.A.
	High	13.54	11.42	107.8	N.A.	6.13	N.A.
UDCA	Low	12.39	13.84	112.9	85.1	11.78	15.86
	Medium	14.34	13.08	117.1	91.1	16.50	10.27
	High	9.35	12.47	115.8	94.4	3.21	5.42
HDCA	Low	12.21	13.41	82.8	112.3	10.74	11.98
	Medium	13.57	12.74	114.9	106.5	18.83	8.47
	High	13.42	10.4	84.2	101.3	5.43	6.70
Mestanolone	Low	3.86	8.19	N.A.	N.A.	N.A.	N.A.
	Medium	8.03	14.60	N.A.	N.A.	N.A.	N.A.
	High	6.97	14.37	N.A.	N.A.	N.A.	N.A.
DCA	Low	13.79	13.87	84.7	115.7	13.48	14.55
	Medium	11.05	13.66	90.0	108.3	3.54	10.77
	High	12.73	10.34	93.2	103.1	1.71	7.98
CDCA	Low	1.86	3.81	91.5	109.4	5.80	11.46
	Medium	4.71	9.82	94.9	104.3	6.24	8.52
	High	13.03	8.12	103.8	100.7	13.73	7.36
20-HETE	Low	11.50	5.23	N.A.	N.A.	N.A.	N.A.
	Medium	9.71	13.98	N.A.	N.A.	N.A.	N.A.
	High	6.61	11.19	N.A.	N.A.	N.A.	N.A.
15-HETE	Low	10.99	13.04	87.5	116.9	10.24	12.44
	Medium	11.78	14.38	83.9	107.1	6.54	8.96
	High	10.28	12.17	106.4	103.1	5.78	8.04
12-HETE	Low	2.41	3.96	81.3	118.6	11.91	12.75
	Medium	5.62	8.15	93.8	110.2	6.81	9.47
	High	13.22	10.92	114.3	107.2	5.69	7.20
5-HETE	Low	6.74	9.49	103.4	90.7	4.29	13.91
	Medium	9.25	13.64	110.2	95.4	10.03	9.54
	High	10.05	10.87	113.4	98.3	15.14	6.47
AA	Low	9.32	10.31	94.4	109.8	6.39	12.63
	Medium	8.01	13.31	87.5	105.6	12.67	8.66
	High	13.44	12.72	79.6	101.3	10.31	5.24

---

3.2.3. Recovery. Following the addition of known amounts (low, medium and high concentration levels) of mixed standard solution into a selected urine sample (FH1 or MR1), whose quantitative profile was characterized, the recovery rates were calculated using the formula aforementioned. Except those unquantifiable components including TCCA, PGD₂, mestanolone, 20-HETE, cygnocholic acid (specific for MR1) and estradiol (specific for MR1), most values located at 79.6–120.0% (RSD% < 20.0%), suggesting satisfactory performance concerning accuracy for the developed method.

Above all, the newly developed method was sensitive, precise, and accurate, and could fulfill the criteria for quantitation of those 28 analytes in mammalian urinary samples. Representative chromatogram of the mixed standard solution is shown in Fig. 2A.

3.3. Simultaneous determination of 28 analytes in mammalian urinary samples using LVDI-online SPE-UHPLC-psMS/MS

Following method validation, the developed LVDI-online SPE-UHPLC-psMS/MS system was applied for simultaneous quantification of 28 analytes in eighteen urinary samples (FH1–FH6, MH1–MH6, and MR1–MR6). The typical chromatogram of urine is shown in Fig. 2B, and all the determined contents are summarized in Table 4. Significant variations regarding the contents occurred among different groups, and even within a single group. TCCA, PGD₂, mestanolone, along with 20-HETE were undetectable in any urinary sample. Isoflavonoids, such as daidzein (44.7–351 ng mL⁻¹), calycosin (15.4–52.4 ng mL⁻¹), and genistein (15.7–63.4 ng mL⁻¹) exhibited primary distributions in rats, however, minor distributions in human urine (≤34.6 ng mL⁻¹ for daidzein, ≤3.21 ng mL⁻¹ for calycosin, and ≤10.9 ng mL⁻¹ for genistein), attributing to the soy-related chow for rats. All urinary samples were rich in AA (1.67–50.1 ng mL⁻¹); nonetheless, minor distributions were disclosed for its derivatives, e.g., 15-HETE (0.0221–0.0795 ng mL⁻¹), 12-HETE (0.00352–0.0504 ng mL⁻¹), and 5-HETE (0.00216–0.0730 ng mL⁻¹). It is worthwhile to mention that significant amounts were detected for PGF_2α (2.33–5.44 ng mL⁻¹) and PGE₂ (2.20–9.75 ng mL⁻¹) in rat urine, rather than human urine (≤1.43 ng mL⁻¹ for PGF_2α, ≤0.478 ng mL⁻¹ for PGE₂ except two values). Apart from testosterone (0.0200–0.787 ng mL⁻¹), cortisol (2.53–21.2 ng mL⁻¹ except one value), cortisone (8.27–61.3 ng mL⁻¹ except two values), and estradiol (0.364–28.2 ng mL⁻¹) were usually detected as primary sterol compounds in human urine. CA (0.12–105 ng mL⁻¹) was usually detected as the most abundant component in urine among various BAs. In comparison with trace distributions of conjugated BAs in mammalian fecal matrices,^29,30 abundant tertiary BAs were detected in urinary matrices, attributing to the hydrolytic potential of intestinal micro-flora.

Table 4 The contents (±SD, ng mL⁻¹) of investigated compounds in the eighteen urinary samples^a

Analyte	FH1	FH2	FH3	FH4	FH5	FH6	MH1	MH2	MH3	MH4	MH5	MH6	MR1	MR2	MR3	MR4	MR5	MR6
a N.D.: not detectable; N.Q.: not quantifiable.
CTCA	17.1 ± 1.3	56.2 ± 8.7	33.2 ± 6.1	36.2 ± 8.8	11.1 ± 1.5	17.6 ± 2.4	55.3 ± 7.9	12.1 ± 1.8	50.3 ± 6.4	9.36 ± 2.2	3.54 ± 0.46	14.7 ± 1.4	5.31 ± 0.87	3.89 ± 0.52	2.82 ± 0.43	1.34 ± 0.14	1.60 ± 0.36	1.68 ± 0.32
Daidzein	1.24 ± 0.31	0.635 ± 0.12	10.5 ± 1.3	0.751 ± 0.19	1.34 ± 0.26	7.45 ± 1.1	60.6 ± 10	0.606 ± 0.14	9.35 ± 1.6	34.6 ± 3.7	31.1 ± 4.1	0.699 ± 0.13	177 ± 31	112 ± 8.9	44.7 ± 7.9	351 ± 35	189 ± 21	165 ± 11
Calycosin	0.302 ± 0.081	N.Q.	0.344 ± 0.088	N.Q.	1.12 ± 0.18	0.307 ± 0.065	3.21 ± 1.0	0.0321 ± 0.011	0.531 ± 0.081	2.05 ± 0.48	2.06 ± 0.39	N.Q.	28.5 ± 4.6	32.5 ± 4.8	40.8 ± 7.3	52.4 ± 6.5	41.1 ± 7.2	15.4 ± 2.7
Cortisol	6.37 ± 1.0	5.64 ± 1.0	2.53 ± 0.72	3.01 ± 0.48	3.11 ± 1.87	21.2 ± 2.3	15.0 ± 2.7	0.150 ± 0.024	48.0 ± 10	11.5 ± 2.4	11.7 ± 1.8	4.32 ± 1.0	0.474 ± 0.081	0.487 ± 0.032	0.371 ± 0.037	0.428 ± 0.014	0.393 ± 0.021	0.536 ± 0.10
Cortisone	30.9 ± 3.7	17.4 ± 3.2	8.27 ± 1.2	16.5 ± 2.1	20.6 ± 3.1	61.3 ± 7.8	37.7 ± 4.3	0.377 ± 0.11	52.3 ± 7.2	34.0 ± 6.8	16.9 ± 4.2	0.0463 ± 0.011	0.199 ± 0.032	0.160 ± 0.015	0.0696 ± 0.012	0.068 ± 0.011	0.2145 ± 0.034	0.185 ± 0.021
Genistein	0.149 ± 0.041	N.Q.	0.587 ± 0.10	0.0945 ± 0.030	0.330 ± 0.10	0.973 ± 0.28	10.9 ± 2.1	0.739 ± 0.13	1.65 ± 0.34	5.73 ± 1.1	6.12 ± 1.8	N.Q.	49.2 ± 6.5	44.9 ± 7.1	37.9 ± 4.9	43.6 ± 6.9	63.4 ± 11	15.7 ± 1.8
Kaempferol	0.131 ± 0.030	0.130 ± 0.021	0.899 ± 0.14	N.Q.	N.Q.	0.511 ± 0.081	0.665 ± 0.13	0.178 ± 0.041	0.217 ± 0.045	0.157 ± 0.021	0.272 ± 0.034	N.Q.	1.34 ± 0.032	0.794 ± 0.11	0.959 ± 0.23	0.593 ± 0.16	1.04 ± 0.21	0.260 ± 0.081
TUDCA	2.12 ± 0.61	14.7 ± 1.9	0.467 ± 0.071	4.83 ± 1.1	4.24 ± 0.78	20.3 ± 4.3	22.8 ± 5.8	0.228 ± 0.071	20.8 ± 4.7	4.99 ± 1.1	0.152 ± 0.036	0.237 ± 0.037	0.109 ± 0.018	0.0663 ± 0.016	0.0628 ± 0.011	0.106 ± 0.024	0.140 ± 0.028	0.142 ± 0.028
TCA	1.71 ± 0.43	0.787 ± 0.12	3.49 ± 0.84	0.371 ± 0.086	0.389 ± 0.11	0.981 ± 0.17	0.426 ± 0.10	0.329 ± 0.084	3.62 ± 0.89	0.586 ± 0.14	0.901 ± 0.22	0.758 ± 0.16	1.82 ± 0.28	1.07 ± 0.26	0.875 ± 0.19	1.41 ± 0.36	1.64 ± 0.41	1.55 ± 0.37
TCCA	N.D.	N.D.	N.D.	N.D.	N.D.	N.D.	N.D.	N.D.	N.D.	N.D.	N.D.	N.D.	N.D.	N.D.	N.D.	N.D.	N.D.	N.D.
Isorhamnetin	14.6 ± 2.8	15.0 ± 3.1	14.7 ± 2.5	14.6 ± 1.7	14.8 ± 1.9	14.6 ± 1.7	14.9 ± 2.1	14.6 ± 2.8	14.7 ± 1.9	14.6 ± 3.1	14.7 ± 2.7	14.6 ± 3.0	14.9 ± 3.3	15.1 ± 2.7	15.1 ± 3.0	14.9 ± 2.8	14.9 ± 2.7	14.7 ± 2.6
PGF2α	0.401 ± 0.11	0.381 ± 0.039	0.887 ± 0.21	0.400 ± 0.11	0.242 ± 0.031	0.965 ± 0.18	0.941 ± 0.23	0.312 ± 0.10	0.786 ± 0.17	0.465 ± 0.11	1.43 ± 0.32	0.322 ± 0.10	5.44 ± 1.1	2.59 ± 0.65	3.55 ± 1.0	2.73 ± 0.65	2.33 ± 0.51	3.98 ± 0.88
PGE2	0.132 ± 0.040	0.0173 ± 0.0051	0.478 ± 0.11	0.0135 ± 0.0021	0.0240 ± 0.010	0.293 ± 0.047	16.3 ± 3.4	0.163 ± 0.035	0.0876 ± 0.022	0.261 ± 0.034	16.0 ± 2.8	0.105 ± 0.026	9.75 ± 1.8	2.55 ± 0.46	3.11 ± 0.87	2.20 ± 0.64	3.22 ± 0.62	13.5 ± 2.4
PGD2	N.D.	N.D.	N.D.	N.D.	N.D.	N.D.	N.D.	N.D.	N.D.	N.D.	N.D.	N.D.	N.D.	N.D.	N.D.	N.D.	N.D.	N.D.
Estradiol	0.510 ± 0.11	8.46 ± 1.84	15.0 ± 2.1	0.621 ± 0.10	0.364 ± 0.11	2.22 ± 0.58	1.95 ± 0.54	0.468 ± 0.11	2.84 ± 0.56	1.62 ± 0.37	2.57 ± 0.46	28.2 ± 5.8	N.D.	N.D.	N.D.	N.D.	N.D.	N.D.
Testosterone	0.0872 ± 0.02	0.0463 ± 0.011	0.265 ± 0.034	0.0799 ± 0.0010	0.0245 ± 0.0081	0.150 ± 0.036	0.607 ± 0.17	0.0500 ± 0.010	0.390 ± 0.10	0.379 ± 0.10	0.787 ± 0.18	0.200 ± 0.0031	0.182 ± 0.041	0.123 ± 0.040	0.091 ± 0.024	0.278 ± 0.064	0.0777 ± 0.018	0.0691 ± 0.017
CA	20.4 ± 4.1	105 ± 26	39.3 ± 6.8	40.5 ± 3.8	2.28 ± 0.74	0.1925 ± 0.041	33.9 ± 5.8	3.62 ± 0.88	37.3 ± 6.5	104 ± 27	16.4 ± 2.7	80.2 ± 25	49.1 ± 3.8	79.8 ± 18	20.6 ± 3.7	46.6 ± 5.9	0.119 ± 0.21	22.4 ± 2.8
Cygnocholic acid	0.817 ± 0.18	N.Q.	2.40 ± 0.51	10.4 ± 1.7	0.493 ± 0.11	6.53 ± 1.2	N.Q.	N.Q.	N.Q.	N.Q.	12.5 ± 3.1	1.39 ± 0.31	N.Q.	N.Q.	N.Q.	N.Q.	N.Q.	N.Q.
UDCA	0.220 ± 0.057	3.97 ± 1.0	1.79 ± 0.47	5.51 ± 1.0	1.46 ± 0.31	7.89 ± 1.7	1.08 ± 0.28	1.52 ± 0.47	4.56 ± 1.0	4.57 ± 1.3	1.96 ± 0.36	0.331 ± 0.10	0.230 ± 0.071	0.384 ± 0.10	0.223 ± 0.038	4.20 ± 1.2	1.40 ± 0.34	2.67 ± 0.65
HDCA	0.143 ± 0.34	0.202 ± 0.055	0.397 ± 0.051	0.741 ± 0.10	1.29 ± 0.19	0.996 ± 0.24	1.95 ± 0.41	3.44 ± 0.89	13.7 ± 2.8	2.42 ± 0.65	9.39 ± 2.8	1.32 ± 0.24	0.751 ± 0.15	0.405 ± 0.11	0.147 ± 0.033	0.854 ± 0.19	1.49 ± 0.28	1.86 ± 0.34
Mestanolone	N.D.	N.D.	N.D.	N.D.	N.D.	N.D.	N.D.	N.D.	N.D.	N.D.	N.D.	N.D.	N.D.	N.D.	N.D.	N.D.	N.D.	N.D.
DCA	1.55 ± 0.24	3.07 ± 0.81	2.54 ± 0.43	7.97 ± 2.31	0.814 ± 0.18	40.8 ± 7.7	95.6 ± 31	2.38 ± 0.64	12.3 ± 2.9	23.9 ± 4.1	3.47 ± 1.1	125 ± 258	63.1 ± 6.1	3.70 ± 1.0	1.48 ± 0.28	1.97 ± 0.31	181 ± 28	3.07 ± 1.0
CDCA	3.45 ± 0.87	7.63 ± 1.3	2.51 ± 0.61	3.23 ± 0.84	1.43 ± 0.28	0.804 ± 0.21	26.2 ± 7.5	2.18 ± 0.51	1.23 ± 0.25	2.22 ± 0.34	8.74 ± 2.1	2.87 ± 0.46	2.31 ± 0.46	1.76 ± 0.41	0.322 ± 0.10	0.312 ± 0.021	4.45 ± 1.1	0.923 ± 0.18
20-HETE	N.D.	N.D.	N.D.	N.D.	N.D.	N.D.	N.D.	N.D.	N.D.	N.D.	N.D.	N.D.	N.D.	N.D.	N.D.	N.D.	N.D.	N.D.
15-HETE	0.0494 ± 0.011	0.0278 ± 0.0047	0.0276 ± 0.0081	0.0224 ± 0.0031	0.0221 ± 0.0031	0.0252 ± 0.0080	0.0363 ± 0.011	0.0224 ± 0.0061	0.0227 ± 0.0046	0.0366 ± 0.010	0.0795 ± 0.021	0.0274 ± 0.0021	0.0391 ± 0.010	0.0277 ± 0.010	0.0263 ± 0.0036	0.0264 ± 0.0034	0.0378 ± 0.0084	0.0278 ± 0.051
12-HETE	0.0504 ± 0.080	0.0102 ± 0.0031	0.0317 ± 0.0072	0.162 ± 0.024	0.0108 ± 0.0033	0.0684 ± 0.0017	0.00570 ± 0.0010	0.0128 ± 0.0021	0.00352 ± 0.0010	0.0151 ± 0.0031	0.0175 ± 0.0031	0.00511 ± 0.0010	0.0265 ± 0.011	0.0138 ± 0.0034	0.0148 ± 0.0027	0.0141 ± 0.0025	0.0317 ± 0.0074	0.0102 ± 0.0031
5-HETE	0.0730 ± 0.015	0.0300 ± 0.010	0.0285 ± 0.0034	0.0227 ± 0.0080	0.0216 ± 0.0051	0.0253 ± 0.0062	0.0274 ± 0.0051	0.0246 ± 0.0036	0.0224 ± 0.0041	0.0381 ± 0.011	0.0324 ± 0.010	0.0221 ± 0.0046	0.0400 ± 0.014	0.0279 ± 0.0027	0.0272 ± 0.0075	0.0274 ± 0.0071	0.0328 ± 0.010	0.0300 ± 0.010
AA	38.0 ± 7.8	24.3 ± 4.5	22.9 ± 1.9	16.3 ± 3.1	1.67 ± 0.12	17.4 ± 3.1	50.1 ± 7.2	0.501 ± 0.11	20.4 ± 3.4	25.2 ± 4.1	28.2 ± 3.6	2.38 ± 0.56	29.1 ± 4.3	7.76 ± 1.4	12.8 ± 1.9	12.9 ± 2.4	15.9 ± 1.8	6.13 ± 0.78

---

Comparisons were carried out between regular injection (50 μL for once) and LVDI (50 μL for ten times, 500 μL in total). As expected, LODs along with most LOQs of multi-injection manner were approximately 10-fold lower than those afforded by single injection. Quantitative results of those primary components, such as CTCA, cortisone, cortisol, TUDCA, estradiol, HDCA, UDCA, CA, AA, CDCA, DCA, etc., exhibited great consistence between those two injection approaches, whereas those trace constituents, in particular those AA derivatives (i.e. PGE₂, PGF_2α, 15-HETE, 12-HETE, and 5-HETE), were only quantifiable with LVDI. In comparison with those data archived in the literature,^24–27,31 the developed method is significantly more sensitive than those existing approaches. In addition, simultaneous monitoring of those compounds with different favoritisms of positive (e.g. sterols) or negative (e.g. phenols, BAs, and eicosanoids) ionization mode was achieved attributing to the employment of polarity-switching program, showing a significant time-saving advantage in comparison with the protocol archived in the literature.³²

Actually, preliminary experiments were carried out to assess the stability of endogenous substances, and significant decrements were observed for BAs (e.g., 51.3% and 52.3% decrements for TCA and TUDCA, respectively), and eicosanoids (e.g., 24.6% and 9.8% decrements for 15-HETE and 5-HETE, respectively) when selected urine sample (MH1) was maintained in auto-sampler at 20 °C for 10 hours. Fortunately, the developed method could exactly fulfill the demands for avoiding, to some extent, the potential degradation of the endogenous substances according to directly subjecting the biological sample to measurement following immediate dilution. Moreover, the whole measurement was free from the time-consuming and laborious sample preparation procedures that were usually involved in the conventional methods,^33,34 suggesting a significant benefit for the achievement of high throughput and automated analysis.

Phenols, especially flavonoid- and isoflavonoid-derivatives, widely served as primary effective components of diverse medicinal plants, such as Scutellaria baicalensis Georgi³⁵ and Astragalus membranaceus (Fisch.) Bge.³⁶ However, they are also usually observed as an important chemical category in many edible herbs, e.g., soybean and rice.^37,38 As a consequence, it is an annoying task to differentiate drug-derived metabolites from diet-related components during the metabolic and pharmacokinetic characterization of the effective compounds of those herbal medicines. Daidzein and formononetin were characterized as the metabolites of calycosin or its glucoside,^17,18 whereas those components were undetectable in Zebrafish following administration of calycosin-7-O-β-D-glucoside that is a primary effective component in A. membranaceus.³⁹ Therefore, extensive accounts should be paid onto those food-derived phenolic compounds, e.g., daidzein and formononetin when characterizing the metabolic profiles of those phenols.

4. Conclusions

A sensitive LVDI-online SPE-UHPLC-psMS/MS method was developed and the method validation results demonstrated that the new approach could meet the quantitative criteria for simultaneous determination of urinary samples in terms of linearity, precision, and accuracy. Particularly, extreme sensitivity was observed judging from LOD and LOQ assays. A total of twenty-eight constituents, including ten BAs, five sterols, eight eicosanoids and five phenolic compounds were simultaneously and directly monitored in eighteen urinary samples (FH1–FH6, MH1–MH6, and MR1–MR6) from human and rats. Significant variations occurred for urine samples from different species, as well as urinary matrices from different individuals within a same group. Above all, the validated method takes advantages of detecting those trace components and preserving those instable substances from chemical degradation, thus offering a meaningful and reliable tool for widely targeted monitoring substances in biofluids.

Acknowledgements

This work was financially supported by National Science Fund of China (No. 81403073 to Y.-L. S., 81530097 to P.-F. T., and 81503286 to J. Z.).

References

1. P. Y. Yin, R. Lehmann and G. W. Xu, Anal. Bioanal. Chem., 2015, 407, 4879–4892.
2. A. Gałuszka, Z. Migaszewski and J. Namieśnik, TrAC, Trends Anal. Chem., 2013, 50, 78–84.
3. M. Tobiszewski and J. Namieśnik, Anal. Chem., 2012, 35, 67–73.
4. Y. L. Song, W. H. Jing, R. Yan and Y. T. Wang, Anal. Methods, 2014, 6, 623–628.
5. N. Zhang, J. Li, Q. Q. Song, S. P. Shi, C. Li, Y. F. Zhao, Y. L. Song and P. F. Tu, Anal. Methods, 2016, 8, 1942–1949.
6. Y. L. Song, W. H. Jing, F. Q. Yang, Z. Shi, M. C. Yao, R. Yan and Y. T. Wang, J. Pharm. Biomed. Anal., 2014, 88, 269–277.
7. A. C. Chiaia, C. Banta-Green and J. Field, Environ. Sci. Technol., 2008, 42, 8841–8848.
8. J. Liu, Q. He, L. Yang, E. Hu and M. Wang, Chin. J. Chromatogr., 2015, 33, 1110–1114.
9. Q. Zhong, L. Shen, J. Liu, D. Yu, S. Li, J. Yao, S. Zhan, T. Huang, Y. Hashi, S. Kawano, Z. Liu and T. Zhou, J. Chromatogr. A, 2016, 1442, 53–61.
10. R. E. Wilson, S. R. Groskreutz and S. G. Weber, Anal. Chem., 2016, 88, 5112–5121.
11. F. Sabermahani and M. A. Taher, J. AOAC Int., 2014, 97, 1713–1718.
12. P. Hušek, Z. Švagera, D. Hanzlíková, L. Řimnáčová, H. Zahradníčková, I. Opekarová and P. Šimek, J. Chromatogr. A, 2016, 1443, 211–232.
13. Q. Su, T. Guan and H. Lv, Sci. Rep., 2016, 6, 24099.
14. T. Zhang and D. G. Watson, J. Chromatogr. B: Anal. Technol. Biomed. Life Sci., 2016, 1022, 199–205.
15. J. E. Kinsella, B. Lokesh, S. Broughton and J. Whelan, Nutrition, 1990, 6, 24–44.
16. A. T. Kicman, E. Houghton and D. B. Gower, Steroid Analysis, Springer, 2010, pp. 743–836.
17. Y. Z. Zhang, J. D. F. Xu, J. Liang, Y. Hashi, M. Y. Shang, D. H. Yang, X. Wang and S. Q. Cai, J. Pharm. Biomed. Anal., 2012, 70, 425–439.
18. L. Chen, Z. Li, Y. Tang, X. Cui, R. Luo, S. Guo, Y. Zheng and C. Huang, J. Pharm. Biomed. Anal., 2011, 56, 382–389.
19. D. Bi, X. Y. Chai, Y. L. Song, Y. Lei and P. F. Tu, Chem. Pharm. Bull., 2009, 57, 528–531.
20. Y. L. Song, N. Zhang, Y. Jiang, J. Li, Y. F. Zhao, S. P. Shi and P. F. Tu, RSC Adv., 2015, 5, 6419–6428.
21. X. Qiao, M. Ye, C. F. Liu, W. Z. Yang, W. J. Miao, J. Dong and D. A. Guo, Steroids, 2012, 77, 204–211.
22. W. Z. Yang, M. Ye, X. Qiao, Q. Wang, T. Bo and D. A. Guo, Eur. J. Mass Spectrom., 2012, 18, 493–503.
23. http://www.absciex.com/Documents/Downloads/Literature/turbo-v-operator-gu%20-ide-en.pdf.
24. X. Qiao, M. Ye, D. L. Pan, W. J. Miao, C. Xiang, J. Han and D. A. Guo, J. Chromatogr. A, 2011, 1218, 107–117.
25. J. Abdel-Khalik, E. Björklund and M. Hansen, J. Chromatogr. B: Anal. Technol. Biomed. Life Sci., 2013, 935, 61–69.
26. W. Gao, T. Stalder and C. Kirschbaum, Talanta, 2015, 143, 353–358.
27. J. Yang, K. Schmelzer, K. Georgi and B. D. Hammock, Anal. Chem., 2009, 81, 8085–8093.
28. Y. L. Song, W. H. Jing, G. Du, F. Q. Yang, R. Yan and Y. T. Wang, J. Chromatogr. A, 2014, 1338, 24–37.
29. G. Kakiyama, A. Muto, H. Takei, H. Nittono, T. Murai, T. Kurosawa, A. F. Hofmann, W. M. Pandak and J. S. Bajaj, J. Lipid Res., 2014, 55, 978–990.
30. C. John, P. Werner, A. Worthmann, K. Wegner, K. Tödter, L. Scheja, S. Rohn, J. Heeren and M. Fischer, J. Chromatogr. A, 2014, 1371, 184–195.
31. G. Du, H. Y. Zhao, Y. L. Song, Q. W. Zhang and Y. T. Wang, J. Sep. Sci., 2011, 34, 2576–2585.
32. Y. L. Song, Q. Q. Song, J. Li, J. Zheng, C. Li, Y. Zhang, L. L. Zhang, Y. Jiang and P. F. Tu, J. Chromatogr. A, 2016, 1454, 58–66.
33. M. H. Sarafian, M. R. Lewis, A. Pechlivanis, S. Ralphs, M. J. McPhail, V. C. Patel, M. E. Dumas, E. Holmes and J. K. Nicholson, Anal. Chem., 2015, 87, 9662–9670.
34. J. Boccard, F. Badoud, N. Jan, R. Nicoli, C. Schweizer, F. Pralong, J. L. Veuthey, N. Baume, S. Rudaz and M. Saugy, Bioanalysis, 2015, 6, 2523–2536.
35. M. Li-Weber, Cancer Treat. Rev., 2009, 35, 57–68.
36. L. Z. Lin, X. G. He, M. Lindenmaier, G. Nolan, J. Yang, M. Cleary, S. X. Qiu and G. A. Cordell, J. Chromatogr. A, 2000, 876, 87–95.
37. L. Coward, N. C. Barnes, K. D. R. Setchell and S. Barnes, J. Agric. Food Chem., 1993, 41, 1961–1967.
38. W. Chen, L. Gong, Z. Guo, W. Wang, H. Zhang, X. Liu, S. Yu, L. Xiong and J. Luo, Mol. Plant, 2013, 6, 1769–1780.
39. G. Hu, S. O. Siu, S. Li, I. K. Chu, Y. W. Kwan, S. W. Chan, G. P. Leung, R. Yan and S. M. Lee, Xenobiotica, 2012, 42, 294–303.

---