Development of a Highly Sensitive and Rapid Liquid Chromatography–Tandem Mass Spectrometric Method Using a Basic Mobile Phase Additive to Determine the Characteristics of the Urinary Metabolites for Niemann–Pick Disease Type C

Abstract

As Niemann–Pick disease type C (NPC) is difficult to diagnose owing to its various clinical symptoms; biomarker tests have been developed. Previously, we revealed urinary sulfated cholesterol metabolites as noninvasive biomarkers for NPC. However, LC/tandem mass spectrometry (LC/MS/MS) requires long separation time and large urine volumes. Recently, a basic mobile phase was reported to increase the MS intensity. Thus, we developed a highly sensitive and rapid LC/MS/MS method for analyzing urinary cholesterol metabolites using a basic mobile phase additive. 3β-Sulfooxy-7β-N-acetylglucosaminyl-5-cholenic acid, its glycine and taurine conjugates, 3β-sulfooxy-7β-hydroxy-5-cholenic acid, and 7-oxo form were measured, with selected reaction monitoring in negative ion mode. Oasis HLB and L-column 3 were used for column-switching LC/MS/MS and urine diluted 10-fold was employed as the sample. After trapping, gradient separation was performed using solutions containing 1% (v/v) ammonium solution. On average, a 16-fold increase in peak areas was observed compared to that obtained at pH 5.5 with the mobile phases. Although the previous method needed 60 min for separation from interference peaks, we succeeded to separate them in 7 min with optimized LC condition. Further, all compounds showed good linearity from 0.3–1000 ng/mL, with satisfactory intra- and inter-day reproducibility. The developed method was applied to the urinalysis of healthy participants and NPC patients. Overall, the concentrations of metabolites correlated with those obtained using the previous method. Therefore, we succeeded to increasing MS intensity and shorten LC running time; and the method is useful for the noninvasive diagnostic screening of patients with NPC.

INTRODUCTION

Niemann–Pick disease type C (NPC) is a progressive and life-limiting autosomal recessive disorder.^1,2) The number of NPC patients is increasing and the estimated prevalence of this condition is approximately 1/80000.³⁾ Although NPC is not related to enzyme deficiency, it is regarded as a lysosomal storage disease. NPC is caused by mutations in NPC1, which codes for the NPC1 cholesterol transporter protein in the lysosomal membrane, or NPC2, which codes for the NPC2 intracellular secreted cholesterol transporter protein.⁴⁾ Because NPC1 and NPC2 transport cholesterol in a coordinated manner, genetic mutation of NPC1 or NPC2 causes a functional deficiency of cholesterol traffic. As a result, cholesterol^5–7) and sphingolipids^8,9) accumulate in NPC. Patients with NPC present a variety of symptoms; however, the detailed mechanisms of this disease are unknown.³⁾ In addition, the onset of symptoms varies from neonatal to old age.^1–3) Because disease prognosis is considered to be better when treatment begins earlier,¹⁰⁾ early diagnosis is desired. Biomarkers^11–13) that can more easily diagnose NPC are attracting attention owing to the rarity of a specialist for this disease and the complexity of the conventional methods.⁷⁾ Since 2010, oxysterols,^14,15) lysosphingomyelin,¹⁶⁾ N-palmitoyl-O-phosphocholine-serine^17–20) (previously called Lyso-SM-509), and abnormal bile acids^21–24) have been reported as NPC biomarkers in the blood. Accordingly, these biomarkers can be measured in blood samples (serum, plasma, and dried blood spots), which can be easily obtained.

Sulfated cholesterol metabolites have been reported as NPC biomarker candidates in urine.^{11–13,24–34)} Various conjugated cholesterols are known to be present in urine,^35–37) and different molecules have been detected using metabolome analysis.²⁷⁾ Previously, our research group found that five urinary sulfated cholesterol metabolites are good biomarker candidates for NPC (Supplelmentary Fig. S1). These biomarkers include 3β-sulfooxy-7β-N-acetylglucosaminyl-5-cholen-24-oic acid, glycine, and taurine conjugates (SNAG-Δ⁵-CA, SNAG-Δ⁵-CG, and SNAG-Δ⁵-CT, respectively), 3β-sulfooxy-7β-hydroxy-5-cholenoic acid (S7B-Δ⁵-CA), and 3β-sulfooxy-7-oxo-5-cholenoic acid (S7O-Δ⁵-CA). In particular, S7B-Δ⁵-CA has displayed excellent diagnostic performance and good sensitivity and specificity.²⁹⁾ To accurately quantify the five metabolites without matrix effects,^26,28) accurate separation from the contaminating peaks was achieved with a 60-min LC gradient.²⁹⁾ However, a large amount of urine (200 µL of 10-fold diluted urine) was injected into the system.

An increase in MS intensity using basic mobile phase additives has been previously reported.^38–49) In fact, the ion abundance in MS seems to increase, regardless of whether the MS polarity is positive^{41,42,44,45,47–49)} or negative.^{39,40,43,44,46–48,50)} Examples of mobile phase additives include ammonium solutions,^39,40,44,50) ammonium hydrogen carbonate,^42,43) ammonium bicarbonate,⁴¹⁾ and ammonium fluoride.^45–49) In addition to improving MS and MS/MS sensitivity, which are advantages of basic mobile phase additives, these additives can shorten the retention time for acidic analytes.^40,44,46) Therefore, basic mobile phase additives were considered to be a suitable solution to the problems faced in the analysis of urinary cholesterol metabolites.

In this study, we developed an LC/tandem mass spectrometry (LC/MS/MS) analytical method using a basic mobile phase additive for analyzing five urinary conjugated cholesterol metabolites. The method was evaluated by comparing the concentrations obtained with this method and those obtained using the previously developed method.²⁹⁾ In addition, the diagnostic performance of the two methods was compared for NPC screening and the usefulness of basic additives for mobile phases in LC/MS/MS was clarified for biomarker analysis in NPC diagnostic screening.

MATERIALS AND METHODS

Chemicals and Reagents

SNAG-Δ⁵-CA, SNAG-Δ⁵-CG, SNAG-Δ⁵-CT, S7B-Δ⁵-CA, S7O-Δ⁵-CA, and 3β-sulfooxy-7β-hydroxy-23-nor-5-cholenoic acid (as an internal standard (IS)) were used as previously described; the structures of these compounds are displayed in Supplementary Fig. S1.^26,51–53) Ultrapure water was prepared using a Puric-α apparatus (Organo Corporation, Tokyo, Japan). Ammonium solution (HPLC grade), ammonium acetate, and methanol were purchased from FUJIFILM Wako Pure Chemical Corporation (Osaka, Japan). Urine samples were collected following retrieval of informed consent from untreated patients diagnosed with NPC and healthy volunteers. Urine samples were collected in the morning and stored at −80 °C until analysis. All experiments were performed according to the protocol approved by the Ethics Committee of the Graduate School of Medicine at Tohoku University (Approval No. 2013-1-293).

LC/MS/MS Equipment

A Nexera ultra-HPLC system (Shimadzu Corp., Kyoto, Japan) was connected to a triple quadrupole tandem mass spectrometer API 5000 equipped with an electrospray ionization probe (SCIEX, Framingham, MA, U.S.A.). MS/MS was performed in selective reaction monitoring (SRM) mode under negative ion detection. Ion spray voltage, turbo spray temperature, curtain gas, nebulizer gas, turbo gas, and collision gas were set to –4500 V, 700 °C, 20, 50, 50 psi, and 6 units, respectively. The SRM conditions are listed in Supplementary Table S1; the conditions before (A) and after (B) selecting the most intense SRM transitions are also listed in this table. Data acquisition was performed using Analyst version 1.5.0 (SCIEX) and the SCIEX OS-Q software (SCIEX) for data integration. Column switching was used as the LC system in this study.^26,27,29,54) After injection of the diluted sample, a 20 mmol/L ammonium acetate solution/methanol (9 : 1, v/v) mixture was loaded onto an OASIS HLB column for pretreatment (2.1 mm i.d. × 20 mm, 5 µm, Waters, Milford, MA, U.S.A.). Sample pretreatment was performed at a flow rate of 2.0 mL/min for 1 min. After the sample was washed and the analytes were concentrated, the sample eluent was loaded on an L-column 3 (2.1 mm i.d. × 30 mm, 2 µm, Chemicals Evaluation and Research Institute, Saitama, Japan) by switching the valve used for changing the liquid flow path. Mobile phase A (water/25% ammonium solution [100 : 1, v/v]) and mobile phase B (methanol/25% ammonium solution [100 : 1, v/v]) were gradually changed from A : B = 90 : 10 to A : B = 65 : 35 over 5.5 min. The LC cycle run time which contains washing and equilibrating, was set at 7 min.

Preparation of the Stock and Working Solutions of Standards and IS

The analytes and IS were prepared at a concentration of 100 µg/mL using water/ethanol (1 : 1, v/v) and stock solutions. The IS was diluted with water/ethanol (1 : 1, v/v) to 33 ng/mL and used as the IS solution. The analytes were mixed and diluted with water/ethanol (1 : 1, v/v) to 0.3, 1, 3, 10, 30, 100, 300, and 1000 ng/mL (working solutions for the calibration curve). For quality control (QC), mixed solutions of 2, 50, and 800 ng/mL were used as low-quality control (LQC), middle-quality control (MQC), and high-quality control (HQC) (working solution for QC), respectively.

Procedure for Preparing the Sample Aliquot

Twenty-five microliters of urine (or water for the preparation of calibration curves) was added to a 1.5 mL polypropylene microtube and 25 µL of a mixture of water/ethanol (1 : 1, v/v) (or standard solutions for the preparation of calibration curves and assays of the quality controls), 25 µL of 33 ng/mL IS solution, and 175 µL of water. The solution was mixed and centrifuged at 15000 × g for 3 min at 4 °C. The supernatants were transferred to glass vials for LC/MS/MS analysis, and 50 µL of each sample was injected into the LC/MS/MS system.

Effect of the Mobile Phases on the Intensity of the Analytes and the Proportion of Ion Types

Five types of mobile phases were used to determine their effect on analyte intensity and the proportion of ion types. The detailed components of the mobile phases are listed in Supplementary Table S2. A standard solution mixture of 100 ng/mL was used in this investigation. The procedure for sample dilution is described in the “Procedure for preparing the sample aliquot.”

Preparation of the Calibration Curves

Twenty-five microliter of water was used as a surrogate matrix, and 25 µL of IS solution, 25 µL of working solution for the calibration curves, and 175 µL of water were added and mixed. The mixture was then centrifuged at 15000 × g for 3 min at 4 °C and 50 µL of the supernatant was injected into the system for LC/MS/MS analysis. The peak area ratio of each analyte to the IS was plotted against the standard concentration, and calibration curves were prepared using the least-squares method with 1/x² weighting.

Matrix Effects

To calculate the matrix effects, 25 µL of the IS solution, 25 µL of water/ethanol (1 : 1, v/v) or QCM solution, and 175 µL of water were added to 25 µL of urine from a healthy control or water. After the same procedure described above was performed, the samples were analyzed. The matrix factors for each analyte were calculated using the following formula, and the ratios of their matrix factors to those of the IS were calculated as the IS normalized matrix factors.^{16,20,51,55,56)}

  

Intra- and Inter-day Assay Reproducibility

To determine the reproducibility of this method, intra- and inter-assay analyses were performed. Twenty-five microliters of QC standard solution (blank, LQC, MQC, and HQC), 25 µL of IS solution, and 175 µL of water were added to 25 µL of urine from healthy participants. Urine from patients with NPC was analyzed using the procedure described above. Every three days, urine samples were prepared and analyzed to measure blank, LQC, MQC, and HQC (N = 6). Generally, accuracy (%) was calculated using the relative error (R.E. (%)). However, because the analytes in this study are endogenous, they were calculated by adding the concentration in the urine of healthy participants (blank). The equation is as follows:

  

Precision (%) was calculated based on relative standard deviation (R.S.D. (%)) using the following equation:

  

Urine Analysis

To evaluate the urine samples, 25 µL of urine from healthy participants (N = 38) and patients with NPC (N = 33) were subjected to analysis. Urine was analyzed using the method developed in this study (method S) and a previously developed method (method L).²⁹⁾ The data were processed using JMP Pro version 16.2.0 software (SAS Institute Inc., NC, U.S.A.). Wilcoxon’s test and receiver operating characteristic (ROC) analysis were used for the intergroup analysis of significant differences and diagnostic performance. Urinary creatinine (Cr) levels were analyzed using an enzymatic creatinine analysis kit (Serotec, Sapporo, Japan). The urinary concentrations of the five metabolites were corrected using urinary Cr concentration. Correlation regression analysis, Passing–Bablok regression analysis and Bland–Altman analysis were performed with JMP Pro 16.2.0 software and Microsoft Excel.

RESULTS AND DISCUSSION

Optimization of the MS/MS Conditions

In this study, we aimed to develop an improved analytical method for the detection of five cholesterol metabolites in urine using a basic mobile phase additive. First, we determined the effect of the basic mobile phase additive by testing five mobile phase combinations (Supplementary Table S1, MP1–MP5). The basic mobile phase additive was employed to increase the ion intensities in MS/MS and shorten the running time in LC. Based on previous reports,^39,40,44) an ammonia solution was selected as the additive. The ion intensities in each mobile phase combination are shown in Fig. 1. The increasing ratios compared to the previous pH 5.5 conditions were 31.1, 23.2, 14.1, 11.6, 7.24, and 9.03 in SNAG-Δ⁵-CA, SNAG-Δ⁵-CG, SNAG-Δ⁵-CT, S7B-Δ⁵-CA, S7O-Δ⁵-CA, and IS, respectively. The calculations for SNAG-Δ⁵-CA, S7B-Δ⁵-CA, S7O-Δ⁵-CA, and IS were performed using the values of [M-2H]²⁻ ions in MP5 and those of [M-H]⁻ ions in MP1. In SNAG-Δ⁵-CG and SNAG-Δ⁵-CT, [M-2H]²⁻ were used for calculation. The analytes and IS have two ionized functional groups (Supplementary Fig. S1). The first group is a sulfate group, which is common to all analytes and IS. Under normal conditions, the group is negatively charged as its pK_a is very low. The other functional groups varied for each compound. SNAG-Δ⁵-CA, S7B-Δ⁵-CA, S7O-Δ⁵-CA, and IS contain carboxyl groups (-COOH) while SNAG-Δ⁵-CA and SNAG-Δ⁵-CT contain glycine (-NHCH₂COOH) and taurine residues, respectively. They have been speculated to have pK_as values of approx. 4.5, approx. 4, and approx. 1.5, respectively. In our previous report in pH 5.5, the precursor ions were set as [M-H]⁻ for SNAG-Δ⁵-CA, S7B-Δ⁵-CA, S7O-Δ⁵-CA, and IS; and [M-2H]²⁻ for SNAG-Δ⁵-CG and SNAG-Δ⁵-CT.^26,29,55) In this study, under weakly acidic (pH 5.5) and neutral conditions (pH 6.8), the peaks derived from [M-2H]²⁻ were not detected while the peaks from [M-H]⁻ were detected (Fig. 1). In addition, when MP3, MP4, and MP5 were used, ions derived from [M-2H]²⁻ could be detected. The intensities increased with an increase in the volume of the ammonium solution and reached the maximum intensity with the use of MP5. In SNAG-Δ⁵-CG and SNAG-Δ⁵-CT, SRM transitions were set as [M-2H]²⁻ to m/z 97 in all mobile phases as they were monitored as [M-2H]²⁻ to m/z 97 transitions in previous studies using weakly acidic or neutral conditions.^26,27,29,55) As a result, the intensities of SNAG-Δ⁵-CG and SNAG-Δ⁵-CT increased according to the increasing proportion of the ammonium solution in the mobile phase and reached the highest intensity when MP5 was used for all compounds. MP5 contained approximately 1% (v/v) ammonium solution and accounted for approximately 147 mmol/L of ammonium. Tan et al. used a mobile phase containing 0.2% (v/v) ammonium solution.⁴⁰⁾ In this study, tetrahydrocurcumin, which has two aromatic hydroxyl groups and is a weak anionic compound, was analyzed under negative conditions. Compared to the acidic conditions, the MS intensity increased by approximately 6-fold and the LC retention time shortened.⁴⁰⁾ Tan also reported negative ion MS/MS improvement of several very weak compounds using an ammonium solution additive. Thus, product ions that have never been detected could be identified owing to the increasing ion abundance.³⁹⁾ In our result, the detailed reason for the increase in the ion intensities of analytes with the use of ammonium solution is known; however, we succeeded at increasing the MS ion abundance using basic mobile phase additive (Fig. 1).

Fig. 1. The Ion Intensities Using Each Mobile Phase Combination

(a) SNAG-Δ⁵-CA, (b) SNAG-Δ⁵-CG, (c) SNAG-Δ⁵-CT, (d) S7B-Δ⁵-CA, (e) SNAG-Δ⁵-CA, and (f) IS, respectively. MP1: 20 mmol/L ammonium acetate buffer (pH 5.5) as mobile phase A and methanol as mobile phase B. MP2: 20 mmol/L ammonium acetate solution as mobile phase A and methanol as mobile phase B. MP3: Water/25% ammonium solution (100 : 0.1, v/v) as mobile phase A and methanol/25% ammonium solution (100 : 0.1, v/v) as mobile phase B. MP4: Water/25% ammonium solution (100 : 0.3, v/v) as mobile phase A and methanol /25% ammonium solution (100 : 0.3, v/v) as mobile phase B. MP5: Water/25% ammonium solution (100 : 1, v/v) as mobile phase A and methanol /25% ammonium solution (100 : 1, v/v) as mobile phase B. SNAG-Δ⁵-CA, non-amidated 3β-sulfooxy-7β-N-acetylglucosaminyl-5-cholen-24-oic acid; SNAG-Δ⁵-CG, glycine-amidated 3β-sulfooxy-7β-N-acetylglucosaminyl-5-cholen-24-oic acid; SNAG-Δ⁵-CT, taurine-amidated 3β-sulfooxy-7β-N-acetylglucosaminyl-5-cholen-24-oic acid; S7B-Δ⁵-CA, 3β-sulfooxy-7β-hydroxy-5-cholen-24-oic acid; S7O-Δ⁵-CA, 3β-sulfooxy-7-oxo-5-cholen-24-oic acid; IS, internal standard; 3β-sulfooxy-7β-hydroxy-23-nor-5-cholenoic acid.

Optimization of the LC Condition for Bile Acids

To evaluate the LC separation performance, standard solutions (10 ng/mL), urine samples from healthy participants, and urine spiked with 10 ng/mL standard solution were analyzed (Fig. 2). In the LC analysis, the retention of the analytes shortened with an increase in the proportion of ammonium solution in the MP5 mobile phases (data not shown), similar to that reported in previous studies.^40,44,46) Although some endogenous peaks were detected in all transitions, all analytes in the optimized LC conditions were separated from the interference peaks (Fig. 2). The separation of analytes was achieved using a gradient of 10 to 32.5% mobile phase B over 4.5 min (Fig. 2). The analytes and IS are acidic compounds. Because the pK_a of the charge of the sulfuric acid group on the compounds is markedly lower than the pH of all mobile phases, the sulfuric acid group on all compounds is always negatively ionized and is not influenced by the pH of the mobile phases. Further, as the pK_a of the carboxy group on SNAG-Δ⁵-CA, S7B-Δ⁵-CA, S7O-Δ⁵-CA, and IS, is approximately 4.5, their charge state must change according to the change from pH 5.5 (MP1) to approximately pH 11.5 (MP5). Ionization of the carboxy group increases the hydrophilicity and decreases the retention of analytes for octadecylsilyl-bonded columns. Ozaki et al. reported the shortening of the retention of negatively ionized fatty acids because the ion repulses the hydrophobic octadecylsilyl group of the column.⁴³⁾ In bile acid sulfates, which have common functional groups to the compounds in this study, the retention shortens with increasing pH of the mobile phase.^57,58) Applicating the LC condition to normal ODS column (L-column 2 ODS), the retention was shortened and the peak shapes were broadened by according to the continuous analyses (data not shown). In L-column 3, the peak shapes and retentions of analytes have been retained (data not shown). The silica particle and bonded ODS in L-column 3 have a resistance to the continuous analysis with a basic mobile phase. In addition, we used the 2 µm particle size of the ultra-high-performance column and performed good separation even during a short run time. Consequently, we succeeded to shorten the LC run time 60 min in the previous method²⁹⁾ to 7 min in this study.

Fig. 2. SRM Chromatograms of the Analytes and IS with Method S

(a) Ten ng/mL of the standard mixture, (b) urine of healthy participants, and (c) 10 ng/mL of standard mixture-spiked urine from a healthy participant. All analytes and the IS were separated from each other and completely separated from the contaminant peaks. SRM, selected reaction monitoring; IS, internal standard. SNAG-Δ⁵-CA, 3β-sulfooxy-7β-N-acetylglucosaminyl-5-cholenoic acid; SNAG-Δ⁵-CG, glycine-amidated 3β-sulfooxy-7β-N-acetylglucosaminyl-5-cholenoic acid; SNAG-Δ⁵-CT, taurine-amidated 3β-sulfooxy-7β-N-acetylglucosaminyl-5-cholenoic acid; S7B-Δ⁵-CA, 3β-sulfooxy-7β-hydroxy-5-cholenoic acid; S7O-Δ⁵-CA, 3β-sulfooxy-7-oxo-5-cholenoic acid.

Linearity, Calibration Curves, and Matrix Effects

Generally, calibration curves for drugs that are not normally contained are generated by adding standard solutions to the sample matrix.^59–63) However, when analytes are endogenous compounds that are found in all participants, an alternative approach must be selected.⁶⁴⁾ In this study, we prepared calibration curves using water as a surrogate matrix based on previous methods.^{18,20,26,29,55)} The calibration curves were prepared in the range of 0.3–1000 ng/mL, as employed in previous studies.²⁹⁾ As a result of optimization, the injection volume was optimized to 50 µL, which is one-quarter that used in a previous study.²⁹⁾ The calibration curves are presented in Supplementary Table S3. these curves showed high linearity over a wide range (Supplementary Table S3a). In addition, calibration curves were generated using standard spiked urine, and their slopes were compared with those of the standard solutions. The slope ratios of the calibration curves ranged from 88.5 to 112% (Supplementary Table S3b). The slope calibration curves obtained by the standard addition method⁶⁵⁾ did not differ from those prepared via the addition of water as a surrogate matrix. Next, the matrix effects were subsequently evaluated using matrix factors.^16,20,29) The matrix factor was calculated as the ratio of the peak intensities in standard solutions spiked with urine and those in neat standard solutions in water. The matrix factors of compounds ranged from 90.4 to 106%. The IS normalized matrix factors, which are the ratios of the matrix factors of the analytes and IS, were calculated; the values were found to range from 88.3 to 103%. IS-normalized matrix factors were calculated, and the values were found to range from 100 ± 15% (Supplementary Table S4b). Therefore, the compounds were not affected by the matrix effects in this method. In this study, online pretreatment and small particle column were adapted. This developed LC system might contribute to eliminate the matrix effect with urine. Accordingly, we concluded that there were no problems with the use of water as a surrogate matrix in this study as the analytes and IS were not influenced by matrix effects.

Reproducibility Test

We performed a reproducibility test, which consisted of intra- and interday assays. The results of the test are summarized in Table 1. Based on the intra-day assay, precision ranged from 0.329 to 11.7% while accuracy ranged from −5.66 to 8.45% (Table 1a). Based on the inter-day assay, precision ranged from 1.25 to 8.39% while accuracy ranged from −5.27 to 4.68% (Table 1b). Therefore, no problems were found. Overall, the developed analytical method was recognized to be highly reproducible and reliable. Of note, as the stability of the analytes was proven in a previous study,²⁹⁾ a stability test was not performed in this study.

Table 1. The Result of the Reproducibility Tests

(a) Intra-day assay
No.	Compound	Precision (%)				Accuracy (%)
		Blank	LQC	MQC	HQC	LQC	MQC	HQC
1	SNAG-Δ5-CA	3.87	2.17	5.62	2.66	1.10	−1.01	−0.328
2	SNAG-Δ5-CG	0.329	0.182	3.92	1.36	0.199	−5.66	−3.72
3	SNAG-Δ5-CT	11.7	1.62	1.98	1.43	4.56	8.45	5.77
4	S7B-Δ5-CA	4.08	2.07	4.47	2.03	−1.78	−0.391	−0.584
5	S7O-Δ5-CA	5.62	1.76	3.45	1.19	3.51	7.26	11.4

(Each QC were tested, N = 6)

(b) Inter-day assay
No.	Compound	Precision (%)				Accuracy (%)
		Blank	LQC	MQC	HQC	LQC	MQC	HQC
1	SNAG-Δ5-CA	1.37	1.25	1.29	1.80	−0.130	1.85	−0.126
2	SNAG-Δ5-CG	2.60	2.07	2.61	2.31	1.05	1.91	−1.06
3	SNAG-Δ5-CT	2.49	1.60	1.70	1.46	−0.421	2.07	0.999
4	S7B-Δ5-CA	3.31	2.86	2.63	3.44	1.29	1.17	−5.27
5	S7O-Δ5-CA	8.39	4.97	2.23	1.49	3.42	4.68	0.986

(Each QC were tested, N = 3) LQC, low quality control (2 ng/mL); MQC, middle quality control (50 ng/mL); HQC, high quality control (800 ng/mL); SNAG-Δ⁵-CA, 3β-Sulfooxy-7β-N-acetylglucosaminyl-5-cholenoic acid; SNAG-Δ⁵-CG, Glycine-amidated 3β-sulfooxy-7β-N-acetylglucosaminyl-5-cholenoic acid; SNAG-Δ⁵-CT, Taurine-amidated 3β-sulfooxy-7β-N-acetylglucosaminyl-5-cholenoic acid; S7B-Δ⁵-CA, 3β-Sulfooxy-7β-hydroxy-5-cholenoic acid; S7O-Δ⁵-CA, 3β-Sulfooxy-7-oxo-5-cholenoic acid.

Analysis of Five Urinary Cholesterol Metabolites in Healthy Controls and Patients with NPC

Urine samples from healthy participants (N = 33) and patients with NPC (N = 38) were analyzed using the method S. The sex ratio of human urine samples was found to significantly differ (p = 0.0385); however, there was no difference in age between patients with NPC and healthy participants (Supplementary Table S5).

The urinary concentrations are summarized in Supplementary Fig. S2 (all data are shown in Supplementary Table S6a). Although some of the samples could not be quantified as they were lower than the limit of quantification, most analytes could be analyzed in most of the samples (Supplementary Table S6a). In a previous report, the researchers faced difficulty in detecting metabolites with an N-acetylglucosamine group in some participants. Mazzacuva et al. reported that the mutation of UGT3A1 affects the conjugation and production of metabolites with N-acetylglucosamine groups,²⁴⁾ which aligns with the results of our previous studies.²⁹⁾ Regarding metabolites without N-acetylglucosamine, only one participant (NPC24) had a value lower than the limit of quantification for S7O-Δ⁵-CA (Supplementary Table S6b).

Significant differences were found in urinary mean concentrations between healthy participants and patients with NPC (Supplementary Table S7a). Next, the diagnostic performance was investigated using ROC analysis. The area under the curve (AUC) values ranged from 0.8418 to 0.9530 (Supplementary Fig. S3). The cutoff concentrations ranged from 12.7 ng/mg Cr to 60.4 ng/mg Cr, sensitivity ranged from 83.8 to 92.1%, and specificity ranged from 84.9 to 90.9% (Supplementary Fig. S4). These metabolites were reported to have good performance; however, the values were slightly lower than those reported previously.²⁹⁾ Therefore, the samples were further analyzed and their performances were verified.

Comparison of the Concentrations Obtained with the Current Method to Those Obtained with the Previously Developed Method

All urine samples were analyzed using the previously developed method (method L),²⁹⁾ and all concentrations are listed in Supplementary Table S6b. The mean concentrations obtained using method S were not significantly different from those obtained using method L (Supplementary Fig. S3, Supplementary Table S8). The diagnostic performance of method L was investigated using ROC analysis and the results are presented in Supplementary Fig. S5. The data from ROC analysis for both methods are summarized in Supplementary Table S9. Although the values showed a small difference between the two methods, generally similar diagnostic performances were identified. Finally, a correlation analysis was performed between the concentrations using the two methods; the results of this analysis are shown in Fig. 3. The results for all analytes had significant correlations between the two methods (p < 0.001). Further, the coefficient of determination (R²) ranged from 0.7007 to 0.9994, while the slope of the linear range ranged from 0.6537 to 1.447. One-third of the five analytes had good coefficient of determination values (SNAG-Δ⁵-CT, S7B-Δ⁵-CA, and S7O-Δ⁵-CA), with SNAG-Δ⁵-CA and SNAG-Δ⁵-CG having values of 0.7769 and 0.7007, respectively. Because the correlation regression of SNAG-Δ⁵-CG provided the high intercept (+190.92, Fig. 3b); and the concentration might differ in two methods. S7B-Δ⁵-CA provided good coefficient of determination and slope. A slope lower than 1.00 was obtained for SNAG-Δ⁵-CG (Fig. 3b) while slopes higher than 1.00 were obtained for SNAG-Δ⁵-CT and S7O-Δ⁵-CA (Figs. 3c, e).

Fig. 3. The Correlation of the Urinary Concentrations between the Two Methods, Method S and Method L

(a) SNAG-Δ⁵-CA, (b) SNAG-Δ⁵-CG, (c) SNAG-Δ⁵-CT, (d) S7B-Δ⁵-CA, and (e) S7O-Δ⁵-CA, respectively. SNAG-Δ⁵-CA, 3β-sulfooxy-7β-N-acetylglucosaminyl-5-cholenoic acid; SNAG-Δ⁵-CG, glycine-amidated 3β-sulfooxy-7β-N-acetylglucosaminyl-5-cholenoic acid; SNAG-Δ⁵-CT, taurine-amidated 3β-sulfooxy-7β-N-acetylglucosaminyl-5-cholenoic acid; S7B-Δ⁵-CA, 3β-sulfooxy-7β-hydroxy-5-cholenoic acid; S7O-Δ⁵-CA, 3β-sulfooxy-7-oxo-5-cholenoic acid.

Next, we performed the Bland–Altman analysis for investigating the concentration depending difference between both methods (Supplementary Fig. S6). As a result, some data were plotted out of ±1.96 S.D. in all analytes. In particular, SNAG-Δ⁵-CT and S7O-Δ⁵-CA provided some plots with less than −1.96 S.D. (Supplementary Figs. S6c, e). In SNAG-Δ⁵-CA and SNAG-Δ⁵-CG, because some plots with larger difference than ±1.96 S.D. were observed (Supplementary Figs. S6a, b); their coefficients of determination values were smaller than other analytes.

Furthermore, Passing–Bablok regression analysis was performed for evaluation of the correlation in two methods. As a result, the 95% confidence interval of slope of regression equations of SNAG-Δ⁵-CA, of SNAG-Δ⁵-CG, and S7B-Δ⁵-CA included 1.0 (y = x). In other, those of SNAG-Δ⁵-CT and S7O-Δ⁵-CA were out of over 1.0 of slopes. The difference of matrix effect in samples with larger concentrations might affect the result by speculating from the Bland-Altman analysis (Supplementary Fig. S6). In particular, two patients had different concentrations of the two analytes. In NPC 13, although the concentration of SNAG-Δ⁵-CA with method S was 16530 ng/mg Cr, that with method L was 6497 ng/mg Cr (Supplementary Table S6). Conversely, the concentration of SNAG-Δ⁵-CG was 7045 ng/mg Cr with method S and 17003 ng/mg Cr with method L. In NPC 9, the concentration of SNAG-Δ⁵-CA was 3938 ng/mg Cr with method S and 7033 ng/mg Cr with method L (Supplementary Table S5). Conversely, the concentration of SNAG-Δ⁵-CG was 8527 ng/mg Cr with method S and 3196 ng/mg Cr with method L. As a possibility, the huge matrix effect difference with two methods in those samples are considered; the detailed investigation will be performed in the future.

Next, we tried to analysis the correlations of two methods by excluding samples with more than 3000 ng/mg Cr in method S (Supplementary Fig. S8). As a result, both slopes and coefficient of determinations of the correlation equations were improved generally, comparing with the whole dataset (Fig. 3). Therefore, the samples with 3000 ng/mg Cr in method S might have the larger matrix effect.

By comparing the diagnostic performance of the two methods, the values for AUC, sensitivity, and specificity were found to be similar (i.e., within ±10%; Supplementary Table S9). However, the cut-off values were found to differ between two methods, with ratios ranging from 30% in SNAG-Δ⁵-CG to 172.3% in SNAG-Δ⁵-CT. A similar cut-off concentration was obtained for S7B-Δ⁵-CA between the two methods, with 57.6 ng/mg Cr using method S and 72.7 ng/mg Cr using method L. The diagnostic performance of S7B-Δ⁵-CA with method S also provided the highest AUC value. Therefore, this analyte is useful for the diagnostic screening of NPC. Although the value is slightly lower than that of other biomarkers, such as plasma oxysterols,^14,15) plasma cholestanoic acids,²³⁾ and plasma N-palmitoyl-O-phosphocholine-serine,¹⁸⁾ urinary biomarkers are advantageous in terms of noninvasive collection. Therefore, these biomarkers might be useful alternatives for NPC diagnostic screening. Further, the developed method S might be useful for the diagnostic screening of NPC using urine specimens, and the basic additive could be useful for increasing ion intensities in negative-ion mode.

CONCLUSION

By using basic mobile phase additives, we developed an LC/MS/MS method for analyzing urinary cholesterol metabolites with increased MS intensity and reduced LC running time. We succeeded at increasing the MS intensities by approximately 30-fold at the maximum compared to that obtained with the previous method that used a weakly acidic mobile phase. In addition, the LC running time was shortened from 60 to 7 min. The developed method had satisfactory reproducibility in analytical method validation and proved useful for the diagnostic screening of NPC. Overall, this method might be useful as a noninvasive alternative diagnostic screening method for other plasma biomarkers of NPC.

Acknowledgments

We thank all donors for providing urine samples. This work was supported in part by JSPS KAKENHI 21K07814.

Conflict of Interest

The authors declare no conflict of interest.

Supplementary Materials

This article contains supplementary materials.

REFERENCES

• 1) Vanier MT. Niemann–Pick disease type C. Orphanet J. Rare Dis., 5, 16 (2010).
• 2) Patterson MC, Hendriksz CJ, Walterfang M, Sedel F, Vanier MT, Wijburg F. Recommendations for the diagnosis and management of Niemann–Pick disease type C: an update. Mol. Genet. Metab., 106, 330–344 (2012).
• 3) Geberhiwot T, Moro A, Dardis A, et al. Consensus clinical management guidelines for Niemann–Pick disease type C. Orphanet J. Rare Dis., 13, 50 (2018).
• 4) Kwon HJ, Abi-Mosleh L, Wang ML, Deisenhofer J, Goldstein JL, Brown MS, Infante RE. Structure of N-terminal domain of NPC1 reveals distinct subdomains for binding and transfer of cholesterol. Cell, 137, 1213–1224 (2009).
• 5) Pentchev PG, Comly ME, Kruth HS, Vanier MT, Wenger DA, Patel S, Brady RO. A defect in cholesterol esterification in Niemann–Pick disease (type C) patients. Proc. Natl. Acad. Sci. U.S.A., 82, 8247–8251 (1985).
• 6) Pentchev PG, Boothe AD, Kruth HS, Weintroub H, Stivers J, Brady RO. A genetic storage disorder in BALB/C mice with a metabolic block in esterification of exogenous cholesterol. J. Biol. Chem., 259, 5784–5791 (1984).
• 7) Pentchev PG, Comly ME, Kruth HS, Vanier MT, Wenger DA, Patel S, Brady RO. A defect in cholesterol esterification in Niemann-Pick disease (type C) patients. Proc. Natl. Acad. Sci. U.S.A., 82, 8247–8251 (1985).
• 8) Vanier MT. Biochemical studies in Niemann-Pick disease. I. Major sphingolipids of liver and spleen. Biochim. Biophys. Acta BBALipids Lipid Metab., 750, 178–184 (1983).
• 9) Svennerholm L, Vanier MT. The distribution of lipids in the human nervous system. IV. Fatty acid composition of major sphingolipids of human infant brain. Brain Res., 55, 413–423 (1973).
• 10) Patterson MC, Vecchio D, Prady H, Abel L, Wraith JE. Miglustat for treatment of Niemann–Pick C disease: a randomised controlled study. Lancet Neurol., 6, 765–772 (2007).
• 11) Maekawa M, Mano N. Identification and evaluation of biomarkers for Niemann–Pick disease type C based on chemical analysis techniques. Chromatography, 41, 19–29 (2020).
• 12) Maekawa M, Iwahori A, Mano N. Biomarker analysis of Niemann–Pick disease type C using chromatography and mass spectrometry. J. Pharm. Biomed. Anal., 191, 113622 (2020).
• 13) Maekawa M, Mano N. Cutting-edge LC–MS/MS applications in clinical mass spectrometry: focusing on analysis of drugs and metabolites. Biomed. Chromatogr., 36, e5347 (2022).
• 14) Jiang X, Sidhu R, Porter FD, Yanjanin NM, Speak AO, te Vruchte DT, Platt FM, Fujiwara H, Scherrer DE, Zhang J, Dietzen DJ, Schaffer JE, Ory DS. A sensitive and specific LC-MS/MS method for rapid diagnosis of Niemann–Pick C1 disease from human plasma. J. Lipid Res., 52, 1435–1445 (2011).
• 15) Porter FD, Scherrer DE, Lanier MH, Langmade SJ, Molugu V, Gale SE, Olzeski D, Sidhu R, Dietzen DJ, Fu R, Wassif CA, Yanjanin NM, Marso SP, House J, Vite C, Schaffer JE, Ory DS. Cholesterol oxidation products are sensitive and specific blood-based biomarkers for Niemann–Pick C1 disease. Sci. Transl. Med., 2, 56ra81 (2010).
• 16) Welford RWD, Garzotti M, Marques Lourenço C, Mengel E, Marquardt T, Reunert J, Amraoui Y, Kolb SA, Morand O, Groenen P. Plasma lysosphingomyelin demonstrates great potential as a diagnostic biomarker for Niemann–Pick disease type C in a retrospective study. PLOS ONE, 9, e114669 (2014).
• 17) Sidhu R, Mondjinou Y, Qian M, Song H, Kumar AB, Hong X, Hsu F-F, Dietzen DJ, Yanjanin NM, Porter FD, Berry-Kravis E, Vite CH, Gelb MH, Schaffer JE, Ory DS, Jiang X. N-acyl-O-phosphocholineserines: structures of a novel class of lipids that are biomarkers for Niemann–Pick C1 disease. J. Lipid Res., 60, 1410–1424 (2019).
• 18) Maekawa M, Jinnoh I, Matsumoto Y, et al. Structural determination of lysosphingomyelin-509 and discovery of novel class lipids from patients with Niemann–Pick disease type C. Int. J. Mol. Sci., 20, 5018 (2019).
• 19) Sidhu R, Kell P, Dietzen DJ, Farhat NY, Do AND, Porter FD, Berry-Kravis E, Vite CH, Reunert J, Marquardt T, Giugliani R, Lourenço CM, Bodamer O, Wang RY, Plummer E, Schaffer JE, Ory DS, Jiang X. Application of N-palmitoyl-O-phosphocholineserine for diagnosis and assessment of response to treatment in Niemann–Pick type C disease. Mol. Genet. Metab., 129, 292–302 (2020).
• 20) Iwahori A, Maekawa M, Narita A, Kato A, Sato T, Ogura J, Sato Y, Kikuchi M, Noguchi A, Higaki K, Okuyama T, Takahashi T, Eto Y, Mano N. Development of a diagnostic screening strategy for Niemann–Pick diseases based on simultaneous liquid chromatography-tandem mass spectrometry analyses of N-palmitoyl-O-phosphocholine-serine and sphingosylphosphorylcholine. Biol. Pharm. Bull., 43, 1398–1406 (2020).
• 21) Jiang X, Sidhu R, Orsini JJ, Farhat NY, Porter FD, Berry-Kravis E, Schaffer JE, Ory DS. Diagnosis of Niemann–Pick C1 by measurement of bile acid biomarkers in archived newborn dried blood spots. Mol. Genet. Metab., 126, 183–187 (2019).
• 22) Sidhu R, Kell P, Dietzen DJ, Farhat NY, Do AND, Porter FD, Berry-Kravis E, Reunert J, Marquardt T, Giugliani R, Lourenço CM, Wang RY, Movsesyan N, Plummer E, Schaffer JE, Ory DS, Jiang X. Application of a glycinated bile acid biomarker for diagnosis and assessment of response to treatment in Niemann–Pick disease type C1. Mol. Genet. Metab., 131, 405–417 (2020).
• 23) Jiang X, Sidhu R, Mydock-McGrane L, Hsu F-F, Covey DF, Scherrer DE, Earley B, Gale SE, Farhat NY, Porter FD, Dietzen DJ, Orsini JJ, Berry-Kravis E, Zhang X, Reunert J, Marquardt T, Runz H, Giugliani R, Schaffer JE, Ory DS. Development of a bile acid-based newborn screen for Niemann–Pick disease type C. Sci. Transl. Med., 8, 337ra63 (2016).
• 24) Mazzacuva F, Mills P, Mills K, Camuzeaux S, Gissen P, Nicoli E-R, Wassif C, te Vruchte D, Porter FD, Maekawa M, Mano N, Iida T, Platt F, Clayton PT. Identification of novel bile acids as biomarkers for the early diagnosis of Niemann–Pick C disease. FEBS Lett., 590, 1651–1662 (2016).
• 25) Alvelius G, Hjalmarson O, Griffiths WJ, Björkhem I, Sjövall J. Identification of unusual 7-oxygenated bile acid sulfates in a patient with Niemann–Pick disease, type C. J. Lipid Res., 42, 1571–1577 (2001).
• 26) Maekawa M, Misawa Y, Sotoura A, Yamaguchi H, Togawa M, Ohno K, Nittono H, Kakiyama G, Iida T, Hofmann AF, Goto J, Shimada M, Mano N. LC/ESI-MS/MS analysis of urinary 3β-sulfooxy-7β-N-acetylglucosaminyl-5-cholen-24-oic acid and its amides: new biomarkers for the detection of Niemann–Pick type C disease. Steroids, 78, 967–972 (2013).
• 27) Maekawa M, Shimada M, Ohno K, Togawa M, Nittono H, Iida T, Hofmann AF, Goto J, Yamaguchi H, Mano N. Focused metabolomics using liquid chromatography/electrospray ionization tandem mass spectrometry for analysis of urinary conjugated cholesterol metabolites from patients with Niemann–Pick disease type C and 3β-hydroxysteroid dehydrogenase deficiency. Ann. Clin. Biochem., 52, 576–587 (2015).
• 28) Maekawa M, Omura K, Sekiguchi S, Iida T, Saigusa D, Yamaguchi H, Mano N. Identification of two sulfated cholesterol metabolites found in the urine of a patient with Niemann–Pick disease type C as novel candidate diagnostic markers. Mass Spectrom. (Tokyo), 5, S0053 (2016).
• 29) Maekawa M, Jinnoh I, Narita A, Iida T, Saigusa D, Iwahori A, Nittono H, Okuyama T, Eto Y, Ohno K, Clayton PT, Yamaguchi H, Mano N. Investigation of diagnostic performance of five urinary cholesterol metabolites for Niemann–Pick disease type C. J. Lipid Res., 60, 2074–2081 (2019).
• 30) Masamitsu M, Anna I, Nariyasu M. Identification and diagnostic performance evaluation of biomarkers for Niemann–Pick disease type C using mass spectrometry. Japanese Journal of Clinical Chemistry, 50, 20–26 (2021).
• 31) Maekawa M, Yamaguchi H, Nariyasu M. Conjugated cholesterol metabolites as urinary diagnostic biomarkers of Niemann–Pick disease type C. JSBMS Lett., 43, 45–50 (2018).
• 32) Maekawa M, Yamaguchi H, Mano N. Discovery of diagnostic markers for Niemann–Pick disease type C by focused metabolomics. Japanese Journal of Clinical Chemistry, 47, 161–168 (2018).
• 33) Mano N, Maekawa M, Yamaguchi H. Clinical chemistry based on highly accurate separation analysis technology. Chromatography, 39, 49–57 (2018).
• 34) Maekawa M, Yamaguchi H, Mano N. Analytical Technologies Used in the Medical Front. Bunseki Kagaku, 66, 413–419 (2017).
• 35) Clayton PT. Disorders of bile acid synthesis. J. Inherit. Metab. Dis., 34, 593–604 (2011).
• 36) Clayton PT. Applications of mass spectrometry in the study of inborn errors of metabolism. J. Inherit. Metab. Dis., 24, 139–150 (2001).
• 37) Clayton PT. Inborn errors of bile acid metabolism. J. Inherit. Metab. Dis., 14, 478–496 (1991).
• 38) Tan A, Fanaras JC. Use of high-pH (basic/alkaline) mobile phases for LC–MS or LC–MS/MS bioanalysis. Biomed. Chromatogr., 33, e4409 (2019).
• 39) Tan A, Gu G, Gui X, Wu Y, Bolger G, Licollari A, Fanaras JC. Improving the selectivity and sensitivity for quantifying 8-α-hydroxy-mutilin in rabbit tissues by using basic mobile phases and negative ionization. J. Chromatogr. B Analyt. Technol. Biomed. Life Sci., 1072, 142–148 (2018).
• 40) Tan A, Wu Y, Wong M, Licollari A, Bolger G, Fanaras JC, Shopp G, Helson L. Use of basic mobile phase to improve chromatography and boost sensitivity for quantifying tetrahydrocurcumin in human plasma by LC-MS/MS. J. Chromatogr. B Analyt. Technol. Biomed. Life Sci., 1028, 86–93 (2016).
• 41) Zhang D, Fu Y, Gale JP, Aubry AF, Arnold ME. A sensitive method for the determination of entecavir at picogram per milliliter level in human plasma by solid phase extraction and high-pH LC-MS/MS. J. Pharm. Biomed. Anal., 49, 1027–1033 (2009).
• 42) Peng L, Farkas T. Analysis of basic compounds by reversed-phase liquid chromatography-electrospray mass spectrometry in high-pH mobile phases. J. Chromatogr. A, 1179, 131–144 (2008).
• 43) Ozaki H, Nakano Y, Sakamaki H, Yamanaka H, Nakai M. Basic eluent for rapid and comprehensive analysis of fatty acid isomers using reversed-phase high performance liquid chromatography/Fourier transform mass spectrometry. J. Chromatogr. A, 1585, 113–120 (2019).
• 44) Nakano Y, Nagasawa E, Uchida T, Sakamaki H. Highly sensitive detection of pesticide residues by LC-MS/MS using ammonia solution as an eluent. Bunseki Kagaku, 69, 141–150 (2020).
• 45) Takkis K, Aro R, Kõrgvee LT, Varendi H, Lass J, Herodes K, Kipper K. Signal enhancement in the HPLC-ESI-MS/MS analysis of spironolactone and its metabolites using HFIP and NH4F as eluent additives. Anal. Bioanal. Chem., 409, 3145–3151 (2017).
• 46) Li XS, Li S, Kellermann G. Simultaneous determination of three estrogens in human saliva without derivatization or liquid-liquid extraction for routine testing via miniaturized solid phase extraction with LC-MS/MS detection. Talanta, 178, 464–472 (2018).
• 47) Gaudl A, Kratzsch J, Bae YJ, Kiess W, Thiery J, Ceglarek U. Liquid chromatography quadrupole linear ion trap mass spectrometry for quantitative steroid hormone analysis in plasma, urine, saliva and hair. J. Chromatogr. A, 1464, 64–71 (2016).
• 48) Häkkinen MR, Heinosalo T, Saarinen N, Linnanen T, Voutilainen R, Lakka T, Jääskeläinen J, Poutanen M, Auriola S. Analysis by LC–MS/MS of endogenous steroids from human serum, plasma, endometrium and endometriotic tissue. J. Pharm. Biomed. Anal., 152, 165–172 (2018).
• 49) Boegh Andersen I, Lohman Brasen C, Skov Madsen J, Schmedes A. Quantitation of vitamin K1 in serum using online SPE-LC-MS/MS and the challenges of working with vitamin K. J. Chromatogr. B Analyt. Technol. Biomed. Life Sci., 1117, 41–48 (2019).
• 50) Abe A, Maekawa M, Sato T, Sato Y, Kumondai M, Takahashi H, Kikuchi M, Higaki K, Ogura J, Mano N. Metabolic alteration analysis of steroid hormones in Niemann–Pick disease type C model cell using liquid chromatography/tandem mass spectrometry. Int. J. Mol. Sci., 23, 4459 (2022).
• 51) Ohnishi T, Balan S, Toyoshima M, et al. Investigation of betaine as a novel psychotherapeutic for schizophrenia. EBioMedicine, 45, 432–446 (2019).
• 52) Iida T, Kakiyama G, Hibiya Y, Miyata S, Inoue T, Ohno K, Goto T, Mano N, Goto J, Nambara T, Hofmann AF. Chemical synthesis of the 3-sulfooxy-7-N-acetylglucosaminyl-24-amidated conjugates of 3β,7β-dihydroxy-5-cholen-24-oic acid, and related compounds: Unusual, major metabolites of bile acid in a patient with Niemann–Pick disease type C1. Steroids, 71, 18–29 (2006).
• 53) Kakiyama G, Muto A, Shimada M, Mano N, Goto J, Hofmann AF, Iida T. Chemical synthesis of 3beta-sulfooxy-7beta-hydroxy-24-nor-5-cholenoic acid: an internal standard for mass spectrometric analysis of the abnormal delta5-bile acids occurring in Niemann–Pick disease. Steroids, 74, 766–772 (2009).
• 54) Maekawa M, Mori M, Fujiyoshi M, Suzuki H, Yanai K, Noda A, Tanaka M, Takasaki S, Kikuchi M, Akasaka K, Kisara S, Matsuura M, Hisamichi K, Sato M, Goto J, Shimada M, Yamaguchi H, Mano N. A direct injection LC/ESI-MS/MS analysis of urinary cyclophosphamide as an anticancer drug for monitoring occupational exposure. Chromatography, 39, 41–47 (2018).
• 55) Maekawa M, Narita A, Jinnoh I, Iida T, Marquardt T, Mengel E, Eto Y, Clayton PT, Yamaguchi H, Mano N. Diagnostic performance evaluation of sulfate-conjugated cholesterol metabolites as urinary biomarkers of Niemann–Pick disease type C. Clin. Chim. Acta, 494, 58–63 (2019).
• 56) Shoji S, Maekawa M, Ogura J, Sato T, Mano N. Identification cholesterol metabolites altered before the onset of nonalcoholic steatohepatitis by targeted metabolomics. Biochim. Biophys. Acta Mol. Cell Biol. Lipids, 1867, 159135 (2022).
• 57) Goto J, Kato H, Nambara T. Separation of monosulfated bile acids by high-performance liquid chromatography. J. Liq. Chromatogr., 3, 645–655 (1980).
• 58) Goto J, Chikai T, Nambara T. Separation of bile acid 7- and 12-sulfates by high performance liquid chromatography with pre-column fluorescence labeling. Anal. Sci., 2, 175–178 (1986).
• 59) Maekawa M, Mano N. Cutting-edge LC–MS/MS applications in clinical mass spectrometry: focusing on analysis of drugs and metabolites. Biomed. Chromatogr., 36, e5347 (2022).
• 60) Hirasawa T, Kikuchi M, Shigeta K, Takasaki S, Sato Y, Sato T, Ogura J, Onodera K, Fukuhara N, Onishi Y, Maekawa M, Mano N. High-throughput liquid chromatography/electrospray ionization–tandem mass spectrometry method using in-source collision-induced dissociation for simultaneous quantification of imatinib, dasatinib, bosutinib, nilotinib, and ibrutinib in human plasma. Biomed. Chromatogr., 35, e5124 (2021).
• 61) Sato Y, Shigeta K, Hirasawa T, Sato T, Ogura J, Maekawa M, Ebata A, Hamanaka Y, Tada H, Ishida T, Kikuchi M, Mano N. Establishment of an analytical method for simultaneous quantitation of CDK4/6 inhibitors, aromatase inhibitors, and an estrogen receptor antagonist in human plasma using LC-ESI-MS/MS. J. Chromatogr. B Analyt. Technol. Biomed. Life Sci., 1173, 122655 (2021).
• 62) Sato T, Suzuka M, Sato Y, Iwabuchi R, Kobayashi D, Ogura J, Takasaki S, Yokota M, Tsukamoto T, Hayakawa Y, Kikuchi M, Maekawa M, Mano N. Development of a simultaneous analytical method for clozapine and its metabolites in human plasma using liquid chromatography/electrospray ionization tandem mass spectrometry with linear range adjusted by in-source collision-induced dissociation. Biomed. Chromatogr., 35, e5094 (2021).
• 63) Takasaki S, Hirasawa T, Sato Y, Maekawa M, Tsukamoto T, Kikuchi M, Ogura J, Hayakawa Y, Matsuda Y, Oishi H, Sado T, Noda M, Okada Y, Yamaguchi H, Mano N. Simultaneous analysis of drugs administered to lung-transplanted patients using liquid chromatography–tandem mass spectrometry for therapeutic drug monitoring. Biomed. Chromatogr., 35, e5067 (2021).
• 64) Thakare R, Chhonker YS, Gautam N, Alamoudi JA, Alnouti Y. Quantitative analysis of endogenous compounds. J. Pharm. Biomed. Anal., 128, 426–437 (2016).
• 65) Bader M. A systematic approach to standard addition methods in instrumental analysis. J. Chem. Educ., 57, 703–706 (1980).