A Method for Simultaneous Determination of 25-Hydroxyvitamin D 3 and Its 3-Sulfate in Newborn Plasma by LC/ESI-MS/MS after Derivatization with a Proton-Affinitive Cookson-Type Reagent

A method for the simultaneous determination of 25-hydroxyvitamin D 3 [25(OH)D 3 ] and its 3-sulfate [25(OH)D 3 S] in newborn plasma, which is expected to be helpful in the assessment of the vitamin D status, using stable isotope-dilution liquid chromatography/electrospray ionization-tandem mass spectrometry (LC/ESI-MS/MS) has been developed and validated. The plasma was pretreated based on the deprotein-ization and solid-phase extraction, then subjected to derivatization with 4-(4-dimethylaminophenyl)-1,2,4-triazoline-3,5-dione (DAPTAD). The derivatization enabled the accurate quantification of 25(OH)D 3 without interference from 3-epi-25(OH)D 3 and also facilitated the simultaneous determination of the two metabolites by LC/positive ESI-MS/MS. Quantification was based on the selected reaction monitoring with the characteristic fragmentation of the DAPTAD-derivatives during MS/MS. This method was reproducible (intra- and inter-assay relative standard deviations of 7.8% or lower for both metabolites) and accurate (analytical recovery, 95.4–105.6%). The limits of quantification were 1.0 ng/mL and 2.5 ng/mL for 25(OH)D 3 and 25(OH)D 3 S, respectively, when using a 20- µ L sample. The developed method was applied to the simultaneous determination of plasma 25(OH)D 3 and 25(OH)D 3 S in newborns; it was recognized that the plasma concentration of 25(OH)D 3 S is significantly higher than that of 25(OH)D 3 , and preterm newborns have lower plasma 25(OH)D 3 S concentrations than full-term newborns.


INTRODUCTION
e plasma/serum concentration of 25-hydroxyvitamin D 3 [25(OH) D 3 ], the major circulating metabolite of vitamin D 3 , is well recognized as an indicator of the vitamin D status. 1) Vitamin D de ciency/insu ciency in a newborn/ infant is associated not only with several bone metabolic diseases, such as rickets, but also with a wide range of adverse health outcomes, such as type 1 diabetes, 2) multiple sclerosis 3) and schizophrenia, 4) in later life. 25-Hydroxyvitamin D 3 3-sulfate [25(OH) D 3 S], the sulfated conjugate of 25(OH) D 3 , is another major metabolite of vitamin D 3 , and its circulating level was found to be much higher than that of 25(OH) D 3 in infants. 5) 25(OH) D 3 S might be the storage form of vitamin D 3 ; 25(OH) D 3 S may be utilized a er deconjugation to 25(OH) D 3 . 5-7) erefore, it is expected that the simultaneous determination of 25(OH) D 3 and 25(OH) D 3 S in plasma/serum is also helpful in the assessment of the vitamin D status and diagnosis for vitamin D de ciency/insufciency of newborns/infants. Liquid chromatography/electrospray ionization-tandem mass spectrometry (LC/ESI-MS/MS) is now the most commonly used method to determine vitamin D metabolites in various biological samples due to its high speci city and accuracy. [8][9][10] Many LC/ESI-MS/MS assays for the serum/ plasma 25(OH) D 3 have been reported, but not without a problem in these assays. One of the complicated problems in the 25(OH) D 3 quanti cation is the potential interference from its inactive epimer, 3-epi-25-hydroxyvitamin D 3 [3-epi-25(OH) D 3 ], leading to overestimation of the true 25(OH) D 3 concentrations 11,12) ; it was reported that the epimer contributed 9-61% of the total 25(OH) D 3 in infants aged 0. 11) To overcome this problem, we have developed 4-(4-dimethylaminophenyl)-1,2,4-triazoline-3,5dione (DAPTAD) as a novel Cookson-type reagent 13) (Fig.  1).
e Cookson-type reagent is the 4-substituted-1,2,4triazoline-3,5-dione, which quantitatively reacts with the s-cis-diene of vitamin D compounds to give the stable Diels-Alder adducts. By using DAPTAD, the near-baseline separation of 25(OH) D 3 from 3-epi-25(OH) D 3 on the reversed-phase LC using an octadecylsilyl-silica gel column was achieved. 13) e reversed-phase LC separation of these epimers was di cult by the derivatization with 4-phenyl-1,2,4-triazoline-3,5-dione (PTAD), a representative Cookson-type reagent. 12) Furthermore, the detection response of the DAPTAD-derivatized 25(OH) D 3 in the positive ESI-MS/MS was 30-fold and twice higher than those of the intact and PTAD-derivatized 25(OH) D 3 , respectively, due to the high proton-a nity of the dimethylamino group of DAPTAD and good fragmentation behavior during MS/MS; the collision-induced dissociation (CID) of the protonated molecule of the DAPTAD-derivative of the vitamin D metabolite produces an intense fragment ion derived from the cleavage of the C6-7 bond of the vitamin D skeleton. is ESI-MS/MS responsive property of the DAPTAD-derivative works well in the 25(OH) D 3 assay when only a limited blood sample volume is obtained from newborns/infants. 13) us, DAPTAD is a promising derivatization reagent for the trace and accurate quanti cation of 25(OH) D 3 in small volume blood samples.
In our previous study, the 25(OH) D 3 S was separately quanti ed by LC/negative ESI-MS/MS from the 25(OH) D 3 quanti cation. 5) Because 25(OH) D 3 S has a strong acidic and fragmentable moiety, i.e., a sulfate ester, it can be quanti ed without any derivatization. However, for the simultaneous quanti cation of 25(OH) D 3 and 25(OH) D 3 S, the DAPTADderivatization has the advantage of being able to remove the interference from 3-epi-25(OH) D 3 and provide the reliable measured values of 25(OH) D 3 .
Based on this background information, the objective of this study was to develop and validate an LC/ESI-MS/MS method for the simultaneous determination of 25(OH) D 3 and 25(OH) D 3 S in newborn plasma with a small sample volume. e method employed the DAPTAD-derivatization for the accurate 25(OH) D 3 quanti cation. e application of the method to the determination of the plasma 25(OH) D 3 and 25(OH) D 3 S in preterm and full-term newborns is also described.

Chemicals and reagents
25(OH) D 3 and 3-epi-25(OH) D 3 were purchased from Wako Pure Chemical Industries, Ltd. (Osaka, Japan) and Cayman Chemical Company (Ann Arbor, MI, USA), respectively. 25(OH) D 3 S and 7-dehydro-25-hydroxycholesterol 3-sulfate were synthesized from 25(OH) D 3 and 7-dehydro-25-hydroxycholesterol, respectively, according to the method of Iida et al. 14) in our laboratories. [ 3 and 25(OH) D 3 S were prepared as 2.5 µg/mL solutions in ethanol, and their concentrations were con rmed by UV spectroscopy using the molar absorptivity (ε) of 18200 at 265 nm. Subsequent dilutions were carried out using ethanol to prepare 2.0, 5.0, 10, 20, 50 and 100 ng/mL solutions. e ethanolic solutions of the ISs (25 ng/mL) were also prepared. DAPTAD was the same as used in a previous study. 13) e DAPTAD precursor (urazole form), which can be easily converted into DAPTAD by the oxidation, is now commercially available from several manufacturers. An Oasis ® HLB cartridge (30 mg adsorbent; Waters, Milford, MA, USA) was successively washed with methanol (1 mL) and water (1 mL) prior to use. All other reagents and solvents were of analytical grade or LC/MS grade.

Pretreatment of plasma
e plasma (20 µL) was added to acetonitrile (50 µL) containing the ISs (0.50 ng each), vortex-mixed for 30 s and centrifuged at 1000×g (10 min). e supernatant was diluted with water (300 µL), and the sample was passed through an Oasis ® HLB cartridge. A er washing with water (1 mL) and methanol-water (1 : 1, v/v) (1 mL), the vitamin D 3 metabolites and ISs were eluted with methanol (1 mL). A er removal of the solvent, the residue was subjected to derivatization with DAPTAD. When the 25(OH) D 3 S concentration was over 50 ng/mL, 10 µL of the plasma was used for the quanti cation.
Derivatization e standard vitamin D 3 metabolites and pretreated plasma samples were dissolved in ethyl acetate (50 µL) containing DAPTAD (20 µg). e mixture was stored at room temperature for 1 h, then ethanol (40 µL) was added to the mixture to terminate the reaction. e solvent was evaporated, then the residue was dissolved in mobile phase A (60 µL), 15 µL of which was injected into the LC/ESI-MS/MS.

Comparison of positive-and negative-ion modes for detection of 25(OH)D 3 S-DAPTAD
e positive-and negative-ion modes were compared for the detection of the DAPTAD-derivatized 25(OH) D 3 S. e limits of detection [LODs; the amount of the derivatized 25(OH) D 3 S per injection giving a signal to noise ratio (S/N) of 5] were determined in both modes. e S/N value was manually calculated by division of the peak height of the derivatized 25(OH) D 3 S by the noise level around the peak. 25(OH) D 3 S (100 pg) was derivatized with DAPTAD and the resulting derivative was dissolved in the mobile phases (100 µL) described below, then subjected to LC/ESI-MS/MS. By stepwise decreasing the injection volume of the resulting solution, the LOD was determined. e mobile phase (isocratic elution) used in the positive-ion mode was methanol-10 mM ammonium formate (4 : 1, v/v) containing 0.05% (v/v) formic acid, and that in the negative-ion mode was methanol-10 mM ammonium formate (4 : 1, v/v).

Precision and accuracy (analytical recovery)
e intra-and inter-assay precisions were assessed by the repeated measurements (n=5) of two plasma samples (A and B) on one day and over ve days, respectively. e precision was determined as the relative standard deviation (RSD, %). e assay accuracy was determined using plasma samples A and B. e plasma (20 µL) was added to acetonitrile (50 µL) containing the ISs (0.50 ng each), 25(OH) D 3 and 25(OH) D 3 S (0.10 or 0.20 ng each; corresponding to 5.0 or 10 ng/mL) (spiked sample), and pretreated, then derivatized in the same way as described in the above sections. e assay accuracy (analytical recovery) was de ned as e recovery rates were calculated from the peak area ratio [25(OH) D 3 / 2 H 3 -25(OH) D 3 or 25(OH) D 3 S/ 2 H 6 -25(OH) D 3 S] in samples I and II as described below; recovery=peak area ratio in sample II/peak area ratio in sample I.
Sample II (n=6): e plasma (20 µL) was added to acetonitrile (50 µL) and pretreated in the same way as already described. A er the addition of the ISs (0.50 ng each) to this pretreated plasma, the resulting sample was subjected to the derivatization.

Newborn plasma sample
e anonymized plasma samples from 59 Japanese newborns (gestational age, 30.1-41.0 weeks) including 39 preterm newborns of both sexes were examined. Blood was collected from their dorsal hand vein within 28 days a er birth at the Shizuoka Saiseikai General Hospital (Shizuoka, Japan). Written informed consent was obtained from their parents. irteen cord blood samples were also obtained at delivery, and then the plasma was separated. e experimental procedures were approved by the Ethics Committees of the Tokyo University of Science and Shizuoka Saiseikai General Hospital.
Although the plasma was used in this study for no particular reason, it was reported that there is no signi cant di erence between the 25(OH) D 3 concentration in plasma and that in serum. 15) 25(OH) D 3 in plasma/serum has also been demonstrated to be extremely stable; when it is stored at room temperature, marked degradation is not observed at least for 24 h. 15)

Derivatization of 25(OH)D 3 and 25(OH)D 3 S with DAPTAD
e DAPTAD-derivatization for the vitamin D 3 metabolites was carried out at room temperature for 1 h by reference to previous studies 13,16) (Fig. 1). Our previous study demonstrated that the derivatization rate was almost quantitative for 25(OH) D 3 under this reaction condition. 13 (Fig. 2b).
e LODs in the positive-ion mode were 0.37 and 6.3 fmol for the derivatized 25(OH) D 3 and 25(OH) D 3 S, respectively. e detection response of 25(OH) D 3 -DAPTAD was approximately 30-fold higher than that of the intact 25(OH) D 3 as previously reported. 13) e use of ammonium acetate and acetic acid as the mobile phase additives instead of ammonium formate and formic acid caused little change in the chromatographic behaviors of the derivatized vitamin D 3 metabolites, but decreased the assay sensitivity to onehalf for 25(OH) D 3  e calculated proton a nity of the N,N-dimethylaniline moiety in DAPTAD is reported to be 941 kJ/mol, which is much higher than that of benzene (750 kJ/mol) in PTAD. 19) is is a major reason why the DAPTAD-derivatization was superior to the PTAD-derivatization for the analysis of 25(OH) D 3 S by the positive ESI-MS/MS.
As already described, two SRM modes with transitions of m/z 699.6→421.2 (positive-ion mode) and m/z 697.6→96.9 (negative-ion mode) were applicable for the detection of 25(OH) D 3 S-DAPTAD, but the LOD in the negative-ion mode (10.4 fmol) was higher than that in the positive-ion mode (6.3 fmol). Based on this result, the positive-ion mode was employed for the detection of the 25(OH) D 3 S as the DAPTAD-derivative in the following studies.

LC behavior of DAPTAD-derivatized 25(OH)D 3 and 25(OH)D 3 S
e DAPTAD-derivatives of the vitamin D 3 metabolites consist of the 6R-(minor) and 6S-isomers (major), 13) therefore, the derivatives sometimes give characteristic two peaks on their chromatograms. Under the LC conditions used in this study, the retention times (t R s) of 25(OH) D 3 -DAPTAD were 7.1 (6R) and 8.2 min (6S). A satisfactory separation of 25(OH) D 3 and 3-epi-25(OH) D 3 , which is a potential inter-fering metabolite in the 25(OH) D 3 assay, 11,12) was achieved by the DAPTAD-derivatization; the resolution (Rs) for the major peak (t R 8.2 min) of 25(OH) D 3 -DAPTAD and 3-epi-25(OH) D 3 -DAPTAD (t R 7.7 min; the 6R/S-isomers of 3-epi-25(OH) D 3 -DAPTAD co-eluted as a single peak) was 1.40. 25(OH) D 3 S-DAPTAD was eluted at 6.1 min as a single peak. 7-Dehydro-25-hydroxycholesterol 3-sulfate, a possible endogenous steroid, has the same molecular weight as 25(OH) D 3 S and also reacts with DAPTAD. e DAPTADderivatized 7-dehydro-25-hydroxycholesterol 3-sulfate (t R 1.8 min) was chromatographically well separated from the derivatized 25(OH) D 3 S under the LC conditions and did not provide a fragment ion at m/z 421.2 during the MS/MS. us, this steroid did not interfere with the 25(OH) D 3 S measurement.

Pretreatment of plasma
e plasma (20 µL) was deproteinized in acetonitrile and the supernatant was then puri ed using an Oasis ® HLB cartridge. e absolute recovery rates [mean±standard deviation (S.D.), n=5] of 2 H 3 -25(OH) D 3 and 2 H 6 -25(OH) D 3 S from the plasma specimen were 83.1±2.1 and 84.8±4.0%, respectively. Because the present method employed a stable isotope dilution technique, the absolute recovery rates of 25(OH) D 3 and 25(OH) D 3 S from a plasma specimen were considered to be similar to those of the ISs. e chromatograms obtained from a newborn with ISs are shown in Fig. 3, in which the peaks corresponding to 25(OH) D 3 (t R 7.1 and 8.2 min) and 25(OH) D 3 S (t R 6.1 min) were clearly observed with satisfactory shapes.

Linearity and calibration curves
In order to assess the linearity, the ISs (0.50 ng each), ese results demonstrated that the plasma matrix had no in uence on the determination of 25(OH) D 3 and 25(OH) D 3 S as the DAPTAD-derivatives. Furthermore, it has been reported that calibration curves constructed with the standard samples (non-matrix sample) can be used for the quanti cation as long as the ISs co-elute with the analytes of interest; the matrix matching is not always necessary for the stableisotope dilution LC/MS/MS. 20) Based on these data, the calibration curves constructed using the standard solutions were used for the quanti cation in the following studies.

Assay precision and accuracy
e intra-assay (n=5) RSDs did not exceed 4.5 and 2.7% for 25(OH) D 3 and 25(OH) D 3 S, respectively, and good interassay (n=5) RSDs [not exceeding 7.8 and 5.9% for 25(OH) D 3 and 25(OH) D 3 S, respectively] were also obtained, as shown in Table 1. A satisfactory assay accuracy (analytical recoveries) ranging from 95.4 to 99.4% for 25(OH) D 3 and from 103.3 to 105.6% for 25(OH) D 3 S was obtained (Table 1). ese data indicated that the present method is highly reproducible and accurate.
Stability 25(OH) D 3 and 25(OH) D 3 S in the plasma were stable up to three additional freeze/thaw cycles; 101.1±2.1% and 99.6±1.3% (mean±S.D., n=3), respectively, of the initial measured values were obtained a er three additional freeze/thaw cycles. Furthermore, it was possible to store the plasma at −20°C without loss of either metabolite for at least 2 months.

Simultaneous determination of newborn plasma 25(OH)D 3 and 25(OH)D 3 S
e 25(OH) D 3 and 25(OH) D 3 S concentrations in the newborn plasma were simultaneously determined based on the developed method (Fig. 4). e plasma concentration of 25(OH) D 3 S was 30.1±12.2 ng/mL (mean±S.D., n=59) with the range of 3.9-65.1 ng/mL and signi cantly higher than that of 25(OH) D 3 (7.1±3.0 ng/mL, 2.7-17.0 ng/mL); these agreed with previously reported results. 5) As mentioned in the introduction, 25(OH) D 3 S might be the storage form of vitamin D 3 and utilized a er deconjugation to 25(OH) D 3 . 5-7) Based on this concept, the simultaneous determination of the plasma 25(OH) D 3 and 25(OH) D 3 S will be more helpful than the determination of 25(OH) D 3 alone in the assessment of the vitamin D status and diagnosis for vitamin D de ciency/insu ciency of newborns. e preterm newborns have lower plasma 25(OH) D 3 S concentrations (Pearson's correlation coe cient test, p<0.01), whereas the plasma  25(OH) D 3 concentration was not related to the gestational age (p=0.36). e low level of 25(OH) D 3 S may be a possible cause of rickets that is more common in the preterm newborns. e anonymized samples were examined in this study; no subject information other than the gestational age was provided. erefore, the gender di erences in the plasma concentrations of the vitamin D 3 metabolites were not examined in this study.
Using the developed method, the concentrations of 25(OH) D 3 and 25(OH) D 3 S in the cord plasma were also determined and compared with those in the plasma of newborns of 0 or 1 day old (n=13). For both metabolites, there were good correlations in the concentrations between the newborn plasma and cord plasma as shown in Fig. 5 (Pearson's correlation coe cient test, p<0.01).
is result indicates that the cord plasma can also be used as the specimen for the assessment of the vitamin D status for newborns.

CONCLUSION
We have developed the stable isotope-dilution LC/ESI-MS/MS method for the simultaneous determination of 25(OH) D 3 and 25(OH) D 3 S in newborn plasma. e method employed the DAPTAD-derivatization, which enabled the accurate quanti cation of 25(OH) D 3 without interference from 3-epi-25(OH) D 3 and also facilitated the simultaneous determination of the two metabolites by positive ESI-MS/ MS. e method was able to quantify 1.0-50 ng/mL of 25(OH) D 3 and 2.5-50 ng/mL of 25(OH) D 3 S with satisfactory accuracy and reproducibility using a 20-µL plasma sample. e developed method was successfully applied to the analysis of newborn plasma; it was recognized that the plasma concentration of 25(OH) D 3 S is signi cantly higher than that of 25(OH) D 3 in newborns, and preterm newborns have lower plasma 25(OH) D 3 S concentrations than full-term newborns.
us, this well-characterized method will prove helpful in the assessment of the vitamin D status for newborns.