Chinese Chemical Letters  2018, Vol. 29 Issue (1): 115-118   PDF    
Chiral derivatization coupled with liquid chromatography/mass spectrometry for determining ketone metabolites of hydroxybutyrate enantiomers
Qing-Yun Chenga, Jun Xionga, Fang Wanga,b, Bi-Feng Yuana, Yu-Qi Fenga    
a Key Laboratory of Analytical Chemistry for Biology and Medicine(Ministry of Education), Department of Chemistry, Wuhan University, Wuhan 430072, China;
b Department of Pharmacy, Dingxi Campus, Gansu University of Traditional Chinese Medicine, Dingxi 743000, China
Abstract: Ketone bodies are small lipid-derived molecules and the metabolism of ketone bodies interfaces with various physiological processes. 3-Hydroxybutyric acid (3HB) is the most stable ketone body and can be employed to supply energy source. 2-Hydroxybutyric acid (2HB), an isomer of 3HB, was demonstrated to be an early biomarker for both insulin resistance and impaired glucose regulation. Both 2HB and 3HB are chiral carboxylic acids and exist in two steric configurations (D/L-2HB and D/L-3HB). It is difficult for these enantiomers to be differentiated by routine analytical methods In the current study, we developed a strategy by chiral derivatization coupled with liquid chromatography/mass spectrometry (LC-ESI-MS) analysis for simultaneous determination of D-2HB, L-2HB, D-3HB and L-3HB enantiomers. (S)-(+)-1-(2-Pyrrolidinylmethyl)-pyrrolidine (PMP) was used for efficient labeling of HBs. Our results showed that the retention behavior of D/L-2HB and D/L-3HB enantiomers was greatly improved after labeling by PMP and the four derivatives can be distinctly separated on C18 reversed-phase column. Moreover, PMP chiral derivatization greatly enhanced the detection sensitivities of HBs up to 55.3 folds because of the introduction of easily ionizable tertiary amino group. Using this method, we simultaneously quantified D-2HB, L-2HB, D-3HB, and L-3HB enantiomers in human renal cell carcinoma (RCC) tissues and the tumor adjacent normal tissues. The result demonstrated that both D-3HB and L-3HB can be detected in human renal tissues, however, only L-2HB was detected in human renal tissues. In addition, the quantification results showed that the contents of D-3HB were approximate 10 folds higher than L-3HB. Taken together, the developed method offered an efficient approach for the sensitive analysis of D/L-2HB and D/L-3HB enantiomers, which may facilitate the in-depth study of the functions of HBs.
Key words: 2-Hydroxybutyric acid     3-Hydroxybutyric acid     Enantiomer     Chiral derivatization     Mass spectrometry    

Ketone bodies are small lipid-derived molecules including acetone, acetoacetic acid, and 3-hydroxybutyric acid (3HB) [1]. Ketone bodies are mainly produced in the liver of mammals through fatty acid oxidation and serve as alternative energy sources for tissues during carbohydrate restrictive diets, starvation, and prolonged exercise [2]. The metabolism of ketone bodies interfaces with the sterol biosynthesis, β-oxidation of fatty acids, de novo lipogenesis, tricarboxylic acid cycle, and intracellular signal transduction [3].

3HB is the most stable ketone body and can be employed to supply energy source when blood glucose is low [4]. In addition to as a carrier of energy from the liver to peripheral tissues during fasting or exercise, 3HB also possesses signaling activities and as an endogenous inhibitor of histone deacetylases [5]. Recently, 3HB was found to be able to block NLRP3 inflammasome-mediated inflammatory disease [6]. Concentration of 3HB was also used for assessment of severity and prevention of diabetic ketoacidosis [7]. 2-Hydroxybutyric acid (2HB), an isomer of 3HB, was demonstrated to be an early biomarker for both insulin resistance and impaired glucose regulation [8]. In addition, accumulation of 2HB in plasma from Leigh syndrome suggested a metabolic signature of mitochondrial dysfunction [9].

Both 2HB and 3HB are chiral carboxylic acids and exist in two steric configurations (D/L-2HB and D/L-3HB) resulting from an asymmetric carbon atom in the carbon backbone. Routine analytical methods are not able to differentiate between enantiomers, therefore, most of the previous studies analyzed the total amount of D-2HB and L-2HB or D-3HB and L-3HB [10-12]. However, despite their similar chemical and physical properties, D-and L-configurations have different biochemical properties. For example, D-3HB was found in human plasma and 3HB dehydrogenase can specially convert D-3HB instead of L-3HB to acetoacetate [13]. Although L-3HB could not be utilized for energy source [14], L-3HB possessed anticonvulsant activity while D-3HB was not effective [15]. Little attention was paid for biological properties of 2HB enantiomers [10].

Recently, Calderón et al. [16] used chiral column with cinchona alkaloid-derived chiral stationary phases to separate chiral short chain aliphatic hydrocarboxylic acids. D/L-2HB and D/L-3HB enantiomers can be separated on the chiral column. However, due to the poor ionization efficiency of 2HB and 3HB in liquid chromatography-mass spectrometry analysis, the detection sensitivities for 2HB and 3HB were low. In addition, this method required specific chiral column that is very expensive. Determination of L-threonine and D-threonine substituted serine octamer was recently also achieved by comparing infrared photo dissociation (IRPD) spectra, which provides a valuable alternative strategy for the chiral differentiation [17].

In the current study, we developed a strategy by chiral derivatization coupled with liquid chromatography/tandem mass spectrometry (LC-ESI-MS) analysis for simultaneous determination of D/L-2HB and D/L-3HB enantiomers. (S)-(+)-1-(2-Pyrrolidinylmethyl)-pyrrolidine (PMP) was used for highly efficient labeling of D/L-2HB and D/L-3HB enantiomers. The materials and reagents are shown in Supporting information.

Upon PMP labeling, D/L-2HB and D/L-3HB enantiomers are converted to the corresponding diastereomers (Fig. 1), which offer the possibility for these diastereomers to be separated on C18 column due to their different chemical properties.

Download:
Fig. 1. (A) Structures of D-2HB, L-2HB, D-3HB, L-3HB; (B) Chemical reaction between HB and PMP.

Shown in Fig. 2 are the fragmentation ions of PMP-labeled D-2HB and D-3HB. The results demonstrated that the desired PMP-labeled D-2HB and D-3HB were obtained. Notably, the difference of the hydroxy group position between 2HB and 3HB resulted in the specific fragment ion of m/z 126.3 from PMP-labeled D-3HB (Fig. 2). Shown in Fig. S1 (Supporting information) is the proposed fragmentation pathways of these two derivatives. Moreover, the PMP-labeled D-2HB and D-3HB were further confirmed by highresolution mass spectrometry analysis (Fig. S2 in Supporting information).

Download:
Fig. 2. Product ions spectra of PMP-labeled (A) D-2HB and (B) D-3HB.

To obtain good derivatization efficiency, we optimized the derivatization conditions, including reaction time and temperature, and concentration of PMP.

We first optimized the reaction time, and the result showed the highest reaction efficiencies were achieved at 90 min for both D/L-2HB and D/L-3HB (Fig. S3A in Supporting information). Therefore, 90 min was used as the reaction time to ensure the complete derivatization. We next optimized the reaction temperature. Our results demonstrated that good labeling can be achieved at 60 ℃ for these four compounds of D/L-2HB and D/L-3HB enantiomers (Fig. S3B in Supporting information). So we chose 60 ℃ for the subsequent experiments.

As for the optimization of the concentration of PMP, the results showed that the peak area ratios of PMP-labeled D/L-2HB and D/L-3HB reach to the plateau when the concentration of PMP is 0.2 mmol/L (Fig. S3C in Supporting information). Therefore, 0.2 mmol/L of PMP was used for the following experiments. We finally evaluated the stability of the PMP-labeled D-2HB and D-3HB. The result demonstrated that the derivatives were stable at least for 16 h (Fig. S3D in Supporting information), which is sufficient for the subsequent LC-ESI-MS analysis.

Taken together, the optimized derivatization conditions for D/L-2HB and D/L-3HB by PMP were at 60 ℃ for 90 min with the concentration of PMP being 0.2 mmol/L. Under optimized derivatization conditions, more than 99% of the analytes can react with PMP by comparing the signals of analytes before and after PMP labeling, suggesting high derivatization efficiencies were achieved.

The main purpose for derivatization is to improve the chromatographic separation and detection sensitivities of D/L-2HB and D/L-3HB enantiomers during LC-ESI-MS analysis. The extracted ion chromatograms showed that the retention of D/L-2HB and D/L-3HB enantiomers was relatively weak and they coeluted (~2 min) on C18 column (Fig. S4 in Supporting information). However, the retention of PMP-labeled D/L-2HB and D/L-3HB enantiomers increased and can be distinctly separated (Fig. 3A), suggesting the formed diastereomers by PMP labeling can efficiently improve the separation of D/L-2HB and D/L-3HB enantiomers.

Download:
Fig. 3. Extracted ion chromatograms of PMP-labeled D-2HB, L-2HB, D-3HB and L-3HB: (A) PMP-labeled standards; (B) Detected PMP-labeled L-2HB, D-3HB, and L-3HB in human RCC tissue.

Chemicallabeling is an effective strategy to increase the detection sensitivities of analytes by mass spectrometryanalysis [18-29]. Here we found PMP derivatization could increase the detection sensitivities by 16.4, 19.7, 55.3 and 43.5 folds for D-2HB, L-2HB, D-3HB and L-3HB, respectively, compared to their native forms (Table S1 in Supporting information). The limits of detection (LODs) of PMPlabeled D-2HB, L-2HB, D-3HB and L-3HB were 28.2, 22.8, 6.8 and 8.0 fmol, respectively (Table S1). The increased detection sensitivities by PMP derivatization can be attributed to several aspects. Firstly, the ionization of native D/L-2HB and D/L-3HB is relatively poor under ESI-MS analysis; however, PMP derivatization can effectively enhance the ionization efficiency, which therefore leads to increased mass spectrometry response. Secondly, PMP derivatization increased the retention of D/L-2HB and D/L-3HB on reversed-phase chromatographic column, resulting in longer retention time and thus elution within a higher ratio of organic solvents. Therefore the analytes could be ionized more effectively in ESI.

The calibration curves of D-2HB, L-2HB, D-3HB and L-3HB in tissue matrix were constructed by plotting the mean peak area ratios of PMP-labeled D/L-2HB and D/L-3HB enantiomers to citalopram (I.S.) versus the corresponding D/L-2HB and D/L-3HB enantiomers concentrations based on data obtained from triplicate measurements. The results showed that good linearities were obtained with the coefficient of determination (R2) being greater than 0.99 for both D/L-2HB and D/L-3HB enantiomers in tissue matrix (Table S2 in Supporting information).

The accuracy of the proposed method was assessed by comparing the measured D/L-2HB and D/L-3HB enantiomers contents to the theoretical D/L-2HB and D/L-3HB enantiomers contents (Table S3 in Supporting information). In addition, the repeatability of the developed method was evaluated by measuring intra-and inter-day precisions. The intra-and inter-day relative standard deviations (RSDs) were calculated at three different concentrations. Five parallel treatments of samples over a day gave the intra-day RSDs, and the inter-day RSDs were determined by treating samples independently for three consecutive days. The results showed that good accuracies were achieved, which is manifested by the relative errors (RE) ranging from -10.6% to 14.9% (Table S3 in Supporting information). The results also showed that the intra-and inter-day RSDs were less than 14.0% and 14.4% for D/L-2HB and D/L-3HB, respectively (Table S3), demonstrating that good repeatability was achieved.

With the developed method, we quantified the amounts of D/L-2HB and D/L-3HB enantiomers in human renal cell carcinoma (RCC) tumor tissues and tumor adjacent normal tissues (Table S4 in Supporting information). A total of 28 RCC tissue samples from 14 RCC patients without preoperative target therapy/chemotherapy were obtained from Department of Urology, Peking University First Hospital and Department of Urology, Peking University People's Hospital and the approvals for the study were obtained from the Ethical Committee. The RCC tissues and tumor adjacent normal tissues were kept at -80 ℃. Extraction of D/L-2HB and D/L-3HB enantiomers from tissue samples was performed according to the preciously described method [30].

Shown in Fig. 3B is the typical extracted-ion chromatogram of PMP-labeled L-2HB, D-3HB, and L-3HB detected in RCC tissues. Notably, only L-2HB, but not D-2HB, was detectable in RCC tissues and tumor adjacent normal tissues, suggesting that endogenous D-2HB may not exist in human renal tissues. As the major ketone body, D-3HB was investigated more intensively than L-3HB, and little was known about the source and metabolic pathway of L-3HB. Here we were able to detect both D-3HB and L-3HB in human renal tissues, suggesting that both D-3HB and L-3HB are endogenous metabolites.

The measured contents of D/L-2HB and D/L-3HB enantiomers were normalized to total protein content. Tissues lysate was generated using RIPA buffer according to previously described procedure [31]. Protein contents were measured using the bicinchoninic acid (BCA) protein assay kit (Beyotime Biotechnology, Shanghai, China). The results showed that the mean measured contents of L-2HB in RCC tissues and tumor adjacent normal tissues were 93.9 and 71.9 pmol/mg total protein, respectively (Fig. 4A). The mean measured contents of D-3HB in RCC tissues and tumor adjacent normal tissues were 364.2 and 383.7 pmol/mg total protein, respectively (Fig. 4B); and the mean measured contents of L-3HB in RCC tissues and tumor adjacent normal tissues were 28.2 and 36.2 pmol/mg total protein, respectively (Fig. 4C). The quantification results showed that the contents of D-3HB were approximate 10 folds higher than L-3HB. In addition, we observed an increase of L-2HB, and decrease for both D-3HB and L-3HB in RCC tissues compared to tumor adjacent normal tissues. However, the changes of these HBs are not statistically significant (Fig. 4). Nevertheless, the developed method offered an efficient approach for the sensitive analysis of D/L-2HB and D/L-3HB enantiomers, which may facilitate the in-depth study of the functions of HBs.

Download:
Fig. 4. Measured contents of L-2HB (A), D-3HB (B), and L-3HB (C) in RCC tissues and tumor adjacent normal tissues.

In conclusion, simultaneous determination of D-2HB, L-2HB, D-3HB and L-3HB enantiomers was achieved by PMP chiral derivatization coupled with LC-ESI-MS analysis. Our results indicated that D/L-2HB and D/L-3HB enantiomers were distinctly separated in C18 reversed-phase column and the retention behavior of the four derivatives was greatly improved upon PMP labeling. In addition, the detection sensitivities of HBs dramatically increased. Using this method, we simultaneously quantified D-2HB, L-2HB, D-3HB, and L-3HB enantiomers in human RCC tissues and the tumor adjacent normal tissues. The result demonstrated that both D-3HB and L-3HB can be detected in human renal tissues, however, only the L-configuration of 2HB was detected in human renal tissues. Taken together, the developed method will contribute to accurate detection of D-2HB, L-2HB, D-3HB and L-3HB and benefit the research of HBs related diseases.

Acknowledgment

The authors thank the financial support from the National Natural Science Foundation of China (Nos. 21522507, 21672166, 21635006).

Appendix A. Supplementary data

Supplementary data associated with this article can be found, in the online version, at http://dx.doi.org/10.1016/j.cclet.2017.06.009.

References
[1]
P. Puchalska, P.A. Crawford, Cell Metab. 25(2017) 262-284. DOI:10.1016/j.cmet.2016.12.022
[2]
J.C. Newman, E. Verdin, Trends Endocrinol. Met. 25(2014) 42-52. DOI:10.1016/j.tem.2013.09.002
[3]
D.G. Cotter, R.C. Schugar, P.A. Crawford, Am. J. Physiol. Heart Circ. Physiol. 304(2013) H1060-H1076. DOI:10.1152/ajpheart.00646.2012
[4]
L. Camberos-Luna, C. Gerónimo-Olvera, T. Montiel, R. Rincon-Heredia, L. Massieu, Neurochem. Res. 41(2016) 600-609. DOI:10.1007/s11064-015-1700-4
[5]
T. Shimazu, M.D. Hirschey, J. Newman, et al., Science 339(2013) 211-214. DOI:10.1126/science.1227166
[6]
Y.H. Youm, K.Y. Nguyen, R.W. Grant, et al., Nat. Med. 21(2015) 263-269. DOI:10.1038/nm.3804
[7]
S. Misra, N.S. Oliver, Diabet. Med. 32(2015) 14-23.
[8]
W.E. Gall, K. Beebe, K.A. Lawton, et al., Plos One 5(2010) e10883. DOI:10.1371/journal.pone.0010883
[9]
J.T. Legault, L. Strittmatter, J. Tardif, et al., Cell Rep. 13(2015) 981-989. DOI:10.1016/j.celrep.2015.09.054
[10]
Q. Zhang, L.A. Ford, K.D. Goodman, et al., J. Chromatogr. B 1038(2016) 101-108. DOI:10.1016/j.jchromb.2016.10.025
[11]
T. Miyazaki, A. Honda, T. Ikegami, et al., Springerplus 4(2015) 494. DOI:10.1186/s40064-015-1304-0
[12]
N. Goudarzi, M. Goodarzi, M.A. Chamjangali, M.H. Fatemi, Chin. Chem. Lett. 24(2013) 904-908. DOI:10.1016/j.cclet.2013.06.002
[13]
D.H. Williamson, J. Mellanby, H.A. Krebs, Biochem. J. 82(1962) 90-96. DOI:10.1042/bj0820090
[14]
Y.C. Tsai, Y.C. Chou, A.B. Wu, et al., Life Sci. 78(2006) 1385-1391. DOI:10.1016/j.lfs.2005.07.013
[15]
J.M. Rho, G.D. Anderson, S.D. Donevan, H.S. White, Epilepsia 43(2002) 358-361. DOI:10.1046/j.1528-1157.2002.47901.x
[16]
C. Calderon, M. Lammerhofer, J. Chromatogr. A 1487(2017) 194-200. DOI:10.1016/j.chroma.2017.01.060
[17]
J. Ren, Y.Y. Wang, R.X. Feng, X.L. Kong, Chin. Chem. Lett. 28(2017) 537-540. DOI:10.1016/j.cclet.2016.10.032
[18]
Y. Tang, S.J. Zheng, C.B. Qi, et al., Anal. Chem. 87(2015) 3445-3452. DOI:10.1021/ac504786r
[19]
H. Zeng, C.B. Qi, T. Liu, et al., Anal. Chem. 89(2017) 4153-4160. DOI:10.1021/acs.analchem.7b00052
[20]
H.Y. Zhang, J. Xiong, B.L. Qi, et al., Chem. Commun. (Camb.) 52(2016) 737-740. DOI:10.1039/C5CC07354E
[21]
W. Huang, M.D. Lan, C.B. Qi, et al., Chem. Sci. 7(2016) 5495-5502. DOI:10.1039/C6SC01589A
[22]
B.L. Qi, P. Liu, Q.Y. Wang, et al., TrAC-Trend Anal. Chem. 59(2014) 121-132. DOI:10.1016/j.trac.2014.03.013
[23]
J.M. Chu, C.B. Qi, Y.Q. Huang, et al., Anal. Chem. 87(2015) 7364-7372. DOI:10.1021/acs.analchem.5b01614
[24]
P. Liu, Y.Q. Huang, W.J. Cai, et al., Anal. Chem. 86(2014) 9765-9773. DOI:10.1021/ac5023315
[25]
Y. Tang, J. Xiong, H.P. Jiang, et al., Anal. Chem. 86(2014) 7764-7772. DOI:10.1021/ac5016886
[26]
N. Guo, P. Liu, J. Ding, et al., Anal. Chim. Acta 905(2016) 106-114. DOI:10.1016/j.aca.2015.12.010
[27]
N. Guo, C.Y. Peng, Q.F. Zhu, et al., Anal. Chim. Acta 967(2017) 42-51. DOI:10.1016/j.aca.2017.03.006
[28]
Q.F. Zhu, Z. Zhang, P. Liu, et al., J. Chromatogr. A 1460(2016) 100-109. DOI:10.1016/j.chroma.2016.07.017
[29]
Q.F. Zhu, Y.H. Hao, M.Z. Liu, et al., J. Chromatogr. A 1410(2015) 154-163. DOI:10.1016/j.chroma.2015.07.100
[30]
Q. Y. Cheng, J. Xiong, W. Huang, et al., Sci. Rep. 5(2015) 15217. DOI:10.1038/srep15217
[31]
J.M. Tolosa, J.E. Schjenken, T.D. Civiti, et al., Biotechniques 43(2007) 799-804. DOI:10.2144/000112594