Addressing the need for biomarker liquid chromatography/mass spectrometry assays: A protocol for effective method development for the bioanalysis of endogenous compounds in cerebrospinal fluid

Authors

  • Yulia Benitex,

    Corresponding author
    • Molecular Sciences and Candidate Optimization, Bioanalytical Research, Bristol-Myers Squibb Company, Research & Development, Wallingford, CT, USA
    Search for more papers by this author
  • Colleen A. McNaney,

    1. Molecular Sciences and Candidate Optimization, Discovery Analytical Sciences, Bristol-Myers Squibb Company, Research & Development, Wallingford, CT, USA
    Search for more papers by this author
  • David Luchetti,

    1. Discovery Medicine and Clinical Pharmacology, Bristol-Myers Squibb Company, Research & Development, Wallingford, CT, USA
    Search for more papers by this author
  • Eric Schaeffer,

    1. Discovery Medicine and Clinical Pharmacology, Bristol-Myers Squibb Company, Research & Development, Wallingford, CT, USA
    Search for more papers by this author
  • Timothy V. Olah,

    1. Molecular Sciences and Candidate Optimization, Bioanalytical Research, Bristol-Myers Squibb Company, Research & Development, Lawrenceville, NJ, USA
    Search for more papers by this author
  • Daniel G. Morgan,

    1. Molecular Sciences and Candidate Optimization, Bioanalytical Research, Bristol-Myers Squibb Company, Research & Development, Wallingford, CT, USA
    Search for more papers by this author
  • Dieter M. Drexler

    1. Molecular Sciences and Candidate Optimization, Discovery Analytical Sciences, Bristol-Myers Squibb Company, Research & Development, Wallingford, CT, USA
    Search for more papers by this author

Correspondence to: Y. Benitex, Molecular Sciences and Candidate Optimization, Bioanalytical Research, Bristol-Myers Squibb Company, Research & Development, 5 Research Parkway, Wallingford, CT 06492, USA.

E-mail: yulia.benitex@bms.com

Abstract

RATIONALE

Research on disorders of the central nervous system (CNS) has shown that an imbalance in the levels of specific endogenous neurotransmitters may underlie certain CNS diseases. These alterations in neurotransmitter levels may provide insight into pathophysiology, but can also serve as disease and pharmacodynamic biomarkers. To measure these potential biomarkers in vivo, the relevant sample matrix is cerebrospinal fluid (CSF), which is in equilibrium with the brain's interstitial fluid and circulates through the ventricular system of the brain and spinal cord. Accurate analysis of these potential biomarkers can be challenging due to low CSF sample volume, low analyte levels, and potential interferences from other endogenous compounds.

METHODS

A protocol has been established for effective method development of bioanalytical assays for endogenous compounds in CSF. Database searches and standard-addition experiments are employed to qualify sample preparation and specificity of the detection thus evaluating accuracy and precision.

RESULTS

This protocol was applied to the study of the histaminergic neurotransmitter system and the analysis of histamine and its metabolite 1-methylhistamine in rat CSF.

CONCLUSIONS

The protocol resulted in a specific and sensitive novel method utilizing pre-column derivatization ultra high performance liquid chromatography/tandem mass spectrometry (UHPLC/MS/MS), which is also capable of separating an endogenous interfering compound, identified as taurine, from the analytes of interest. Copyright © 2013 John Wiley & Sons, Ltd.

An imbalance in the levels of specific endogenous neurotransmitters has been implicated in the etiology of many disorders of the central nervous system. For example, reduced levels of monoamines have been suggested to play a role in major depressive disorders,[1] and aberrant levels of dopamine and glutamate have been hypothesized to underlie schizophrenia.[2] It is therefore not surprising that many of the medications developed to treat these disorders target mechanisms intended to restore neurotransmitter balance in an effort to ameliorate disease symptoms. In preclinical research, neurotransmitter levels can be measured either in isolated brain tissue from terminal experiments, or in vivo using microdialysis to quantify the levels of these molecules in awake and freely moving animals. However, these methods are not applicable to the clinical setting, and thus there is an interest in developing methods to detect neurotransmitters and their metabolites as potential biomarkers in the cerebrospinal fluid (CSF) of animals and humans, to enable translation of preclinical research into early clinical development.

To address the need for specific and sensitive bioanalytical analysis of these potential endogenous biomarkers in CSF, the following protocol for effective assay development has been established.

  • Due to the low concentrations of the neurotransmitters and their metabolites in biofluids, the analytical methods are typically based on in-line liquid chromatography/tandem mass spectrometry (LC/MS/MS) on triple quadrupole (QqQ) instruments often employing pre-column derivatization to improve chromatographic behavior and detection limits.[3] Notably, recent advances in time-of-flight (TOF) or Orbitrap systems makes quantitation by high-resolution accurate mass spectrometry (HRMS) a viable alternative technique.[4]
  • After the target analyte has been identified via literature search or screening efforts, a search in the Human Metabolome Database[5] is performed to identify potential endogenous interfering compounds within a molecular weight window of ±5 Da. to just not only realize isobaric compounds, but also to be aware of the isotopic envelope (e.g. contribution of 13C, 34S, 37Cl, 81Br, …) of lower mass non-isobaric compounds.
  • An interfering compound within a ±1 Da. window should be chromatographically resolved from the analyte of interest since the QqQ MS instrument is usually operated at unit resolution.
  • If required, the pre-column derivatization and chromatographic conditions are optimized.
  • The MS conditions are optimized.
  • The initial analysis of naïve CSF samples establishes a baseline value of the analyte.
  • Various known concentrations of the compound of interest are spiked into naïve CSF samples (standard-addition method) to qualify sample preparation and specificity of the detection thus evaluating accuracy and precision.

This protocol has been applied to the study of the histaminergic neurotransmitter system, which, in recent years, has become an increasingly important target in drug development, as its role in attention, cognition and sleep has become better understood.[6] In addition, studies have shown the levels of histamine (HA) and its major metabolites 1-methylhistamine (1-mHA, also referred to as tele-methylhistamine (t-mHA)) and tele-imidazolacetic acid (t-MIAA) may be altered in some brain disorders including schizophrenia,[7] and Alzheimer's disease.[8] Also, studies have demonstrated that modulation of brain HA levels may have beneficial effects on attentional processes with the potential for therapeutic benefit in a variety of disorders including attention deficit hyperactivity disorder and Alzheimer's disease.[6] This growing interest in the role of HA has led to the development of methods for measuring this neurotransmitter and its metabolites in the CSF, with potential utility as a biomarker of disease as well as to assess the pharmacodynamic effects of experimental drugs.

Techniques for the analysis of HA and its metabolites include gas chromatography, capillary electrophoresis, and LC employing hydrophilic interaction LC (HILIC) without pre-column derivatization or reversed-phase LC (RPLC) with pre-column derivatization.[9-14] However, no systematic approach has been reported on the separation and analysis of HA and its metabolite 1-mHA, including its stereoisomer 3-mHA, in CSF in the presence of other potentially interfering compounds.

A protocol was established to address the challenges of the bioanalysis of a potential biomarker for an in vivo study with low CSF sample volume and potential interferences from other endogenous compounds, resulting in a novel method with high specificity and sensitivity for the analysis of HA, 1-mHA and 3-mHA by utilizing pre-column derivatization ultra high performance liquid chromatography/tandem mass spectrometry (UHPLC/MS/MS).

EXPERIMENTAL

Materials and reagents

Artificial CSF (Perfusion fluid CNS) was purchased from CMA Microdialysis Inc. (North Chelmsford, MA, USA). Analytical standards of histamine (HA), 1-methylhistamine (1-mHA), 3-methylhistamine (3-mHA), taurine (Tau) and all other reagents and chemicals were purchased from Sigma-Aldrich (St. Louis, MO, USA) unless otherwise stated. The internal standard (d3-1-mHA) was obtained from C/D/N Isotopes Inc. (Quebec, Canada). Acetaldehyde-d4 was purchased from Cambridge Isotope Laboratories, Inc. (Andover, MA, USA). Rat Sprague Dawley CSF was purchased from Bioreclamation (Melville, NY, USA).

The molecular structures of 1-mHA, 3-mHA, and, d3-1-mHA were confirmed by 2D-NMR.

Preparation of samples

Calibration curve

The stock solutions (1 mM) of HA, 1-mHA, 3-mHA, and Tau were freshly prepared in HPLC-grade water and serially diluted with artificial CSF to make twelve standard concentration levels from 250 to 0.2 nM. The internal standard (IS) solution, d3-1-mHA, was prepared at 1 mM in HPLC-grade water and diluted with artificial CSF to make a 0.5 μM IS solution.

Derivatization of analytes

The analytes and IS were derivatized with acetaldehyde-d4 according to a published procedure prior to UHPLC/MS/MS analysis (Scheme 1).[15] Briefly, 20 μL of standard solution or sample was mixed with 10 μL of IS solution. Next, to the mixture, 20 μL of coupling reagent was added, which consisted of 25% acetylaldehyde-d4 in acetate buffer (pH 5) and cyanoborohydrate coupling buffer (1:3, v/v). The solution was incubated at 37 °C for 30 min. The efficiency of this procedure was 100% as determined via monitoring of the respective MS channels of the underivatized and derivatized analytes after derivatization.

Scheme 1.

Derivatization reaction.

LC/MS/MS conditions

A CTC HTS PAL autosampler (Leap Technologies, Carrboro, NC, USA), a 1200 series UHPLC system (Agilent Technologies, Santa Clara, CA, USA), and an API 4000 triple quadrupole mass spectrometer (Applied Biosystems, Foster City, CA, USA) were used as the LC/MS system. Reversed-phase (RP)LC was performed using an Acquity BEH C18 column (2.1 mm × 50 mm, 1.7 µm; Waters, Milford, MA, uSA). The column temperature was set at 55 °C. Chromatographic separation was achieved using the mobile phases consisting of 30 mM ammonium bicarbonate in water (mobile phase A) and acetonitrile (mobile phase B). The UHPLC was performed at a flow rate of 600 μL/min and the sample injection volume was 5 μL. The initial condition was 0% B which was held for 0.5 min and a linear gradient was performed with mobile phase B increasing to 70% within 3.5 min and held for 0.5 min, then the system was returned to initial condition in 0.2 min and held for 0.3 min. The mass spectrometer was operated in positive ion electrospray mode with source conditions set as follows: ion spray voltage 4800 V, ion source temperature 500 °C, declustering potential 80 V, and collision energy 25 eV. Table 1 shows the selected reaction monitoring (SRM) transitions for the analytes.

Table 1. SRM transitions of acetaldehyde-d4-derivatized analytes
AnalyteQ1Q3
HA176.194.1
1-mHA190.194.1
3-mHA190.194.1
Tau190.194.1
d3-1-mHA193.1112.1

RESULTS AND DISCUSSION

UHPLC/MS/MS analysis of derivatized HA, 1-mHA, and 3-mHA

To address the challenge of a specific and sensitive assay for the bioanalysis of HA and its metabolites in CSF, a protocol was established for effective method development. Due to the polar character and low endogenous levels of the analytes in CSF, pre-column derivatization with acetaldehyde-d4 according to a previously published procedure was chosen for sample preparation to improve chromatographic behavior and MS detection limits.[3, 15] Then, following the protocol, the initial search in the Human Metabolome Database[5] for potential interfering endogenous compounds in CSF (molecular weight (MW) 111.1 ± 5 Da, similar to HA, and 125.1 ± 5 Da, similar to 1/3-mHA, containing a derivatizable amino group) highlighted the compound taurine (2-aminoethanesulfonic acid, Tau, MW 125.1 Da).

The analysis of CSF samples employing pre-column derivatization and a generic RPLC method (A: 0.1% formic acid in water/B: 0.1% formic acid in acetonitrile), yielded data on 1-mHA with levels ~20-fold higher than what was reported previously on non-derivatized 1-mHA.[10] It appears that there was a contribution to the SRM/MS channel of 1-mHA from an analyte which could not be separated by the generic acidic mobile phase. Subsequent spiking experiments confirmed the interfering compound to be Tau as indicated by the database search. Figure 1 shows the representative chromatograms under generic acidic LC conditions of acetaldehyde-d4-derivatized analytes in rat CSF indicating no chromatographic resolution (total ion current, TIC) of HA, 1-mHA, and Tau and no mass spectrometric resolution (extracted ion current, XIC) of 1-mHA and Tau due to their isobaric nature and similar MS/MS product ions.

Figure 1.

Representative chromatograms of acetaldehyde-d4-derivatized analytes in rat CSF using a generic RPLC method (acidic condition). The TIC indicates no chromatographic resolution of HA, 1-mHA & IS d3-1-mHA, and Tau. The XIC shows no mass spectrometric resolution of 1-mHA and Tau.

To separate these endogenous compounds, a basic mobile phase was developed (30 mM ammonium bicarbonate, pH 9.5). Using this LC method and further spiking experiments, the interfering compound was again confirmed as Tau. Additionally, to ensure that the correct molecular species of 1-mHA was quantified and not its isomer 3-mHA, and to investigate potential interferences from the isomers, 3-mHA was included in the analysis. Figure 2 shows the representative chromatograms under the developed basic LC conditions showing separation of the HA, 1-mHA, 3-mHA, and Tau analytes in artificial CSF. The lower limits of quantitation (LLOQs) for HA, 1-mHA, and 3-mHA were 1.0, 0.2, and 0.5 nM, respectively.

Figure 2.

Representative TIC chromatogram of acetaldehyde-d4-derivatized HA (retention time (RT) = 1.7 min), 1-mHA & IS d3-1-mHA (RT = 1.8 min), 3-mHA (RT = 1.9), and Tau (RT = 1.3 min) in artificial CSF using the modified RPLC method (basic condition). The concentrations were 62.5 nM for HA, 1-mHA, 3-mHA and 1000 nM for Tau.

Evaluation of HA and m-HA assays

The calibration curves were obtained by plotting the known concentration of each analyte standard against the peak area ratios of analyte relative to the IS. All calibration curves exhibited acceptable linearity with dynamic ranges of 0.2–250 nM (r2 >0.998). The samples were quantified using 1/x2 weighted linear regression. Precision and accuracy were evaluated by spiking naive rat CSF samples with two different concentration levels (10 and 20 nM). The coefficients of variance (CVs) were calculated between the initial pre-spiking and the spiked values. This was performed on five replicates per level assayed in triplicate (Table 2). The accuracies and precisions were in the range of 87–109% and 3.2–7.4%, respectively. The levels of HA and 1-mHA in rat CSF were 7.6 and 4.4 nM, respectively. Figure 3 displays the separation and detection of endogenous HA, 1-mHA, and Tau, while 3-mHA was not observed in rat CSF.

Table 2. HA and 1-mHA levels in rat CSF before and after spiking with known concentrations (n = 5)
AnalyteMean value before spiking (nM)Spiking concentration (nM)Mean theoretical spiking value (nM)Mean experimental spiking value (nM)
HA7.61017.617.9
HA7.62027.627.4
1-mHA4.41014.414.7
1-mHA4.42024.424.3
Figure 3.

Representative TIC chromatogram of acetaldehyde-d4-derivatized endogenous analytes in rat CSF showing the target analytes HA (RT = 1.7 min), 1-mHA & IS d3-1-mHA (RT = 1.8 min), and the now chromatographically resolved Tau (RT = 1.3 min).

In the method development for the accurate quantitation of endogenous molecules, it is important to be able to separate any endogenous interfering compounds, especially when a derivatization method is used, since common fragmentations of derivatized products usually occur resulting in identical product ions as evident with the analytes 1-mHA and Tau.

CONCLUSIONS

To address the challenge of the bioanalysis of endogenous compounds as potential biomarkers in CSF, a protocol has been established for effective method development. This protocol was applied to the study of the histaminergic neurotransmitter system and the analysis of histamine and its metabolites resulting in a specific and sensitive novel method utilizing pre-column derivatization UHPLC/MS/MS for the analysis of endogenous HA and its metabolite 1-mHA in rat CSF. This method is also capable of separating an endogenous interfering compound, identified as Tau, from the analytes of interest.

Acknowledgement

The authors would like to thank Dr. Stella Huang for NMR support.

Ancillary