Article pubs.acs.org/ac
Online, Absolute Quantitation of Propranolol from Spatially Distinct 20- and 40-μm Dissections of Brain, Liver, and Kidney Thin Tissue Sections by Laser Microdissection−Liquid Vortex Capture−Mass Spectrometry John F. Cahill,† Vilmos Kertesz,† Taylor M. Weiskittel,‡ Marissa Vavrek,§ Carol Freddo,§ and Gary J. Van Berkel*,† †
Mass Spectrometry and Laser Spectroscopy Group, Chemical Sciences Division, Oak Ridge National Laboratory, Oak Ridge, Tennessee 37831-6131, United States ‡ ORISE HERE Intern, University of Tennessee, Knoxville, Tennessee 37996, United States § Department of Pharmacokinetics, Pharmacodynamics, and Drug Metabolism, Merck Research Laboratories, West Point, Pennsylvania 19486, United States S Supporting Information *
ABSTRACT: Spatial resolved quantitation of chemical species in thin tissue sections by mass spectrometric methods has been constrained by the need for matrix-matched standards or other arduous calibration protocols and procedures to mitigate matrix effects (e.g., spatially varying ionization suppression). Reported here is the use of laser “cut and drop” sampling with a laser microdissection-liquid vortex capture electrospray ionization tandem mass spectrometry (LMD-LVC/ESI-MS/MS) system for online and absolute quantitation of propranolol in mouse brain, kidney, and liver thin tissue sections of mice administered with the drug at a 7.5 mg/kg dose, intravenously. In this procedure either 20 μm × 20 μm or 40 μm × 40 μm tissue microdissections were cut and dropped into the flowing solvent of the capture probe. During transport to the ESI source drug related material was completely extracted from the tissue into the solvent, which contained a known concentration of propranolol-d7 as an internal standard. This allowed absolute quantitation to be achieved with an external calibration curve generated from standards containing the same fixed concentration of propranolold7 and varied concentrations of propranolol. Average propranolol concentrations determined with the laser “cut and drop” sampling method closely agreed with concentration values obtained from 2.3 mm diameter tissue punches from serial sections that were extracted and quantified by HPLC/ESI-MS/MS measurements. In addition, the relative abundance of hydroxypropranolol glucuronide metabolites were recorded and found to be consistent with previous findings.
T
homogeneous matrix effect throughout the whole sample surface.6−9 Nearly every other common surface sampling/ ionization method utilized in conjunction with mass spectrometry has inherent sampling biases (e.g., varying extraction efficiencies),10 are susceptible to sample induced ionization suppression (i.e., matrix effects),11−13 and/or suffer from other factors that limit quantitative capabilities. This includes secondary ion MS (SIMS),14,15 desorption electrospray ionization-MS (DESI-MS),16,17 laser ablation-inductively coupled plasma-MS (LA/ICP-MS),18−21 laser ablation-ESIMS (LAESI-MS),22−24 nanospray DESI-MS (nano-DESIMS),10,25 liquid extraction surface analysis-MS (LESA-
he spatially resolved, quantitative measurement of drug in small regions of tissue represents a difficult analytical challenge, with important ramifications toward many pharmacological studies. Quantitative whole-body autoradiography (QWBA) is the standard method used to quantitate drug tissue distribution, but the development cost of a radiolabeled drug can often be quite high and the technique cannot differentiate between the drug and its associated metabolites.1−4 Matrixassisted laser desorption/ionization mass spectrometry (MALDI-MS) is another common technique to measure drug distributions in tissue. However, with MALDI-MS the need for homogeneous application of a chemical matrix to the sample can present several difficulties.5 MALDI-MS is also not particularly suitable for the detection of fragile phase II metabolites and quantitation requires the application of calibration droplets across the sample surface and assumes a © XXXX American Chemical Society
Received: March 23, 2016 Accepted: May 14, 2016
A
DOI: 10.1021/acs.analchem.6b01155 Anal. Chem. XXXX, XXX, XXX−XXX
Article
Analytical Chemistry MS),26−29 and many others. For example, laser ablation of a surface is known to preferentially segregate elements based on ablation conditions, a process termed chemical fractionation. This has been a long-standing problem in the LA/ICP-MS field.30−33 In direct liquid extraction techniques, complete extraction of drug from the sample is usually not achieved, and extraction efficiencies can vary across different tissue types complicating quantitative analysis.28,34,35 For more detail on the challenges of quantitative mass spectrometry methods we recommend the reader to several reviews covering the subject.4,8,12,33,35−40 To overcome the quantitative limitations in these techniques, several novel calibration methods have been developed. Most commonly, the matrix-matched strategy is employed in which an internal standard spiked homogenate of the sample is used to correct biases because of the matrix of the sample, assuming that the homogenate will exhibit similar matrix effects as in the sample.35,41 This is often used for quantitation in SIMS and LA/ICP-MS measurements35,37,42 and has had some success for MALDI-MS.9,43 However, use of this method is hindered by availability of sample homogenate and the homogenate may not necessarily mimic the matrix of specific microenvironments in the sample. Many other methods involve depositing standards under, on top of, or both under and on top of tissue samples.9 Recently, Chumbley et al.9 compared these methods for MALDI-MS measurements and found that depositing standards on top of tissue followed by matrix application yielded the most accurate results when compared to high performance liquid chromatography (HPLC)-MS/MS measurements of tissue extracts. In the same publication, mass spectral images were quantified pixel-by-pixel by spotting standards and matrix on top of the tissue for every pixel in the image. However, high resolution sampling (
