ARTICLE pubs.acs.org/ac
Qualitative and Quantitative MALDI Imaging of the Positron Emission Tomography Ligands Raclopride (a D2 Dopamine Antagonist) and SCH 23390 (a D1 Dopamine Antagonist) in Rat Brain Tissue Sections Using a Solvent-Free Dry Matrix Application Method Richard J. A. Goodwin,*,†,§ C. Logan Mackay,‡ Anna Nilsson,§ David J. Harrison,^ Lars Farde,||,# Per E. Andren,§ and Suzanne L. Iverson† †
Global Distribution Imaging, DMPK IM, AstraZeneca R&D, S€odert€alje, Sweden SIRCAMS, School of Chemistry, University of Edinburgh, Edinburgh, U.K. § Medical Mass Spectrometry, Department of Pharmaceutical Biosciences, Uppsala University, Uppsala, Sweden ^ Division of Pathology, Institute of Molecular Medicine, University of Edinburgh, Edinburgh, U.K. Karolinska Institutet, Department of Clinical Neuroscience, Karolinska University Hospital, Stockholm, Sweden # iMed CNS/Pain, AstraZeneca, S€odert€alje, Sweden
)
‡
bS Supporting Information ABSTRACT: The distributions of positron emission tomography (PET) ligands in rat brain tissue sections were analyzed by matrix-assisted laser desorption/ ionization mass spectrometry imaging (MALDI MSI). The detection of the PET ligands was possible following the use of a solvent-free dry MALDI matrix application method employing finely ground dry α-cyano-4-hydroxycinnamic acid (CHCA). The D2 dopamine receptor antagonist 3,5-dichloro-N-{[(2S)1-ethylpyrrolidin-2-yl]methyl}-2-hydroxy-6-methoxybenzamide (raclopride) and the D1 dopamine receptor antagonist 7-chloro-3-methyl-1-phenyl-1,2,4,5-tetrahydro-3-benzazepin-8-ol (SCH 23390) were both detected at decreasing abundance at increasing period postdosing. Confirmation of the compound identifications and distributions was achieved by a combination of mass-tocharge ratio accurate mass, isotope distribution, and MS/MS fragmentation imaging directly from tissue sections (performed using MALDI TOF/TOF, MALDI q-TOF, and 12T MALDI-FT-ICR mass spectrometers). Quantitative data was obtained by comparing signal abundances from tissues to those obtained from quantitation control spots of the target compound applied to adjacent vehicle control tissue sections (analyzed during the same experiment). Following a single intravenous dose of raclopride (7.5 mg/kg), an average tissue concentration of approximately 60 nM was detected compared to 15 nM when the drug was dosed at 2 mg/kg, indicating a linear response between dose and detected abundance. SCH 23390 was established to have an average tissue concentration of approximately 15 μM following a single intravenous dose at 5 mg/kg. Both target compounds were also detected in kidney tissue sections when employing the same MSI methodology. This study illustrates that a MSI may well be readily applied to PET ligand research development when using a solvent-free dry matrix coating.
T
he ability to spatially resolve the distribution of pharmacological compounds in organs and tissues is vital in all stages of drug research and development. Of particular importance to CNS disease and therapy development is the ability to examine biomarkers for central neurotransmission systems in animal models and humans by positron emission tomography (PET).1 7 PET imaging, along with similar in vivo imaging techniques such as single photon emission computed tomography (SPECT) imaging, enables quantification of radioligand binding to receptors, transporters, enzymes, and other proteins in brain. Such assays that map the distribution of a target compound or ligands are typically reliant on markers or probes that are complex in their production and require extensive molecular manipulation. r 2011 American Chemical Society
Furthermore, and most significantly, it is the distribution of the probe that is detected and mapped even if any metabolism/ modification of the original compound occurs. However, mass spectrometry imaging (MSI),8 with the ability to acquire multiplex data with limited a priori information, is on the verge of becoming a main-stay in the early stages of drug metabolism and pharmacokinetics (DMPK) research. The number of reports demonstrating that MSI has the sensitivity and selectivity Received: October 11, 2011 Accepted: November 11, 2011 Published: November 11, 2011 9694
dx.doi.org/10.1021/ac202630t | Anal. Chem. 2011, 83, 9694–9701
Analytical Chemistry sufficient to accurately map pharmaceutical compound distributions are increasing year on year.9 19 The applicability of MSI to the development of new PETradioligands for CNS-disorders is clear if it can provide information on compound distribution and temporal information on concentration in brain. [11C] SCH23390, a benzazepine, was the first radioligand developed for PET imaging of the D1 receptor,20 and the most commonly used radioligand for the PET imaging of D2 receptors is [11C] raclopride.21,22 The development of suitable PET-radioligands is sometimes hampered by the formation of radioactive metabolites that may obscure the binding of the radioligand itself and confound the modeling of such data. Comprehensive reviews describing the development of the MSI technique, highlighting the suitability for proteomic and pharmaceutical research have been published.23 31 With the efficacy, limitations and practical considerations relating to MSI being widely discussed in the literature,32 35 a detailed overview of MSI methodology will not be provided here. However, a recent modification of the matrix application process has involved the use of a solvent-free dry matrix coating (which we will now simply refer to as a dry matrix). This technique, whereby finely ground MALDI matrix is dusted over tissue sections, was first employed for the analysis of phospholipids36 but has also been shown to be beneficial in the detection of compounds that could not previously be detected when using a solvent based wet matrix.21,37 We present here the use of the dry matrix application enabling MALDI MSI of two widely used reference PET ligands, raclopride and SCH 23390, from rat brain and kidney tissue sections. Exploitation of three different types of MALDI mass spectrometer was performed. Compound identification and distribution were confirmed through mass accuracy and MS/MS fragmentation data. Both dose-dependent and a time course of distribution and abundance were measured. Semiquantitative data was also obtained by MALDI MSI analysis via spotting known amounts of the target compound on to the surface of control tissue. The ability to directly measure the distribution of PET ligands, without relying on radiolabeling, will increase the pace at which such compounds can be discovered and developed.
’ MATERIAL AND METHODS Materials and Reagents. All chemicals used were of analytical reagent grade. Raclopride was obtained from AstraZeneca R&D S€odert€alje (Sweden). SCH 23390 and MALDI-MS grade α-cyano-4-hydroxycinnamic acid (CHCA) were purchased from Sigma Aldrich (Sweden). Animals. All animal tissues were obtained from AstraZeneca (DMPK R&D, S€odert€alje, Sweden ethical approval no. 50/09). Adult Sprague Dawley rats (approximate weight 400 g) were given, intravenously into tail vein, a single dose of either 2 or 7.5 mg/kg of 3,5-dichloro-N-{[(2S)-1-ethylpyrrolidin-2-yl]methyl}2-hydroxy-6-methoxybenzamide (raclopride) or a single dose of 5 mg/kg of 7-chloro-3-methyl-1-phenyl-1,2,4,5-tetrahydro-3benzazepin-8-ol (SCH 23390). Compounds were formulated as an aqueous solution with sterile saline. Vehicle controls rats were also prepared with a single dose of saline solution. Animals were euthanized after being anaesthetized with isoflurane (Abbott Laboratories, U.S.A.). Organs were removed before snap-freezing in liquid nitrogen. Animals were sacrificed at 1, 5, or 30 min post dose (min p.d.) Tissues collected for short-term
ARTICLE
storage and transportation were maintained at below 25 C, while long-term storage was at 80 C. Tissue Processing. Sagittal and coronal sections were cut at a thickness of 14 μm and thaw mounted onto indium tin oxide (ITO) coated MALDI target slides (Bruker Daltonics, Germany Cat. # 237001). Sections were taken at approximately equal depth from the midline to allow visualization of major structures such as cerebral cortex, corpus callosum, hippocampus, striatum, and cerebellum. Tissue sections from dosed animals were mounted adjacent to vehicle control sections to minimize variability caused through variations in matrix application and when analyzing separate MALDI targets. Tissue sections were analyzed nonsequentially to limit the risk of any observed variation in relative abundance being as a result in loss of analyzer sensitivity during the course of the analysis. Mounted tissue sections were stored at 80 C until required. Matrix Coating. Optical images were taken using a standard flat bed scanner (Seiko Epson, Japan) prior to MALDI matrix application. Dry matrix coating was applied as previously described for the analysis of small molecules.37 In summary, CHCA was finely ground using a pestle and mortar. Tissue sections were taken from freezer storage and gently dried under stream of oxygen free N2 gas. Tissue sections were then dusted with a large excess of the finely ground CHCA. The MALDI target was then tipped and tapped to remove loosely adhering particles of MALDI matrix. Cycles of dusting and tapping were repeated approximately 10 times. Finally all the excess MALDI matrix not adhering to the section was removed by briefly placing target under a fast stream of N2 gas. Wet matrix coating was performed as previously described,34 using a pneumatic TLC sprayer (Sigma Aldrich). Once samples were coated in matrix subsequent transportation was performed with samples sealed in container to limit effects of light and humidity on sample and matrix adhesion. MALDI-MSI Analysis. MSI was initially performed using an Ultraflex II TOF/TOF (Bruker Daltonics) in positive ion reflectron mode using a Smartbeam II 200 Hz laser. The mass spectrometer parameters were as per the manufactures recommended settings adjusted for optimal imaging performance. Laser spot size was set at medium focus (∼50 μm laser spot diameter) and laser power optimized at the start of each run and then fixed for the MSI experiment. Further MSI was performed using a MALDI q-TOF MS (MALDI quad/TW-1M/oa-TOF, SYNAPT G2 HDMS, Waters Corporation, U.K.) and an 12T solariX MALDI FT-ICR MS (Bruker Daltonics, Germany). For analysis by the G2 SYNAPT, the region selected for imaging was defined using the MALDI Imaging Pattern Creator (Waters Corporation, U.K.), where the spatial resolution was also defined (typically 100 μm). The data was acquired in positive ion mode, operated in sensitivity mode over the range of m/z 200 600 with 300 laser shots per raster position using a 1 kHz laser. Analysis by FT-ICR employed a Smartbeam 1 kHz laser, with instrument control using SolariX control, version 1.5.0 (build 42.8), Hystar 3.4 (build 8). Each analysis was the result of 200 laser shots, using a laser spot diameter of ∼60 μm and a power level of 20%. Ions were detected between m/z 200 and 3000, yielding a 1 Mword time-domain transient, and with a laser spot raster spacing of 100 μm. On-tissue MS/MS fragmentation MSI was performed using both the Ultraflex II and the SYNAPT G2 while in positive ion modes (using the manufactures recommended settings). Optimization of the mass spectrometers was achieved by tuning acquisition settings while collecting data from 9695
dx.doi.org/10.1021/ac202630t |Anal. Chem. 2011, 83, 9694–9701
Analytical Chemistry a manually deposited control spot of the target compounds (1 μL of drug standard solution at approximately 0.1 mg/mL manually spotted onto the ITO target and overspotted with 1 μL of CHCA matrix in solution of 10 mg/mL in 50/50 v/v acetonitrile/water 0.1% TFA). MALDI-MSI Data Interpretation. Bruker MSI data was analyzed and normalized using FlexImaging, versions 2.0 through to 3.0 (Bruker Daltonics). Regions of interest were manually defined in the analysis software using both the optical image and MSI data image. Masses were selected with a mass precision of (0.1 Da for Ultraflex data sets and (0.001 Da for FT-ICR data sets. Average abundances were determined by defining each tissue section in an experiment as a separate region of interest (ROI) using FlexImaging software. A measure of the average abundance can then be taken from the summed spectra generated by the software. G2 SYNAPT raw data was converted into analyze file format using MALDI imaging converter (Waters Corporation, U.K.) and normalized by total ion current using an in-house written script. Subsequent image analysis was performed using BioMap (Novartis, Switzerland). Mass filter windows were selected with a precision of (0.1 Da. Quantification. Vehicle control tissue sections were thawmounted adjacent to the tissue to be quantified on MALDI targets and dried under a fast stream of N2 gas. A serial dilution of the target compound was prepared in a solution of 50/50 v/v methanol/water. A 0.1 μL deposition of quantitation standards were manually spotted onto vehicle control tissue sections. Once the required range of quantitation droplets had been spotted (and allowed to dry) the slide was returned to 80 C. Matrix application was then as described earlier.
’ RESULTS AND DISCUSSION Solvent-Free Dry Matrix. Many small molecule pharmacological and endogenous compounds can diffuse readily in the aqueous/organic solution used for wet matrix application, thus a major benefit of the dry matrix method is the elimination of this risk of analyte diffusion. Also, as a result of the sample not being wetted or rehydrated during the matrix coating, the risk of enzymatic or physiochemical modifications that have been reported are reduced.34 Moreover, the time needed to apply a matrix coating is substantially reduced (from hours using robotic spotter or spray coaters to a matter of minutes). Therefore the pace at which MSI screening can be performed is greatly increased following the removal of the matrix coating period. However, such benefits are an aside to the main reason for employing a dry matrix application, which is that is has been demonstrated to be successful for allowing the detection of small molecules that could not be detected using a conventional solution-based matrix coating.21,37 Rat brain tissue sections were prepared from whole rat brains taken from animals treated with raclopride. MALDI MSI analysis was performed using MALDI matrix (CHCA) applied either uniformly as an organic/aqueous solution or using the dry matrix application method. Spectra were collected directly from tissue sections over a mass range of m/z 150 1000. Detection of the PET ligand standards was readily achieved when they were manually spotted with a standard CHCA matrix solution. However, we were unable to detect the compound from the drug treated tissue sections when the matrix coating was applied as a wet organic/aqueous solution. Figure 1 highlights the broad effect that matrix application has on detected masses. Figure 1a is
ARTICLE
a spectrum collected from dry matrix coated raclopride containing tissue. Figure 1b is a spectrum collected from wet matrix coated raclopride dosed tissue. When comparing the spectra we observed that for dry matrix application a lower overall number of masses were detected (
