Visualizing spatial distribution of alectinib in murine brain using quantitative mass spectrometry imaging

In the development of anticancer drugs, drug concentration measurements in the target tissue have been thought to be crucial for predicting drug efficacy and safety. Liquid chromatography-tandem mass spectrometry (LC-MS/MS) is commonly used for determination of average drug concentrations; however, complete loss of spatial information in the target tissue occurs. Mass spectrometry imaging (MSI) has been recently applied as an innovative tool for detection of molecular distribution of pharmacological agents in heterogeneous targets. This study examined the intra-brain transitivity of alectinib, a novel anaplastic lymphoma kinase inhibitor, using a combination of matrix-assisted laser desorption ionization–MSI and LC-MS/MS techniques. We first analyzed the pharmacokinetic profiles in FVB mice and then examined the effect of the multidrug resistance protein-1 (MDR1) using Mdr1a/b knockout mice including quantitative distribution of alectinib in the brain. While no differences were observed between the mice for the plasma alectinib concentrations, diffuse alectinib distributions were found in the brain of the Mdr1a/b knockout versus FVB mice. These results indicate the potential for using quantitative MSI for clarifying drug distribution in the brain on a microscopic level, in addition to suggesting a possible use in designing studies for anticancer drug development and translational research.

In the development of anticancer drugs, drug concentration measurements in the target tissue have been thought to be crucial for predicting drug efficacy and safety. Liquid chromatography-tandem mass spectrometry (LC-MS/MS) is commonly used for determination of average drug concentrations; however, complete loss of spatial information in the target tissue occurs. Mass spectrometry imaging (MSI) has been recently applied as an innovative tool for detection of molecular distribution of pharmacological agents in heterogeneous targets. This study examined the intra-brain transitivity of alectinib, a novel anaplastic lymphoma kinase inhibitor, using a combination of matrix-assisted laser desorption ionization-MSI and LC-MS/MS techniques. We first analyzed the pharmacokinetic profiles in FVB mice and then examined the effect of the multidrug resistance protein-1 (MDR1) using Mdr1a/b knockout mice including quantitative distribution of alectinib in the brain. While no differences were observed between the mice for the plasma alectinib concentrations, diffuse alectinib distributions were found in the brain of the Mdr1a/b knockout versus FVB mice. These results indicate the potential for using quantitative MSI for clarifying drug distribution in the brain on a microscopic level, in addition to suggesting a possible use in designing studies for anticancer drug development and translational research.
During the drug development process, determination of whether a new compound can reach its intended target or not is one of the key factors for obtaining the proof of mechanism and concept 1 . A previous study has shown that plasma drug concentrations do not accurately predict the exposure in some organs and tissues 2 . To determine drug exposure in the target tissues, today, homogenized tissues are now commonly used for quantitative assessments by liquid chromatography tandem mass spectrometry (LC-MS/MS).
Although LC-MS/MS can accurately determine the average levels of a compound in a target organ or tissue, there is a disadvantage when using this technique, as there is a complete loss of the spatial information for the compound in the target tissue. When estimating the efficacy and safety of a compound, particularly for anti-tumor drugs, it is very important to know both the specific location of the compound and the concentration that is present at these points within the target tissue. In order to understand the complex biochemical interactions that occur between a drug and its target tissue, it is necessary to detect not only the level of exposure, but also the spatial distribution within the target tissue. Thus, the use of only LC-MS/MS analysis is insufficient for clarifying the behavior of compounds such as anti-tumor drugs 3 .
Matrix-assisted laser desorption ionization mass spectrometry imaging (MALDI-MSI) is one of the newer innovative technologies that are highly sensitive and can provide mass spectrum information in two dimensions. This technology makes it possible to measure the distribution of diverse molecular species in a tissue section, including endogenous molecules at a specific time, without destroying the target tissue or requiring the use of target-specific molecular labeling reagents 2,4-9 . Even though this technique has proved to be beneficial, some challenges remain regarding the quantitative capability and reproducibility.
Assessment of the drug concentration and distribution is not only of interest in tumors but also in the brain as well. The blood-brain barrier (BBB), in particular the multidrug resistance protein 1 (MDR1, also known as P-glycoprotein or ATP-binding cassette, sub-family B, member 1: ABCB1) has a central role in the drug distribution and transition to the brain parenchyma 10 . MDR1 is an important transporter protein located in the cell membrane that pumps many endogenous substances out of the cells. Moreover, this protein is highly expressed in various tissues, including the endothelial cells of the brain blood vessels and the intestinal epithelium 10 . While MDR1 has been shown to regulate the intestinal absorption of a wide variety of drugs, it has been frequently reported that the blood concentration of a drug does not necessarily correlate with the MDR1-determined substance properties [11][12][13][14][15][16] . For instance, a pharmacokinetics (PK) study showed that after administration of the anti-cancer drug, docetaxel, in FVB (wild-type, WT) and the Mdr1a/b knockout (KO) mice, there was no difference in the plasma concentrations. However, the brain concentrations in the Mdr1a/b KO mice were much higher than that observed in the FVB mice 15 .
In our current study, we employed a quantitative MSI (qMSI) method in a preclinical mouse model to evaluate the intra-brain distribution of a new anti-cancer drug, the echinoderm microtubule-associated protein-like 4 and anaplastic lymphoma kinase (EML4-ALK) inhibitor, alectinib [17][18][19] , which was approved for use in Japan on July 4, 2014. Using serial sections, we used the combination of MALDI-MSI and LC-MS/MS techniques to facilitate inter-sample comparisons and improve reproducibility of the analysis, in addition to investigating the effect of MDR1 on the intra brain transitivity of alectinib in Mdr1a/b KO mice. The aim of this study was to examine the potential of utilizing qMSI to translate drug delivery evaluations for use in new drug development and practical applications.

Pharmacokinetics of alectinib in mice.
Recently, there has been an increased interest in the pharmacokinetics of anti-tumor drugs in the brain from an efficacy and toxicity viewpoint 20,21 . Therefore, our current study focused on alectinib, which was recently approved for the treatment of lung cancer in Japan (Fig. 1A). We performed a pharmacokinetic analysis of alectinib in FVB and Mdr1a/b KO mice; Plasma, cerebrospinal fluid (CSF), and brain tissue were collected 1, 1.5, 2, and 4 hours after a single oral administration of a 4 and 20 mg/kg dose. In mice, a 20 mg/kg administration of alectinib resulted in higher plasma concentrations as compared to a 4 mg/kg administration (Supplemental Table 1). There were no significant differences in the alectinib plasma concentrations observed between the FVB and Mdr1a/b KO mice at each dosage (Fig. 1B).
The majority of the alectinib concentrations in the CSF were lower than the limits of quantitation (4 mg/kg group: 16/16 samples in FVB mice, 15/16 samples in Mdr1a/b KO mice; 20 mg/kg group: 9/16 samples in FVB  Table 2).
Use of qMSI to visualize alectinib distribution in the mouse brain. To determine alectinib transitivity into the mouse brain, we performed qMSI analysis (Methods and Supplemental Figure 1) using FVB and Mdr1a/b KO murine brain sections. Brain tissue collected included the cerebrum, cerebellum, and choroid plexus (Fig. 2). Alectinib qMSI was performed on mouse brains that were collected at 1, 1.5, 2 and 4 hours after the administration of alectinib (4 mg/kg, 20 mg/kg). There was a higher level of drug accumulation and a greater diffusion in the brain tissue section after the administration of 4 mg/kg of alectinib in the Mdr1a/b KO versus the FVB group ( Fig. 2) (Supplemental Figure 2). After administration of 20 mg/kg of alectinib, there was a significantly increased alectinib exposure and diffused distribution of alectinib in Mdr1a/b KO versus FVB murine brains (Fig. 3).
Quantitative distribution of alectinib shown by the qMSI was confirmed by using laser microdissection (LMD) of additional serial sections (Supplemental Figure 4). The calculated concentrations of the arbitrary regions R1 to R8 from the images of qMSI correlated well with the concentration of the corresponding LMD R1 to R8 regions measured by LC-MS/MS (Supplemental Table 5).
Localization of alectinib in the brain blood vessels. The final step of the experiment focused on the histological intra brain localization of alectinib. After treating the FVB and Mdr1a/b KO mice with 20 mg/kg of alectinib, we used MSI at a 20-micrometer resolution and immunohistochemistry (IHC) to examine the localization of alectinib in the brain blood vessels (Fig. 5). In the FVB mice, MSI showed that the distribution of alectinib was mostly co-localized with the heme that was derived from the red blood cells (Fig. 5B), while IHC demonstrated there was distribution in the CD31 and MDR1 (Fig. 5A), which indicated the location of the blood vessels in the choroid plexus. On the other hand, a diffuse distribution of alectinib was observed in the Mdr1a/b KO mouse brain (Fig. 5B). By using MSI, it was possible to visualize the true distribution of alectinib in the brains of both the Mdr1a/b KO and the FVB mice. Thus, we were able to evaluate the precise pharmacokinetics of the drug in target tissues or organs at the micro level.

Discussion
In this study, we propose application of this quantitative MSI method, which would add the quantitative capability to conventional MSI by using concurrent LC-MS/MS analysis of serial sections, into the visualizing study of spatial distribution of pharmaceutical agent in the brain. We evaluated the intra-brain distribution of alectinib and  the effect of MDR1 (p-glycoprotein) on pharmaceutical profiles with MSI concurrently with a general LC-MS/ MS approach by using an in vivo murine model. The alectinib PK study using Mdr1a/b KO mice showed that while the alectinib brain distribution was affected by MDR1, it was not associated with the blood concentration.
MSI can directly detect diverse molecular species on tissue sections simultaneously and utilize this information to create a drug distribution map without the use of target-specific molecular labeling reagents 22 . This technique could potentially be useful for detecting drug concentrations and distributions within target tumors and organs, which have complicated structures. Furthermore, maturation and standardization of this procedure could also contribute to the prediction of efficacy and safety during the drug development process. Several studies have reported on the development of quantitative MSI methodologies 23,24 . One quantitative approach is to create a standard curve by determining known concentrations of a target drug in sections of control tissue, after which this curve is then compared to the signal intensity observed for adjacent dosed tissue sections [25][26][27][28] ; however, some problems may occur with this technique including issues with choosing an appropriate control tissue, difficulties when trying to examine cases with heterogeneous tissues that have been determined to be the cause of differences in ion suppression, or difficulties in deciding what the concentrations may be in dried spot areas. To avoid these potential problems, in this study, we selected a quantitative method using LC-MS/MS of adjacent serial sections 23,29 and combined this with the use of an internal standard that was sprayed within the matrix.
Recent studies have used MSI to analyze the activity of various compounds with regard to their pharmaceutical transit into the brain 16,30-32 . However, there have been few reports that have mentioned the reproducibility and quantitative capacity of targeting compounds. Even when a compound is a substrate of MDR1, which closely regulates intestinal absorption as well as brain distribution, there have been other cases reported, similar to our current study, where the brain transitivity differed from the blood concentrations found after an oral administration of a drug [13][14][15] . Generally, the drug transport and transit into the brain has been evaluated through the use of radioisotopes or LC-MS/MS analyses of homogenized brain tissue, precluding accurate determination of where the drug localized in the brain tissue, brain parenchyma, CSF, or brain blood. In addition, the calculation of drug concentrations in the CSF is currently a controversial subject 33 . Therefore, our current approach was able to precisely evaluate drug distribution at the microscopic level in wild-type and Mdr1a/b KO mice and could be a powerful tool in drug development.
While the resolution of the MSI images was a limitation of our current study, a conflict still exist between the analytical resolution and sensitivity in our MSI system 34 . There are recent reports that have described single cell imaging through the use of different MALDI-MSI system 35,36 . Although our data showed the positional relationship between alectinib and the brain blood vessels, further improvements in the specificity, sensitivity, and image resolution may give us further novel insights into PK research at a cellular level. Another limitation of our study is that even though our results showed an association between the brain distribution of alectinib and the substrate property of MDR1, we did not specifically examine the efficacy of alectinib with regard to brain metastasis. The reason for this is that we did not use a tumor xenografted model. Since it has been reported that alectinib has exhibited promising activity in lung cancer patients with brain metastasis 37 , our data might support the speculation that the intact brain transitivity of drugs or the substrate property of MDR1 do not directly correlate with the drug penetration and treatment efficacy in brain metastatic tumors 17,38 . Interestingly, a recent study showed that even macromolecule such as antibody-drug was possible to localize at metastatic brain lesions of breast cancer by PET imaging of 64 Cu-labeled trastuzumab 39 . Further translational research using brain metastatic samples and MSI of drugs might be able to resolve this controversial clinical question.
In conclusion, the spatial measurement of drug concentrations in target tissues and organs is crucial for predicting efficacy and safety during drug development and clinical application, and MSI technology is widely expected to contribute to this evaluation. It is hoped that present approach of using quantitative MSI and Mdr1a/b KO mice during the evaluation of brain drug distribution might be helpful in the determination of reproducible PK and pharmacodynamic studies of investigational new compounds.

Methods
Chemicals. Alectinib (Alecensa ® ) was purchased from Chugai Pharmaceutical Co., Ltd. (Tokyo, Japan). Pharmacokinetic studies in mice. Alectinib was administered orally to the FVB and Mdr1a/b KO mice at doses of 4 and 20 mg/kg (n = 4). At 1, 1.5, 2 and 4 hours after the drug administration, brain, blood and cerebrospinal fluid (CSF) samples were collected from each mouse. Blood samples were collected in heparinized tubes, and centrifuged at 1500 × g at 4 °C for 10 min to obtain the plasma. After freezing the brains in liquid N 2 , they were stored at − 80 °C until analyzed. Plasma and CSF were also stored at − 80 °C until analyzed.

LC-MS/MS analysis.
Chromatographic separation of the alectinib, the plasma internal standard, erlotinib-d6, CSF, and the tissue sections was performed using an XBridge C18 HPLC column (2.1 × 5.0 mm, 3.5 μm) maintained at 40 °C. Mobile phases A and B consisted of 0.1% FA aqueous solution and methanol containing 0.1% FA, respectively. Separation was performed using an isocratic elution (A:B = 1:1) at a flow rate of 0.2 mL/min with a Nexera X2 series (Shimadzu, Co., Kyoto, Japan). The column was equilibrated with the mobile phase for 10 min, with a run time of 6 min for each 2 μL injection volume.
Quantitation was conducted by selected reaction monitoring on a QTRAP4500 mass spectrometer (AB SCIEX, Framingham, MA, USA) with electrospray ionization in the positive mode. The optimized electrospray ionization parameters were as follows: ion source temperature, 700 °C; curtain gas, setting 30; nebulizing gas (GS1), setting 60; turbo-ionspray gas (GS2), setting 60; ionspray voltage, 4500 V; declustering potential, 121 V; collision energy, 35 V. The selected reaction monitoring transitions were m/z 483.2 to 396.0 for alectinib and m/z 400.0 to 338.9 for erlotinib-d6. The dwell time was 500 ms for each transition channel. All data were acquired and analyzed using the Analyst version 1.6.1 software (AB SCIEX). qMSI. Sectioning. qMSI requires at least three serial sections (8 μm). Tissue was sliced at − 20 °C with a cryomicrotome (Leica CM 1950, Tokyo, Japan). Brain tissues were sliced in both the coronal and sagittal directions (Supplemental Figure 1). The first and third sections were used for the LC-MS/MS measurements of the amount of alectinib contained in the tissue, while the second section was used for MALDI-MSI. The MSI sections were placed on indium-tin-oxide coated glass slides (ITO glass, Matsunami Glass Ind., Ltd., Tokyo, Japan). Extraction. For the alectinib extraction, after vortexing the first and third tissue sections in 300 μL of 50% methanol/water for 1 min, sections were centrifuged at 12,000 × g for 10 min at 4 °C and the supernatants were collected. A portion of the supernatant was used to measure the protein concentration using a Qubit 2.0 Fluorometer (Thermo Fisher Scientific, MA, USA) and was also used as a reference for the quantitation of alectinib. A standard addition method 42 was used for quantitating alectinib in the samples to avoid the matrix effect. Alectinib in the pieces that were cut off by LMD (Leica LMD 6500, Tokyo, Japan) from parts of the three additive serial sections of the tissue was also extracted using the above procedure. The processed samples were then analyzed by LC-MS/MS.

MALDI-MSI.
The two-step matrix technique was used 43,44 . α -CHCA was first coated on the slide using a homebuilt sublimation apparatus at 250 °C for 8 min, after which 10 mg/mL of α -CHCA solution containing 30% acetonitrile, 10% isopropanol, and 0.1% FA was sequentially sprayed onto the slide using a sprayer (PS270, GSI Creos Corp., Tokyo, Japan) according to the following 10 cycles: first three cycles of spraying for 3 seconds at 90-second intervals, next seven cycles of spraying for 1 second at 30-second intervals. For examination of the ionization efficiency of alectinib on the tissue section, erlotinib-d6 was used as the internal standard (IS) in the α -CHCA solution (2 μg/mL).
MALDI-MSI acquisition was performed using an iMScope (Shimadzu). This instrument consisted of an optical microscope and a quadrupole ion trap time-of-flight (qIT-TOF) analyzer with an atmospheric pressure MALDI source consisting of a 355-nm Nd:YAG laser (1,000 Hz, minimum spot diameter < 10 μm) 34 . MSI data for the brain sections were acquired with a pixel step size for the surface raster set to a spatial resolution of 80 or 20 μm. A mass spectrum was obtained with 150 (50 × 3 times) laser shots each over a mass range of 50-500 Da at a resolution of 10,000 at m/z 1,000.
The optimal transitions for the alectinib and erlotinib-d6 as IS used for the MSI were m/z 483.3 to 396.2 (Supplemental Figure 5) and m/z 400.2 to 339.2, respectively. The IS data were acquired with 40 μm shift from the laser spot obtained with alectinib. Heme was measured as a surrogate marker of the vasculature by monitoring the ion of m/z 616.2 ± 0.02 30 , and the specificity of heme was confirmed by MS/MS analysis (m/z 616.2 to 557.2, data not shown).
Molecular images were visualized using the Imaging MS Solution ver. 1.11 (Shimadzu) with a scale bar of absolute ion intensity. Biomap 3.8.0.4 (Novartis Institutes for BioMedical Research, Basel, Switzerland) was also used to display the divided images by the IS or qMSI images, in which we converted the ion intensity of the scale bar into the amount of alectinib using the sum of intensity in MSI and the total quantity of alectinib using LC-MS/ MS analysis of serial sections.
Immunohistochemistry. Brain tissue sections were serially sectioned at an 8 μm thickness for immunohistochemical detection of vascular endothelial cells and MDR1. After fixing the sections with cold methanol at − 20 °C for 10 min, sections were dried at ambient temperature, and then incubated with the primary anti-CD31 antibody (#ab28364, Abcam, Cambridge, MA, USA) at a 1:50 dilution or with anti-MDR1 (#ab170904, Abcam) at a 1:50 dilution overnight at 4 °C. After the incubation, sections were washed with 1 × TBS buffer and then further incubated with SignalStain Boost IHC Detection Reagent (anti-rabbit, Cell Signaling Technology, Tokyo, Japan) at room temperature for 30 minutes. BZ-X710 microscope (Keyence, Itasca, IL, USA) was used for the observation of stained sections and calculation of the section area. Statistical analysis. Unless stated otherwise, all data are presented as the mean ± standard deviation of the mean (SD) from three or more animals. All quantification and calibration data were statistically processed using JMP software version 12.0.1 for MAC (SAS Institute Japan, Tokyo, Japan). The significance of differences was evaluated using a unilateral t-test, with a P value < 0.05 considered to be statistically significant. Pharmacokinetic parameters for alectinib in the mouse plasma were calculated using the non-compartmental analysis of the WinNonlin software (version 6.3, Pharsight Corp., Mountain View, CA, USA).