**3.1.1 Sample condition for MALDI-IMS**

Collection and treatment procedures need to be sufficiently fast to prevent rapid tissue degradation, because the sample degradation process starts immediately after the cessation of blood flow. The most preferred sample for MALDI-IMS is a chemically unmodified freshfrozen one. Fresh-frozen samples can be prepared using dry ice, liquid nitrogen, or liquid nitrogen-chilled isopentane, and can be preserved in a deep freezer until required. The samples should be well sealed to prevent drying during storage, and it is important to ensure that the tissue section morphology is well preserved before MALDI-IMS.

Application of Matrix-Assisted Laser Desorption/Ionization Imaging Mass Spectrometry 441

Attachment of adhesive film to the sample block (a). The end of the adhesive film must be anchored with tweezers to prevent adhesion of the film to the sample stage (b). After the sample section is obtained (c), the sample section on the adhesive film is attached to a glass

Washing is required for peptide or protein analysis because their detection is often prevented by large amounts of easily ionized lipid species. Lipid removal simplifies mass spectra in the range of *m/z* 400–1000; thus, lipid removal enables the detection of low-mass peptides that are usually masked by lipid peaks. The washing method should be optimized for the target imaging molecules. Several washing protocols using organic solvents have been reported (Aerni et al., 2006; Andersson et al., 2008; Groseclose et al., 2007; Lemaire et

Washing is also used for removing the matrix from the tissue section after MALDI-IMS analysis. The matrix can be removed using the solvent that is used for preparing the matrix solution. For example, 2,5-dihydroxybenzoic acid (DHB) can be rapidly removed by methanol. Matrix removal enables the microscopic observation of a tissue section followed by pathological staining, such as hematoxylin and eosin (HE) staining, toluidine blue

Fig. 3. Procedure for sectioning using adhesive film.

slide (d).

**3.1.4 Washing** 

staining etc.

al., 2006; Schwartz et al., 2003).

#### **3.1.2 Fixation and embedding**

Fixation of samples, such as formalin fixation, is preferably avoided because the protein crosslinking introduced by formalin fixation makes MALDI-IMS analysis difficult. However, many medical samples are routinely formaldehyde-fixed and paraffin-embedded (FFPE) just after dissection. To address this problem, the on-tissue proteolytic digestion method, in which proteins are denatured and digested by enzymes, has been developed (Djidja et al., 2009; Groseclose et al., 2007; Lemaire et al., 2007; Morita et al., 2010). The ontissue proteolytic digestion method includes a paraffin removal step using xylene and ethanol. In the paraffin removal step, lipophilic molecules are lost; therefore, FFPE samples cannot be used for lipid imaging. When the samples are formaldehyde-fixed without paraffin-embedding, lipid imaging can be performed (Zaima et al., 2011c). However, the detected ion intensities of lipids in formaldehyde-fixed samples are lower than those in fresh-frozen ones are.

Embedding of the tissue samples in supporting material, such as an optimal cutting temperature (OCT) compound, allows the maintenance of tissue morphology and precise sample sectioning. However, supporting materials are often ionized during MALDI-MS analysis and sometimes act as ion suppressors of molecules of interest (Schwartz et al., 2003). Therefore, samples should not be embedded if precise sample sections can be prepared without embedding. When it is difficult to prepare a sample section, the use of carboxymethylcellulose (CMC) or gelatin as embedding material is recommended. Sodium CMC (2%) is reported to be used as an alternative embedding compound that does not interfere with the detection sensitivity of biomolecules in MALDI-IMS analysis (Stoeckli et al., 2006; Zaima et al., 2010a). Chen et al. reported that gelatin provides a cleaner signal background than OCT (Chen et al., 2009). Researchers should ensure compatibility between the supporting material and the biomolecules of interest.

#### **3.1.3 Sectioning**

The basic sectioning procedure for MALDI-IMS samples is same as that for pathological examination. Sections for MALDI-IMS can be prepared using a cryostat. The sample stage temperature is typically maintained between -5 and -20°C. To obtain high quality sections from tissues with high fat content (e.g., brain), or atherosclerotic lesions, breast tissue, or lipid storage disease samples lower temperatures are required. In general, 5–20-μm-thick sections are prepared for the analysis of low-molecular-weight molecules. The use of thinner tissue sections (2–5 μm thick) has been recommended for the analysis of high-molecularweight molecules (range, 3–21 kDa) (Goodwin et al., 2008). Sections are usually thawmounted on a stainless steel conductive stage or on commercially available indium-tin oxide (ITO)-coated glass slides. We recommend the use of ITO-coated glass slides because these transparent slides enable microscopic observation of the section after MALDI-IMS. Use of adhesive film is suitable for samples for which thaw-mounted preparation of sections is challenging (e.g., bone or whole-body sections) (Stoeckli et al., 2006; Zaima et al., 2010a). The procedure for sectioning using adhesive film is shown in Figure 3. The prepared section should be immediately dried in a vacuum desiccator to avoid moisture condensation that could cause delocalization of analyte molecules in the tissue. Moisture condensation can be avoided by placing the prepared section in a dry and cold container until return to room temperature.

Fig. 3. Procedure for sectioning using adhesive film.

Attachment of adhesive film to the sample block (a). The end of the adhesive film must be anchored with tweezers to prevent adhesion of the film to the sample stage (b). After the sample section is obtained (c), the sample section on the adhesive film is attached to a glass slide (d).

#### **3.1.4 Washing**

440 Pharmacology

Fixation of samples, such as formalin fixation, is preferably avoided because the protein crosslinking introduced by formalin fixation makes MALDI-IMS analysis difficult. However, many medical samples are routinely formaldehyde-fixed and paraffin-embedded (FFPE) just after dissection. To address this problem, the on-tissue proteolytic digestion method, in which proteins are denatured and digested by enzymes, has been developed (Djidja et al., 2009; Groseclose et al., 2007; Lemaire et al., 2007; Morita et al., 2010). The ontissue proteolytic digestion method includes a paraffin removal step using xylene and ethanol. In the paraffin removal step, lipophilic molecules are lost; therefore, FFPE samples cannot be used for lipid imaging. When the samples are formaldehyde-fixed without paraffin-embedding, lipid imaging can be performed (Zaima et al., 2011c). However, the detected ion intensities of lipids in formaldehyde-fixed samples are lower than those in

Embedding of the tissue samples in supporting material, such as an optimal cutting temperature (OCT) compound, allows the maintenance of tissue morphology and precise sample sectioning. However, supporting materials are often ionized during MALDI-MS analysis and sometimes act as ion suppressors of molecules of interest (Schwartz et al., 2003). Therefore, samples should not be embedded if precise sample sections can be prepared without embedding. When it is difficult to prepare a sample section, the use of carboxymethylcellulose (CMC) or gelatin as embedding material is recommended. Sodium CMC (2%) is reported to be used as an alternative embedding compound that does not interfere with the detection sensitivity of biomolecules in MALDI-IMS analysis (Stoeckli et al., 2006; Zaima et al., 2010a). Chen et al. reported that gelatin provides a cleaner signal background than OCT (Chen et al., 2009). Researchers should ensure compatibility between

The basic sectioning procedure for MALDI-IMS samples is same as that for pathological examination. Sections for MALDI-IMS can be prepared using a cryostat. The sample stage temperature is typically maintained between -5 and -20°C. To obtain high quality sections from tissues with high fat content (e.g., brain), or atherosclerotic lesions, breast tissue, or lipid storage disease samples lower temperatures are required. In general, 5–20-μm-thick sections are prepared for the analysis of low-molecular-weight molecules. The use of thinner tissue sections (2–5 μm thick) has been recommended for the analysis of high-molecularweight molecules (range, 3–21 kDa) (Goodwin et al., 2008). Sections are usually thawmounted on a stainless steel conductive stage or on commercially available indium-tin oxide (ITO)-coated glass slides. We recommend the use of ITO-coated glass slides because these transparent slides enable microscopic observation of the section after MALDI-IMS. Use of adhesive film is suitable for samples for which thaw-mounted preparation of sections is challenging (e.g., bone or whole-body sections) (Stoeckli et al., 2006; Zaima et al., 2010a). The procedure for sectioning using adhesive film is shown in Figure 3. The prepared section should be immediately dried in a vacuum desiccator to avoid moisture condensation that could cause delocalization of analyte molecules in the tissue. Moisture condensation can be avoided by placing the prepared section in a dry and cold container until return to room

the supporting material and the biomolecules of interest.

**3.1.2 Fixation and embedding** 

fresh-frozen ones are.

**3.1.3 Sectioning** 

temperature.

Washing is required for peptide or protein analysis because their detection is often prevented by large amounts of easily ionized lipid species. Lipid removal simplifies mass spectra in the range of *m/z* 400–1000; thus, lipid removal enables the detection of low-mass peptides that are usually masked by lipid peaks. The washing method should be optimized for the target imaging molecules. Several washing protocols using organic solvents have been reported (Aerni et al., 2006; Andersson et al., 2008; Groseclose et al., 2007; Lemaire et al., 2006; Schwartz et al., 2003).

Washing is also used for removing the matrix from the tissue section after MALDI-IMS analysis. The matrix can be removed using the solvent that is used for preparing the matrix solution. For example, 2,5-dihydroxybenzoic acid (DHB) can be rapidly removed by methanol. Matrix removal enables the microscopic observation of a tissue section followed by pathological staining, such as hematoxylin and eosin (HE) staining, toluidine blue staining etc.

Application of Matrix-Assisted Laser Desorption/Ionization Imaging Mass Spectrometry 443

be adequately extracted from the tissue section. The operation should be performed at a constant room temperature and humidity. Beginners are recommended to practice spraying until homogeneous matrix spraying can be reproducibly achieved. Sublimation is a new method for applying matrix to tissue sections (Hankin et al., 2007). Using this technique, a matrix can be applied uniformly over a large sample plate in a few minutes without solvents. Additionally, previous reports demonstrated that this method increases analyte signal and that the fine microcrystals formed from the condensed vapor reduce the image resolution

MALDI-IMS should be performed as soon as possible after matrix application, regardless of the coating method. The procedure to obtain a good spectrum in MALDI-IMS is almost the same as that for traditional MALDI-MS; mass range, detector gain, and laser power must be optimized. From the mechanical setting perspective, there are 3 differences between MALDI-MS and MALDI-IMS. The first difference is the above-mentioned matrix application. The second difference is the need for focusing of the laser beam. To obtain meaningful biological images by MALDI-IMS, the laser spot size should be reduced to 10–50 μm. The third difference is that a two-dimensional region must be set for analyses. The scan pitch, which signifies the distance between laser irradiation spots, must be fixed. The limitation of the scan pitch, which decides the spatial resolution of the image, depends on the laser spot size and mechanical movement control of the mass spectrometer sample stage. We have developed a new instrument (Mass Microscope) that can move the sample stage by 1 μm, and in which the finest size of the laser diameter is approximately 10 μm (Harada et al., 2009). The measurement time depends on the number of data spots, the frequency of the laser, the number of shots per spot, and the time required to move the sample stage. For example, when researchers select the region of interest as a 1 × 1 mm2 area with a 10-μm scan pitch (10,000 data points), it takes about 1 h to complete the measurement using a mass

MALDI-IMS ionizes numerous compounds in a tissue at the same time. Sometimes, we detect multiple molecules with the same *m/z* value. In such cases, a new imaging technique, "MS/MS imaging," is effective. Using this technique, we can separate each ion derived from their specific fragment ions. Some reports have described the use of MS/MS imaging for IMS of endogenous metabolites and an exogenous drug (Khatib-Shahidi et al., 2006; Porta et al., 2011). Additionally, the combination of ion-mobility separation with MALDI-IMS provides a unique separation dimension to further enhance the capabilities of IMS (Jackson et al., 2007; McLean et al., 2007; Stauber et al., 2010). It can be used to produce images without interference from background ions of similar mass, and this can remove ambiguity from imaging experiments

A large amount of data (a few gigabytes) is obtained from MALDI-IMS; therefore, visualization software packages that can rapidly and efficiently analyze enormous spectra have been developed. BioMap (a free software; Novartis, Basel, Switzerland), FlexImaging

limitation caused by crystal size (Dekker et al., 2009; Vrkoslav et al., 2010).

microscope equipped with a 1000-Hz laser (100 shots/data point).

and lead to a more precise localization of the compound of interest.

**3.3 Measurement and data analysis** 

**3.3.1 Measurement** 

**3.3.2 Data analysis** 

#### **3.2 Matrix application**

The matrix plays a central role in MALDI-MS soft ionization (Karas & Hillenkamp, 1988; Karas & Kruger, 2003). Biomolecules are softly ionized in the cocrystal with the matrix, which absorbs the laser beam energy and protects biomolecules from the disruptive energy. Protonated ion ([M + H]+) or deprotonated ion ([M − H]−) molecules are generally detected. Sodium adduct ion ([M + Na]+) and potassium adduct ion ([M + K]+) are often observed by biological sample analysis. It is very important to choose appropriate matrices for obtaining meaningful biomolecular images. An overview of the matrices used for IMS can also be found in other reviews (Chughtai & Heeren, 2010; Kaletas et al., 2009).
