Mass spectrometric imaging (MSI) is a powerful tool that can be used to discover and identify various chemical species in intact tissues, preserving the compounds in their native environments, which can provide new insights into biological processes. Herein a MSI method developed for the analysis of small molecules is described.
Most techniques used to study small molecules, such as pharmaceutical drugs or endogenous metabolites, employ tissue extracts which require the homogenization of the tissue of interest that could potentially cause changes in the metabolic pathways being studied1. Mass spectrometric imaging (MSI) is a powerful analytical tool that can provide spatial information of analytes within intact slices of biological tissue samples1-5. This technique has been used extensively to study various types of compounds including proteins, peptides, lipids, and small molecules such as endogenous metabolites. With matrix-assisted laser desorption/ionization (MALDI)-MSI, spatial distributions of multiple metabolites can be simultaneously detected. Herein, a method developed specifically for conducting untargeted metabolomics MSI experiments on legume roots and root nodules is presented which could reveal insights into the biological processes taking place. The method presented here shows a typical MSI workflow, from sample preparation to image acquisition, and focuses on the matrix application step, demonstrating several matrix application techniques that are useful for detecting small molecules. Once the MS images are generated, the analysis and identification of metabolites of interest is discussed and demonstrated. The standard workflow presented here can be easily modified for different tissue types, molecular species, and instrumentation.
The growing field of metabolomics has many important biological applications including biomarker discovery, deciphering metabolic pathways in plants and other biological systems, and toxicology profiling4,6-10. A major technical challenge when studying biological systems is to study metabolomic pathways without disrupting them11. MALDI-MSI allows for direct analysis of intact tissues that enables sensitive detection of analytes in single organs12,13 and even single cells14,15.
Sample preparation is a crucial step in producing reproducible and reliable mass spectral images. The quality of the images greatly depends upon factors such as tissue embedding medium, slice thickness, MALDI matrix, and matrix application technique. For imaging applications, ideal section thickness is the width of one cell (8-20 µm depending on the sample type). MALDI requires deposition of an organic, crystalline matrix compound, typically a weak acid, on the sample to assist analyte ablation and ionization.16 Different matrices provide different signal intensities, interfering ions, and ionization efficiencies of different classes of compounds.
The matrix application technique also plays a role in the quality of mass spectral images and different techniques are appropriate for different classes of analytes. Three matrix application methods are presented in this protocol: airbrush, automatic sprayer, and sublimation. Airbrush matrix application has been widely used in MALDI imaging. The advantage of airbrush matrix application is that it is relatively fast and easy. However, the quality of the airbrush matrix application greatly depends on the skill of the user and tends to be less reproducible and cause diffusion of analytes, especially small molecules17. Automatic sprayer systems have similar mechanics to airbrush matrix application, but have been developed to remove the variability seen with manual airbrush application, making the spray more reproducible. This method can sometimes be more time-consuming than traditional airbrush matrix application. Both manual airbrush and automatic sprayer systems are solvent-based matrix application methods. Sublimation is a dry matrix application technique that is becoming more and more popular for mass spectral imaging of metabolites and small molecules because it reduces analyte diffusion; however, it lacks the solvent necessary to extract and observe higher mass compounds18.
Confident identification of metabolites typically requires accurate mass measurements to obtain putative identifications followed by tandem mass (MS/MS) experiments for validation, with MS/MS spectra being compared to standards, literature, or theoretical spectra. In this protocol high resolution (mass resolving power of 60,000 at m/z 400), liquid chromatography (LC)-MS is coupled to MALDI-MSI to obtain both spatial information and confident identifications of endogenous metabolites, using Medicago truncatula roots and root nodules as the biological system. MS/MS experiments can be performed directly on the tissue with MALDI-MSI or on tissue extracts with LC-MS and used for the validation of metabolite identifications.
This protocol provides a simple method to map endogenous metabolites in M. truncatula, which can be adapted and applied to MSI of small molecules in various tissue types and biological systems.
1. Instrumentation
2. Tissue Preparation
3. Matrix Application
4. Image Acquisition
5. Image Generation
6. Metabolite Identification
An experimental overview of MSI is shown in Figure 1. At the very beginning of the experiment, sample preparation is a critical step. Nodules are trimmed from the plant root and embedded in gelatin. The tissue must be pressed flat against the cryostat cup, with no bubbles, while it is being frozen; this will ensure easier and proper alignment of the tissue while it is being sectioned. When the tissue is being sliced, it is important to cut the tissue at the proper thickness; too thin of sections will tear, ruining the tissue integrity, while too thick of sections will reduce the number of analytes extracted and detected from the tissue. Selection of matrix and application technique will determine the types of analytes detected. Using a combination of matrices could provide complementary results. Three matrix application techniques are presented in this work. The airbrush technique is fast, but typically not suitable for MSI of small molecules because of analyte diffusion. Automatic sprayer systems and sublimation provide smaller matrix crystals, better reproducibility, and less analyte diffusion. Figure 2 shows an optical image of a Medicago truncatula root nodule section. Figure 3 shows an optical image example of the matrix coverage and crystal sizes using airbrush, automatic sprayer, and sublimation respectively.
Conventional matrices, like DHB, produce many ions in the lower mass range (m/z 100-400)20. These matrix ions can interfere with the detection of metabolites in this range. Figure 4 shows MS spectra of just DHB matrix compared to root nodule tissue coated with DHB matrix. DHB matrix peaks are indicated in red and root nodule tissue covered with DHB is shown in blue. Novel matrices such as TiO2 nanoparticles21, 1,5-diaminonapthalene (DAN)22, 2,3,4,5-Tetrakis(3′,4′-dihydroxylphenyl)thiophene (DHPT)23, and 1,8-bis(dimethyl-amino) naphthalene (DMAN)24,25 have been reported that reduce the interference of matrix ions in the low mass range, and also enhance the detection of certain classes of metabolites5. Matrix peaks can be distinguished from real metabolites using the MS images. When a peak is clicked on, the ion image is extracted and displayed overlapping the optical image. Those peaks that generate images with distinct localization to the tissue, and are not present in the matrix only area imaged, are considered metabolites. Figure 5a shows several representative ion images of metabolites found in root nodule tissue, while Figure 5b shows examples of MS images corresponding to matrix related peaks. In Figure 5a the image shows distinct localization to the root nodule tissue and a lack of signal in the matrix only area that was imaged. In Figure 5b the signal shows little localization and is present over the entire tissue; the signal is also seen in the matrix only area that was imaged (square areas in the top right corners). Sample lists of analytes of interest from M. truncatula root nodule tissue is listed in Table 1 (positive mode) and Table 2 (negative mode)4.
The end goal of untargeted metabolomics experiments is to identify the compounds that were detected. When performing MSI on a medium resolution instrument (MRMS), such as a TOF/TOF, it is necessary to obtain accurate mass measurements in a different way. This can be done with a multifaceted MS approach in which MALDI-MSI results are compared to high resolution LC-MS data using tissue extracts. There are many possible tissue extraction protocols to choose from depending on the analytes of interest. High resolution MS (HRMS) can be performed in the positive or negative ionization modes and with normal or reversed phase LC depending upon the analytes of interest. Once an accurate mass is obtained with high resolution LC-MS, the resulting mass can be searched with several databases, listed previously, to obtain a putative identification. Next MS/MS data is collected and the characteristic fragmentation pattern of the analyte of interest can be compared to standards, literature spectra, or theoretical fragmentation patterns. Figure 6 shows an example of one of the metabolites detected with MSI and LC-MS. This metabolite was identified as heme based on the MS/MS spectrum collected with high resolution LC-MS. This MS/MS data was compared to the MS/MS spectra previously published by Shimma and Setou26. The two MS/MS spectra match, therefore the identity of m/z 616.2 was confidently assigned as heme based on the accurate mass database searching and MS/MS data compared to literature MS/MS data.
Figure 1. Overview of MSI workflow. Root nodules are trimmed from the plant, embedded in gelatin, sectioned with the cryostat and mounted onto an ITO-coated glass slide. Matrix is applied to the slide using one of three matrix application techniques. MSI acquisition is performed with a MALDI-TOF/TOF mass spectrometer and MS spectra are compiled into images with MSI software.
Figure 2. Tissue sample optical image. Representative optical image of a Medicago truncatula root nodule tissue section.
Figure 3. Example matrix depositions. Representative images showing the different consistencies of matrix deposition with a) airbrush, b) automatic sprayer, and c) sublimation techniques. The airbrush application method generates large and small crystals while the automatic sprayer method produces small evenly sized crystals. Sublimation produces one even layer of matrix. Please click here to view a larger version of this figure.
Figure 4. MS spectrum of plain DHB matrix vs. root nodule tissue covered with DHB. DHB matrix peaks are indicated in red and root nodule tissue covered with DHB is shown in blue. Matrix peaks can be distinguished from real metabolites using the MS images. When a peak is clicked on, the ion image is extracted and displayed overlapping the optical image. Those peaks that generate images with distinct localization to the tissue, and are not present in the matrix only area imaged, are considered metabolites. Please click here to view a larger version of this figure.
Figure 5. MS images of M. truncatula root nodules. a) Representative ion images of metabolites found in wild-type M. truncatula root nodules. b) Representative ion images of matrix species that would not be considered metabolites. Please click here to view a larger version of this figure.
Figure 6. Example MS/MS data for metabolite identification. MS/MS spectrum of m/z 616.2, identified as heme [M+]. The chemical structure is shown and fragmentation structures are assigned. The experimental MS/MS data is compared to the MS/MS spectrum of heme reported in literature by Shimma and Setou26.
Table 1. Analytes of interest – positive mode. Sample list of analytes of interest from M. truncatula root nodule tissue in positive ion mode4.
Name of metabolite | Theoretical [M+H]+ | MRMS Measured [M+H]+ | HRMS Measured [M+H]+ | Δm (mDa) |
γ-aminobutyric acida | 104.0706 | 104.1 | 104.0706 | 0 |
cholinea,* | 104.1070 | 104.1 | 104.1071 | 0.01 |
proline | 116.0706 | 116.07 | 116.0706 | 0 |
valine | 118.0863 | 118.09 | 118.0865 | 0.02 |
leucinea | 132.1019 | 132.1 | 132.1022 | 0.03 |
asparaginea | 133.0608 | 133.06 | 133.0602 | -0.06 |
adeninea | 136.0618 | 136.07 | NA | NA |
proling betainea | 144.1019 | 144.1 | 144.1024 | 0.05 |
glutaminea | 147.0764 | 147.09 | 147.0768 | 0.04 |
histidinea | 156.0768 | 156.06 | 156.0771 | 0.03 |
argininea | 175.1190 | 175.13 | 175.1167 | -0.23 |
sucrose+K | 381.0794 | 381.05 | 381.0791 | -0.03 |
hemea,* | 616.1768 | 616.15 | NA | NA |
NADa | 664.1164 | 664.1 | NA | NA |
formononetina | 269.0808 | 269.08 | 269.0803 | 0.05 |
chrysoseriol GlcAa | 477.1028 | 477.1 | 477.1008 | 0.2 |
formononetin MalGlca | 517.1341 | 517.13 | 517.1337 | 0.04 |
aformosin MalGlca | 547.1446 | 547.16 | 547.1427 | 0.19 |
* [M+]
a MS/MS performed for identification.
Table 2. Analytes of interest – negative mode. Sample list of analytes of interest from M. truncatula root nodule tissue in negative ion mode4.
Name of metabolite | Theoretical [M-H]– | MRMS Measured [M-H]– | HRMS Measured [M-H]– | Δm (mDa) |
pyruvic acid | 87.0077 | 87 | NA | NA |
alaninea | 88.0393 | 88.01 | 88.0374 | -0.19 |
lactic acida | 89.0233 | 89.01 | 89.0296 | 0.63 |
phosphoric acid | 96.9685 | 96.96 | 96.9625 | -0.6 |
2-ketobutyric acid | 101.0233 | 101.02 | 101.0191 | -0.42 |
γ-aminobutyric acida | 102.055 | 102.04 | 102.0528 | -0.22 |
serine | 104.0342 | 104.02 | 104.0228 | -1.14 |
maleic/fumaric acida | 115.0026 | 115.03 | 115 | -0.26 |
succinic acida | 117.0182 | 117.01 | 117.0125 | -0.57 |
threonine | 118.0499 | 118.04 | 118.0456 | -0.43 |
oxalacetic acid | 130.9975 | 131.03 | 130.991 | -0.65 |
aspartic acida | 132.0291 | 132.03 | 132.0256 | -0.35 |
malic acida | 133.0132 | 133.02 | 133.0221 | 0.89 |
salicyclic acid | 137.0233 | 137.02 | 137.0349 | 1.16 |
α-ketoglutaric acid | 145.0132 | 145.02 | 145.0063 | -0.69 |
glutamic acida | 146.0448 | 146.01 | 146.0408 | -0.4 |
pentose | 149.0445 | 149.04 | 149.0414 | -0.31 |
aconitic acida | 173.0081 | 173.03 | 173.0057 | -0.24 |
ascorbic acida | 175.0237 | 175.05 | 175.0341 | 1.04 |
hexose | 179.055 | 179.05 | 179.0484 | -0.66 |
citric/isocitric acida | 191.0186 | 191.02 | 191.0166 | -0.2 |
palmitic acid | 255.2319 | 255.22 | 255.2246 | -0.73 |
hexose-6-phosphatea | 259.0213 | 259.04 | 259.014 | -0.73 |
stearic acid | 283.2632 | 283.26 | 283.2552 | -0.8 |
sucrosea | 341.1078 | 341.07 | 341.0972 | -1.06 |
a MS/MS performed for identification.
As discussed above, sample preparation is the most critical step in the MSI workflow. Embedding the tissue unevenly will cause sectioning to be difficult or not possible in some cases. The section size and adequate equilibration time are crucial to maintaining the tissue integrity and avoiding folding and tears. Selection of matrix and application technique will play a role in determining the types of analytes to be detected, the spatial resolution, and reproducibility of the results. Using a combination of matrices or application techniques could provide complementary results.
This method was designed specifically for untargeted MSI of endogenous metabolites in M. truncatula root nodule tissue, but can easily be adapted to other tissue types and biological questions. The recommended matrix application methods for small molecule MSI are sublimation and automatic sprayer. Increasing the amount of solvent deposited on the tissue, by adjusting the automatic sprayer method, will increase the extraction of analytes and allow for detection of higher mass compounds should one choose to perform MSI of lipids, peptides, etc. When using other types of tissue, the main adjustment will be the section thickness. Ideally the tissue should be sliced to the thickness of one cell; therefore thicker sections maybe appropriate for plant tissue and thinner sections appropriate for animal tissue. If folding or tearing occurs, typically a longer equilibration time is needed before sectioning or a thicker section size could be necessary. When performing LC-MS to obtain accurate masses, the tissue extraction protocol, mobile phase solvents and gradient, column stationary phase, and MS ionization mode can all be adjusted and optimized for the analytes of interest.
The main advantage of MALDI-MSI is its ability to provide not only mass information, but also spatial information for a given sample without the need for prior knowledge of the target analytes. Other imaging techniques require the use of derivatizations or tags5. As discussed previously, one limitation of this technique is the abundance of interfering matrix ion peaks. Novel matrices have been reported to address this limitation. Alternatively, secondary ion mass spectrometry (SIMS)-MSI is a matrix-free imaging option; however, it has less sensitivity than MALDI-MSI3. Another limitation of this technique is the lower mass resolution of MALDI-TOF/TOF instruments. Because of the low mass resolving power, it is necessary to also perform high resolution MS to obtain the accurate masses for metabolite identification, which means more experiments and more time. This problem could be solved by performing MALDI-MSI on a high resolution instrument platform such as the MALDI-Orbitrap. The final limitation of performing MSI experiments is the lack of software tools available for analysis of MSI data, although some recent advances in MSI software have been made27,28. Typically the MS spectrum for each imaging region must be manually analyzed and ion images extracted by hand. MSI experiments produce an abundance of data and can be incredibly time consuming to analyze. Overall, MALDI-MSI offers unique advantages for obtaining spatial information of many compounds simultaneously within a single experiment that can be extremely useful for the untargeted analysis of small molecules and other compounds with many biological applications.
The authors have nothing to disclose.
The authors would like to acknowledge Dr. Jean-Michel Ané in the Department of Agronomy at UW-Madison for providing Medicago truncatula samples. This work was supported in part by funding from the National Science Foundation (NSF) grant CHE-0957784, the University of Wisconsin Graduate School and the Wisconsin Alumni Research Foundation (WARF) and Romnes Faculty Research Fellowship program (to L.L.). E.G. acknowledges an NSF Graduate Research Fellowship (DGE-1256259).
Gelatin | Difco | 214340 | heat to dissolve |
Cryostat- HM 550 | Thermo Scientific | 956564A | |
indium tin oxide (ITO)-coated glass slides | Delta Technologies | CB-90IN-S107 | 25-75-0.8 mm (width-length-thickness) |
2,5-Dihydroxybenzoic acid (DHB) matrix | ICN Biomedicals | PI90033 | |
Airbrush | Paasche Airbrush Company | TG-100D | coupled with 75 ml steel container |
Automatic matrix sprayer system- TM-Sprayer | HTX Technologies, LLC | HTX.TMSP.H021-U | Specific start-up and shut-down instructions will be given when the instrument is installed |
Sublimation apparatus | Chemglass Life Science | CG-3038-01 | |
Vaccum pump- Alcatel 2008 A | Ideal Vacuum Products | P10976 | Ultimate Pressure = 1×10-4 Torr |
ultrafleXtreme MALDI-TOF/TOF | Bruker Daltonics | 276601 | |
FlexImaging | Bruker Daltonics | 269841 | One example of "vender specific software" |
MALDI LTQ Orbitrap | Thermo Scientific | IQLAAEGAAPFADBMASZ | High resolution MALDI instrument for accurate mass measurements |
Q Exactive | Thermo Scientific | IQLAAEGAAPFALGMAZR | High resolution LC-MS instrument for accurate mass measurements |