Challenges and recent advances in mass spectrometric imaging of neurotransmitters

Abstract

Mass spectrometric imaging (MSI) is a powerful tool that grants the ability to investigate a broad mass range of molecules, from small molecules to large proteins, by creating detailed distribution maps of selected compounds. To date, MSI has demonstrated its versatility in the study of neurotransmitters and neuropeptides of different classes toward investigation of neurobiological functions and diseases. These studies have provided significant insight in neurobiology over the years and current technical advances are facilitating further improvements in this field. neurotransmitters, focusing specifically on the challenges and recent Herein, we advances of MSI of neurotransmitters.

Background

Mass spectrometric imaging (MSI) shows great promise for biological analyses because it allows for molecular analysis of tissue while retaining information about the spatial distribution of different analytes, including proteins, peptides, lipids and small molecules [1]. During an MSI experiment, a multitude of mass spectra are collected from a tissue slice in a predefined raster, resulting in a 2D distribution map for each mass measured. One of the advantages of MSI is that it allows for the analysis of thousands of analytes at once, without the need of labels or prior knowledge of the analytes, and provides spatial information along with the mass analysis. MSI is often used in tandem with other techniques to obtain more molecular information, while using MSI to visualize the results [2–7]. Many excellent reviews have already been published on the subject of MSI [8–17]. Herein, we review current publications that highlight the critical role that MSI plays in the study of neurotransmitters and neuropeptides, focusing on the challenges and recent advances in the field.

Ionization techniques: advantages & disadvantages

There are three main ionization methods used for MSI: MALDI [18,19], secondary ion MS (SIMS) [20–22] and DESI [23,24].

MALDI-MSI has proven to be a valuable technology with numerous applications for analyzing proteins [25,26], lipids [27], neuropeptides [28,29] and small molecules [2] at both organ and cellular levels. One advantage of using MALDI for MSI on biological samples is that it allows for the generation of larger ions, such as peptides and proteins, which is one of the reasons why MALDI is the most widely used method for MSI [1,30–32]. Sample preparation is extremely important for this ionization technique. Immediately after the sacrifice of an animal specimen, rapid molecular degradation occurs. To limit the degradation of analytes, the tissue can be embedded in supporting media such as gelatin [28,29] or sucrose [33], and snap-frozen in dry ice or liquid nitrogen. Polymer-containing material, such as optimal cutting temperature compound, Tissue-Tek^® and carboxymethylcellulose, should be avoided as they often introduce polymer interferences into the mass spectrometer [34]. The embedded and frozen tissue should be stored at −80°C until use [35]. Prior to long-term storage, the tissue sample can be stabilized using various methods, such as microwave irradiation [36] or heat denaturation by Denator Stabilizor^® T1 (Gothenburg, Sweden), to deactivate proteolytic enzymes, preventing postmortem degradation of proteins or peptides of interest [37–39].

Next, the frozen tissue can be cut into thin sections with a cryostat. MSI experiments typically require 10–20 μm thick tissue sections [40]. The previous step of embedding tissue in supporting media allows for precise sectioning of tissue samples. Once sliced, tissue sections are then transferred and mounted onto a target plate or glass slide [41,42] by thaw-mounting [34]. When working with higher mass analytes, such as peptides and proteins, washing tissue sections with organic solvents is a recommended step to fix tissues, and remove ion-suppressing salts and lipids, thus increasing the analyte signal [43–45].

The next step of the MALDI-MSI sample-preparation process involves coating the tissue with a thin layer of matrix and irradiating the sample with a laser beam. The matrix is designed to absorb much of the incident laser, providing ‘soft’ ionization for analyte compounds, which allows for the ionization of larger molecules (m/z over 100 kDa) [46]. Matrix selection and application is a critical step for MALDI-MSI, greatly impacting the sensitivity, spatial resolution and selective analyte ionization of the experiment [43]. Conventional matrices include α-cyano-4-hydroxycinnamic acid (CHCA) and 2,5-dihydroxy benzoic acid (DHB) [28,29,34]. One disadvantage of using these conventional matrices is that they produce ions themselves, which can interfere or mask analyte ions when looking at small molecules. High-resolution instrumentation may negate some of the interference; however, less conventional matrices, such as ionic matrices, made by mixing conventional matrices with organic bases [47–49], TiO₂ nanoparticles [50], 1,5-diaminonapthalene [51], 2,3,4,5-Tetrakis(3′,4′-dihydroxylphenyl)thiophene [52], 1,8-bis(dimethyl-amino) naphthalene [53,54] and ferulic acid [55] are being used more frequently, and are reported to improve spectral quality, crystallization and vacuum stability. The matrix application technique plays a crucial role in the quality of mass-spectral images, especially when obtaining high spatial resolution images [10,51,56].

Spatial resolution and reproducibility of results are limited by the matrix crystal size and application consistency, among other instrumental parameters such as raster step size and laser beam diameter [10]. Matrix application methods can be divided into two categories: solvent-based and solvent-free. Solvent-based methods apply matrix dissolved in solution and are typically performed by manual airbrush or automated systems [56]. For solvent-free methods, dry matrix can be sublimated onto tissue under low pressure [57], filtered onto tissue through a small (~20 μm) sieve [58] or ground through fine mesh (~10–1 μm) above a tissue slice using a ball-mill [59]. These solvent-free methods were developed to limit spatial delocalization of soluble analytes that often occurs from excessive amounts of solvent during matrix application from solvent-based methods [60]. In comparison with the solvent-based matrix application methods, solvent-free methods yield very fine crystals amenable to high spatial-resolution MSI, but may also suffer from relatively low sensitivities due to the limited analyte–matrix interactions [51]. Several researchers have modified these dry-matrix application techniques by rehydrating the sections following dry-coating to improve the detection sensitivity of analytes [25,61].

SIMS is a long-established technique in which a sample is put under high vacuum and bombarded with high energy primary ions, which facilitates the ionization of analytes [62,63]. The ionized analytes are referred to as secondary ions and are sputtered from the sample surface and then drawn into the mass analyzer for analysis. The highly focused ion beam in SIMS provides excellent spatial resolution; however, the primary ion source is fairly limited to small molecules, under 1000 Da [2], due to its high energy that is prone to fragment large molecules in the ablation/ionization process. Recent advances in SIMS, such as using beams of gold trimers (Au₃⁺) [64] or buckminsterfullerenes (C₆₀) [65], can provide significantly greater yields of higher mass ions. SIMS imaging does not require special preparation after sectioning and mounting the tissue, but there are some optional methods that improve the imaging results. The tissue can be coated with a thin layer of gold, silver or matrix (e.g., DHB, α-CHCA or sinapinic acid) to improve the ionization and reduce the fragmentation of larger molecules [66–68].

In contrast to traditional MALDI and SIMS, which require a vacuum (although atmospheric pressure-MALDI has been developed), DESI is a simple, ambient ionization technique [69–72] that channels charged solvent droplets and ions from an electrospray source onto the sample surface [73]. The surface is impacted by the charged particles, yielding gaseous analyte ions. MSI had been mainly viewed as an invasive process until the development of ambient ionization techniques, such as DESI. DESI’s ambient nature and softness allow for the examination of various natural surfaces with no need for matrices [74]. An advantage of DESI, is its capability for high-throughput analysis with minimal tissue adulteration, again due to its ambient nature [13]. Since a vacuum is not required for DESI, it has a greater advantage of being used for clinical or field application, as the removal of the vacuum makes it more portable [75]. The ability to perform in situ analysis and the convenience of portable mass spectrometers suggests the potential role of DESI-MSI and other ambient ionization techniques lie in histology guided diagnoses and therapies in the clinical setting [75]. A disadvantage of DESI when compared with vacuum MS methods, SIMS and MALDI, is that it has lower spatial resolution of approximately 180–200 μm [74]. Recently, the nano-DESI probe was developed and reported an increase in the spatial resolution of approximately 12 μm. With the nano-DESI probe, controlled desorption of analytes present in a restricted region of specimen is achieved by using a minute amount of solvent between two capillaries [71]. Table 1 summarizes the optimal analytes, mass range, and spatial resolutions of MALDI, SIMS and DESI.

Table 1.

Comparison of MS imaging approaches detailing ionization methods, mass analyzers, optimal analytes, mass range and spatial resolution.

Type of MSI	Ionization source	Common mass analyzer	Optimal analytes	Mass range (Da)	Spatial resolution (μm)
MALDI	UV/IR laser Soft ionization	TOF	Lipids, peptides, proteins, small molecules	0–50,000	30–50
SIMS	Ion gun Hard ionization	TOF Magnetic sector Orbitrap	Lipids, small peptides, small molecules	0–2000	0.5–1
DESI	Solvent spray Soft ionization	Orbitrap	Lipids, peptides, small molecules	0–2000	100

MSI: MS imaging; SIMS: Secondary ion MS.

Reprinted with permission from [16] © Elsevier (2013).

A fourth ionization technique worth mentioning is nanostructure-initiator MS (NIMS) [76]. NIMS is a matrix-free ionization technique that shows promise for the analysis of small molecules. NIMS uses a liquid initiator to facilitate desorption instead of a crystalline MALDI-matrix. NIMS initiators are different from MALDI matrices in that the initiators do not absorb UV energy and most do not ionize, thus, do not produce interfering background ions. The advantages of the NIMS initiator are that the NIMS surface is stable in ambient air, has improved reproducibility, enables direct biofluid analysis and tissue imaging, and allows for a significantly expanded mass range [77]. In NIMS desorption/ionization, the analyte is adsorbed on top of the NIMS surface and laser irradiation results in rapid surface heating, causing the NIMS initiator to vaporize, triggering the desorption of adsorbed materials [76]. Several other MSI techniques have been developed more recently, including IR matrix-assisted laser DESI [78–80], liquid extraction surface analysis [81] and laser ablation ESI [82–86], but are beyond the scope of this review.

Neurotransmitters

Neurotransmitters are endogenous substances released by synaptic terminals upon depolarization, which transmit signals between nerve cells [87,88]. More than 100 substances have been identified as neurotransmitters since 1914, when the first neurotransmitter, acetylcholine, was discovered [89]. In the nervous system, neurotransmitters are the most common chemical messengers [90]. Neurotransmitters can be divided into six groups: acetylcholine, purines, amino acids, biogenic amines, peptides and unconventional neurotransmitters. Except for peptides, all other neurotransmitters are considered to be small-molecule neurotransmitters due to their lower molecular weights (<1000 Da). Most small-molecule neurotransmitters modulate activities requiring fast responses. Therefore, they are rapidly inactivated by highly specific transporter systems or enzymes in order to prevent continuous activation of the postsynaptic nerve cells [90]. Peptides and some small neurotransmitters tend to modulate slower and ongoing actions, and these neurotransmitters have slower turnover rates [87]. Due to differences in modes of action, neuropeptides are sometimes considered to be neuromodulators or neurohormones. Neuropeptides with neurotransmitter functions are defined as putative neurotransmitters or co-transmitters [90].

Neurotransmitter MSI

Recently, many studies have successfully mapped the distributions of neurotransmitters using MSI by developing new sample preparation techniques, matrix application methods, ionization methods and mass analyzers, which are summarized in Table 2. MSI studies of different classes of neurotransmitters are reviewed in the following sections.

Table 2.

Summary of neurotransmitter groups, sample preparation, suitable matrices, MSI acquisition methods and related studies.

Neurotransmitter groups	Examples	Sample preparation	Matrices	MSI methods	Related studies
Small molecule neurotransmitters
Acetylcholine	Acetylcholine	In situ freezing	Deuterated matrix (D4-CHCA)	MALDI tandem mass MSI; high-resolution and mass-accuracy MSI	[92–94]
Purines	ATP, AMP, adenosine	In situ freezing	9-AA	MALDI, high-resolution and mass-accuracy MSI	[7,93,95,96]
Amino acids	Glutamate, γ-aminobutyric acid, glycine	NA	DHB	MALDI, high-resolution and mass-accuracy MSI	[93,103–105]
Biogenic amines	Dopamine, epinephrine, serotonin	DPP derivatization	CHCA	DESI MALDI	[105,109]
Peptide neurotransmitters
Neuropeptides	Substance P, somatostatin	NA	CHCA, DHB and so on	MALDI, SIMS, NIMS	[26,114,128,129,132]

9-AA: 9-aminoacridine; CHCA: α-cyano-4-hydroxycinnamic acid; D4-CHCA: Deuterated α-cyano-4-hydroxycinnamic acid; DHB: 2,5-dihydroxy benzoic acid; DPP: 2,4-diphenyl pyrylium; MSI: MS imaging; NIMS: Nanostructure-initiator MS; SIMS: Secondary ion MS.

MSI of small-molecule neurotransmitters

MSI of small-molecule neurotransmitters is much more challenging than MSI of other bio-molecules due to the following factors: conventional matrix compounds generate interference in the mass range of small molecules; small molecule neurotransmitters have rapid turnover rates, which introduces difficulties for MSI sample preparations; and small-molecule neurotransmitters have low in vivo concentrations, which requires a sensitive detection method. Despite the difficulties, several MSI studies have analyzed the distribution of small molecule neurotransmitters.

MSI of acetylcholine

Acetylcholine acts as an excitatory transmitter to modulate motor systems in the CNS and PNS, except in the heart, where it acts as an inhibitory transmitter [91]. Both DHB and CHCA, two common MALDI matrices used in small molecule profiling and imaging, produce strong interference for acetylcholine. Therefore, innovative MALDI matrices, high-mass accuracy and high-resolution mass analyzers, or alternative data acquisition methods are needed to solve the matrix interference problem. Additionally, due to the rapid turnover rate of acetylcholine, a well-developed sample preparation procedure is needed to minimize postmortem degradation. Several recent studies have successfully detected endogenous acetylcholine in brain tissues [92–94].

Shariatgorji et al. developed a deuterated matrix, D⁴-α-CHCA (D⁴-CHCA), to solve the signal interference problem associated with small molecule MSI [94]. D⁴-CHCA has mass shifts of +4, +8 and +12 Da. As shown in Figure 1, by switching between D⁴-CHCA and CHCA, matrix peaks can be differentiated from acetylcholine as well as other small molecules. The distribution of acetylcholine was also confirmed by MS/MS with a characteristic fragmentation peak at m/z 87.0 [94]. Sugiura et al. developed a MSI method combining both tandem mass imaging and in situ freezing techniques to visualize acetylcholine distribution in mouse brain [92]. In this study, acetylcholine transition signatures m/z 146→87 and 146→92 were monitored to avoid matrix interference at the parent ion mass. The observed acetylcholine signal distribution at m/z 87 overlapped with the acetylcholinesterase in situ hybridization, proving the validity of this method. Furthermore, using in situ freezing techniques, which freeze anesthetized mouse brain in liquid nitrogen during dissection, postmortem changes were minimized, and dynamic range and sensitivity were improved [92].

Figure 1. D⁴-α-cyano-4-hydroxycinnamic acid has been synthesized for use as a matrix for MALDI-MS and MALDI-MS imaging of small-molecule drugs and endogenous compounds.

MALDI-MS analysis of small molecules has historically been hindered by interference from matrix ion clusters and fragment peaks that mask signals of low molecular weight compounds of interest. By using D⁴-α-cyano-4-hydroxycinnamic acid, the cluster and fragment peaks of α-cyano-4-hydroxycinnamic acid, the most common matrix for analysis of small molecules, are shifted by +4, +8 and +12 Da, which expose signals across areas of the previously concealed low mass range. Here, obscured MALDI-MS signals of a synthetic small molecule pharmaceutical, a naturally occurring isoquinoline alkaloid, and endogenous compounds including the neurotransmitter acetylcholine have been unmasked and imaged directly from biological tissue sections. Reprinted with permission from [94] © American Chemical Society (2012).

The problem of matrix peaks interfering with the acetylcholine signal in MSI can also be solved by using a high-resolution and high-accuracy mass analyzer. Ye et al. successfully mapped acetylcholine in rat brain tissue with standard DHB matrix with an LTQ Orbitrap^™ mass analyzer (Thermo Fisher Scientific, MA, USA) [93]. The high resolving power of the LTQ Orbitrap made it possible to separate the acetylcholine peaks from the matrix peaks. Ye et al. have also mapped several other amino acid and biogenic amine neurotransmitters, such as histidine and AMP in rat and crab brain tissue using this method [93]. Figure 2 summarizes the metabolites and neurotransmitters detected in rat brain tissue using MSI [93]. The successful detection of acetylcholine by MSI is very useful for biologists and neuroscientists to rapidly monitor the distribution and relative quantitation changes during various physiological states.

Figure 2. MS images of metabolites and neurotransmitters identified from rat CNS.

(A) Optical image of a rat brain section subjected to MS imaging sample preparation. MS images of metabolites and neurotransmitters, including: (B) choline at m/z 104.1070; (C) acetylcholine at m/z 146.1175; (D) AMP at m/z 348.0706; (E) GMP at m/z 364.0655; (F) cholesterol with neutral loss at m/z 369.3519; (G) potassiated phosphatidylcholine (32:0) at m/z 772.5258; and (H) nicotinamide adenine dinucleotide at m/z 644.1177. Other than low MW molecules, (I) MS image of the acetylated peptide ASQKRPSQRHGSKYLATA at m/z 2028.076 was also shown. (J) Overlaid image of a neurotransmitter, acetylcholine, as in (C) and cholesterol-derived ion as in (F). Adapted with permission from [93] © American Chemical Society (2013).

MSI of purines

Purine neurotransmitters, including ATP, AMP, adenosine and other purines are small signaling molecules that contain purine rings. Most purines have similar molecular structures, similar molecular weights of 300–1000 Da and similar in situ concentrations on the micromolar order of magnitude [95]. ATP was first thought to be a co-transmitter with an energy supply, but was later found to have signal transmitting capability. It acts on spinal cord motor neurons and sensory and autonomic ganglia as an excitatory neurotransmitter. Adenosine, which is derived from ATP by enzymatic activities, is also widespread in the CNS as an excitatory neurotransmitter [87].

Similar to acetylcholine, MSI of purine neurotransmitters also faces the challenges of rapid turnover rates, low concentrations and matrix interference. Nevertheless, several MSI studies have successfully mapped the distributions of purines, mainly ATP, ADP and AMP, in brain tissues [7,93,95,96].

Benabdellah et al. first reported the detection and identification of 13 primary metabolites, including ADP, AMP, GTP and UDP, in rat brain sections using MALDI-TOF/TOF-MSI [95]. Nine common MALDI matrices were tested on tissue surfaces in either positive or negative ion mode. Only 9-aminoacridine did not generate interfering matrix peaks on the target mass range (m/z 300→1000) under negative ion mode. A robotic sprayer was also used for matrix application in order to generate a homogenous and thin layer of matrix, thus improving reproducibility. However, the distribution of ATP was not detected in the tissue due to low abundance of ATP [95]. Hattori et al. developed a quantitative MSI technique for purines by combining in situ freezing, MALDI-MS and quantitative CE-ESI-MS on ischemic penumbra mouse brain samples [96]. In this study MSI results of samples prepared by in situ freezing and conventional postmortem freezing were compared. Both samples were coated with 9-aminoacridine and analyzed in negative ion mode. ATP could only be detected on in situ frozen samples as postmortem freezing reduced ATP autolysis. Insitu metabolite concentrations were quantified in this study by parallel CE-ESI-MS experiments [96]. Sugiura et al. also utilized a similar quantitative MSI technique by combining MALDI-MSI with quantitative CE-ESI-MS for purines for monitoring spatiotemporal energy dynamics of hippocampal neurons in a kainite-induced seizure mouse model during the seizure process [7]. They reported the ATP depletion and recovery process, along with the changes in other metabolites in the tricarboxylic acid cycle during the seizure, using quantitative MSI. The successful detection and quantitation of purine neurotransmitters, especially ATP, not only allows researchers to monitor the changes of neurotransmitters, but also allows the observation of the energy consumption at different histological regions under different conditions.

MSI of amino acid neurotransmitters

Amino acid neurotransmitters are individual amino acids that act as signaling molecules between neurons, including glutamate, γ-aminobutyric acid (GABA), glycine, D-serine and D-aspartate. Glutamate is the most common neurotransmitter in the CNS with an excitatory postsynaptic effect [97]. GABA is a common inhibitory neurotransmitter in the nervous system [98], although a recent study reported the existence of GABA excitatory synapses in the cerebellar interneuron network [99]. Glycine is also a common, but more localized, inhibitory neurotransmitter in the CNS, especially in the spinal cord. These amino acid neurotransmitters can be removed from the synaptic cleft via reuptake under the actions of various transporters [87]. For instance, GABA can be removed by the highly specific GABA transporters [90].

MSI of amino acid neurotransmitters also faces similar challenges to other small-molecule neurotransmitters. Even though several studies have analyzed amino acid neurotransmitters in various nervous systems using GC [100], LC [101] or microfluidic devices [102] coupled with ESI-type mass spectrometers, few MSI studies have successfully visualized and quantified the distribution of amino acid neurotransmitters [93,103–105].

Goto-Inoue et al. detected GABA (m/z 104.07) in eggplant by MALDI-MSI [103]. DHB was chosen as matrix and produced no interference at the target mass. No specific sample-handling procedures or detection methods were used due to the higher concentration of GABA in eggplant when compared with the nervous system. Toue et al. visualized several amino acids in human colon cancer xenograft tissues using quantitative MSI, combining MALDI-MS with CE-ESI-MS [104]. Due to the low concentration of endogenous amino acids, p-N,N,N-trimethylammonioanilyl N′-hydroxysuccinimidyl carbamate iodide was applied on the surface of the tissue as a derivatization reagent to enhance ionization efficiency. Tandem MSI was used to solve the matrix interference problem. The rest of the tissues were homogenized for quantitative CE-ESI-MS analysis. Several amino acids, including glutamate and glycine, were detected on the tissue surface by MALDI-MS and their concentrations were reported by CE-ESI-MS. Shariatgorji et al. has successfully quantified GABA brain concentrations by MALDI-MSI using their in-house imaging software [105]. In this study, a standard curve was established by measuring GABA standard intensities with various concentrations on tissue surface; the GABA concentration on brain tissue could then be calculated from the standard curve. This method was also successfully applied to many other drugs and neurotransmitters, such as acetylcholine and ATP.

MSI of biogenic amine neurotransmitters

Biogenic amine neurotransmitters are small signaling molecules derived from amino acids [98]. Three catecholamine neurotransmitters: dopamine, epinephrine and norepinephrine, are derived from tyrosine; histamine is derived from histidine, and serotonin and 5-hydroxytryptamine are derived from tryptophan [87]. Dopamine plays an important role in modulation of arterial blood-flow, cognition, motor control and anxiety-related behavior [106]. Parkinson’s disease is caused by a lack of dopamine in the brain [107]. Norepinephrine regulates sleep, attention, and feeding in both the CNS and the PNS, while epinephrine is mostly restricted to the PNS [108]. Epinephrine and norepinephrine also act as hormones, regulating heart rate, respiratory rate, muscle contraction and other actions. Histamine regulates attention and arousal, and is also related to allergic reactions. Serotonin works on CNS neurons that control circadian rhythm, emotions and motor behaviors [87].

Similar to MSI of amino acid neurotransmitters, few studies have successfully performed MSI of biogenic amine neurotransmitters due to the previously mentioned challenges. Wu et al. detected epinephrine and norepinephrine in the adrenal gland of adult porcine by DESI-MSI, which is a matrix free method that avoids the matrix interference problem [109]. Recently, Shariatgorji et al. reported that dopamine was visualized and quantified by MALDI-MSI [105]. 2,4-diphenyl pyrylium was used to derivatize dopamine, which could then be detected by high sensitivity MALDI-MS using CHCA as the matrix. The derivatized dopamine concentration on tissue could be determined by the in-house imaging software mentioned above.

MSI of neuropeptides

Peptide neurotransmitters, or neuropeptides, such as substance P, endorphin and opioid peptides, are larger signaling molecules that are typically composed of three to 70 amino acids. They are derived from polypeptides, prepropeptides and propeptides [87]. Substance P, the first discovered neuropeptide [58], is an excitatory neurotransmitter in pain perception. In contrast, endorphin is an inhibitory neurotransmitter in pain reduction and euphoria production [98]. Other neuropeptides are reported to be involved in anxiety, autism and panic attacks [110–113]. Since the development of MSI for biological systems, MSI analysis of neuropeptides has been an intensive research topic and widely applied to various biological systems. MSI has been used to map the distribution of endogenous neuropeptides within the nervous systems of various organisms, including the medicinal leech [114], terrestrial and fresh water snails [115,116], the sea slug Aplysia californica [117,118], numerous species of crustaceans [29,119–121] and rodents [122]. These MSI studies provide valuable information on the neuropeptide distributions and relative quantitation within various nervous systems, thus providing a new approach for neuroscience research in monitoring neuropeptidomic changes under different physiological conditions. Various ionization techniques have been used in neuropeptide MSI, such as MALDI, SIMS [122] and NIMS [123]; several advanced techniques have been developed for MSI, such as 3D MSI [124,125] and quantitative MSI [126]. Additionally, MSI has been used for biomarker discovery studies in several disease models, including Parkinson’s disease [127], Alzheimer’s disease [128] and cortical spreading depression [125]. Numerous excellent reviews and book chapters have summarized the MSI workflow, techniques and applications to neuropeptides [47,129–132]. Ye et al. used a high-mass-resolution and high-mass-accuracy MALDI-Fourier transform ion cyclotron resonance instrument to map neuropeptide distribution in the crustacean stomatogastric nervous system at cellular resolution [121]. Hanrieder et al. used MSI to study L-DOPA-induced dyskinesia in rat brain [127]. They noted elevated levels of neuropeptides, including substance P and dynorphin B, in the dorsolateral striatum of high-dyskinesia rats compared with control rats.

Challenges of neurotransmitter MSI

Matrix interference

Choosing an appropriate matrix and its application method is one of the main challenges in MALDI-MSI of small molecules, such as neurotransmitters, and there have been extensive method developments in this area. When the sample is ablated conventional matrices produce matrix ions that potentially mask analyte ions in the low mass region (<500 Da). Some matrix application methods, such as manual airbrush, produce matrix crystals of varying sizes, which leads to nonreproducible results. The use of solvent-based matrix application methods can also cause analyte diffusion when working with small molecules. As mentioned previously, Shariatgorji et al. developed a different strategy for reducing matrix interference by using a deuterated matrix [94]. The research group employed the traditional matrix, CHCA, but also synthesized a deuterated version, D⁴-CHCA, giving a 4, 8, or 12 Da mass shift of interfering matrix peaks between the two matrices. When these two matrices are used in conjunction, the mass shift characteristic of matrix interference allows analyte peaks in the lower mass range to be un-masked. In this work, the neurotransmitter, acetylcholine, was detected when employing D⁴-CHCA, but was masked by matrix cluster peaks when employing traditional CHCA [94]; thus, researchers were able to eliminate matrix interferences while still keeping the advantages of CHCA ionization efficiency. Other strategies for eliminating matrix interference have been developed for small-molecule MSI and could be applied to MSI of neurotransmitters in the future. For targeted analysis, Porta et al. demonstrated the use of MALDI-SRM/MS [133]. Using this technique, the analyte was fragmented and several of the unique fragment ions were monitored; thus, the matrix ions were no longer being detected and therefore did not interfere with the analyte signal. Recently, Tang et al. used solvent-free argon ion sputtering to apply gold nanoparticles to mouse brain tissue and were able to directly detect neurotransmitters and other small molecules via MALDI-MSI [134]. Argon ion sputtering is a solvent-free matrix application method that has been reported to produce a thin, even layer of matrix coating onto the sample surface [134]. The chemical interference from the gold nanoparticles was very minimal, thus alleviating the common problem of matrix ions masking analyte ions in the lower molecular weight region. Using this technique, the researchers were able to differentiate five unique regions of mouse tumor tissue based on the distributions of characteristic ions known to be metabolic markers of tumors. Water ice has been used as a matrix in IR-MALDI-MS, but would be difficult to use on MS systems requiring high vacuum pressure, such as TOF instruments, which are most commonly used for MALDI-MSI. Pirkl et al. employed a cooling stage to generate ions at high vacuum conditions using water ice as a matrix [135]. Ice is an ideal matrix for water-rich biological tissue because it produces matrix-ion-free spectra while still facilitating analyte ionization. Thus far, ice has been demonstrated as a suitable matrix for the detection of peptides, such as substance P [135], from tissue sections with IR-MALDI-MS profiling. MSI has yet to be performed on tissue samples utilizing ice as the matrix. Another matrix-free imaging strategy was developed by Qian et al. [136]. In this research, a graphene paper was engineered using a pulse-laser process to form densely packed nanospheres. This graphene paper was used as a substrate for matrix-free laser desorption/ionization-MSI. This method is reported to enhance the detection of small molecules because the method used for engineering this material creates a highly homogenous surface, eliminating the variability that is often seen when matrix is applied to the sample, producing more reliable and reproducible results.

Reproducibility in sample preparation

Aside from matrix selection and application, there are several other challenges of MALDI-MSI sample preparation that are being improved upon for enhanced MSI of small molecules (neurotransmitters) and neuropeptides. Consistency and reproducibility of techniques during sample preparation can be a challenge. Chang et al. developed a technique to obtain more reproducible sample preparations for SIMS-MSI of small molecules [137]. In this method, powerful magnets were used to fracture cells of interest. Typically cells are fractured via cryomicrotoming and freeze–fracture techniques that damage the cells and complicate SIMS-MSI results. By using powerful magnets to fracture the cells, the surface environment can be more controlled, which could provide more reproducible and reliable sample preparations. A comparison between the standard freeze–fracture technique and the powerful magnet method showing the differences in sample surface uniformity is shown in Figure 3. Another challenge in MSI of neurotransmitters is the low signal intensity caused by low abundance of analytes and/or signal suppression from other endogenous compounds in the tissue. Typically sample washing causes delocalization of small molecules and is therefore not part of the MSI workflow for neurotransmitters and small molecules. Shariatgorji et al. demonstrated a washing protocol with a pH-controlled buffer system for the enhanced detection of small molecules [138]. This approach can only be used for targeted analyses because the protocol employs an aqueous, buffered wash solution with the pH adjusted to the point where the target analyte is insoluble. This washing step removes soluble, endogenous salts and other compounds that suppress signal, thus increasing the detection of small molecules. Goodwin et al. developed a method for stabilizing delicate tissue sections for MALDI-MSI [139]. Double-sided conductive carbon tape was used to collect, stabilize and mount delicate, heat-treated rat brain tissue. Neuropeptides are rapidly modified postmortem; heat stabilization, as mentioned previously, can halt the postmortem degradation process, but can also result in more fragile tissues. Sectioning thin or delicate tissues using a cryostat can be challenging and may often result in tearing or folding of the tissue. The carbon tape did not interfere with the detection of small molecules or neuropeptides and could be used to reduce the variability in sample quality when attempting to section delicate tissues. The researchers used this method to detect and identify many endogenous compounds and show their unique distributions in a variety of rat tissues.

Figure 3. TOF-secondary ion MS.

(A–D) Using the freeze–fracture method; (E–H) with the powerful magnet method. (A) and (E) are bright fields; (B) and (F) are Ca, m/z 40; (C) and (G) are Cd, m/z 48; (D) and (H) are total ions.

Reprinted with permission from [137] © John Wiley & Sons, Ltd (2012).

Quantitative MSI

The intensity of an MS signal is related to the analyte’s ionization efficiency as well as its concentration. With MALDI-MSI the smallest difference in matrix inhomogeneity can cause signal suppression and inaccurate quantitation. One technique that is being used for quantitative imaging is multi-isotope imaging MS (MIMS), which has been reviewed previously [140]. Steinhauser et al. demonstrate this technique for the quantification of stem cell division and metabolism on the sub-micrometer scale [141]. MIMS is based on SIMS technology. The instrument can simultaneously measure data from multiple isotopes in the same region of the tissue. MIMS uses isotopic tags to localize and measure their incorporation into intracellular compartments and quantitative data is extracted based on the isotope ratios [142]. The research team used MIMS to test the ‘immortal strand hypothesis’, which predicts that during stem cell division, chromosomes containing older template DNA are segregated to the daughter cell destined to remain a stem cell. When the isotopic tag is introduced, if the hypothesis held true, there would be one completely unlabeled daughter cell after the first cell division. The results showed no unlabeled daughter cells and, thus, do not support the ‘immortal strand hypothesis’. Reviews specifically focused on quantitative MSI have recently been published [13,15,140]. Another, more typical, approach to quantitative MSI employs an IS for quantification, which involves generating a standard curve by measuring standards spotted on the tissue surface and calculating the concentration of the analyte in the tissue based on this standard curve [105]. Pirman et al. quantified cocaine from brain tissue using MALDI-MSI by employing deuterated cocaine as the IS [143]. In order to limit the background interference during the experiment, MS/MS MSI was performed with a wide isolation window to incorporate the analyte and the deuterated IS. A calibration curve was generated using control tissue and used to determine concentrations of the analyte. The quantitative MSI data were compared with quantitative LC–MS/MS and the results were comparable. Furthermore, several semi-quantitative studies have been successfully conducted [127,144–151]. Recently, new software tools have been developed to aid in quantitative MSI and will be discussed below.

Low mass resolution & spatial resolution of MSI instruments

The main, on-going challenges for MSI instrumentation are obtaining faster acquisition speeds, and higher spatial and chemical resolution. Spraggins and Caprioli report high-speed MALDI-MSI using a prototype TOF mass spectrometer [152]. This work incorporates a neodymium-doped yttrium lithium fluoride solid state laser capable of repetition rates up to 5 kHz, which reportedly produce more than a tenfold increase in sample throughput when compared with commercially available instruments. This technique was applied to MSI of lipids, but should translate to MSI of other chemical species, including other small molecules or peptides. As mentioned previously, Ye et al. focused on increased mass resolution of MSI experiments [93]. HRMS and high-accuracy MS is able to distinguish analytes from interfering matrix ions. In this study, Ye et al. performed MALDI-MSI of neurotransmitters in rodent and crustacean nervous systems and were able to resolve peaks that were merged in lower resolution TOF/TOF instruments. Schober et al. tackled the problem of high-spatial resolution imaging, while still maintaining high-mass-accuracy and high-mass-resolution [153]. In this work, HeLa cells were imaged with optical fluorescence imaging and with a high-resolution atmospheric-pressure-MALDI source attached to an Exactive^™ Orbitrap mass spectrometer. A spatial resolution of 7 μm was achieved in this work as shown in Figure 4.

Figure 4. HeLa cells on indium tin oxide slide.

(A) Optical fluorescence of DIOC₆(3) stained HeLa cells, λ = 501 nm. (B–D) MALDI imaging (28 × 21 pixels, pixel size 7 μm). (B) Selected ion image of staining agent DIOC₆(3) ([M⁺]). (C) Selected ion image of PC(32:1) ([M + Na]⁺). (D) Selected ion image of PC(34:1) ([M + Na]⁺).

Reprinted with permission from [153] © American Chemical Society (2012).

Data processing

MSI data analysis software, such as BioMap [201] and proprietary programs for MSI systems (e.g., FlexImaging^™ from Bruker Daltonics [MA, USA], ImageQuest^™ from Thermo Fisher Scientific and TissueView^™ from Applied Biosystems/MDS SCIEX [CA, USA]), are primarily used to produce distribution maps for selected analytes. These software packages have different features but mostly allow the user to adjust color scales, overlay ion density maps and integrate MS images with acquired histological pictures. MSI data typically requires manual extraction of ion images, which can be extremely time-consuming for large-scale analyses, such as untargeted studies. Paschke et al. have developed a free software package, called Mirion^™, for automatic processing of MS images [154]. With Mirion, the user selects an m/z range of interest and the software creates the ion images. Manual image generation is also available with Mirion, since no software algorithm is as good as the eye of the MS expert. The capabilities of this software were demonstrated using peptide standards and tissue from mice. Another open-source MSI data processing software has been developed, called MSiReader [155]. This software allows the user to select a region of interest and a reference region. The software can generate a mass list of compounds found in the region of interest, excluding peaks in the reference region. The created mass list or a manually made mass list, can then be imported and an ion imaged will be extracted for every ion in the list. Software such as Mirion and MSiReader significantly improve MSI data processing in comparison to manual processing by drastically decreasing the amount of time it takes to work through a data set. Recently, software tools have been developed for quantitation of analytes from MSI data, such as Quantinetix^™ from ImaBiotech (Loos, France). Källback et al. have developed a novel MSI software for labeled quantitation and normalization of endogenous neuropeptides [126]. Shariatgorji et al. have developed in-house software for MSI quantitation based on signal intensities. This software can be used to define region of interest, process spectra, normalize spectra and quantify analytes based on pre-established calibration curves. Several neurotransmitters, drugs and neuropeptides have been successfully quantified using this software [105].

Conclusion

Over the past decade, MSI has obtained increasing attention from biologists and become more routinely employed to map various classes of bio-molecules from biological specimens. Its novel applications to the study of neurotransmitters and neuropeptides have provided valuable knowledge regarding disease mechanisms and related reparative processes. These studies offer significant insights into neurobiology and hold promise for our continuous search of effective treatment for diseases. We anticipate MSI for neurotransmitter mapping will lead to a better understanding of the biology of neurological diseases and improvements in clinical diagnostics.

Future perspective

Exciting technical advances are further improving the capabilities of MSI for biological applications from various aspects, such as sample preparation and instrumentation. The development and use of novel matrices and matrix-free imaging techniques, along with improvements in sample preparation are generating more reproducible, clean and reliable results with improved analyte detection. The advancements in instrumentation for shorter acquisition times are increasing the throughput of MSI techniques, making MSI more applicable to large-scale study of complex biological systems. The improved mass accuracy, mass resolution and spatial resolution are exciting developments that could greatly enhance the sensitivity of MSI and make this technique capable of studying biological problems on a cellular scale. Data processing and analysis is the current bottleneck of MSI, but progress is certainly being made in this area. Again, once software becomes commercially available and used enough to inspire confident, reliable results, MSI has the potential to be a powerful analytical tool, expanding into many new areas of research and industries.

Executive summary.

Small molecule neurotransmitter mass spectrometric imaging

• Mass spectrometric imaging (MSI) of small molecules is challenging due to matrix interference, rapid turnover rate and low in situ concentration. Innovative MSI techniques, such as in situ freezing, novel matrices and matrix free techniques, high-resolution and high-mass-accuracy mass analyzer have been used to overcome these difficulties.

Neuropeptide MSI

• MSI of neuropeptides is well developed and widely applied to many species and disease states. Advanced MSI techniques such as 3D imaging and quantitative imaging have been developed to further improve neuropeptide detection and visualization.

Novel matrices

• Advancements in the use of novel matrices and matrix-free methods enhance signal detection and produce cleaner spectra in the low mass region, which will allow for improved metabolomics and small molecule studies.

Sample preparation

• Challenges in sample preparation are being solved in novel ways, leading to more reproducible results that could even turn MSI into a quantitative technique.

Data processing

• There have been several recent advances for improved mass resolution, spatial resolution, acquisition time and data processing. Making these advancements commercially available could produce significant progress in many different fields of study.

Key Terms

Mass spectrometric imaging,

Analytical technique that allows for molecular analysis of tissue while retaining information about the spatial distribution of different analytes by collecting mass spectra in a predefined raster across a tissue sample, resulting in a 2D distribution map for each mass measured

Neurotransmitters

Endogenous substances released by synaptic terminals to transmit signals between nerve cells

Neuropeptides

Endogenous peptides synthesized in nerve cells or peripheral organs that are involved in cell–cell signaling. Often considered as peptide neurotransmitters and peptide hormones with sizes ranging from three to 70 amino acids

Matrix

Typically a small organic compound that is used to coat MALDI samples and assist in the ionization of analytes