Mass spectrometry in animal health laboratories: recent history, current applications, and future directions

Abstract

Mass spectrometry (MS) has long been considered a cornerstone technique in analytical chemistry. However, the use of MS in animal health laboratories (AHLs) has been limited, however, largely because of the expense involved in purchasing and maintaining these systems. Nevertheless, since ~2020, the use of MS techniques has increased significantly in AHLs. As expected, developments in new instrumentation have shown significant benefits in veterinary analytical toxicology as well as bacteriology. Creative researchers continue to push the boundaries of MS analysis, and MS now promises to impact disciplines other than toxicology and bacteriology. I include a short discussion of MS instrumentation, more detailed discussions of the MS techniques introduced since ~2020, and a variety of new techniques that promise to bring the benefits of MS to disciplines such as virology and pathology.

Animal health laboratories (AHLs) are complex organizations, consisting of groups dedicated to various disciplines in veterinary medicine, all working together to detect disease hazards and diagnose diseases and toxicoses in animals. Such laboratories are essential to the surveillance and control of diseases that have the potential to devastate agricultural animal production and affect human health and food safety. Although diagnostic and detection methods understandably vary between disciplines, there are commonalities in their requirements. These methods should be sensitive, accurate, precise, fast, efficient, and cost-effective. Within the discipline of veterinary toxicology, mass spectrometry (MS) has long been considered an essential technique for identifying toxicants in animal-related samples. MS analytical methods provide extremely high confidence in chemical identification, high accuracy and precision in quantification, a broad range of detectable analytes, and high sensitivity. They are also fast and efficient compared to other methods of chemical analysis. The most significant barrier to the use of MS is the high cost of acquiring the instrumentation. Despite this issue, the value of MS instrumentation to toxicologists is such that it has long been used in their laboratory sections.

Since ~2020, there have been significant developments in MS. New types of instruments offering broad capabilities in chemical analysis are widely used in toxicology sections. Other mass spectrometers that have been developed have been adopted in bacteriology sections, offering speed and efficiency in the identification of bacteria. Furthermore, a number of new applications in MS may offer new capabilities for other disciplines such as virology and pathology in the near future. These recent developments and those on the horizon make this an exciting time in the field of MS.

I present here a survey of the developments in MS over the last 20 y that have changed veterinary testing and MS techniques and that show significant promise for future use in the field.

Mass spectrometry basics

According to the International Union of Pure and Applied Chemistry (IUPAC), a mass spectrometer may be defined as an “instrument that measures the m/z values and abundances of gas-phase ions.” ⁵⁵ The term m/z refers to the mass-to-charge ratio of an ionized atom or molecule. (For the purposes of this article, the term “mass” will be used in referencing mass-to-charge ratio.) This simple definition belies the diversity of instrumentation used for this purpose. A time-of-flight (TOF) mass spectrometer used for identifying bacteria has little in common, functionally, with a gas chromatograph–mass spectrometer used for toxicologic analysis, except that they both measure the masses of chemicals. This broad array of MS instrumentation largely accounts for the flexibility of the technique. Mass spectrometers are used for such diverse purposes as measuring atmospheric aerosols, ⁹¹ detecting depleted uranium on battlefields,^89,90 determining the presence of organic chemicals in Martian soil, ⁷⁰ and detecting trace quantities of explosives on luggage at airports. ²⁴

MS is a preferred technique for these various applications for several reasons:

• High selectivity. MS instrumentation is considered a “gold standard” technique for the identification of chemicals, particularly in complex matrices. A properly designed MS method with properly interpreted data offers an extremely high degree of confidence in chemical identification.

• High sensitivity. As each new generation of MS instrumentation is developed, sensitivity improves. MS may be used to detect therapeutic drugs in wastewater,^4,87 to detect perfluoroalkyl substances (PFAS) in serum and cosmetics,^28,56 and to detect oleandrin in bovine tissue and biological fluids,^81,82 all at parts per billion to low parts per trillion concentrations. In all of the above-listed examples, the detection of chemicals at such low concentrations is required to generate relevant data. Even when parts per trillion measurements are not required, high sensitivity is still advantageous. For example, a less sensitive instrument might require lengthy sample extract concentration procedures that are not necessary with a more sensitive instrument.

• Versatility. Various types of MS systems are used to detect analytes ranging from atoms to small molecules to intact proteins in the hundreds of kilodaltons.

All mass spectrometers are comprised of 3 basic components: an ionization source, a mass analyzer, and a detector ²⁶ (Fig. 1). In the ionization source, neutral atoms or molecules are given positive and/or negative charges. The ionization source is necessary because most chemicals of interest are not ionic in their natural state, and mass spectrometers measure ions rather than neutral species. Therefore, the mass spectrometer requires a way to place a positive or negative charge on these chemicals. An extremely broad array of ionization sources has been designed over the past 100 y or so, since the invention of the first mass spectrometer, but only a few are routinely used for veterinary testing. These include primarily electron ionization sources used for gas chromatography–mass spectrometry (GC-MS), electrospray ionization sources used for liquid chromatography–mass spectrometry (LC-MS), and matrix-assisted laser desorption/ionization (MALDI) sources used for bacteriologic analysis.

Figure 1.

Block diagram of a mass spectrometer. Molecules are ionized in the ion source and separated based on atomic or molecular mass in the mass analyzer. The separated ions are then detected and the signals from the detector are acquired by the data system, where they are stored and presented in a manner that allows interpretation.

The mass analyzer separates ions based on their masses (atomic or molecular weights). As with ionization sources, there are various different types of mass analyzers, but only a few that are used in veterinary testing. These include quadrupole analyzers used for GC-MS and LC-MS, Orbitrap and TOF systems used for high-resolution LC-MS, and MALDI-TOF systems used in bacteriologic analysis.

Once the chemicals have been ionized and separated based on their masses, a detector provides signals for the different ions that are then digitized and sent to a data system or display. The signal intensity for each ion will vary as a function of the abundance of that ion, which allows for quantification of chemicals.

Mass spectrometry in veterinary testing, pre-2000

Gas chromatography–mass spectrometry

MS was in common use in veterinary toxicology sections well before 2000 in the form of GC-MS. In these systems, the GC is used to separate components in a sample extract immediately before their introduction into a mass spectrometer. This separation process first involves flash vaporization of the extract. The vaporized components from the extract then move through a separation column in a constant flow of inert gas with separation occurring as the components interact differently with the column. This process introduces the many components from the sample extract into the mass spectrometer in a controlled fashion, which allows for more effective MS analysis. This process works well for the analysis of chemicals that do not degrade under the high temperatures required for vaporization (generally in the 200–350°C range). The GC-MS technique provides analyses targeted towards specific chemicals or classes of chemicals with accurate and precise quantification in parts per million to parts per billion concentrations, depending on the sample preparation method used. It also allows for the identification of chemicals on a non-targeted basis, in which spectra from samples are compared with libraries of hundreds of thousands of mass spectra. By the mid-1990s, publications from veterinary toxicology laboratories included a survey of chemicals identified by non-targeted analysis, ⁶⁴ analysis of fixed tissue for methenamine, ²¹ detection of embutramide in tissue samples, ⁷ detection of drugs from injection sites, ⁷⁴ anticoagulant rodenticide analysis, ⁶³ cantharidin analysis, ⁶² and aldrin analysis in cattle-related samples. ¹²

Although these systems are expensive to acquire, their usefulness is such that laboratories found ways to purchase them. Today, most veterinary toxicology sections use these systems routinely.

Developments in mass spectrometry in veterinary testing since 2000

Liquid chromatography–mass spectrometry

A significant limitation with GC-MS analysis involves detection of thermally labile analytes. Many chemicals of interest to toxicologists degrade at the high temperatures required for GC, never making it to the mass spectrometer in intact form. Although there are ways to circumvent this issue (often involving chemical modification of the analyte ⁸⁵ ), these methods are typically time consuming and may be difficult to perform reliably on a routine basis. It was recognized early on in GC-MS development that a way around this problem would be to interface a liquid chromatograph to a mass spectrometer. With liquid chromatographs, the sample extract is introduced into a flow of solvent rather than a high-temperature injection port, eliminating thermal breakdown issues. The challenges involved with interfacing LC and MS were significant, however, and it took decades to overcome them. It wasn’t until the development of electrospray ionization, which was widely adopted in the late 1990s, that LC-MS became a technique reliable enough for routine laboratory use. ³² These systems revolutionized analytical chemistry by greatly expanding the range of chemicals that may be analyzed by MS, from small molecules to (for the first time) intact proteins, ²² with very high sensitivity.

Veterinary testing laboratories were slow to take up LC-MS, primarily because of the cost of acquiring the systems and the need for analysts with specific training in the technique. Unlike GC-MS systems, LC-MS requires that each test be specifically optimized for the chemical or chemicals being analyzed. This in turn requires analysts with a high degree of knowledge of the chemicals and the instrumentation in order to provide accurate and sensitive testing. Nonetheless, the number of laboratories with these systems has steadily grown to the point that many veterinary toxicology laboratories now use them routinely in their diagnostic work.

Inductively coupled argon plasma–mass spectrometry

In the early 2000s, the primary systems used for elemental analysis in veterinary testing were atomic absorption spectrometry (AAS) and inductively coupled argon plasma–optical emission spectrometry (ICP-OES). ICP-OES systems were considered particularly useful as they allow for the detection of many elements in a single analysis as opposed to AAS, in which only one element can be analyzed at a time. In 1980, researchers successfully interfaced an ICP system with a mass spectrometer rather than an OES. ³⁸ The first commercial ICP-OES systems became available in the early 1980s, but as with other analytical technologies, it took many years for systems to be developed that were reliable and simple enough to use for routine analysis. By the late 2000s, veterinary laboratories began shifting their elemental analysis, including trace elements and heavy metal analyses, from ICP-OES to ICP-MS systems. ICP-MS offers higher sensitivity and more reliable detection than ICP-OES. Today, ICP-MS is a standard technique in veterinary toxicology groups; this was a significant development given the importance of elemental analysis in veterinary testing.

Benchtop high-resolution mass spectrometry

Mass spectrometers typically measure mass on an integer basis (e.g., they can distinguish an ion with mass of 200 from one of 201, or one of 560 from 561). These are commonly termed “nominal mass” or “unit mass” systems. High-resolution MS systems measure mass with far greater precision (e.g., distinguishing an ion of mass 301.5423 from one of 301.5433). Such systems have been commercially available since the middle of the 20th century but were too large, expensive, and complex to be used in routine veterinary testing. Thermo Fisher changed this situation in 2005 with the release of an instrument based on the Orbitrap mass spectrometer (LTQ Orbitrap; Thermo Fisher), ³⁹ the first benchtop LC-MS system offering extremely high mass resolution. High-resolution MS systems offer several advantages over nominal mass systems. Chemicals with the same nominal mass but different chemical formulas are easily distinguished from each other, thus providing enhanced capabilities for targeted analysis in complex sample matrices. These systems also provide the unique capability for non-targeted LC-MS analysis, given that mass measurements of chemicals identified in a sample may be compared with databases of accurate masses of chemicals of interest.

Early Orbitrap systems were oriented toward research groups and were complex and extremely expensive. By 2010, systems were available that were less expensive and suitable for routine analytical work. The success of Orbitraps spurred manufacturers of TOF mass spectrometers to improve their products as well, and there are now high-performance liquid chromatography (HPLC)-TOF MS systems that also provide benchtop high-resolution MS capabilities. The prices for high-resolution instruments are now comparable to nominal mass LC-MS systems, and as their utility becomes more apparent, more laboratories may adopt them.

MALDI-TOF MS for bacteriologic analysis

Routine bacterial identification has traditionally been based on growth characteristics and morphology, along with biochemical tests to systematically determine the identity of a bacterial isolate. These tests often take days to provide results. MS was recognized as a promising technique for rapid identification of bacteria as far back as the mid-1970s, ³ but it was not until researchers began using MALDI-TOF systems in the mid-2000s that its potential for routine analysis was recognized. ⁷¹

In MALDI-TOF analysis, MALDI refers to the ionization source and TOF to the mass analyzer. Isolated bacterial culture samples are spotted onto target plates after which a matrix solution is applied, the purpose of which is to enhance ionization of sample components. Each spot is then subjected to pulses from a laser, which causes ionization of sample components and release of these ions. The ions are then directed into the TOF mass analyzer (Fig. 2). In the case of bacteria, the ions represent bacterial peptides primarily corresponding to ribosomal and reference proteins in the mass range m/z of 2–20 kDa. ⁷²

Figure 2.

Diagram of a MALDI-TOF MS system. The sample is mixed with a matrix that enhances ionization and is then spotted onto a plate. The plate is placed in the ionization chamber, where a laser pulse ionizes molecules via matrix-assisted laser desorption/ionization (MALDI). The ionized molecules, indicated by the + signs, are directed into the time-of-flight (TOF) analyzer via electromagnetic fields. The molecules drift through a long tube with a detector at the end. Separation of ions by molecular mass occurs as smaller molecules arrive at the detector more quickly than larger molecules. Reproduced from ⁵⁸ under the Creative Commons License.

Measurement of the masses of the proteins and their abundances provides a mass spectral “fingerprint” that is unique to a bacterial genus or species. Bacteria are identified by comparison of that peptide mass fingerprint (PMF) of an unknown isolate to libraries of mass spectra from different bacteria. Sample target plates typically allow for 96 or 384 spots per plate, facilitating high-throughput analysis. These systems are quite fast, with each sample taking only seconds to analyze, and 96 samples taking ~30 min. ⁴⁴ The instruments fit on benchtops (Fig. 3) and do not require analysts to have advanced knowledge of mass spectrometry to operate them.

Figure 3.

A workflow for bacterial identification using the Bruker MALDI Biotyper as an example. Colony-scraping is performed using a sterile toothpick. The colony adhering to the toothpick is directly transferred onto a MALDI target plate for spotting. Immediately, 1 µL of 70% formic acid and matrix are spotted onto the MALDI target plate, and bacterial ID is performed with MALDI-TOF MS analysis using a MALDI Biotyper. Reproduced from ⁸³ under the Creative Commons License.

The low per-sample costs and obvious rapid speed of analysis compared to traditional phenotypic testing have therefore made MALDI-TOF MS common among veterinary testing laboratories.

The future of mass spectrometry in veterinary testing

Advancements in the use of MS in veterinary testing over the last 20 y have clearly been significant. Developments in MS instrumentation continue, with creative researchers working on many different applications for many different types of analysis. The next 20 y are likely to bring major new developments that could benefit several veterinary testing disciplines. A few of these are considered below.

Botulinum toxin analysis

Botulism is a paralytic disease caused by neurotoxins produced by various species of Clostridium bacteria. Botulinum neurotoxins (BoNTs) are among the most toxic substances known, with as little as 50 ng being sufficient to cause the disease in humans. ⁵⁹ Botulism is a common disease in animals, with waterfowl being particularly susceptible.^14,86 Cattle^19,65 and horses^78,84 are also commonly affected.

BoNTs exert their effects by cleaving specific proteins, known as SNAP receptor (SNARE) proteins, at neuromuscular junctions, which prevents the release of acetylcholine and causes paralysis. There are 8 recognized botulinum serotypes, named A–H. C/D and D/C chimeric variants and 41 subtypes have also been identified. ⁵⁹ Each serotype cleaves a SNARE protein at a different specific location along its amino acid sequence. Although human botulism typically involves the A, B, or E serotypes, animals are usually affected by the C or D serotypes or the chimeric C/D or D/C variants. Horses are a notable exception, with the A and B serotypes also involved in equine botulism cases. ⁸⁴

Definitive diagnosis of botulism requires detection of the involved BoNT. The current gold standard technique for this detection is the mouse bioassay, although there are a number of problems with this technique. It is time-consuming, difficult to standardize, and causes ethical concerns given that it relies on the use of live animals. ⁶⁰ The sensitivity of the mouse bioassay may also be inadequate to detect BoNT in horses, given their exquisite sensitivity to these toxins. ⁷⁸

MS has long been seen as a promising technique for use in botulism diagnosis. Despite the tremendous improvements in MS instruments, routine direct detection of intact BoNTs at clinically relevant concentrations has not been achieved. This is largely because of the difficulties in detecting these 150-kD BoNT molecules at extremely low concentrations in complex biologic matrices. A breakthrough came in 2005 when researchers from the U.S. Centers for Disease Control and Prevention reported a method they termed the Endopep method, in which MS was used to detect the products of peptide substrate cleavage by BoNTs. ⁶ The researchers synthesized shortened versions of the various SNARE proteins, centered around the locations in which the cleavage occurs. When BoNT was added to these peptide substrates, cleavage occurred at the predicted locations and MS, either MALDI-TOF or LC-MS, was used to detect the cleavage products.

In real world samples, sample extracts were first combined with magnetic beads coated with BoNT antibodies. The beads were then washed, and the peptide substrates were incubated with the beads. After incubation, the substrate solution was transferred from the beads and analyzed by MS to detect the cleavage products. Detection limits were equivalent to 0.1–0.5 mouse LD₅₀, depending on the serotype, for BoNTs A, B, E, and F, ⁴² and could demonstrate clearly the difference in MALDI-TOF spectra between a sample that was negative for BoNT-A and one that was positive (Fig. 4).

Figure 4.

Mass spectra of the Endopep-MS reaction for A. negative control tested for BoNT-A; B. diagnostic sample tested for BoNT-A; C. negative control tested for BoNT-B; and D. diagnostic sample tested for BoNT-B, -E, and -F. The cleavage products of peptides indicating the presence of BoNT-A or -B are circled in red. From ⁴¹ reproduced under the Creative Commons License.

Development of the Endopep method has proceeded since its introduction, resulting in a more sensitive procedure that has been applied to various detection situations. Most of this development has involved analysis for BoNT-A, -B, -E, and -F, but there have been reports of its successful use in detecting BoNT-C, -C/D, -D, and -D/C in animal samples.^27,30,54,80 Note that the cleavage sites for BoNT-C and -C/D are identical to those for BoNT-D and -D/C; hence, the Endopep method cannot distinguish between the members of each of these pairs of variants. In a 2020 study, researchers showed that a MALDI-TOF system commonly used for bacterial testing could be used for identification of BoNT-C, -D, -C/D, and -D/C in animal serum and fecal samples. ²⁰ The limits of detection (LODs) for BoNT-C and -D were 0.5–2 mouse LD₅₀ for culture broth and serum, as well as for BoNT-D in fecal samples. The LOD for BoNT-C in feces was 4 mouse LD₅₀, likely because of nonspecific proteases causing degradation of the peptide substrate before the BoNT-C could cleave it. Another report indicated that the use of a salt wash and protease inhibitors greatly improved results for BoNT-C in liver samples; ⁷⁹ such an approach could be evaluated to potentially improve results in feces or other problematic matrices.

The Endopep technique promises to revolutionize veterinary analysis for BoNTs, but there are a number of challenges to be met before this method can be adopted for routine use. Challenges include access to antibodies for BoNT-C and -D, evaluation of sensitivity in real-world samples from different species, and standardization and validation of the method among laboratories. These issues are resolvable and should soon result in retirement of the mouse bioassay.

Virus detection using mass spectrometry

Widespread outbreaks of viral diseases in animals such as virulent Newcastle disease and highly pathogenic avian influenza are common, often resulting in the need for rapid analysis of large numbers of samples by veterinary laboratories. The current gold standard for viral testing is real-time PCR (rtPCR), but this technique is not without problems. Molecular analysis can be time consuming, particularly when sequencing for phylogenetic analysis is considered. Furthermore, PCR probes and primers must be periodically updated to account for viral evolution. ³⁴ There has been ongoing research into the use of MS methods for viral detection that would be at least as sensitive and specific as RT-PCR while providing faster analysis and enhanced capabilities in detecting viral variants. Detection of viruses at diagnostically relevant concentrations by MS is challenging, however, because viruses have a low protein content compared to the high concentration of other proteins in viral cultures. Therefore, signals from the targeted viral proteins become buried in much more intense signals from other proteins. ⁷³ Although there are reports in the literature of the use of MALDI-TOF MS to detect a variety of viruses, ⁷³ the technique has not yet been widely adopted for routine analysis. The methods that have been developed may be seen as proof of concept and perhaps an indication of the future of viral detection.

In one approach, MALDI-TOF MS is used to detect PCR amplicons. This is an attractive technique given that PCR amplification overcomes problems with low concentrations of viral proteins. Issues involving variants may be mitigated by the use of multiplexed PCR. Examples of methods using this technique include identifying up to 60 hepatitis B variants,^37,51 18 high-risk human papilloma virus variants, ⁹ and 10 different viruses in ducks. ⁴⁸ A commercial platform, the MassARRAY SARS-CoV-2 Panel (Agena Bioscience), has been developed and uses PCR–MALDI-TOF MS for the detection of SARS-CoV-2 infection in humans and has been granted emergency use authorization by the U.S. Food and Drug Administration (FDA).^1,35,67 These methods all show sensitivity and specificity equal to or greater than established methods as well as shorter analysis times.

A newer approach involves direct detection of viral proteins. As noted above, this is difficult given their low concentrations among much higher concentrations of other proteins in samples. The severe demands placed on human testing laboratories by the SARS-CoV-2 pandemic have spurred researchers to develop approaches for direct analysis by MS that might overcome this problem. In one method, a characteristic SARS-CoV-2 protein was extracted, purified, enzymatically digested from nasopharyngeal swabs, and detected by MALDI-TOF MS. ⁸⁸ The method was fast, and results had good concordance with RT-PCR results, but with lower sensitivity. The technique was successful in identifying contagious patients. In another report, machine learning was used to determine MALDI-TOF signals from SARS-CoV-2 viral proteins in transport media from nasopharyngeal swabs.^62,83 The authors developed datasets consisting of direct MALDI-TOF analyses of swab transport media from subjects both positive and negative for the COVID-19 virus by rtPCR. They then applied automated machine-learning platforms to these datasets to determine MALDI-TOF MS signals that corresponded with the presence of the virus. These signals were then used to analyze swab transport media for SARS-CoV-2 and compared to rtPCR results from the same samples. Their MS results compared well with the rtPCR results. The primary benefits of this method are the very short time required to analyze large sets of samples and the minimal need for supplies (Fig. 5).

Figure 5.

Comparison of molecular, antigen, and ML-enhanced MALDI-TOF MS workflows illustrating the tradeoffs of A. high-throughput molecular platforms, B. point-of-care molecular platforms, and C. proposed MALDI-TOF MS method. From ⁶¹ reproduced under the Creative Commons License.

Clearly, there is much work to be done before MS can be considered for routine viral detection in veterinary laboratories. The existence of a commercial turnkey system for SARS-CoV-2 detection that has been FDA authorized is promising, and similar systems for veterinary testing could speed the testing process significantly. The use of machine learning to enable direct detection of viruses from swabs offers the potential of even higher throughput analysis. The difficult experience of the COVID-19 pandemic will likely continue to spur development of these techniques and their potential adoption by veterinary testing laboratories.

Mass spectrometry imaging

In typical MALDI-TOF MS analysis, a plate on which samples are spotted is placed within the instrument. The plate is analyzed by pulsing a laser over the first spot, then moving sequentially to the other spots on the plate and analyzing each of them. If a tissue section is placed on a plate, then the entire section may be analyzed by pulsing the laser over a location on the section, moving the plate very slightly, pulsing the laser again, and so forth until data have been acquired for the entire section. Each spot on the section that has been analyzed can be considered a pixel containing information regarding the molecular weights of chemicals located on that spot. The data for the pixels can then be combined and visualized as a chemical image of the section. Signals from different chemicals may be highlighted to show spatial differences in chemical composition of the tissue (Fig. 6).

Figure 6.

Schematic outline of a typical workflow for an imaging mass spectrometry experiment. Sample pretreatment steps include cutting, mounting the sample on a target, and matrix application. Mass spectra are generated in an ordered array at each x,y coordinate. Individual spectra can be visualized within the tissue to generate protein images. Reproduced from ⁶⁹ under the Creative Commons License.

MS imaging was first developed in the late 1960s for use in mapping semiconductor surfaces. ⁴⁷ Since then, the field has seen significant growth; the annual number of PubMed citations involving imaging mass spectrometry increased from 1 in 2003 to 103 in 2021. ²³ Various ionization techniques have been used for MS imaging, including secondary ion mass spectrometry, desorption electrospray ionization, and MALDI. MALDI-TOF MS was first used for imaging in the 1990s and has since become the most popular MS technique for this application. ¹¹ MALDI-TOF MS systems are well suited to imaging studies given the ease of use of these systems and the capability of detecting a very broad range of chemicals.

Recent reviews of MALDI imaging include applications in neurology, ⁶⁸ protein analysis, ⁶⁶ pharmacokinetic studies, ⁵⁷ food science, ³⁶ and carbohydrate analysis, ²⁹ reflecting the versatility of the technique. MALDI imaging has been used to determine biomarkers for diseases such as osteoarthritis, ¹³ cardiovascular disease, ⁷⁷ mycobacterial infections, ⁵ and neurodegenerative diseases. ² The ability to compare protein, peptide, lipid, and metabolite distributions between healthy and disease-state tissue samples has been of particular advantage in cancer studies. MALDI-TOF MS imaging has been used for classification of tumor cells,^18,52,53 for staging cancer and to help determine prognosis of patients,^50,76 for research into a broad variety of cancer types, ⁴⁵ and for the differentiation of histologically similar tumors. ¹⁵ In a study of human pancreatic cancer, MALDI imaging was used to determine protein biomarkers indicative of lymphatic invasion and nodal metastasis, 2 prognostic indicators for this disease. ⁴⁹ Using single protein biomarker images for tissues positive and negative for these prognostic indicators, positive and negative H&E-stained tissue sections can be differentiated clearly by MS imaging (Fig. 7).

Figure 7.

MALDI MS imaging of chemical biomarkers of the prognostic features of A. lymphatic invasion (pL– and pL+) and B. nodal metastasis (pN– and pN+) in human pancreatic cancer tissue. H&E-stained sections are on the left. The heat maps on the right show the distribution and intensity of MALDI signals corresponding to collagen alpha-2(l), indicative of lymphatic invasion, and histone H1.3, indicative of nodal metastasis. Collagen alpha-2(l) is significantly increased in pL+ tissue; histone H1.3 is significantly decreased in pN+ tissue. Reproduced from ⁴⁹ under the Creative Commons License.

MALDI imaging workflows involve only a few basic steps, but a variety of options can be involved with each step. Detailed information regarding each of the sample preparation steps, including tissue sample preparation and choice of the MALDI matrix, is beyond the scope of my article but is included in a recent review. ⁸ Most commonly, tissue sections are prepared by cryosectioning or as formalin-fixed, paraffin-embedded (FFPE) sections. Cryosectioned tissue is preferred over FFPE sections because the FFPE process alters the chemical composition of the tissue more significantly than cryosectioning. That said, the wealth of information contained in archives of FFPE tissue samples has spurred ongoing research into developing methods of FFPE tissue analysis, with significant progress being made.^25,33 After the tissue section is mounted, the MALDI matrix is applied to the tissue. A wide variety of matrices have been evaluated, ⁸ and the choice of a matrix solution and application technique depends on the goal of the analysis and whether detection of a specific chemical class (lipids, proteins, etc.) is desired. After drying, the sample is ready for analysis.

One of the critical parameters for MALDI-TOF imaging is spatial resolution, or pixel size. Current “off-the-shelf” systems routinely achieve 10-µm resolution, with some more advanced instruments reaching ≤1 µm. ¹⁰ Higher resolution enhances the ability to discriminate the molecular distribution within tissue samples, but it comes with tradeoffs. As resolution increases, the sensitivity at each pixel decreases, the time spent performing the analysis increases, and the data file size increases. ¹⁰ For resolution of <10 µm, a complete section may take hours or days to image using typical MALDI-TOF systems. ⁴³

A technique that promises to mitigate these tradeoffs for diagnostic applications is known as histology-directed mass spectrometry imaging.^16,31 In most versions of this technique, serial tissue sections are produced. One of these is stained and an image of the stained tissue is examined by a pathologist and annotated to indicate regions of interest (ROIs) in which chemical information might be of use. An image of the second section is produced using a flatbed scanner, and that image is then registered with the stained section. Registration involves the use of specialized software to map the annotated image of the stained section to the image of the unstained one. The mapped image of the unstained section is then used by the MALDI-TOF system for accurate MS imaging of the ROIs within that section. One issue with this approach is the lack of resolution of flatbed scanners compared to microscopes, which can result in loss of accuracy in the registration process.

A report describes 2 potential workflows for histology-directed MS imaging that mitigate this problem. ⁵⁸ In one workflow, a single section is imaged using autofluorescence (AF). A pathologist then examines the AF image, annotates ROIs, and that same section is used for MS imaging. “Direct AF microscopy imaging,” as the authors call this, eliminates registration and has the additional advantage of avoiding occasional situations in which there are differences between serial sections that hinder the ability to target specific tissue areas (Fig. 8). The disadvantage to this approach is that it does not allow for staining of the section examined by the pathologist. The second workflow involves the use of serial sections. One section (section 1) is first imaged by AF and then prepared using traditional histologic staining techniques. A pathologist examines the stained image, annotates ROIs, and the stained image is then registered with the AF image of that same section in a process called “intra-section automated registration.” Section 2 is imaged via AF, and its AF image is then registered with the AF image from section 1, with this process termed “inter-section mono-modal automated registration.” The MALDI-TOF system uses the mapped AF image of section 2 to image the ROIs on that section. Despite the requirement for 2 separate registration processes, AF imaging offers significantly higher resolution than that obtained by flatbed scanners, affording more precise registration and more accurate MS imaging of ROIs. This workflow also allows the pathologist to choose any appropriate stain for their examination of section 1. The various scanning and registration steps add up to ~1 h of instrument and computation time but significantly less user time.

Figure 8.

Two autofluorescence (AF)-driven workflows for histology-directed MS imaging. The upper workflow involves annotation of a stained section and then registration of the AF image of that same section with the stained image. A second registration maps the AF image of the first section with the AF image of a serial section. Regions of interest determined by viewing the stained section are then targeted for MS imaging of the second section. In the lower workflow, an AF image of a section is used to determine regions of interest which are then analyzed by MS imaging of that same section. From ⁵⁸ reproduced with permission.

MS imaging is increasingly being used for small molecule applications such as drug development.^17,23,46,75 MS imaging is particularly valuable in mapping the spatial distribution of therapeutic drugs within tissues and can provide reliable quantification of these drugs. MS imaging has also been used for investigations in drug toxicology. ⁴⁰ This work raises the intriguing possibility of using MS imaging to detect toxicants in tissue specimens for postmortem testing. Such an application could provide rapid screening for toxicants without the lengthy purification and separation procedures currently used in toxicologic analysis. To my knowledge, however, there is not yet any proof of concept published for this application.

As with other techniques that are early in development, there is much work to be done to introduce histology-directed MS imaging to routine veterinary testing. Biomarkers for diseases must be identified and validated, and imaging systems evaluated. Nonetheless, the work done so far can be considered as proof of concept, and with developments in MS imaging proceeding at a rapid pace, it may not be long before chemical information begins to augment microscopic examination for routine veterinary pathology.

Conclusion

Prior to 2000, the use of MS for routine veterinary testing was limited to GC-MS in toxicology sections. Since then, use of the technique has expanded to include LC-MS, ICP-MS, and high-resolution ICP-OES in toxicology, and MALDI-TOF MS in bacteriology. With continuing developments in instrumentation and applications, it appears inevitable that MS will make its way into disciplines such as virology and pathology as well as finally allowing for diagnosis of botulism without the use of mice. Instrumentation costs will be high relative to other systems used in veterinary testing, but they will likely be adopted provided that gains in accuracy and efficiency justify those costs. There is a bright future ahead for MS in veterinary testing.