Alasdair Matheson

Mid-infrared (mid-IR) spectroscopy excels in the speed of data acquisition, and this feature has led spectroscopists to pursue mid-IR spectral imaging in particular as a valuable future tool for medicine, in areas such histopathology and cytology for disease research, and clinical diagnostics. Nick Stone of the University of Exeter, UK, spoke to 

 about the advantages of the technique, and the latest technological development that are advancing the capabilities of mid-IR spectroscopy for use in medicine.

You are well known for your work in Raman spectroscopy. Recently, however, you have collaborated on some projects using mid-IR spectroscopy for biological samples. Are there circumstances where mid-IR offers specific advantages for biological or clinical applications?

	Absolutely. Raman and IR spectroscopy are complementary techniques that are able to probe the vibrational modes of biomolecules of interest, and therefore are useful to study the composition and microenvironment of tissues and cells as disease develops (1). We have worked on Raman and IR in parallel for many years to ensure that complementary information is available to us in our work developing methods for clinical diagnostics. Working with both techniques has allowed us to assess the relative pros and cons of each approach for particular clinical applications. We previously undertook a parallel study on IR and Raman for identification of secondary cancers in mediastinal lymph nodes. Both approaches provided an almost identical performance when using multivariate prediction of disease (2).

	There are, of course, pros and cons for each approach, particularly around sample preparation and use in vivo and ex vivo. Raman usually allows for little sample preparation, and IR usually requires dry, thin sections. However, mid-IR imaging excels in the speed of acquisition of the data, and this advantage has led us and other groups to consider mid-IR spectral histopathology as a key future tool for medicine (3). Current "gold standard" cancer diagnostics use formalin-fixed paraffin-embedded tissue sections for staining and histopathological analysis. These sections, when mounted on the appropriate IR transmitting slide, can be used unstained for analysis of disease-specific molecular changes.

	My team in Exeter have been collaborating recently on two European projects seeking to develop IR technologies to further maximize the technique's potential for spectral histopathology. Rohit Bhargava of the University of Illinois at Urbana-Champaign, Wolfgang Petrich of the University of Heidelberg and pharmaceutical company Roche, in Frankfurt, Germany, and others, have shown the value of discrete frequency imaging with quantum cascade lasers (QCLs) as a method to speed up the collection of images at a set of specific wavelengths of interest (4,5).

	We worked as part of the MINERVA consortium, a European Commission funded project that aims to develop photonic technology in the mid-IR to improve early cancer detection, to develop mid-IR supercontinuum sources with acousto-optical tunable filters for application in discrimination of malignancies in the colon (6). We are also working, as part of the Mid-TECH consortium, a consortium of academic and industrial partners in Europe to advance mid-IR spectroscopy, on a combination of tunable light sources (QCLs, or optical parametric oscillators [OPOs]), coupled with upconversion technologies, to enable detection of IR-converted light with silicon-based charge-coupled device (CCD) detectors, with the objective of simplifying systems and reducing costs by removing the need for cryogenic cooling of detectors (7).

Have there been any recent technological advances that have made mid-IR spectroscopy more viable as a technique in medicine?

	Certainly, there are a number of recent developments that will revolutionize the use of IR for real-time microscopic imaging of biological specimens. As I mentioned previously, tunable laser light sources-QCLs, supercontinuum, or OPO systems-allow for significantly improved intensity and optimized high-resolution imaging, compared with globar, thermal IR sources.

	The major clinical benefit of discrete frequency imaging is that clinicians can view molecular images in real time, much as they would now, but with the ability to dial up the molecular "stain" of interest by changing the wavelength.

	Furthermore, chalcogenide fibers for transmission of mid-IR light and upconversion imaging are likely to have a major impact on this field (8).

	The issue of coherence can limit image quality, and a number of strategies have been developed to overcome this. One approach is simply to measure all of the light in a single detector, and raster the illumination point; that is, to map it rather than image it. This overcomes variations in intensities measured at each pixel caused by speckle when using coherent sources to obtain wide-field images.

One of your recent collaborations involved the use of mid-infrared multispectral tissue imaging using a supercontinuum source. What was the aim of this project, and what were the benefits of using a chalcogenide fiber supercontinuum source? 

	The aim of this project was to provide a bright and tunable source as an alternative to QCLs. QCLs have been demonstrated to be most dominant at long wavelengths (5–12 µm), and following recent developments in the MINERVA project, supercontinuum sources are now dominant at shorter wavelengths (2–5 µm). However, the MINERVA consortium were able to deliver a supercontinuum source able to cover much of the fingerprint region of the spectrum (1000–1800 cm-1 or around 5–10 µm), albeit with low power at longer wavelengths (9). Further developments in this area are ongoing. One output from the project was a shorter wavelength supercontinuum source covering the high wavenumber region of the IR spectrum, and providing an intense source able to be used for a number of biomedical applications (10). Successful development of a broadband supercontinuum light source covering the fingerprint and high wavenumber region is likely to be much cheaper than the requirement for numerous QCLs to cover the same range. That said, a tunable filter is still needed, unless the supercontinuum source is used as the source for the Fourier-transform (FT) interferometer, which then removes the benefit of rapid discrete frequency imaging.

What were the main obstacles you had to overcome?

	Technology! It is difficult to make chalcogenide fibers of sufficient quality. Angela Seddon's group in Nottingham were able to do this, and, with Ole Bang's group in DTU in Denmark, they were able to deliver the world's first supercontinuum light source in the fingerprint region (9). Christian Pederson and Peter Tideman-Lichtenberg's group, also at DTU, have been driving forward the upconversion detection instrumentation and methodology (11).

	Coherence in wide-field imaging is also an issue that a number of groups have been seeking solutions for. As mentioned above, the simplest solution is to merge all the photons into one detector and remove the speckle problem, but this reduces other strengths of the approach for real-time imaging.

Could the approach using a chalcogenide fiber supercontinuum source be useful in other tissue imaging applications?

	Yes. This approach is applicable to all biological tissues, although, to date, really only the high wavenumber region of the IR spectrum has had sufficient photons generated with a supercontinuum source to provide data of similar quality to an FT-IR instrument. Clinical diagnostics are possible in this wavelength range (12), but more subtle changes, such as early cancers, usually need the fingerprint data to improve the accuracy of discrimination.

You also used mid-IR for hyper-spectral imaging for label-free histopathology and cytology. What specific advantages does mid-IR offer for this these applications?

	As mentioned before, mid-IR spectroscopy allows molecular analysis of unstained tissues, rapid analysis, and, potentially, the automated prediction of disease, and analysts only need to view a single section for numerous molecular stains of interest (13). Cytology is trickier because of Mie scattering effects, but Achim Kohler and others have worked hard to overcome this problem with post-measurement corrections (14).

	Discrete frequency imaging approaches and upconversion detection can be readily used to provide molecular distributions across tissue sections. There is still some way to go to ensure that these instruments can deliver the spectral quality or reproducibility of FT-IR instruments, but this does not appear to be an impossible task (15). Ongoing work is showing the viability of generating similar IR spectral histopathological images using various instrument configurations (16).

Are you planning to explore your research in mid-IR further? 

	Yes. Now that we have systems that can provide the signals we need, we will begin to explore the clinical needs where this will be most useful, and seek to translate it to the clinic, if possible, for the benefit of patients.

In what specific area of clinical use do you think mid-IR spectroscopy has the greatest potential? Are there any misconceptions that are stopping mid-IR spectroscopy from being used in this field?

	Linking mid-IR imaging with digital pathology is the next key step. Many experts in machine learning have been able to provide pretty good predictions of disease based on pattern recognition and two stains in an RGB image. A few more channels, albeit at lower spatial resolution, should provide much improved performance, particularly for the pathologies that are most difficult to reliably identify, such as dysplasias and early cancers.

	The real power of vibrational spectroscopic measurements lies in their ability to measure the biochemistry of the tissues of interest; this is the downstream expression of any genetic mutation, and will provide us with information relating to the future outcome of that disease. We are at the birth of this field in one sense, even though IR and Raman have been around for decades. The real breakthroughs will come when tens to hundreds of thousands of patients' samples, with particular conditions and their outcomes, are measured, and machine learning approaches are used to extract the key prognostic markers. Then, and only then, will IR really be providing the significant added value that it has promised for so long.

	(1) M.J. Baker, H.J. Byrne, J. Chalmers, P. Gardner, R. Goodacre, A. Henderson, S.G. Kazarian, F.L. Martin, J. Moger, N. Stone, and J. Sulé-Suso 

	(2) M. Isabelle, N. Stone, H. Barr, M. Vipond, N. Shepherd, and K. Rogers, 

	(3) J. Nallala, G.R. Lloyd, M. Hermes, N. Shepherd, and N. Stone, 

	(4) K. Yeh, S. Kenkel, J.N. Liu, and R. Bhargava, 

(1):485–493 (2015 Jan 6). doi: 10.1021/ac5027513. Epub 2014 Dec 22.

	(5) N. Kröger, A, Egl, M. Engel, N. Gretz, K. Haase, I. Herpich, B. Kränzlin, S. Neudecker, A Pucci, A. Schönhals, J. Vogt, and W. Petrich. 

	(8) L. Sójka, Z. Tang, D. Furniss, H. Sakr, E. BereÅ -Pawlik, A.B. Seddon, T.M. Benson, and S. Sujecki, 

(21). (2017). doi: 10.1007/s11082-016-0827-0.

	(9) C.R. Petersen, U.V. MØller, I. Kubat, B. Zhou, S. Dupont, J. Ramsay, T Benson, S. Sujecki, N. Abdel-Moneim, Z. Tang, D. Furniss, A. Seddon, and O. Bang, 

	(10) C.R. Petersen, N. Prtljaga, M. Farries, J. Ward, B. Napier, G.R. Lloyd, J. Nallala, N. Stone, and O. Bang, 

	(11) S. Junaid, J. Tomko, M.P. Semtsiv, J. Kischkat, W.T. Masselink, C. Pedersen, and P. Tidemand-Lichtenberg, 

	(13) F Penaranda, V Naranjo, R Verdu-Monedero, GR Lloyd, J Nallala, and N. Stone, 

	(14) T. Konevskikh, R. Lukacs, R. Blümel, A. Ponossova, and A. Kohlera 

, 235-257 (2016). Doi: 10.1039/C5FD00171D.

	(15) M. Hermes, R.B. Morrish, L. Huot, L. Meng, S. Junaid, J. Tomko, G.R. Lloyd, W. T. Masselink, P. Tidemand-Lichtenberg, C. Pedersen, F. Palombo, and N. Stone,

	(16) Y.P. Tseng, P. Bouzy, C. Pedersen, N. Stone, and P. Tidemand-Lichtenberg, Biomed.

 holds the position of professor of Biomedical Imaging and Biosensing and NHS Consultant Clinical Scientist at the University of Exeter. He recently led the Department of Physics and Astronomy for three years after holding the role of director of research. Stone has worked to pioneer the field of novel optical diagnostics within the clinical environment, moving from the NHS (Gloucestershire Hospitals), after almost 20 years of working closely at the clinical–academic–commercial interface to pull through novel technologies to be used where they have most clinical need. He is an internationally recognized leader in biomedical applications of vibrational spectroscopy (Raman and infrared).

	Stone graduated with a BSc (Hons) from Bath University in Applied Physics with Industrial Training in 1992. Since then he has undertaken numerous studies which include an MSc with Distinction from Heriot-Watt and St Andrews Universities in Laser Engineering with Applications; an MSc with Distinction in Applied Radiation Physics with Medical Physics at Birmingham University; a PhD in the application of Raman spectroscopy for cancer diagnostics at Cranfield University and an MBA (Health Executive) at Keele University.

	Stone has received numerous awards for his research both personally and within his research group. He recently won the International Raman Award for Most Innovative Technological Development 2014 and was the runner up in the 2013 NHS Innovation Challenge Prize. He won the Chief Scientific Officer's National R&D Award for 2009. He has published over 150 papers and book chapters.

Articles

Nick Stone of the University of Exeter explains why mid-infrared (mid-IR) spectroscopy is so valuable for spectral imaging in disease research and clinical diagnostics, and discusses his own recent work in this area.

Moving Mid-IR Spectroscopy Forward in Medicine

Interest in data-independent acquisition (DIA) in mass spectrometry is growing in many fields, including in metabolomics. The approach makes it possible to re-interrogate data from earlier analyses, such as when a researcher wishes to determine if newly identified compounds appeared in a sample analyzed previously. In this interview, Craig Wheelock, the head of the Integrative Molecular Phenotyping Laboratory at the Karolinska Institute (Sweden), discusses the use of DIA in metabolomics. 

Q. What aspects of metabolomics is your group focusing on?

 The analytical work in our group has two distinct focuses. We initially developed targeted methods for metabolic profiling of lipid mediators, such as eicosanoids in respiratory disease. A few years ago, we naively decided to expand into metabolomics, but quickly realized that there are numerous analytical challenges in performing a metabolomics experiment. Therefore, we decided to start from scratch, and construct our own workflow, similar to many other laboratories. The primary aim of our method was to acquire as much high quality data in a single experiment as possible.

	The goal was to then use these metabolite data for omics integration models in systems medicine studies. We wanted a long list of accurately identified metabolites, while simultaneously reducing the use of putative identifications. We were concerned that incorrect metabolite identifications would be problematic for our integrative modeling efforts, and lead to inaccurate interpretation of the observed biology. The last few years have been spent on developing our metabolomics method, which was finally published last year (1). Our focus on metabolite annotation has led us to be strong proponents of the efforts of the Metabolomics Standards Initiative to provide clear criteria on the accuracy of metabolite identification (2).

Q. What are the main challenges for chromatographers engaged in metabolomics at the moment?

 We are currently using liquid chromatography (LC) for our metabolomics work in biofluids, primarily urine and blood. When developing an LC method for metabolomics, it is clearly a trade-off between separation and throughput. Ideally, we would have three hour long gradients using two-dimensional (2D) nanoflow methods. This would optimize our metabolite separation, which would greatly help in unequivocal metabolite identification. Run times of this length are not feasible for our applications, and there is a question of the long-term robustness and retention time stability of such an approach. As with most methods, our current approach represents a compromise. The field appears to have currently coalesced around 15–20 min gradients. However, there is a strong case to be made for short, fast methods, in the order of 2–3 min, using microbore technologies. An additional obstacle is the variability associated with hydrophilic-interaction chromatography (HILIC). It has been challenging to obtain reproducible retention times for HILIC chromatography, especially for larger studies. However, while this has historically been an issue, new HILIC columns have greatly improved in performance. We have successfully used HILIC chromatography for the analysis of more than one thousand samples in a single study. In many ways, gas chromatography (GC) offers several advantages over LC approaches, including extremely consistent retention times, robustness, reduced ion suppression, and excellent spectral libraries. Conversely, analysis of samples using GC generally requires a derivatization step, and the compounds need to be thermally stable. In order to capture the structural diversity of a metabolome, multiple methods are necessary. There is, unfortunately, no single analytical platform capable of simultaneously acquiring a metabolome. I therefore see the current optimal solution to involve multiple methods, including both LC– and GC– based approaches.

Q. You recently developed a method using liquid chromatography high resolution mass spectrometry (LC-HRMS) for metabolite identification using a data-independent acquisition (DIA) approach? What was the aim of this research?

 The aim of this research (1,3) was to develop a method suitable for our laboratory goals, as I noted previously. We wanted to be able to accurately identify as many metabolites as possible in a single acquisition. Using DIA provided us with fragmentation from three different collision energies (0, 10 eV, and 30 eV), which greatly helps with compound identification.

	We are able to identify metabolites based upon our in-house library of standards. However, the DIA approach enables the simultaneous acquisition of targeted and nontargeted data. We can then go through our nontargeted data, and mine it for interesting metabolites that are not included in our database. If we want to conclusively identify a new metabolite, we can acquire the standard, add it to our database, and reprocess the data to characterize the compound. One of the primary challenges for us with the analysis of the nontargeted data is the ability to perform metabolite identification and annotation across thousands of samples in a single study.

Q. What is DIA? What benefits does it offer?

 To improve metabolite identification, and reduce the requirement for multiple analytical runs for structural confirmation, two different tandem mass spectrometry (MS/MS) strategies have been implemented: with selection of the precursor ion (data-dependent acquisition [DDA]), and without selection of the precursor ion (DIA). DIA-based MS generates MS/MS spectra containing a mixed population of product ions together with their precursor ions, and the extracted ion chromatogram (EIC) of each product ion needs to be mapped to its parent compound.

	DIA approaches have been successfully used to conduct multiple fragmentation experiments in a single acquisition. One useful application of the DIA approach is to identify coeluting isobaric compounds, where DIA data are combined with software deconvolution algorithms that merge precursor ions from low-energy experiments and product ions from high-energy experiments. An advantage of the current DIA approach is the concurrent collection of full scan data, enabling identification of metabolites not included in the database. Our data acquisition strategy enables a simultaneous mixture of database-dependent targeted and nontargeted metabolomics in combination with improved accuracy in metabolite identification, increasing the quality of the biological information acquired in a metabolomics experiment.

Q. What were the main analytical challenges you had to overcome in this project, and how did you overcome them?

 A lot of the work performed for this study was time-consuming, and rather repetitive. The characterization of more than four hundred standards to create the in-house library and measure the ion ratios took a very long time. The development of custom standard libraries is not efficient, and it does not make sense for laboratories to do this independently. Happily, there are now multiple metabolite libraries available that have the advantage of being well organized, and all metabolites have known solubilities.

 The aim of our work was to establish a comprehensive analytical workflow for the application of LC–HRMS to nontargeted metabolomics with a high level of accuracy in metabolite identification. Our application of DIA mode includes three sequential full scans, at 0, 10 eV, and 30 eV collision energies. In the subsequent data analysis, EIC from any precursor or associated product ions of interest can be extracted from the low- or high-energy scan data. One EIC is chosen for relative quantification (the quantifier ion) of the metabolite, and further product ions from the same compound are used as qualifier ions. This approach enables us to potentially distinguish coeluting isobaric pairs (provided that unique fragments can be identified). In addition, we have added an ion ratio approach, which means that the ratios of qualifier–quantifier ion intensities are established from analytical standards, and should therefore be preserved when measured in a biological sample, increasing the accuracy of the identification. The same acquired data (0 eV) can be used in parallel for a global metabolite profiling workflow, enabling a combined database-dependent targeted and nontargeted metabolomics experiment. The combination of the DIA-based data acquisition with the ion ratio confirmation and deconvoluted coeluting isobaric pairs provides a useful method for increasing the accuracy of metabolite identification in a metabolomics experiment.

Q. Can you describe a practical example to illustrate how this approach would benefit the analyst in practice?

 The strength of the current method was demonstrated in urine, using the homoserine and threonine isobaric pair as an example. We analyzed a clinical cohort of urine samples from asthmatics using our published metabolomics methods. The homoserine and threonine peak areas were then each integrated separately using their specific product ion as a quantifier ion (homoserine, 

 74.0248 at 10 eV), and their combined peak area (homoserine + threonine, 

118.0509 at 0 eV) was also integrated. The samples were then stratified by the abundance of the threonine value (obtained from threonine-specific peak integration) and the homoserine value (obtained from homoserine-specific peak integration). Following stratification, the 25% and 75% quantile of each sample set were selected in an extreme value approach to test for significance. The combined homoserine + threonine integrated peak was not significant (p = 0.2), but the threonine-specific peak was significantly different between the 25% and 75% quantile (p < 0.0001). In addition, the percentage relative standard deviations (RSDs) decreased in the homoserine- and threonine-specific integrated peaks between the two quantiles relative to the combined peak integration, increasing the precision of the measurement. This metabolite-dependent deconvolution example demonstrates the strength of the current method for increasing the accuracy of metabolite annotation by targeted ion selection, which can have a significant effect upon the observed biological shifts.

Q. What is the aim of the Metabolomics Standards Initiative?

 The field of metabolomics is growing rapidly, and it is exciting to see the recognition of the importance of metabolomics in integrative omics studies. However, with this recognition comes a need for increased standardization in the field. The work by the Metabolomics Standards Initiative is an important step in this direction. In addition, repositories, such as MetaboLights (4), are vital for the science. I think that the field would benefit from a set of standardized recommendations on how to perform a metabolomics experiment. An example can be taken from the development of microarray methods. It is now considered as essentially obligatory that the Minimum Information About a Microarray Experiment (MIAME) (5,6) standards are followed in an experiment, and that the data are deposited in a database such as GEO. It would be preferable if journal editors and reviewers insisted that metabolomics experiments follow this same format.

	Another helpful movement in metabolomics is the need for defined reference materials. The National Institute of Standards and Technology (NIST) group has been active in developing these materials, characterizing them in ring trials, and making them available to the research community. This can be quite helpful in benchmarking methods against a known metabolic profile in a well-described biofluid. Defined parameters for metabolite identification, public availability of acquired datasets, and clear experimental descriptions will help to expand the field and the utility of metabolomics as an informative platform for understanding metabolism.

 We are currently expanding upon the metabolomics method we published last year (1). One of the primary aims is to automate the method as much as possible. We have, therefore, formatted the sample preparation for both urine and plasma using an automated liquid handling platform. We can then prepare all samples in 96-well plates for analysis, which reduces both sample handling and preparation time. As part of these efforts, we are focusing on fully annotating the observed urinary metabolome with our current methodology. These efforts include evaluating shifts in observed metabolites associated with glucuronidase and urease treatment, as well as concentration and fractionation steps.

	We envision developing a series of methods enabling us to capture different metabolic fractions of the urinary metabolome depending upon the biological question. One of the weaknesses of metabolomics is that it is less sensitive compared to targeted approaches. There are multiple metabolites, such as lipid mediators and halogenated tyrosine derivatives, that are of extreme interest in respiratory biology, but cannot be detected by our metabolomics methods, because of their low endogenous concentration. We would like to develop a concentration and clean-up step that would provide us with a urinary metabolomics platform to detect these low abundant compounds. The presence of high abundant metabolites, such as creatinine, will be a challenge. We are also working on developing high-throughput metabolomics applications. There are multiple challenges associated with these efforts, but there is the potential to offer extremely fast analysis times. We are not there yet, but the field is making major advances, and it will be exciting to see where we are in another 5–10 years.

	(1) S. Naz, H. Gallart-Ayala, S.N. Reinke, C. Mathon, R. Blankley, R. Chaleckis, and C.E. Wheelock, 

	(3) R. Chaleckis, S. Naz, I. Meister, and C.E. Wheelock, 

	(6) A. Brazma, P. Hingamp, J. Quackenbush et al., 

(4), 365–371 (2001). doi:10.1038/ng1201-365.

 is head of the Integrative Molecular Phenotyping Laboratory at the Karolinska Institute (Sweden) and an associate professor within the Centre for Allergy Research (CfA). He is also a distinguished visiting professor of metabolomics at the Gunma Institute for Advanced Research (GIAR) at Gunma University (Japan). He is currently a member of the European Respiratory Society Scientific Events Working Group and board member of the International Metabolomics Society.

Data-independent acquisition (DIA) makes it possible to re-interrogate data from earlier analyses to determine if new compounds
have appeared in a sample previously analyzed. In this interview, Craig Wheelock of the Karolinska Institute discusses the use of DIA
in metabolomics.

The Benefits of Data-Independent Acquisition in Metabolomics

In recent years, more and more academic research has been conducted on the use of Raman spectroscopy for biomedical applications. Effectively moving the technique from the research laboratory to the clinic, however, requires overcoming a number of challenges, ranging from dealing with fluorescence interference to automating the process enough to make it easy for medical staff to implement. Juergen Popp and his team at the Leibniz Institute of Photonic Technology and the Friedrich-Schiller University, in Jena, Germany, have been working to address these challenges, with the aim of making Raman spectroscopy a powerful, comprehensive tool for cell biology and clinical studies. He recently spoke to 

What are the main challenges associated with using Raman spectroscopy in biomedical applications? What specific benefits can the technique offer?

	Raman spectroscopy has become one of the most academically researched analytical methods in the last decade. Raman spectra depict molecular vibrations that are characteristic of the molecular system under investigation. In other words, Raman spectra can be seen as a kind of "molecular fingerprint." Raman spectroscopy enables label-free detection of the molecular composition of complex samples with little or no sample preparation. The major challenge in the application of Raman spectroscopy to investigate biomedical samples, such as cells and tissue, is the possible occurrence of an interfering fluorescence background and the low signal yields, compared to, for example, fluorescence spectroscopy.

	However, decisive technical advances have been made to overcome these limitations in recent years, particularly through the realization of highly efficient spectrometers and filters, and the implementation of novel detector technologies. These advances have led to a significant reduction in the recording times of Raman spectra for biological samples. Further challenges toward a routine application of Raman spectroscopy for biomedical analysis are the automation of the complete process chain, that is, from sampling to measurement toward an automated analysis of the Raman spectra. The last point is of great importance, because the differences in the Raman spectra of different biological cells or tissue types are very small and invisible to the naked eye, and therefore require research on tailored statistical evaluation methods (chemometric models). In the last few years, my research team and I have successfully addressed all of these challenges.

You recently developed a simple and fast Raman spectroscopy method to identify antimicrobial susceptibilities and determine minimal inhibitory concentration (1). Can you tell us more about the aims of this project and the benefits of the approach you devised?

	Infectious diseases are on the rise globally, and threaten the health of people all over the world. Using currently approved microbiological methods, the causative pathogens and their potential antibiotic resistance can usually only be unambiguously identified when it is too late to treat them accurately. This is why broad-spectrum antibiotics are often used. However, this type of therapy leads to more and more pathogens becoming resistant to one or more antibiotics. The challenges for the healthcare system, as well as for society and the economy, are enormous. The fight against infections in Germany, for example, costs taxpayers billions of euros every year. To successfully treat an infection, doctors need information about the type of pathogen and its antibiotic resistance, and they need this information as quickly as possible. Microbiological standard procedures normally take more than 24 h because of the cultivation period required. With the chip-based Raman microspectroscopic solution we have developed, this time can be reduced to between 2 to 3.5 h (1). In this scenario, doctors are in a position to find the optimal therapy for their patients.

	My team, together with the team of Prof. Dr. Ute Neugebauer, also from the Leibniz Institute of Photonic Technology, developed an approach for the automated identification of single bacteria. This approach also allows the characterization of a possible (phenotypic) antibiotic resistance, as well as the determination of the minimum inhibitory concentration (MIC) that is necessary to kill infectious agents. The unique selling point of this chip-based Raman approach is a dramatic reduction in the diagnosis time. The germs can be unambiguously identified, and their resistance properties determined in only 2–3.5 h with a very small number of microbial pathogens (a few hundred bacteria are sufficient). Based on the information obtained, the physician can tailor the patient therapy to the specific pathogen.

	We had to automate the approach to such an extent that it could be applied directly in a hospital; this is the basis for the development of chip-based Raman microscopy. The combination of a Raman apparatus and an optical microscope allows for a label free, noninvasive, culture-free, optical and spectroscopic characterization, and the identification of bacteria down to the single-cell level. Each molecule generates an individual signature in the Raman spectrum, resulting in a specific molecular fingerprint for each bacterium. Statistical evaluation algorithms enable automated classification and identification of bacteria and antibiotic resistance.

	Additionally, the integration of chip-based enrichment methods in miniaturized microfluidic structures covers the entire process chain from sampling to the final result. The uniqueness of the approach lies in its culture independence and the scalability of the approach.

You also developed a high-throughput screening Raman spectroscopy (HTS-RS) platform for rapid and label-free macromolecular fingerprinting of tens of thousands of eukaryotic cells. (2). Why is the analysis of eukaryotic cells important?

	In single cell research, Raman spectroscopy has proven itself in many applications, including the identification of circulating tumor cells, the characterization of drug–cell interactions, and the differentiation of apoptotic and necrotic cells. So far, however, many Raman studies have been performed on a limited number of cells, that is, in the order of 100 cells, which does not completely cover the statistical variability of cell experiments. In contrast, for typical cytometric experiments based on scattering or fluorescence, 10,000 cells for each sample can be analyzed. By fully automating the process chain-that is, cell detection, Raman spectrum recording, and chemometric analysis for online cell classification, in multiwell plates or microfluidic environments-this limitation can be overcome. Such an automated Raman platform enables the measurement of a large number of cells, for example 10,000, within 30 min without sample preparation (2).

What benefits does this HTS-RS method offer and what is novel about your approach?

	The advantage of Raman spectroscopy over other traditional staining-based methods, such as fluorescence microscopy, is that no sample preparation is required, and living cells can be studied without perturbing the native cellular environment. Moreover, different types of analysis can be performed on the same device without any particular changes. Here, we specifically exploit these advantages of Raman spectroscopy for the analysis and characterization of single cells, by developing a device that can differentiate apoptosis and necrosis, identify circulating tumor cells, and evaluate the cell cycle. This extensive variety of applications opens new avenues for Raman spectroscopy as a convenient "go-to" tool in single-cell research and clinical investigations.

	Traditionally, researchers in the field of Raman spectroscopy acquire Raman images to characterize individual eukaryotic cells, or acquire individual Raman spectra from subcellular regions in cells. While the former results in a very long acquisition time and low throughput, the latter results in very high spectral variation as a result of the significant intracellular variation, which is commonly higher than interclass differences between cells. To overcome this problem, my team, led by Dr. Iwan Schie, has developed the new high-throughput screening Raman spectroscopy (HTS-RS) platform, which allows the acquisition of Raman spectra from the entire cell at once, overcoming the problem of intracellular heterogeneity. By combining this approach with a self-developed, fully automated acquisition and data processing strategy, we can significantly increase the throughput of the Raman measurements. These improvements help move Raman spectroscopy from being a niche technology to a powerful all-around tool for cell biology and clinical studies.

What other applications would HTS-RS be useful for?

	Because Raman spectroscopy can be used with living cells without affecting them, it allows the investigation of cells in their native environment, opening novel and intriguing research opportunities in single cell research. Now, individual cells can be tracked over an extended period of time to investigate changes of the molecular makeup as a result of environmental effects at any timepoint. Drug-induced effects on an individual cell can be monitored for a better understanding of the pathways, onset of action, and other pharmacodynamic parameters. New therapeutic methods, such as stem-cell therapies, will require the injection of enriched and purified cells into the human body, and in this type of therapy efficacy largely depends on the purity of the injected cells. Fluorescence-based methods are at a clear disadvantage, because they require labeling of cells for differentiation. Raman spectroscopy can perform the identification and the enrichment of these cells label-free.

What other projects have you been working on using Raman spectroscopy for biomedical applications? What do you regard as the most innovative developments in this area of research?

	Our other projects focus on researching Raman approaches to solve current challenges in clinical pathology, such as the determination of the tumor type and grade and intraoperative delineation of tumor margins (3–8). Reliable cancer diagnosis is a complex process and requires a number of diagnostic approaches, including endoscopy, ultrasound, and multidetector computed tomography.

	However, the current "gold standard" for making a definitive diagnosis is the histopathological examination, that is, the microscopic analysis of specially stained tissue biopsies. For a fast and safe in vivo or near in vivo intraoperative diagnosis, however, new methods and approaches are urgently needed. In this context, we explored the potential of Raman spectroscopy to address these challenges through the implementation of optical molecular pathology. We introduced novel Raman fiber probes for in vivo tissue screening to differentiate between healthy and malignant tissue (3). However, scanning relatively large areas is quite difficult because of the relatively long acquisition times of Raman spectroscopy.

	The combination of Raman spectroscopy with fast imaging techniques such as optical coherence tomography (OCT) or fluorescence lifetime imaging (FLIM) can circumvent this problem (4). In addition, we work on multicontrast imaging with nonlinear imaging techniques exhibiting similar imaging times. In particular, we focus on multimodal nonlinear coherent anti-Stokes Raman scattering (CARS), two-photon excited fluorescence (TPEF), and second harmonic generation (SHG) imaging with the potential for a reliable intraoperative tumor margin detection. Here, we developed a compact and easy-to-use CARS–SHG–TPEF multimodal nonlinear microscope combined with a novel fiber laser source for use in clinics (5). The application of this microscope, combined with advanced image processing algorithms, offers great potential to overcome current limitations of frozen section analysis with respect to achieving a reliable intraoperative tumor margin detection (6–8).

	(1) J. Kirchhoff, U. Glaser, J.A. Bohnert, M.W. Pletz, J. Popp, and U. Neugebauer, 

	(2) I.W. Schie, J. Rüger, A.S. Mondol, A. Ramoji, U. Neugebauer, C. Krafft, and J. Popp, 

	(3) C. Matthäus, S. Dochow, G. Bergner, A. Lattermann, B.F. Romeike, E.T. Marple, C. Krafft, B. Dietzek, B.R. Brehm, and J. Popp, 

	(4) S. Dochow, D. Ma, I. Latka, T. Bocklitz, B. Hartl, J. Bec, H. Fatakdawala, E. Marple, K. Urmey, S. Wachsmann-Hogiu, M. Schmitt, L. Marcu, and J. Popp, 

	(5) T. Meyer, M. Chemnitz, M. Baumgartl, T. Gottschall, T. Pascher, C. Matthaus, B.F.M. Romeike, B.R. Brehm, J. Limpert, A. Tunnermann, M. Schmitt, B. Dietzek, and J. Popp, 

	(6) S. Heuke, O. Chernavskaia, T. Bocklitz, F.B. Legesse, T. Meyer, D. Akimov, O. Dirsch, G. Ernst, F. von Eggeling, I. Petersen, O. Guntinas-Lichius, M. Schmitt, and J. Popp, 

	(7) T. Bocklitz, F.S. Salah, N. Vogler, S. Heuke, O. Chernavskaia, C. Schmidt, M. Waldner, F.R. Greten, R. Bräuer, M. Schmitt, A. Stallmach, I. Petersen, and J. Popp, 

	(8) O. Chernavskaia, S. Heuke, M. Vieth, O. Friedrich, S. Schuermann, R. Atreya, A. Stallmach, M. F. Neurath, M. Waldner, I. Petersen, M. Schmitt, T. Bocklitz, and J. Popp, 

Juergen Popp studied chemistry at the universities of Erlangen and Würzburg. After completing his PhD in chemistry, he joined Yale University to do postdoctoral work. He then returned to Würzburg University, where he finished his habilitation in 2002. Since 2002, he has held a chair for physical chemistry at the Friedrich-Schiller University in Jena, Germany. Since 2006, he has also been the scientific director of the Leibniz Institute of Photonic Technology in Jena, Germany. His research interests are mainly concerned with biophotonics. In particular, he focuses on the development and application of innovative Raman techniques for biomedical diagnosis.

He has published more than 800 journal papers, and has been named as an inventor on 12 patents. He is the Editor-in-Chief of the 

. In 2012, he received an honorary doctoral degree from Babes-Bolyai University in Cluj-Napoca, Romania. Professor Popp is the recipient of the 2013 Robert Kellner Lecture Award and the 2016 Pittsburgh Spectroscopy Award. In 2016, he was elected to the American Institute for Medical and Biological Engineering (AIMBE) College of Fellows. In 2018, Popp was awarded the renowned Johannes Marcus Marci Medal of the Czechoslovak Spectroscopy Society. He also won the third prize of the Berthold Leibinger Innovationspreis and was also awarded the "Kaiser-Friedrich-Forschungspreis".

Raman spectroscopy is promising some dramatic breakthroughs in biomedical applications. Juergen Popp and his team are determined to realize that promise, by working to make the technique a powerful tool for cell biology and clinical studies.

The Rise of Raman Spectroscopy in Biomedicine

LCGC spoke to Rudolf Krska from the University of Natural Resources and Life Sciences in Vienna, Austria about the latest analytical techniques, and challenges, facing analysts involved in the evolving field of mycotoxin analysis.

Q. Mycotoxin analysis is a major field in food analysis at the moment. Why has mycotoxin analysis become important?

 The occurrence of mycotoxins (secondary fungal metabolites) in various crops is a global concern, because it has significant implications for food and feed safety, food security, and international trade. Despite huge research investments, prevention and control of these toxic secondary metabolites remains difficult, and the agriculture and food industries continue to be vulnerable to problems of contamination. In addition, extreme weather conditions because of climate change are increasingly affecting the mycotoxin map in Europe and worldwide. The EU's Rapid Alert System for Food and Feed (RASFF) (EC, 2016) showed that of the total border rejections in 2015, 18.3% were due to mycotoxin contamination exceeding the EU legislative limits, accounting for the most frequently reported chemical hazard (1).

	In recent years, research on cumulative risks, exposure, and long-term effects has raised awareness for the control of these health risks. As a result of the potential danger of mycotoxins to humans and livestock, strict regulatory controls determine the sale and use of contaminated food and feeds. Thus, grain and other foodstuff buyers increasingly demand more rigorous and timely food safety testing. Failure to achieve a satisfactory performance may lead to unacceptable consignments being accepted or satisfactory batches being unnecessarily rejected. Recent reports have also demonstrated that the range of mycotoxins present in the food and feed chains goes beyond the list of regulated toxins. In fact, some 140 different fungal metabolites have been found in feed and feed ingredients by the use of a multi-toxin method based on liquid chromatography coupled to tandem mass spectrometry (LC–MS/MS). As a result of such findings, the demand for quantitative and rapid screening tools for the determination of (multiple) mycotoxins has increased tremendously.

Q. What are the biggest challenges in mycotoxin analysis? 

 The biggest challenge in mycotoxin analysis is still the sampling issue. Despite recent available guidance (2), it is still a difficult and tedious task to obtain a representative sample. Appropriate extraction solvents matching the range of multiple mycotoxins to be determined is another crucial step followed by proper cleanup. The latter is dependent on the final determination step. The use of LC–MS/MS instruments, combined with optimized chromatographic separation, reduces the need for sample cleanup. This is especially true for regulated toxins, for which fully 

C labeled internal standards can be used to compensate for matrix effects. Ensuring comparability of measurement results is another challenge, particularly for mycotoxin-commodity combinations for which no certified reference materials exist.

Q. What novel approaches have your group contributed to the field of mycotoxin analysis, and what advantages do they offer?

 Within the frame of international efforts to obtain new and comprehensive data on the occurrence of multiple mycotoxins in food and feed chains, we have developed a unique multi-analyte approach based on LC–MS/MS. The developed analytical method, which does not require any cleanup, has continuously been extended and fully validated for the quantification of more than 300 fungal and bacterial metabolites, including all regulated mycotoxins in various cereals, food, and feed matrices. This highly cited mass spectrometric method is crucial, both to elucidate the occurrence of various mycotoxins potentially being present in agricultural and food commodities and to study their metabolization by plants, animals, and humans. Our mass spectrometric work has also included the development of a fast, easy to handle, and highly accurate isotope dilution mass spectrometric assay.

	The latter has become feasible through the production of fully isotopically labeled mycotoxins by means of a patented technology developed in our laboratory, in cooperation with Romer Labs. We have also provided the first scientific proof of the occurrence of glucose-bound (masked) deoxynivalenol in naturally contaminated wheat and maize, as part of the plant's defense strategy to detoxify this most prevalent Fusarium mycotoxin. The potential threat to consumer safety from masked mycotoxins has, as a result, been addressed by institutions such as ILSI Europe (Washington, D.C., USA) and the European Food Safety Authority (EFSA) (Parma, Italy). Utilizing the power of the advanced multi-biomarker LC–MS/MS method, we have also been able to provide new insights into the human metabolism of Fusarium mycotoxins by identifying and quantifying appropriate biomarkers in human urine.

	Within the frame of the EU-funded project MYCOSPEC, we have developed a mid-infrared spectroscopic sensing method using tunable quantum cascade lasers and thin-film waveguides obtained from the University of Ulm (Ulm, Germany). This novel method showed its great potential for the rapid on-site classification of contaminated wheat, maize, and peanuts at the concentration levels established in the EU for deoxynivalenol and aflatoxin B1, respectively.

Q. You are coordinator of an EU project www.mytoolbox.eu dealing with integrated management strategies to tackle mycotoxins. Can you tell us more about your role in this project and what this project involves?

 In fact, there is still a pressing need to mobilize the wealth of knowledge from the international mycotoxin research conducted over the past 25–30 years, and to perform cutting edge research where knowledge gaps still exist. We believe that this knowledge needs to be integrated into affordable and practical tools for farmers and food processors along the chain to reduce the risk of mycotoxin contamination of crops, feed, and food. This is the mission of MyToolBox, a four-year project that has received funding from the European Commission, and which I have the pleasure to coordinate. It mobilizes a multi-actor partnership of academia, farmers, technology SMEs, food industry, and policy stakeholders to develop novel interventions aimed at achieving a significant reduction in crop losses caused by mycotoxin contamination. Besides a "field to fork" approach, MyToolBox also considers safe use options of contaminated batches, such as the efficient production of biofuels. Within a range of novel preharvest interventions, we will investigate the genetic resistance to fungal infection, cultural control, the use of novel biopesticides suitable for organic farming, competitive biocontrol treatment, and the development of novel modeling approaches to predict mycotoxin contamination. Research into postharvest measures includes real-time monitoring during storage, innovative sorting of crops using hyperspectral vision technology, novel milling technology, and the study of the effects of baking on mycotoxins at an industrial scale. Again, our multi-toxin screening method will be crucial to verify the success of all these intervention strategies.

Q. What areas of mycotoxin analysis will your group focus on next?

 In the past few years, our team has continuously moved from the target analysis of individual mycotoxins to untargeted metabolite profiling and metabolomics of, ideally, all secondary metabolites that are involved in plant-fungi interactions. This methodology is based on in vivo stable isotopic 

C-labeling and subsequent measurement of biological samples by full scan high-resolution LC–MS.

	We plan to continue and expand our interdisciplinary and comprehensive strategy to study plant-fungi interactions and the metabolism of mycotoxins by moving our research to the next level, a fully integrated "omics-based" approach. In this context, we intend to pursue our efforts to characterize the analytically ascertainable metabolome of wheat and maize genotypes differing in their Fusarium resistance level, and ideally to link their metabolite profile to resistance criteria or markers. A major goal within this endeavor is the development of a standardized metabolomics platform to study primary and secondary metabolites produced by microorganisms and plants, and to understand the interactions between plants, fungi, mycotoxins, and other secondary metabolites at a molecular level.

Q. Does your group focus on other areas of food analysis using chromatography?

 In the area of food and feed safety, my colleague, Professor Rainer Schuhmacher, and I aim for the fingerprinting of food and feed samples. The development of standardized conditions is a prerequisite for the development and establishment of robust fingerprinting methods. For the description of defined conditions, again in vivo 

C-labeling of the matrix-for example, grains-is a promising approach, which can be achieved through growing plants under 

 atmosphere. We also intend to combine this approach with the labeling of tracers; for example, different food contaminants. This would facilitate the recognition of changes of the labeled contaminant itself, but also of the tracers or contaminants on the matrix as a result of, for example, food or feed processing. Moreover, the in vivo 

C-labeling of the food or feed matrix would enable endogenous and exogenous compounds to be differentiated between. Stable isotopic labeling can be used to detect deviations of secondary metabolites of fungi, plants, and bacteria from normal patterns, flagging suspicious samples for further analysis and confirmation, and for a more accurate quantification and identification of compounds.

Q. Do you think the time will come when there will be no need for the chromatography component in food analysis?

 In view of the amazing sensitivity and high resolution achievable with novel mass spectrometry, this is certainly a valid question. Nonetheless, chromatography will probably stay forever until, or unless, someone develops a radically different approach to separate complex mixtures. With the advent of small particles and ultrahigh-pressure LC (UHPLC), we can now process smaller amounts of samples faster than ever. And with the wealth of potential compounds, which we aim to quantify in our food and feed chain in highly complex matrices, separation remains as important as ever.

	(1) European Commission (EC), 2016. The Rapid Alert System for Food and Feed 2015 Annual Report. Available at: 

 is full professor of bioanalytics and organic trace analysis at the University of Natural Resources and Life Sciences, Vienna (BOKU), Austria. He is head of the Centre for Analytical Chemistry at the Department of Agrobiotechnology (IFA-Tulln) at BOKU, with a staff of over 50. From 2010–2015, he also served as head of the department. As a member of JECFA (Joint Expert Committee for Food Additives) of the FAO/WHO, he has evaluated the impact of trichothecene mycotoxins on humans. Krska was also member of the Working Group Fusarium of the European Food Safety Authority. He has received 10 scientific awards and is co-author of more than 280 SCI publications (h-index: 49). Currently, Professor Krska is coordinator of the European Commission funded project MyToolBox (Safe Food and Feed through an Integrated ToolBox for Mycotoxin Management), with 23 partners and a funding volume of more than 5 million Euro. Direct correspondence to: 

A discussion of challenges in mycotoxin analysis, a new multi-analyte method that does not require sample cleanup, and how analytical chemistry can help reduce contamination by these toxins along the food chain.

Alasdair Matheson

Articles

Moving Mid-IR Spectroscopy Forward in Medicine

The Benefits of Data-Independent Acquisition in Metabolomics

The Rise of Raman Spectroscopy in Biomedicine

Measuring Mycotoxins