Molecular imaging by mass spectrometry: application to forensics

Tiffany Porta,^a Emmanuel Varesio,^b Thomas Kraemer^b and Gérard Hopfgartner^a

^aSchool of Pharmaceutical Sciences, University of Geneva, University of Lausanne, Life Sciences Mass Spectrometry, Quai Ernest-Ansermet 30, 1211 Geneva, Switzerland
^bDepartment of Forensic Pharmacology and Toxicology, Institute of Legal Medicine, University of Zurich, Building Y52, Winterthurerstrasse 190, 8057 Zurich, Switzerland

Introduction

Imaging mass spectrometry-based techniques are label-free and rely on the desorption of molecules present on the surface of a solid flat sample. The interest in molecular mass spectrometric imaging (MSI) has grown over the past few years in the field of forensic sciences, because it allows the conservation of sample spatial resolution. By recording the mass-to-charge ratio (m/z) signal of molecules of interest at each spatial location (i.e. pixel), a 2-D map is reconstituted as the image of their distribution within the sample.

This article focuses on matrix-assisted laser desorption/ionisation (MALDI) MS imaging and also illustrates desorption electrospray ionisation (DESI) and secondary ion mass spectrometry (SIMS) as alternative desorption/­ionisation techniques. Their main applications in forensics are described, as well as the advantages provided in terms of sample preparation over approaches routinely used in toxicological laboratories.

Technical considerations and general MSI performances

Matrix-assisted laser ­desorption/ionisation (MALDI)

Karas et al. introduced MALDI in its current form in the late 1980s.¹ It is the most widely used ionisation technique for MS imaging and can be ­operated either at atmospheric pressure or under vacuum conditions. MALDI is based on the homogeneous application of a matrix in excess on the top of the sample, resulting in the co-crystallisation of the matrix and analytes. Matrix coating is critical and must keep the analyte localisation intact within the sample. The matrix is usually a small organic compound (for example, alpha-cyano-4-hydroxycinnamic acid, sinapinic acid, 2,5-dihydroxybenzoic acid) that plays an important role in the ionisation process by absorbing the photon energy emitted by the laser (Figure 1). By using an appropriate matrix, MALDI-MS is suitable for characterising a wide range of compounds, including large (i.e. proteins, peptides) and small molecules (i.e. ­pharmaceuticals, lipids) at low femtomole to attomole levels.

Nowadays, most MS instruments can be used for MALDI MS imaging (see, for example, www.maldi-msi.org). Hybrid instruments such as triple quadrupole linear ion trap (QqQ_LIT), quadrupole-time-of-flight (QqToF), ToF/ToF, ion trap-ToF (QIT-ToF) or linear ion trap-orbitrap, offer several possibilities to perform MS and/or tandem mass spectrometry (MS/MS) imaging experiments. Accurate mass is attractive for selectivity, however, QqQ_LIT operated in selected ion monitoring (SRM) mode remains essential for sensitive and fast detection. When comparing MS instrument performance, several aspects have to be considered in terms of resolution, selectivity, speed of acquisition and sensitivity. In all cases, the spatial resolution is dependent on the dispersion and the size of matrix crystals and on the laser beam characteristics. Recent developments in MS instrumentation enable rapid data generation with high repetition rate lasers (i.e. frequency up to 1000 Hz) and focused laser beam to achieve a resolution in the order of 10–50 µm.

Desorption electrospray ionisation (DESI)

DESI is a recent ambient desorption ionisation technique, which has been extensively described by Cooks and co-workers since 2004.² Desorption and ionisation processes take place at atmospheric pressure, where charged droplets generated by a pneumatically-assisted electrospray interface are directed towards the flat sample at a certain angle. The spray extracts the molecules present on the surface and a second generation of charged droplets is produced and analysed by the mass spectrometer. The sample is directly inserted into the sample holder, maintained under ambient pressure and temperature conditions and does not require any sample preparation (Figure 1).

DESI-MS is suitable for the analysis of small and large molecules with moderate sensitivity compared to MALDI-MS, but allows the analysis of a broader range of analytes. However, due to the large surface area sampled by the electrospray beam, the imaging spatial ­resolution achieved with DESI is above several hundred µm.

Secondary ion mass spectrometry (SIMS)

SIMS was introduced in 1977 by Benninghoven³ and has been shown as a powerful technique for molecular surface analysis. The ionisation mechanism occurs under ultra-high vacuum where the surface of the sample is sputtered with a pulsed high-energy (5 keV to 25 keV) primary ion beam (for example, Ga⁺, In⁺, Au_n⁺, C₆₀⁺, Bi₃⁺ ions).⁴ The resulting secondary atomic or molecular ions are ejected and collected with high-voltage extraction lenses (Figure 1). The focused ion beam enables ultra-high resolution images for atomic ions and low molecular mass fragments but the relevant ion yield from pixels in the sub-1 µm region is very limited since less than 1% of the surface is removed during the analysis.

SIMS is suitable for elemental and low molecular weight compounds (i.e. up to 1000 Da) analysis. A significant lack of sensitivity of the mass range above m/z 1000 is observed due to in-source fragmentation of complex molecules.⁵ A matrix, as in MALDI, can be applied onto the sample to prevent extensive ­fragmentation and increase the ionisation efficiency (i.e. matrix-enhanced secondary ion mass spectrometry, ME-SIMS).⁶ As for DESI, no particular sample preparation is required.

Application to forensics

Single hair samples

Body fluids or hair can reveal recent drug of abuse intake, but only hair provides information about chronic consumption. Current hair analysis is mainly based on gas or liquid chromatography coupled to MS/(MS/MS). It can reveal the user's drug intake history but requires multiple sample preparation steps beforehand (for example, washing, pulverisation, overnight hair digestion, extraction, derivatisation). In addition, each hair sample has to be cut into several 1 cm-long segments, which are pulverised and analysed individually to reconstitute the drug consumption from scalp to hair tip (i.e. from present to past).

Recent studies have proposed imaging the distribution of illicit drugs in single hair samples by MALDI-MS(MS/MS), either longitudinally sectioned^7,8 or as intact hair.⁹ For longitudinal section analysis,⁷ the time required for imaging the hair sample in pixel mode was about 2 h with a high-resolution, high laser frequency ToF/ToF instrument. In the intact hair study, we described how to generate an image of the cocaine distribution over several months (i.e. using a hair sample of ca 10 cm) in less than 10 min. This was achieved with a QqQ_LIT instrument operated in SRM mode combining a 1000 Hz-frequency laser with a relatively high MALDI plate speed [i.e. rastering performed every 1 mm at 1 mm s^–1, Figure 2(a)].⁹ Qualitative and quantitative data were acquired to confirm the cocaine intake: (i) by using targeted MS/MS experiments to confirm the drug identity [Figure 2(b)] and (ii) with a relative quantitation from the scalp to the tip and a simultaneous determination of the benzoylecgonine (i.e. a cocaine metabolite) to cocaine ratio (BZE/COC) which must be greater than 0.05 to prevent false positive results due to external contamination [Figure 2(c)]. The method allowed monitoring the consumption from chronic users with a sensitivity of 5 ng mg^–1, which covers a wide range of real cases, typical cocaine concentrations in hair range from 0.5 (cut off from the Society of Hair Testing) to over 200 ng mg^–1.¹⁰ Similarly, the direct MALDI-ToF analysis of single intact hair (4 cm segments) containing 100 ng mg^–1 of cocaine was investigated¹¹ but did not provide MS imaging results due to a lack of sensitivity.

In addition to MALDI, DESI was used for the rapid screening of anabolic steroid esters in bovine-incurred hair for hormone and veterinary drug analysis and forensics.¹² However, the inefficient desorption ionisation and the low quantities of steroid esters present in the surface sampled did not give any satisfactory results and necessitated simplified ultrasonic liquid extraction. ToF-SIMS has also been successfully applied to provide molecular distribution of external chemicals such as hair dye ingredients in longitudinally-sectioned single hair.⁸

Tissue samples

Detection and quantitation of drugs in tissue usually require homogenisation, extraction and/or derivatisation of the analytes of interest and other optional steps to remove lipids or other interfering endogenous compounds prior to the MS analysis. As a consequence, the analyte spatial distribution is lost. For MS imaging, sample preparation for direct analysis of post-mortem tissue only consists of a few steps: (i) organ/tissue deep freezing, (ii) cutting of 10–20 µm-thickness slices, (iii) transfer of tissue slices onto a sample holder (for example, glass slide, stainless steel plate) and (iv) applying the matrix on the top when MALDI-MSI analyses are performed. This rapid ­protocol allowed the quantitation of cocaine in autopsied brain or kidney of human cocaine or morphine users.^13,14,15

Latent fingerprinting

Finally, another forensic domain where MS imaging techniques are emerging is the reconstitution of latent fingerprints (LFP). LFP can help to identify an individual and give potential evidence of prior contact with explosives or illicit substances to prove his/her connection with a crime scene. LFP can be reconstituted from different surfaces (for example, glass, paper, plastic) without any sample pre-treatment in its native environment using DESI¹⁶ or SIMS,^17,18 or even by applying a MALDI matrix.¹⁹ Figure 3(a) shows the DESI image distribution of cocaine on a LFP blotted onto a glass slide obtained with a spatial resolution of 150 µm × 150 µm. This resolution is sufficient to clearly distinguish the ridges and the minutiae that can be exported into a fingerprint recognition software [Figure 3(b)]. Another advantage of this technique is being able to distinguish between overlapping fingerprints belonging to different individuals, relying on the fact that they present distinct exposure to chemicals [Figures 3(c) and 3(d)]. The potential of MALDI to study latent finger marks of endogenous lipids (for example, oleic acid, stearic acid, cholesterol)¹⁹ and the ability of SIMS to detect exogenous contaminants (for example, amphetamine drugs, gunpowder residues, arsenic)^17,18 has also been demonstrated.

Conclusion

Selective and sensitive imaging mass spectrometry-based approaches are presented as very powerful tools for forensic investigations. The sample preparation prior to MS analyses is considerably reduced enabling an enhanced analytical throughput. The material consumption is very low and the spatial distribution of the drugs and their metabolites is kept intact within the samples, which can reveal decisive information in the case of latent fingerprints or hair analysis to monitor, for example, cocaine ­consumption.

The development of ambient mass spectrometry with and without a matrix is still growing with the emergence of techniques such as direct analysis in real-time (DART), plasma-assisted desorption ionisation (PADI) or extractive electrospray ionisation (EESI)²⁰ just to name a few. In forensic investigation, a spatial resolution of several hundred micrometres is sufficient to generate relevant data. However, direct ambient analysis suffers from the lack of a chromatographic step prior to MS detection, which is particularly critical for isobaric or phase II metabolites (in case of in-source fragmentation). In that respect, differential ion mobility spectrometry (DMS) offers an interesting alternative to chromatographic separation because it enables the separation of isobaric analytes and was recently successfully applied in combination with liquid extraction surface analysis (LESA) for the analysis of drugs in human tissues.¹⁵

References

1. M. Karas, D. Bachmann and F. Hillenkamp, “Influence of the ­wavelength in high-irradiance ultraviolet laser desorption mass spectrometry of organic molecules”, Anal. Chem. 57, 2935–2939 (1985). doi: 10.1021/ac00291a042
2. Z. Takats, J.M. Wiseman, B. Gologan and R.G. Cooks, “Mass spectrometry sampling under ambient conditions with desorption electrospray ionization”, Science 306, 471–473 (2004). doi: 10.1126/science.1104404
3. A. Benninghoven and W. Sichtermann, “Secondary ion mass spectrometery: A new analytical technique for biologically important compounds”, Org. Mass Spectrom. 12, 595–597 (1977). doi: 10.1002/oms.1210120914
4. B.K. Kaletas¸ I.M. van der Wiel, J. Stauber, L.J. Dekker, C. Güzel, J.M. Kros, T.M. Luider and R.M.A. Heeren, “Sample preparation issues for tissue imaging by imaging MS”, Proteomics 9, 2622–2633 (2009). doi: 10.1002/pmic.200800364
5. A.F. Altelaar, J. van Minnen, C.R. Jiménez, R.M.A. Heeren and S.R. Piersma, “Direct molecular imaging of Lymnaea stagnalis nervous tissue at subcellular spatial resolution by mass spectrometry”, Anal. Chem. 77, 735–741 (2005). doi: 10.1021/ac048329g
6. K.J. Wu and R.W. Odom, “Matrix-enhanced secondary ion mass spectrometry: a method for molecular analysis of solid surfaces”, Anal. Chem. 68, 873–882 (1996).
7. A. Miki, M. Katagi, T. Kamata, K. Zaitsu, M. Tatsuno, T. Nakanishi, H. Tsuchihashi, T. Takubo and K. Suzuki, “MALDI-TOF and MALDI-FTICR ­imaging mass spectrometry of methamphetamine incorporated into hair“, J. Mass Spectrom. 46, 411–416 (2011). doi: 10.1021/ac950717i
8. B.-J. Chen, P.-L. Lee, W.-Y. Chen, F.-D. Mai and Y.-C. Ling, “Hair dye ­distribution in human hair by ToF-SIMS”, Appl. Surf. Sci. 252, 6786–6788 (2006). doi: 10.1016/j.apsusc.2006.02.230
9. T. Porta, C. Grivet, T. Kraemer, E. Varesio and G. Hopfgartner “Single hair cocaine consumption monitoring by mass spectrometric imaging”, Anal Chem 83, 4266–4272 (2011). doi: 10.1021/ac200610c
10. P. Kintz, V. Cirimele, C. Sengler and P. Mangin, “Testing human hair and urine for anhydroecgonine methyl ester, a pyrolysis product of cocaine”, J. Anal. Toxicol. 19, 479–482 (1995).
11. S. Vogliardi, D. Favretto, G. Frison, S.D. Ferrara, R. Seraglia and P. Traldi, “A fast screening MALDI method for the detection of cocaine and its ­metabolites in hair”, J. Mass Spectrom. 44, 18–24 (2009). doi: 10.1002/jms.1463
12. M.W.F. Nielen, A.W.J.M. Nijrolder, H. Hooijerink, A.A.M. Stolker, “Feasibility of desorption electrospray ionization mass spectrometry for rapid ­screening of anabolic steroid esters in hair”, Anal. Chim. Acta 700, 63–69 (2011). doi: 10.1016/j.aca.2010.08.009
13. R.F. Reich, K. Cudzilo, J.A. Levisky and R.A. Yost, “Quantitative MALDI-MSⁿ analysis of cocaine in the ­autopsied brain of a human cocaine user employing a wide isolation window and internal standards”, J. Am. Soc. Mass Spectrom. 21, 564–571 (2010). doi: 10.1016/j.jasms.2009.12.014
14. T. Porta et al., in 58th ASMS Conference on Mass Spectrometry and Allied Topics (2010).
15. T. Porta et al., in 59th ASMS Conference on Mass Spectrometry and Allied Topics (2011).
16. D.R. Ifa, N.E. Manicke, A.L. Dill and R.G. Cooks, “Latent fingerprint ­chemical imaging by mass ­spectrometry”, Science 321, 805 (2008). doi: 10.1126/science.1157199
17. M.I. Szynkowska, K. Czerski, J. Rogowski, T. Paryjczak and A. Parczewski, “Detection of exogenous contaminants of fingerprints using ToF-SIMS”, Surf. Interface Anal. 42, 393–397 (2010). doi: 10.1002/sia.3193
18. M.I. Szynkowska, K. Czerski, J. Rogowski, T. Paryjczak and A. Parczewski, “ToF-SIMS application in the ­visualization and analysis of fingerprints after contact with amphetamine drugs”, Forensic Sci. Int. 184, e24–e26 (2009).
19. R. Wolstenholme, R. Bradshaw, M.R. Clench and S. Francese, “Study of latent fingermarks by matrix-assisted laser desorption/ionisation mass spectrometry imaging of endogenous lipids”, Rapid Commun. Mass Spectrom. 23, 3031–3039 (2009). doi: 10.1002/rcm.4218
20. F.M. Green, T.L. Salter, P. Stokes, I.S. Gilmore and G. O'Connor, “Ambient mass spectrometry: advances and applications in forensics”, Surf. Interface Anal. 42, 347–357 (2010). doi: 10.1002/sia.3131

​