Multimodal MSI in Conjunction with Broad ... - ACS Publications

Subscriber access provided by READING UNIV

Technical Note

Multimodal MSI in conjunction with broad coverage spatially resolved MS2 increases confidence in both molecular identification and localization Dusan Velickovic, Rosalie K. Chu, Alyssa A. Carrell, Mathew Thomas, Ljiljana Paša-Toli#, David J. Weston, and Christopher R. Anderton Anal. Chem., Just Accepted Manuscript • DOI: 10.1021/acs.analchem.7b04319 • Publication Date (Web): 06 Dec 2017 Downloaded from http://pubs.acs.org on December 11, 2017

Just Accepted “Just Accepted” manuscripts have been peer-reviewed and accepted for publication. They are posted online prior to technical editing, formatting for publication and author proofing. The American Chemical Society provides “Just Accepted” as a free service to the research community to expedite the dissemination of scientific material as soon as possible after acceptance. “Just Accepted” manuscripts appear in full in PDF format accompanied by an HTML abstract. “Just Accepted” manuscripts have been fully peer reviewed, but should not be considered the official version of record. They are accessible to all readers and citable by the Digital Object Identifier (DOI®). “Just Accepted” is an optional service offered to authors. Therefore, the “Just Accepted” Web site may not include all articles that will be published in the journal. After a manuscript is technically edited and formatted, it will be removed from the “Just Accepted” Web site and published as an ASAP article. Note that technical editing may introduce minor changes to the manuscript text and/or graphics which could affect content, and all legal disclaimers and ethical guidelines that apply to the journal pertain. ACS cannot be held responsible for errors or consequences arising from the use of information contained in these “Just Accepted” manuscripts.

Analytical Chemistry is published by the American Chemical Society. 1155 Sixteenth Street N.W., Washington, DC 20036 Published by American Chemical Society. Copyright © American Chemical Society. However, no copyright claim is made to original U.S. Government works, or works produced by employees of any Commonwealth realm Crown government in the course of their duties.

Page 1 of 8 1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60

Analytical Chemistry

Multimodal MSI in conjunction with broad coverage spatially resolved MS2 increases confidence in both molecular identification and localization Dušan Veličković1, Rosalie K. Chu1, Alyssa A. Carrell3, Mathew Thomas2, Ljiljana Paša-Tolić1, David J. Weston3, Christopher R. Anderton1* 1

Environmental Molecular Sciences Laboratory and 2Physical Sciences Division, Pacific Northwest National Laboratory, Richland, Washington 99352 3 Biosciences Division, Oak Ridge National Laboratory, Oak Ridge, Tennessee 37830 ABSTRACT: One critical aspect of mass spectrometry imaging (MSI) is the need to confidently identify detected analytes. While orthogonal tandem MS (e.g., LC-MS2) experiments from sample extracts can assist in annotating ions, the spatial information about these molecules is lost. Accordingly, this could cause mislead conclusions, especially in cases where isobaric species exhibit different distributions within a sample. In this Technical Note, we employed a multimodal imaging approach, using matrix assisted laser desorption/ionization (MALDI)-MSI and liquid extraction surface analysis (LESA)-MS2I, to confidently annotate and localize a broad range of metabolites involved in a tripartite symbiosis system of moss, cyanobacteria, and fungus. We found that the combination of these two imaging modalities generated very congruent ion images, providing the link between highly accurate structural information offered by LESA and high spatial resolution attainable by MALDI. These results demonstrate how this combined methodology could be very useful in differentiating metabolite routes in complex systems.

INTRODUCTION The technique of mass spectrometry imaging (MSI) has grown into a powerful tool for understanding biological systems, due in part to its ability to provide spatial information about molecules (from small metabolites to proteins) and their relative abundance within a sample.1 A significant limitation in most MSI measurements is the capacity to attain molecular annotations with high confidence. The use of high resolution mass analyzers in MSI workflows, like that of a Fourier transform ion-cyclotron resonance mass spectrometer (FTICR-MS), can provide the necessary mass resolving power and mass accuracy to enable more precise assignments of a detected analyte’s molecular formula.1,2 However, the ability to confidently identify ions within a sample remains a significant challenge when using MSI as a standalone methodology.3 Standard chromatography-based omics methods, in conjunction with tandem MS modes (e.g., LC-MS2) of homogenates and/or extracts from the same sample can afford the required orthogonal measurement to confidently identify an ion of interest.4,5 Yet, there still remains a blind-spot in implementation of these types of multi-technique approaches. Specifically, in the ability to spatially resolve mass isomers, which would provide both improved confidence in the identity of a molecule and its distribution within a sample. Matrix-assisted laser desorption/ionization (MALDI) is the most commonly employed MSI technique for molecular imaging of complex organic samples.6,7 This ‘soft-ionization’ method can provide broad molecular coverage with relatively high lateral resolution (< 2 µm).8 To date, there continues to be efforts in developing MALDI tandem MS imaging (MS2I) modalities.3,9,10 However, a major setback that has resulted in

the scarcity of these endeavors is the fast depletion of a parent signal during an imaging run, which effectively limits MS2 measurements to targeted molecules or the most abundant signals in the MS. Liquid extraction surface analysis (LESA)-MS is an example of a technique capable of providing spatial molecular information of a sample, but that does not suffer from limited analyte concentration issues during probing like MALDI. The sensitivity of LESA-MS is through a combination of efficient analyte extraction and an increase in the relative area probed (i.e., low lateral resolution).11 Further, the desorption and ionization steps are decoupled in LESA-MS, which enables a constant flow of parent molecules of interest into mass analyzer over an extended period of time and, thus, can provide an efficient route for performing MS2 of a broad range of ions within a single probing area. There are previously reported studies that have used MALDI and LESA in combination to address complex analytical and biological questions.12,13 These multimodal MS approaches can complement each other by leveraging the strengths of each technique. For example, Quanico et al. used the increased sensitivity of LESA to get a complete identification of proteins imaged by MALD-MSI.12 As of yet, there have been no reports on employing LESA-MS2 to provide both spatial and molecular insights into MALDI-MSI data of sample. In this Technical Note, we exploit LESA-MS2I to assist in confident molecular annotations and their localization within previously MALDI-MS imaged sample. Here, we used a model peatland system as our sample that is composed of cocultures of a peat moss (Sphagnum fallax), a cyanobacterium (Nostoc muscorum) and a fungus (Trizodia spp.) cultivated on

ACS Paragon Plus Environment

Analytical Chemistry 1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60

a solid agar medium. Given that three different biological entities are interacting in this tripartite system, there is an increased probability of isomeric metabolites existing with different localizations. Consequently, this spatial information would be missed using both conventional bulk-based MS2 approaches in combination with MALDI-MSI or with targeted MALDI-MS2I approaches that do not provide broad MS2 coverage of detected ions. Both of which could lead to misannotating specific ions and their spatial distribution. Our results demonstrate that the high lateral resolution ion images acquired from MALDI-MSI can be more confidently annotated by LESA-MS2I, which provides the identity and spatial localization of the detected ions. Specifically, results of resolving spatially distinct structural isomers across this tripartite interaction demonstrates the value of this approach. EXPERIMENTAL SECTION Tripartite culturing. Nostoc muscorum UTEX 1037 was cultivated in a 250 mL flask with a final volume of 100 mL at 24 °C in BG-110 medium at pH 8.214 and shaken at 125 rpm with a 16 hr/8 hr (day/night) cycle at 150 PAR for 21 d. Trizodia acrobia M277 was cultured on potato dextrose agar (PDA) (Sigma) plates for 21 d at 24 °C in the dark. Axenic Sphagnum fallax cultures were maintained on Knop’s medium15 at pH 5.7 with a 16 hr/8 hr (day/night) cycle at 150 PAR. For MSI experiments, 100 mg (fresh weight) of Nostoc muscroum UTEX 1037, 2 mm plug of Trizodia m277 with agar removed, and/or 2 cm of the upper most portion of an axenic Sphagnum fallax individuals were used to inoculate each plate for individual or symbiotic tripartite interaction assays. For tripartite interactions, each organism was positioned approximately 2 cm from one another and for individual controls, a single organism was placed in the center of the plate. All samples to be analyzed by MSI were grown on ~1.5 mm thick BG-110 (pH 5.5) 1.5% agar plates at 24 °C with a 16 hr/8 hr (day/night) cycle at 150 PAR. The petri dishes were sealed with Parafilm to minimize dehydration of thin agar during incubation and incubated for 12 d. MALDI-MS Imaging. After the incubation period, agar areas of the tripartite interaction and of isolated culture controls were excised from petri dishes and placed together onto an indium tin oxide (ITO)-coated glass slide (Bruker Daltonics; Fig1A). Note, the moss excessively protruded from the surface of the agar, so plant tissue was carefully removed prior to dehydration to assist in MS analyses. Samples were then dried at 40 oC,16 after which MALDI matrix was applied using an HTX TM-Sprayer (HTX Technologies).17 Universal MALDI matrix (1:1 DHB:CHCA), 20 mg/mL in 70% MeOH,18 was applied using the following spraying conditions: 12 passes with track spacing 3 mm, flow rate 0.1 mL/min, spray velocity 1200 mm/min, spray pressure of 10 psi (N2), and a 40 mm sprayer nozzle distance from the sample. Imaging was performed on a 15 Tesla MALDI-FTICR-MS (SolariX, Bruker Daltonics) equipped with SmartBeam II laser source (355 nm, 2 kHz) in positive mode (m/z 92- m/z 500), using 200 shots at 2kHz and a 200 µm step size. FTICR-MS was operated to collect m/z 92-1000, using a 209 ms transient, which translated to a mass resolution of R ~ 130,000 at m/z 400. Data was acquired using FlexImaging (v 4.1, Bruker Daltonics), and image processing and visualization were performed using SCiLS Lab. LESA-MS2 Imaging. The MALDI-imaged sample was then analyzed via LESA-MS using a Triversa Nanomate (Advion) coupled to Velos Orbitrap MS (Thermo Scientific) oper-

ated in positive ion mode. LESA sampling conditions were as follows:

Figure 1. Correlative MALDI-MS and LESA-MS images of model peatland cultures. (A) Overview of agar regions that were excised and mounted on an ITO slide, and then were dehydrated. Note, that moss is removed before dehydration. (B) Ion image of m/z 136.0618, identified as adenine. (C) Ion image of m/z 232.1543, identified as butyryl-carnitine. (D) Ion image of m/z 206.0459, identified as choline sulfate.

8.0 µL of 70% MeOH was aspirated in the pipette tip, then 0.3 µL of this volume was dispensed at a height of 0.2 mm above the agar. After a 1 s post-dispense delay (extraction) time, the volume was re-aspirated into the pipette tip and then infused into the MS via nano-ESI. LESA sampling occurred serially with a 3 mm step size between probing areas. Each sample extraction location was sprayed into the MS for 4 min for iterative MS and top 6 MS2 analyses. As such, an overall MS scan was collected and then another every seventh scan thereafter. For the interim six MS scans, CID fragmentation was performed on the top six peaks detected (using a new top six every cycle). LESA-MS data was processed by mMass software, and LESA MS and LESA MS/MS images were created using both in-house-developed MSI visualization tool (MSIquickview)19 and SCiLS Lab. RESULTS AND DISCUSSION High performance MALDI-MS, using a 15T FTICR-MS, enabled us to map and assign molecular formulas for 125 out of 204 ions (after deisotoping) distributed across the tripartite community and isolated cultures grown on agar. However, because of natural isomers present in both plant and microbial systems, molecular annotations of only a few of these ions could be made with fairly high confidence upon consulting the KEGG database. Thus, in an attempt to obtain more confident identification of these metabolites we performed LESA-MS2I on MALDI-imaged samples. By matching the (highly) accurate mass data generated by both MS imaging modalities (within 10 ppm divergence), we found that 87 out of the total 204 ions had similar spatial distributions in the MALDIFTICR-MS and LESA-MS images. Of these, LESA-MS2 information was attained for 49 ions, and 33 of them we were able to annotate based on high accurate mass and LESA-MS2 data acquired in the same LESA-MS imaging experiment (refer to Supporting Figures S1-S5 for further information). Although LESA-MSI could be performed on fresh agar plates


Page 2 of 8

Page 3 of 8 1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60

Analytical Chemistry with inoculated cultures, here we chose to acquire LESA data on previously MALDI-imaged samples for several reasons.

First, agar dehydration becomes a significant factor during these long image

Figure 2. Illustrating the benefits of multimodal imaging with MALDI-MS + LESA MS2 in revealing metabolic pathways. (A) Distribution of m/z 189.1598 ion. This ion could correspond to two mass isomers: N6,N6,N6-trimethyllysine and 7,8-diaminononanoate, whose structures and CID fragmentation patterns are presented the center panel. The red and black circles in the ion image denote LESA pixels where the MS2 data were different (as seen in the bottom panel, red and black spectra, respectively). The numbers inside the red and black circles below the structural formulas show the MetFrag scores from matching experimental MS2 data with theoretical ones. (B) Distribution of biotine amide around cyanobacteria in tripartite system. Matching of experimental MS2 data of biotin amide with a theoretical one by MetFrag (center panel) together with biosynthesis of biotin amide from 7,8-diaminononanoate (pathway noted in the bottom panel). (C) Distribution of carnitine molecule. Matching of experimental MS2 data of carnitine with a theoretical one by MetFrag (center panel) together with biosynthesis of carnitine from trymethyllysine (pathway noted in the bottom panel). scans, and can change both the concentration of analytes being extracted and the sampling conditions (e.g., tip-to-sample height must be recalibrated). Second, serial sampling is extremely

difficult, as one would have to produce the same tripartite interaction on two different agar plates, and given that interaction area is quite small, LESA/MALDI image correlation could be problematic. Lastly, we wanted to enhance the chance of observing the same molecules in the LESA measurements as the in MALDI analysis. Considering these three factors, the LESA experiments served to provide validation of ion localization and relative abundances, and, more importantly, more confidence in the identification of metabolites in MALDI imaged areas. An example of our general workflow is shown in Figure 1, along with ion maps of a few metabolites annotated using the KEGG database. In every example, there was complete congruence in MALDI and LESA MS images comparing metabolite abundance and distribution. For example, adenine (Figure 1B) is found colocalized with the cyanobacteria (in both the isolated and tripartite culture), which is expected as this nucleobase is important for maintaining energy balance during nitrogen fixation process.20 Butyryl-carnitine, described as a signaling molecule in plant-fungus symbiosis,21 was almost exclusively mapped around the fungi in isolation and the tri-

partite interaction (Figure 1C). Finally, choline-sulfate, a plant osmoprotectant,22 showed colocalization with plant part of the tripartite community. However, many other MALDI-mapped ions cannot be unambiguously identified according to output results of KEGG database. For example, m/z 215.0163 can be ascribed to 14 compounds with common formula C6H8O7 (Supporting Figure S6). The intense fragment at m/z 183 in LESA tandem MS spectrum, however, reveals that this molecule is most likely diketogulonate, a product of enzymatic pentose interconversion.23 Interestingly, this molecule has been found dominantly in the isolated fungus culture, while during interaction with plant and cyanobacteria it seems that some of the enzymatic processes that regulate its biosynthesis are inhibited. The benefit of LESA imaging modality on MALDI imaged sample could be further visualized in Figure 2. The ion m/z 189.1598 is identified as [C9H20N2O2+H]+, showing distribution around cyanobacteria and fungus in the tripartite interaction. This molecular formula can be ascribed to two previously identified natural compounds: 7,8-diaminononanoate and trimethyllysine, metabolites with related, but distinct biological roles in fatty acid metabolism (Figure 2A).24,25 In particular, trimethyllysine is a precursor of carnitine biosynthesis, a molecule that serves as a fatty acid shuttle during lipid catabolism,24 while 7,8-diaminononanoate is part of the biotin bio-



synthetic route, an essential carbon dioxide carrier cofactor for fatty acid synthesis.25 Our results show that LESA-MS2 spectrum in the pixels around cyanobacteria is quite different than the MS2 spectrum from the fungus region, implying that both

molecules are participants of tripartite interaction, however with obvious spatial segregation. Although the predicted fragmentation of these

Figure 3. Spatial-structural relationship of disaccharides revealed by MALDI-MSI + LESA-MS2I. (A) Distribution of total disaccharides (m/z 365.1052) throughout tripartite system as revealed by MALDI- and LESA-MSI. (B) Spatial segmentation of disaccharide LESAMS2I spectra reveals different complexity in disaccharide composition among tripartite members. (C) Imaging of characteristic disaccharides fragments by LESA-MS2I reveals information about linkage and anomer distribution throughout sample. (D) CID fragmentation of different disaccharides structures based upon reference 26.

two mass isomers is quite similar and they cannot be significantly resolved in the fungus region, MetFrag shows that in cyanobacteria this ion is most likely 7,8- diaminononanoate (Fig 2A). If we go further and take a look into the downstream metabolic outcome of two compounds, we can see how the prediction from MetFrag was consistent with corresponding ion images. Predicted 7,8-diaminononanoate in cyanobacteria colocalizes with biotin amide (Figure 2B), while carnitine distribution around fungus in tripartite culture (Figure 2C) could be a good indication that indeed trimethyllysine was used as carnitine biosynthetic substrate. These findings also impose that biotin amide biosynthesis is intensified in cyanobacteria during the tripartite interaction, as both diaminononanoate and biotin amide were not imaged in isolated cyanobacteria culture. Furthermore, high abundance of carnitine in noninteracting fungus imply that its biosynthetic route could vary depending on the microenviroment surrounding the fungus given that trimethyllysine was not observed in the isolated fungus culture. While the functional significance of these differences remains to be elucidated, this example highlights the great analytical advantage of using the correlative LESA-MS2I modality, which can provide spatially resolved MS2 of a number of ions within a single pixel, to draw more insightful conclusions from MALDI-MSI data. Imaging of MS2 discriminate ions could additionally resolve or at least hint at the exact biological nature of isobaric parent ion. Disaccharides ([M+Na]+, m/z 365.1052) are mapped in every member of the tripartite community (Figure 3A). However, spectral segmentation regarding six main fragment ions in CID spectra of [dissacharide+Na]+- m/z 185, 203, 245, 305, 347, and 365 -26 reveals that there is at least five different disaccharide families distributed across the tripartite system and the isolated controls (Figure 3B). Isolated fungus shows the highest variability in disaccharides distribution profile, while cyanobacteria is the most uniform in the appearance of pro-

duced disaccharides. Furthermore, it seems that profile of synthesized disaccharides substantially changes during cyanobacteria-fungus communication, since cyanobacteria produces the additional type of disaccharide, whereas the fungus disaccharide profile becomes simpler during the tripartite interaction. Although it is very reasonable to assume that in single pixel there is a mixture of different disaccharides, some general conclusions about the linkage-distribution relationship can be derived from discriminant peaks and/or ratios between fragments (Figure 3C).26 In particular, the general absence of m/z 305 fragment in the fungus and moss indicate that they are dominant in non-reducing sugars like trehalose or sucrose, since this peak corresponds to cleavage of [C2H4O2] from the reducing end. At the same time, the presence of m/z 245 in cyanobacteria, but also in some fungus correlated pixels, indicates the presence of 1-6 linked disaccharides in these areas. Finally, the high relative ratio of 203/185 ions in the fungus and moss shows that they are rich in α-bonded sugars, and conversely the very low ratio of these fragments in cyanobacteria implies more β-anomers in its disaccharides profile. Finally, in some cases, even highly accurate m/z results and matching the isotopic pattern cannot resolve the molecular formula of imaged ions (Figure 4). For example, ion m/z 221.0422, localized around the fungus, can be ascribed either as [C6H14O6+K]+ or [C9H10O5+Na]+, within 1 ppm accuracy, which could correspond to hexitols or vanillylmandelic acid, respectively (Figure 4A). Both molecules are known to be metabolized by fungi and are involved in fungi-plant interactions,2,27 which additionally complicates ion image interpretation. Nonetheless, perfect matching in the distribution of this ion with m/z 205.0681, which corresponds to Na adduct of hexitols, can draw the wrong conclusion about vanillylmandelic acid absence in the imaged area (Figure 4B). However, LESA-MS2 data show that there are two discriminant vanillylmandelic acid fragment ions, m/z 83 and 175 (Figure 4C),


Page 4 of 8

Page 5 of 8 1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60

Analytical Chemistry mapped around the fungus region (Figure 4D), pointing out that the observed m/z 221.0422 parent ion reflects vanillylmandelic acid distribution, too. CONCLUSIONS This Technical Note exemplifies how the combination of two imaging modalities, MALDI-MSI and LESA-MS2I, enables resolving of isobaric species with high spatial resolution. Although we combined MALDI and LESA modalities, this could probably be done with any number of combination of tools (Nano-DESI + LESA for example). Note, whatever the combination might be, the higher lateral resolution method would ideally be performed first, as the lower-lateral resolution probing methodology might delocalize analytes or irreparably damage the sample. Our protocol was tested on a microbial-plant community system, providing us distinct biological organisms in isolated and interacting conditions to compare to, but this same principle can be applied to other biological samples. While the MS2 workflow presented here was untargeted, a targeted approach could be applied if high-confidence molecular mapping of specific species is necessary. Furthermore, the addition of pre-mass analysis ion mobility separation to MALDI or LESA could further elucidate structural isomers that MS2 cannot resolve alone.

Fig S4. Tentative annotation of MALDI and LESA imaged ions, Fig S5. Non-identified MALDI and LESA imaged ions, Fig S6. Annotation of MALDI ion based on LESA-MS2: example 215.0163 m/z.

AUTHOR INFORMATION Corresponding Author * E-mail: [email protected], +1 (509) 371-7970

Notes The authors declare no competing financial interest

ACKNOWLEDGMENT The authors would like to thank Andrew Palmer and Theodore Alexandrov of the European Molecular Biology Laboratory, as well as Dennis Trede and Jan Kobarg of SCiLS Lab, for their advice and help with processing the MSI data. This work was performed at EMSL, a national scientific user facility sponsored by the Office of Biological and Environmental Research (BER), U.S. Department of Energy (DOE). EMSL is located at PNNL, a multidisciplinary national laboratory operated by Battelle for the U.S. DOE. Preparation of the biological samples was performed at ORNL as part of US DOE-BER, Early Career Research Program. ORNL is managed by UT-Battelle, LLC, DE-AC0500OR22725.

REFERENCES

Figure 4. Multimodal MALDI-MS + LESA-MS2 imaging to determine the molecular formula of a MALDI imaged ion. (A) MALDI-MSI + LESA-MSI of ion m/z 221.0422. (B) MALDIMSI + LESA-MSI of ion m/z 205.0681. (C) Results of matching between experimental LESA-MS2 spectrum of 221.04 parent ion in fungus pixels and MetFrag predicted theoretical spectra for vanillylmandelic acid and hexitol. D) LESA-MS2I of characteristic fragments of vanillylmandelic acid.

ASSOCIATED CONTENT Supporting Information The Supporting Information is available free of charge on the ACS Publications website. Fig S1. Overlap of MALDI detected ions with LESA detected ions during imaging of tripartite system, Fig S2. Identified MALDI imaged ions based on LESA-MS2I information, Fig S3. Non-identified MALDI imaged ions with LESA-MS2 information,

(1) Dilillo, M.; Ait-Belkacem, R.; Esteve, C.; Pellegrini, D.; Nicolardi, S.; Costa, M.; Vannini, E.; Graaf, E. L.; Caleo, M.; McDonnell, L. A. Scientific reports 2017, 7, 603. (2) Patel, T. K.; Williamson, J. D. Trends in plant science 2016, 21, 486-497. (3) Hansen, R. L.; Lee, Y. J. J Am Soc Mass Spectrom 2017. (4) Boya, C. A.; Fernandez-Marin, H.; Mejia, L. C.; Spadafora, C.; Dorrestein, P. C.; Gutierrez, M. Scientific reports 2017, 7. (5) Sumner, L. W.; Amberg, A.; Barrett, D.; Beale, M. H.; Beger, R.; Daykin, C. A.; Fan, T. W.; Fiehn, O.; Goodacre, R.; Griffin, J. L.; Hankemeier, T.; Hardy, N.; Harnly, J.; Higashi, R.; Kopka, J.; Lane, A. N.; Lindon, J. C.; Marriott, P.; Nicholls, A. W.; Reily, M. D., et al. Metabolomics : Official journal of the Metabolomic Society 2007, 3, 211-221. (6) Palmer, A.; Trede, D.; Alexandrov, T. Metabolomics 2016, 12. (7) Velickovic, D.; Anderton, C. R. Rhizosphere 2017, 3, 254-258. (8) Khalil, S. M.; Pretzel, J.; Becker, K.; Spengler, B. Int J Mass Spectrom 2017, 416, 1-19. (9) Perdian, D. C.; Lee, Y. J. Anal Chem 2010, 82, 9393-9400. (10) Lunsford, K. A.; Peter, G. F.; Yost, R. A. Anal Chem 2011, 83, 6722-6730. (11) Eikel, D.; Vavrek, M.; Smith, S.; Bason, C.; Yeh, S.; Korfmacher, W. A.; Henion, J. D. Rapid Commun Mass Sp 2011, 25, 3587-3596. (12) Quanico, J.; Franck, J.; Dauly, C.; Strupat, K.; Dupuy, J.; Day, R.; Salzet, M.; Fournier, I.; Wisztorski, M. Journal of proteomics 2013, 79, 200-218. (13) Randall, E. C.; Race, A. M.; Cooper, H. J.; Bunch, J. Anal Chem 2016, 88, 8433-8440. (14) Rippka, R.; Deruelles, J.; Waterbury, J. B.; Herdman, M.; Stanier, R. Y. J Gen Microbiol 1979, 111, 1-61. (15) Reski, R.; Abel, W. O. Planta 1985, 165, 354-358. (16) Hoffmann, T.; Dorrestein, P. C. J Am Soc Mass Spectr 2015, 26, 1959-1962. (17) Anderton, C. R.; Chu, R. K.; Tolic, N.; Creissen, A.; PasaTolic, L. J Am Soc Mass Spectr 2016, 27, 556-559. (18) Yang, J. Y.; Phelan, V. V.; Simkovsky, R.; Watrous, J. D.; Trial, R. M.; Fleming, T. C.; Wenter, R.; Moore, B. S.; Golden, S. S.; Pogliano, K.; Dorrestein, P. C. J Bacteriol 2012, 194, 6023-6028.



(19) Lanekoff, I.; Heath, B. S.; Liyu, A.; Thomas, M.; Carson, J. P.; Laskin, J. Anal Chem 2012, 84, 8351-8356. (20) Privalle, L. S.; Burris, R. H. J Bacteriol 1983, 154, 351-355. (21) Laparre, J.; Malbreil, M.; Letisse, F.; Portais, J. C.; Roux, C.; Becard, G.; Puech-Pages, V. Mol Plant 2014, 7, 554-566. (22) Rivoal, J.; Hanson, A. D. Plant Physiol 1994, 106, 1187-1193. (23) Karkonen, A.; Dewhirst, R. A.; Mackay, C. L.; Fry, S. C. Arch Biochem Biophys 2017, 620, 12-22.

(24) Franken, J.; Burger, A.; Swiegers, J. H.; Bauer, F. F. Appl Microbiol Biot 2015, 99, 6377-6389. (25) Wang, M. L.; Moynie, L.; Harrison, P. J.; Kelly, V.; Piper, A.; Naismith, J. H.; Campopiano, D. J. Nat Chem Biol 2017, 13, 660-+. (26) Lee, S.; Valentine, S. J.; Reilly, J. P.; Clemmer, D. E. Int J Mass Spectrom 2012, 309, 161-167. (27) Fewson, C. A. FEMS microbiology reviews 1988, 4, 85-110.

TOC


Page 6 of 8

Page 7 of 8 1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60

Analytical Chemistry

Figure 1. Correlative MALDI-MS and LESA-MS images of model peatland cultures. (A) Overview of agar regions that were excised and mounted on an ITO slide, and then were de-hydrated. Note, that moss is removed before dehydration. (B) Ion image of m/z 136.0618, identified as adenine. (C) Ion image of m/z 232.1543, identified as butyryl-carnitine. (D) Ion image of m/z 206.0459, identified as choline sulfate. 84x68mm (300 x 300 DPI)



Figure 2. Illustrating the benefits of multimodal imaging with MALDI-MS + LESA MS2 in revealing metabolic pathways. (A) Distribution of m/z 189.1598 ion. This ion could correspond to two mass isomers: N6,N6,N6trimethyllysine and 7,8-diaminononanoate, whose structures and CID fragmentation patterns are presented the center panel. The red and black circles in the ion image denote LESA pixels where the MS2 data were different (as seen in the bottom panel, red and black spectra, respectively). The numbers inside the red and black circles below the structural formulas show the MetFrag scores from matching experimental MS2 data with theoretical ones. (B) Distribution of biotine amide around cyanobacteria in tripartite system. Matching of experimental MS2 data of biotin amide with a theoretical one by MetFrag (center panel) together with biosynthesis of biotin amide from 7,8-diaminononanoate (pathway noted in the bottom panel). (C) Distribution of carnitine molecule. Matching of experimental MS2 data of carnitine with a theoretical one by MetFrag (center panel) together with biosynthesis of carnitine from trymethyllysine (pathway noted in the bottom panel). 177x99mm (300 x 300 DPI)


Page 8 of 8

Multimodal MSI in Conjunction with Broad ... - ACS Publications

Recommend Documents