Chapter 1. Introduction
1.6. Goals of the work
In this work, we were aiming to develop two analytical methods, one preserving spatial resolution, and the other one preserving temporal resolution.
The goal of the first study was to develop a new type of MALDI matrix, based on
hybrid nanopatricles, which would enable performing MALDI imaging of single cells at a high spatial resolution. The goal of the second study was to develop a method useful in the monitoring of
heterogeneous dynamic chemical systems while preserving temporal resolution. The method should accommodate optical as well as mass spectrometric detection.Chapter 2
Hybrid nanoparticles for mass spectrometric imaging of single cells
2.1...Introduction
As outlined in section 1.2, Matrix-assisted laser desorption/ionization (MALDI) is an analytical technique in which laser light is used to desorb and ionize molecules – previously co-crystallized with a chemical matrix – to enable mass spectrometric (MS) detection of the resulting gas-phase ions.1,2 One of the interesting features of this technique is the possibility of mapping chemical distributions of analytes in biological specimens. In fact, MALDI-MS has widely been used for mapping lipids,43-45 proteins,46-48 and small molecules49-51 in the samples such as tissues or single cells.17,20,23,52-54
Examples of powerful chemical matrices used in MALDI include α-cyano-4-hydroxycinnamic acid (CHCA), 2,5-dihydroxybenzoic acid (2,5-DHB), and 9-aminoacridine (9-AA). Before analysis, concentrated solution of a selected chemical matrix is applied onto the surface of the biological sample. The biggest nuisance in MALDI imaging is the heterogeneous crystallization of MALDI matrices on the sample surface, which considerably decreases lateral resolution of the resulting images, and often disables the possibility of performing single-cell studies. Homogeneous and reproducible application of MALDI matrices is critical for obtaining high-quality results using this technique.
Here we demonstrate a new type of hybrid inorganic-organic nanomaterial which enables in-situ delivery of a chemical matrix for mass spectrometric imaging with high lateral
resolution. Application of the hybrid matrix to biological specimens poses less threat to the experimenters and the environment since the toxic matrix compound does not need to be sprayed by using a gas-powered sprayer, which could lead to the contamination of the laboratory environment with toxic aerosols. The hybrid nanomaterial binds to the surface of the cells while still in the liquid phase, which is followed by a controlled release of the organic matrix molecules trigerred by alkaline vapors delivered to the specimen in a closed chamber.
2.2...Materials and methods 2.2.1...Materials
Acetyl coenzyme A, adenosine triphosphate, 9-aminoacridine (9-AA), ammonium hydroxide (33% solution), bradykinin acetate, guanosine triphosphate, tetraethoxysilane (TEOS), and uridine diphosphate glucose were purchased from Sigma-Aldrich (St. Louis, USA). 1,2-Dipalmitoyl-sn-glycero-3-[phospho-rac-1-glycerol] was purchased from Avanti Polar Lipids (Alabaster, USA). Ethanol was purchased from Echo Chemical (Miaoli, Taiwan).
Water (18.2 MΩ·cm at 25 °C) was obtained from a Milli-Q water purification system (Merck Millipore, Billerica, USA). Indium tin oxide (ITO) glass slides were purchased from Bruker Daltonics (Part No. 237001; Bremen, Germany). Aluminum plate (thickness: 200 μm) was obtained from a local supplier and cut into small pieces in house.
2.2.2...Synthesis of SiO
2/9-AA and SiO
2nanoparticles
In order to synthesize SiO2/9-AAnanoparticles, 54.4 mg of 9-AA and 97 µL of 33%
NH3(aq) solution were mixed with 15.25 mL 50% ethanol solution (Figure 2.1). The mixture was sonicated until almost all 9-AA crystals were dissolved, and saturation of the solution with 9-AA was reached. The mixture was further stirred for 30 min, and 117.3 µL of TEOS were subsequently added. Stirring continued for 5 min, and the suspension was then left to
settle for 12 h. The color of the resulting suspension was bright-yellow. The as-prepared SiO2/9-AA nanoparticle suspension was centrifuged (5500 rpm, 20 min). The SiO2/9-AA nanoparticles were recovered from the pellet, and they were rinsed with 50% ethanol. After that, they were resuspended in water and stored in the refrigerator at 4 C.
To synthesize SiO2 nanoparticles (without 9-AA), 3 mL of 30% ammonia solution were mixed with 7.63 mL 99.5% ethanol and 6.13 mL water. The mixture was further stirred for 30 min, and 117.3 µL of TEOS were added. Stirring continued for 5 min, and the product suspension was left to settle for 12 h. The as-prepared SiO2 nanoparticle suspension was centrifuged (5500 rpm, 20 min), followed by rinsing the nanoparticles with 50% ethanol. The SiO2 nanoparticles were also resuspended in water and stored in the refrigerator at 4
C.
Samples of SiO2/9-AA and SiO2 nanoparticles were analyzed using a scanning electron microscope (SEM; JEOL JSM-7401 F, Tokyo, Japan).
Figure 2.1 Synthesis of the hybrid SiO2/9-AAnanoparticles.
2.2.3...The controlled release of 9-AA from SiO
2/9-AA nanoparticles by alkali
In order to demonstrate the feasibility of the controlled release of the 9-AA MALDI matrix from the SiO2/9-AA nanoparticles, 200 µL (an equivalent of 0.5 mg mL-1) of SiO2/9-AA nanoparticle suspension in 50% ethanol solution was prepared. Subsequently, an aliquot of 10 µL 33% NH3(aq) was added, and the suspension was stirred for 5 min, and
centrifuged at 10000 rpm for 10 min. The supernatant was diluted 100× with 50% ethanol, and used as a sample to obtain the fluorescence spectra by FluoroMax-3 spectrofluorometer (Horiba Jobin Yvon, Edison, USA). The excitation wavelength was set to 400 nm, and the emission spectrum was recorded in the wavelength range of 410-600 nm.
In another experiment – which was designed to demonstrate the controlled release of MALDI matrix – an aliquot of 200 µL 33% NH3(aq)) was mixed with 200 µL (an equivalent of 5 mg mL-1) SiO2/9-AA suspension, and stirred for 30 min. An aliquot of 200 µL of 95.5%
ethanol was subsequently added to solubilize the 9-AA matrix released from the SiO2/9-AA nanoparticles. The suspension of the residue nanoparticles was then centrifuged at 10000 rpm for 10 min, and the pellet fraction containing nanoparticles (after 9-AA release) was re-suspended in 100 µL of pure water. The suspension was then spotted onto an SEM target, and imaged by SEM.
2.2.4...Preparation of algal cells for analysis
The protocol used to prepare single cells of Closterium acerosum (Carolina Biological Supply Company, Burlington, USA) for analysis is outlined in Figure 2.2. Initially, 0.5 mL of the cell suspension was mixed with 1.5 mL of water, and the resulting suspension was centrifuged at 2000 rpm for 5 min. The supernatant was removed, and the pellet was resuspended in 100 µL of water. A droplet of the resulting dilute suspension of cells was placed in a Petri dish, and several (~ 5) cells were picked up using a micropipette. Cells were washed several times by subsequent transfers into 2-µL droplets of fresh water. A small amount of the suspension of SiO2/9-AA nanoparticles(2 µL, 5 mg mL-1) was pipetted onto the cells, and the excess suspension of the nanoparticles was quickly removed by pipetting. The unbonded SiO2/9-AA nanoparticles were subsequently washed away with water. Finally, 2 µL of water were used to re-suspend the cells, and transfer them onto an aluminum plate. The optical and fluorescence images were captured using an upright fluorescence microscope
(Eclipse 80i; Nikon, Yokohama, Japan) fitted with a digital camera (DS-Ril; Nikon, Tokyo, Japan). Fluorescence images were obtained using the excitation filters: UV-2A (ex = 330-380 nm) and G-2A (ex = 510-560 nm)
Figure 2.2 Preparation of cells for mass spectrometric imaging.
2.2.5...Controlled release of 9-AA by exposure to ammonia vapors
In order to induce the controlled release of 9-AA from the SiO2/9-AA nanoparticles by exposure to ammonia vapors, a simple incubation system was constructed (Figure 2.3). A hole was drilled in the lid of a plastic Petri dish; one side of a silicon tube (length, 40 cm; ID, 2 mm, OD, 4 mm) was slid through that hole, and epoxy glue was used to seal the junction.
The other side of the silicon tube was passed through the rubber septum mounted on the top of a 20-mL glass vial acting as the ammonia-vapor generator. The vial was filled with ~15 mL 33% NH3(aq) solution and placed in a water bath set to 50 C. The release of ammonia vapors occurred instantly – as verified by inserting a wet pH-indicator strip into the Petri dish used as
the incubation chamber. An aluminum target with the sample (algal cells pre-coated with the SiO2/9-AA nanoparticles) was incubated inside the Petri dish containing ammonia vapors for 30 min. Parafilm was used to seal the slit in between the top and bottom parts of the Petri dish, thus to prevent possible escape of gaseous ammonia from the system.
Figure 2.3 Setup for the controlled release of 9-AA in the presence of gaseous ammonia.
2.2.6...MALDI-MS and MALDI imaging
For MALDI-MS detection and MALDI imaging experiments, we used the Autoflex III Smartbeam instrument (Bruker Daltonics) fitted with a solid-state laser (λ = 355 nm). The settings of this instrument were as follows: negative ion mode, ion source 1, -19 kV; ion source 2, -16.7 kV; lens, -9 kV; delay time, 0 ns. During the MALDI imaging routine, the laser beam was focused to 10 μm, the scan raster was set to 10-15 µm (in different experiments), and 35 laser shots were fired at every raster point at a frequency of 50 Hz. The mass range was set to 300-1100 Da, and a cut-off limit of 300 Da was applied. Data were acquired using the flexControl software (version 3.0; Bruker Daltonics). The MALDI imaging data were collected and viewed using the flexImaging software (version 2.0; Bruker Daltonics).
2.3...Results and discussion
2.3.1...Synthesis and initial testing of the hybrid nanoparticles
It is known that hybrid inorganic-organic materials can be synthesized using the sol-gel reaction system.55,56 For example, Laperriere et al.57 doped 9-AA into glass produced in the course of a sol-gel process. In the present work, we have made an attempt to synthesize SiO2/9-AA nanoparticles with the prospect of using them as a matrix suitable for high-spatial-resolution MALDI imaging (Figure 2.1). To achieve this goal, we substituted a fraction of the alkaline component (NH3(aq)) of the sol-gel synthesis process with 9-AA. This yielded a suspension of hybrid nanoparticles with the inorganic framework of SiO2, and the organic filling of 9-AA. Figure 2.4 presents scanning electron micrographs of the resulting hybrid SiO2/9-AA nanoparticles as well as SiO2 nanoparticles (without 9-AA). On the nanoscopic level, the nanoparticles loaded with 9-AA resemble those without 9-AA, however the hybrid nanomaterial has yellow color due to the presence of the 9-AA ―cargo‖. Since the active MALDI matrix compound (9-AA) is embedded within the inorganic structure of SiO2, at this stage, the as-prepared hybrid nanomaterial does not yet fulfill the function of MALDI matrix: MALDI-MS detection of four standard compounds mixed with this nanomaterial gave poor results (Figures 2.5A and 2.5C). However, following the release of 9-AA in a chamber saturated with gaseous ammonia (Figure 2.3), the recorded MS signals were very high (Figures 2.5B and 2.5D). We have also noticed that – unlike most conventional MALDI matrices – the SiO2/9-AA nanoparticles form an extremely homogeneous layer on the MALDI target, without almost any signs of heterogeneous crystallization (Figure 2.6). This can facilitate application of the matrix by simple incubation of the hybrid nanoparticles suspended in an aqueous solution – compatible with biological specimens.
Figure 2.4 Scanning electron micrographs of the hybrid SiO2/9-AA nanoparticles (A and B) as well as single-component SiO2 nanoparticles (C and D). (A) SiO2/9-AA nanoparticles (as synthesized). (B) SiO2 nanoparticles (without 9-AA) synthesized using 3 mL 33% NH3(aq). (C) SiO2 nanoparticles (without 9-AA) synthesized using 4 mL 33% NH3(aq).
2.3.2...Evaluation of the controlled matrix release process
The controlled release process was further studied in a series of experiments: in one of them, we suspended the SiO2/9-AA nanoparticles in 50% ethanol, added 33% NH3(aq), and stirred the resulting suspension for 5 min, centrifuged, and measured the fluorescence of the supernatant (Figure 2.7A). Fluorescence intensity of the supernatant decreased in the subsequent washing steps using NH3(aq) as the washing solvent. In another test, pure water was used instead of NH3(aq) as the washing solvent; in this case, fluorescence intensity of the supernatant did not decrease as much as when using NH3(aq) (Figure 2.7B). In addition, after the washing with NH3(aq), the supernatant had yellow color (Figure 2.7C). After three
Figure 2.5 The effect of ammonia vapors (NH3(g)) on signal intensity and signal-to-noise (S/N) ratio in MALDI-MS analysis of a chemical standard solution (1 µL) containing adenosine triphosphate (ATP), guanosine triphosphate (GTP), uridine diphosphate glucose (UDP-Glc) and 1,2-dipalmitoyl-sn-glycero-3-[phospho-rac-α-glycerol] (PG) using SiO2/9-AA nanoparticles as matrix.
The sample/nanoparticle deposits in (A) and (C) were not incubated with NH3(g) while the sample/nanoparticle deposits in (B) and (D) were incubated with NH3(g). Analyte concentrations in (A and B): ATP, GTP, and UDP-Glc – each 2.5 × 10-5 M; PG – 3.6 × 10-5 M. Analyte concentrations in (C and D):
ATP, GTP, and UDP-Glc – each 2.5 × 10-6 M; PG – 3.6 × 10-6 M. Peak identities: ATP, m/z 506.0; GTP m/z 522; UDP-Glc, m/z 565; PG, m/z 721.5.
Figure 2.6 Optical micrographs of dry deposits of different samples on an aluminum plate used as MALDI target: (A) 2 µL of 1:1 (v/v) mixture of 9 mg mL-1 9-AA solution in acetone and 10-5 M adenosine triphosphate in water; (B) 2 µL 10-5 M solution of adenosine triphosphate was allowed to dry, followed by deposition of 0.5 µL suspension of the SiO2/9-AA nanoparticles (5 mg mL-1 in water). (C) Micrograph showing the edge of the SiO2/9-AA nanoparticle deposit on the aluminum plate (same as in (B)). Hybrid SiO2/9-AA nanoparticles provide unprecedented homogeneity of the matrix deposit on the microscopic level (B and C), as compared with the heterogeneous crystalline deposit of 9-AA (A). Scale bars: 200 µm.
consecutive wash steps with NH3(aq), the supernatant was clear, and the amount of pellet was less than in the control series without NH3(aq); this is due to the fact that some of the SiO2 material got dissolved in the alkaline solution (cf. scanning electron micrograph in Figure 2.8). These results prove that the structure of the hybrid SiO2/9-AA nanoparticles is degraded in the alkaline environment (NH3(aq)), and this degradation process is accompanied by the release of 9-AA.
2.3.3...Application in the imaging of single cells
Equipped with the controlled release hybrid SiO2/9-AA nanoparticles, we further aimed to apply this material as a matrix in the mass spectrometric imaging of microscopic specimens such as single cells. As a model sample, we selected unicellular algae – Closterium acerosum.
Figure 2.7 Controlled release of 9-AA in the liquid phase. (A and B) Fluorescence spectra (ex = 400 nm) of supernatants collected from the SiO2/9-AA nanoparticle suspensions. (A) Leaching 9-AA from the SiO2/9-AA nanoparticles by addition of 10 µL of 33% NH3(aq) to 200 µL of the 0.5 mg mL-1 nanoparticle suspension in 50% ethanol. The pH of the resulting suspension was 14. (B) Leaching 9-AA from the SiO2/9-AA nanoparticles with 10 µL of pure water. The black lines in (A) and (B) correspond to the supernatant from the 1st washing step, and the red and blue lines correspond to the supernatants obtained after the 2nd and the 3rd washing step, respectively. (C) Photographs of 0.6-mL microcentrifuge tubes containing SiO2/9-AA nanoparticles – after washing with NH3(aq)/water, and centrifugation (10000 rpm, 10 min). Note that the pellets became pale after the treatment with NH3(aq) (C), which indicates the 9-AA had been leached from the SiO2/9-AA nanoparticles.
Figure 2.8 Scanning electron micrographs of SiO2/9-AA nanoparticles following incubation in NH3(aq) solution for 30 min.
This species has rod-shaped cells (~ 350 × 40 µm). Since cell wall in algal cells is rich in anion-forming biomolecules (e.g. galacturonic acid – a component of pectin),58 the SiO2/9-AA nanoparticles can readily attach onto the outer surface of the cell due to electrostatic interactions (Figure 2.9). Figure 2.10 presents optical and fluorescence micrographs of individual C. acerosum cells with/without SiO2/9-AA nanoparticles. Red fluorescence originates from the natural dyes present in the cell, while yellow fluorescence originates from SiO2/9-AA nanoparticles attached onto the cell surface. We also tested the new matrix with another species of algae – Anabaena sp.: also in this case, the SiO2/9-AA nanoparticles readily attached to the outer surface of the cells (Figure 2.11). We noted that, following the incubation with the SiO2/9-AA nanoparticles, heterocysts of Anabaena sp. – which normally do not fluoresce at the
ex = 330-380 nm – started to fluoresce with blue light. One can speculate the cell wall in heterocysts contains more anionic components which can bind a greater amount of the SiO2/9-AA nanoparticles, as compared with the photosynthesizing cells.Following the treatment of C. acerosum cells with SiO2/9-AA nanoparticles (Figures 2.10 and 2.12A), and subsequent exposure of the sample to ammonia vapors, we used
Figure 2.9 Putative mechanism of the adsorption of SiO2/9-AA nanoparticles on algal cells followed by the controlled release of 9-AA.
Figure 2.10 Optical and fluorescence micrographs of individual Closterium acerosum cells with/without SiO2/9-AAnanoparticles. Scale bars: 200 µm.
Figure 2.11 Optical and fluorescence micrographs of Anabaena sp. with/without SiO2/9-AA nanoparticles (0.5 µL, 5 mg mL-1). The fluorescence micrographs were obtained using two different excitation wavelengths (λex = 510-560 and 330-380 nm – middle and right, respectively). Blue arrows indicate heterocysts. Scale bars: 50 µm.
MALDI-MS to obtain spectra of single cells. Figure 2.12B shows a mass spectrum of C.
acerosum with three high signals at the m/z 709.4, 779.5, and 815.5. We further conducted
MS/MS analysis of these three ions by analyzing their post-source decay products using the laser-induced fragmentation technology (LIFT) cell. The presence of three fragments – PO3
-(m/z 97), H2PO4- (m/z 97), and C3H6O5P- (m/z 153) – suggests that the three signals are related to analytes from the group of phospholipids (Figure 2.14). We confirmed the identities of these peaks by performing MALDI-MS analysis with internal calibrants (Table 2.1).
Matching the measured and the predicted m/z values led to a tentative identification of the three metabolites as phosphatidylglycerols.
Figure 2.12 Single-cell MS imaging with SiO2/9-AAnanoparticles used as matrix. (A) Optical and fluorescence micrographs of a single Closterium acerosum cell with the attached SiO2/9-AA nanoparticles. (B) MALDI mass spectrum of a single cell of Closterium acerosum – following the occlusion with SiO2/9-AA nanoparticles, and the release of 9-AA induced by gaseous ammonia. The blank spectrum of SiO2/9-AA nanoparticles shows no peaks overlapping with the three sample-related peaks (Figure 2.13). (C) Mass spectrometric images of a single cell of Closterium acerosum (same as in (A)). The MS images were
obtained in the negative-ion mode by MALDI-time-of-flight (TOF)-MS. Laser beam wavelength: 355 nm;
frequency: 50 Hz; diameter: 10 μm; raster spacing: 15 μm. The red-color dashed line in (A) approximately delimits the MS imaging area in (C). Scale bars: 200 µm.
Figure 2.13. MALDI mass spectrum of SiO2/9-AA nanoparticles deposited on an aluminum plate, and incubated with gaseous ammonia (blank). Note that the signal at the m/z: 814.9 – in this blank spectrum – does not completely overlap with the signal at the m/z: 815.5 – recorded when analyzing the sample of Closterium acerosum cells (cf. Figure 2.12).
The three prominent MS signals (m/z 709.5, 779.5, and 815.5) were subsequently monitored in a MALDI imaging sequence (Figure 2.12) using a 10-µm UV laser beam, and a raster with 15-µm spacing. The MS signals follow the contours of single cells as observed in the optical and fluorescence images (Figure 2.12A). Although the coverage of the cell with MALDI matrix is highly homogeneous (Figures 2.12A, 2.10 and 2.14), the distribution of metabolites within the cell – as visualized by MALDI-MS – is seen to represent some heterogeneity (Figure 2.12C): this points out the advantage of performing MALDI-MS imaging with subcellular resolution. For example, the metabolite corresponding to the MS signal at the m/z 779.5 seems to be present in the whole cell, while the metabolites corresponding to the MS signals at the m/z 709.5 and 815.5 appear to be more concentrated in one of the two semicells of the C. acerosum cell under investigation (Figure 2.12C).
Figure 2.14 MALDI-MS/MS spectra of the three ions corresponding to the MALDI images depicted in Figures 2.12 and 2.15: (A) m/z 709.4, (B) m/z 779.5, (C) m/z 815.5. Precursor ions are marked with asterisks (*).
The traces in the left-side column represent full m/z-range spectra whereas the traces in the right-side column display the low-m/z range (within the same spectra). The tandem mass spectra were obtained from the samples of Closterium acerosum (co-crystallized with 9-AA) in the negative-ion mode, using the laser-induced fragmentation technology (LIFT).
Table 2.1 Matching the observed and the predicted m/z values after the analysis of Closterium acerosum cells by negative-ion MALDI-TOF-MS using 9-aminoacridine as matrix, and in the presence of internal calibrants.
The internal calibrant mixture contained adenosine triphosphate, guanosine triphosphate, uridine diphosphate glucose, acetyl coenzyme A and bradykinin acetate (each at the concentration of 8.33 × 10-6 M).
Observed m/z Predicted formula Predicted m/z [M-H]- |∆m| (Da)
709.379 C38H63O10P 709.40861 0.03
779.510 C43H73O10P 779.48686 0.02
815.500 C46H73O10P 815.48484 0.02
High-quality images of single C. acerosum cells were obtained using either 15 and 10-µm laser scan rasters (Figures 2.12C and 2.15, respectively). In the case of Anabaena sp., the method allowed imaging individual chains of cells (thickness < 10 µm; Figure 2.16). Further improvement in lateral resolution is expected after combining the sample preparation using the SiO2/9-AA nanoparticle matrix (proposed here) with state-of-the-art MALDI-MS instruments which use laser beams with diameters smaller than 10 µm.
Coating specimens with chemical matrices is seen to be a big challenge in MALDI imaging. Common coating methods involve pneumatic spray and electrospray deposition [for reviews, see refs 16, 17, 21]. Since matrix compounds are normally dissolved in organic solvents, spray-based methods cause analyte to spread over the sample surface.17,21 Several alternative ways of applying matrix prior to MALDI-MS imaging have been proposed to date:
for example, matrix can be deposited on the sample by using inkjet printing, which produces an array of matrix spots, the method which is mainly applicable for imaging at low lateral resolution.59 The matrix sublimation/recrystallization method ensures small crystal size, high
Figure 2.15 Optical and fluorescence micrographs (top) as well as mass spectrometric images of a single cell of Closterium acerosum (bottom). The fluorescence micrograph (ex = 330-380 nm) shows the presence of SiO2/9-AA nanoparticles (yellow color) on the cell surface (before the exposure to gaseous ammonia).
MS images were obtained in the negative-ion mode by MALDI-TOF-MS. Laser beam wavelength: 355
MS images were obtained in the negative-ion mode by MALDI-TOF-MS. Laser beam wavelength: 355