Large-Area Graphene Films as Target Surfaces for Highly Reproducible Matrix-Assisted Laser Desorption Ionization Suitable for Quantitative Mass Spectrometry
- 309 Downloads
Due to the known sweet-spot issues that intrinsically arise from inhomogeneous formation of matrix-analyte crystals utilized as samples in matrix-assisted laser desorption ionization (MALDI) mass spectrometry, its reproducibility and thus its applications for quantification have been somewhat limited. In this paper, we report a simple strategy to improve the uniformity of matrix-analyte crystal spots, which we realized by adapting large-area graphene films, i.e., non-inert, interacting surfaces, as target surfaces. In this example, the graphitic surfaces of the graphene films interact with excess matrix molecules during the sample drying process, which induces spontaneous formation of optically uniform MALDI sample crystal layers on the film surfaces. Further, mass spectrometric imaging reveals that the visible uniformity is indeed accompanied by reproducible MALDI ionization over an entire sample spot, which greatly suppresses the appearance of sweet spots. The results of this study confirm that the proposed method achieves good linear responses of ion intensity to the analyte concentration (R2 > 0.99) with small relative standard deviations (σ < 10%), which is a range applicable for quantitative measurements using MALDI mass spectrometry.
KeywordsLarge-area graphene films MALDI mass spectrometry Sample preparation Quantification
For decades, matrix-assisted laser desorption ionization (MALDI) has been established as a versatile method for intact ionization of thermally labile molecules, for which broad applications involve mass spectrometry (MS) of large biomolecules, such as peptides, glycans, and lipids, as well as synthetic polymers . In recent years, MALDI MS has been further highlighted for new bio-clinical research opportunities with imaging mass spectrometry (IMS), which allows label-free chemical mapping of biomolecules on the surfaces of specimens such as single cells, tissues, and microarrays . In the MALDI method enabling intact desorption ionization of large molecules, co-crystallization of analytes with the excess matrix molecules, such as α-cyano-4-hydroxycinnamic acid (CHCA) and 2,5-dihydroxybenzoic acid (DHB), is essential. However, using matrix-analyte crystals as MALDI samples also results in certain drawbacks; for example, tiny yet inhomogeneous sample crystals create noticeable position-to-position variations in ionization efficiency, the so-called sweet-spot issues, even in the same sample spot. This makes the MALDI method not generally amenable to quantification, which is still a subject of further research [3, 4, 5].
In the early days, various methods to improve the homogeneity of matrix crystals were utilized; examples include vacuum drying, fast-solvent evaporation, electrospray deposition, and vacuum sublimation of the matrix, as well as multilayer methods [6, 7, 8, 9], some of which are still used as good lab practice to maintain MALDI reliability. Later, to avoid the use of such solid crystals, various room-temperature ionic liquid (RTIL) matrixes were also examined. RTILs for MALDI were formed as acid-base ion pairs with extremely low vapor pressures of acidic matrixes, such as CHCA and DHB, and organic bases including 1-methylimidazole, which exhibited good linearity suitable for quantification [10, 11]. Furthermore, to avoid using the matrix itself, researchers extensively explored LDI methods that utilized various nanostructured surfaces , such as nanostructured Si , ZnO-nanowire surfaces , and Au nanoparticle-implemented surfaces  as well as special flat surfaces that exhibited LDI capability, such as calcinated films on gold  and amorphous Si  as matrix-free target surfaces. Unfortunately, despite these extensive investigations, the proposed methods were still not robust enough to fully replace the conventional MALDI method utilizing common organic matrixes.
In fact, the most recent endeavors to achieve MALDI reproducibility to a quantitative level have still focused on sample preparation methods, where sophisticated instruments are often employed. Examples include a new interface that couples liquid chromatography with MALDI IMS by employing an automated matrix sprayer produced uniform matrix layers suitable for quantification  and excitation of evaporating MALDI sample drops using AC-electrowetting with specially patterned on-plate electrodes that led to formation of smaller and more homogeneous MALDI sample crystals . Furthermore, self-aliquoting plates combined with a specially designed sliding device, which divided a sample drop of microliters into 20 tiny replicates, offered sufficient statistical sampling for accurate quantification even without the use of internal standards .
Graphene, a two-dimensional allotrope of carbon, is a single layer of carbon atoms arranged in a hexagonal lattice. Due to the intriguing material properties resulting from structural confinement as well as its wide applications, including composites, nanoelectromechanical systems, and various sensing devices, graphene has been the special material of extensive research in recent years [21, 22]. With its unique structural properties, graphene offers atomically thin interfaces when employed in molecule-on-substrate systems, suppressing direct interaction between adsorbed molecules and underlying substrates . In the fields related to MS, flakes and nanoparticles of graphene and graphene oxides have been demonstrated as efficient matrix alternatives or additives for MALDI MS and IMS [23, 24, 25, 26, 27, 28, 29, 30]. Recently, production of large-area and high-quality graphene films became facile with the chemical vapor deposition (CVD) method, and the large-area graphene films on Cu foils are also commercially available [31, 32].
Herein, we report a simple strategy to induce spontaneous formation of highly uniform matrix-analyte crystals by exploiting large-area graphene films as attractive target surfaces, which leads to highly reproducible MALDI ionization that is suitable for quantitative measurements.
For MALDI experiments using the graphene films, target surfaces were prepared by attaching the graphene foils to stainless steel (SUS) target plates using conductive carbon tapes. For IMS of CHCA sample spots, the matrix solution was prepared by dissolving CHCA in a solution of 50% acetonitrile (MeCN) with 0.1% trifluoroacetic acid (TFA) to a final concentration of 5 mg/mL. For analytes, an aqueous mixture of standard peptides (Peptide mixture II), which includes substance P (monoisotopic mass = 1346.7 Da), renin (1759.9), and ACTH 18-39 (2464.2), was prepared to contain about 20 pmol of each peptide in a 1-μL drop. It was then mixed with the same volume of matrix solution just prior to experiments. The drops loaded on target surfaces were allowed to dry in ambient conditions. For IMS of DHB sample spots, the matrix solution of DHB (20 mg/mL) was prepared in a solution (MeCN, 20 mM NaCl(aq) = 8:2), which was then mixed with the same amount of the analyte solution. A 1-μL drop of the sample solution loaded on the SUS and graphene surfaces was dried in a vacuum desiccator. NaCl was included in the matrix solution as a cationization agent. However, the cationization reagent was not used in MALDI MS of LPC(18:0) using a DHB matrix. The analytes for IMS, N-glycans, were released directly from human serum, of which detailed procedure can be found elsewhere . In MALDI MS using a DHB matrix, commercial self-focusing MALDI plates (μfocus plate; ASTA Corp., Suwon, Korea) were utilized.
MALDI MS experiments were performed using an imaging MALDI-TOF mass spectrometry of reflectron type (IDSys IM; ASTA Corp., Suwon, Korea), which is equipped with a Nd:YLF laser (349 nm; 1 kHz). The ion acceleration energy was 19 keV, and 23 kV was applied to the reflectron. Using the specially designed laser optics module, the beam diameter of laser at the sample surface is adjustable, which can be set to be from 15 to 200 μm. In this study, mass spectrometric imaging was carried out at a spatial resolution of 50 μm in the positive ion mode. Typically, IMS data was acquired by accumulating ion signals for 100 consecutive laser shots (100 Hz) at each position. For data processing, imaging softwares, IDsys2.0 and MALDIVision (PREMIER Biosoft, Palo Alto, CA, USA), were utilized.
Results and Discussion
MALDI Using CHCA on the Graphene Films
In this study, commercially obtained large-area graphene films on Cu foils produced by the CVD method were utilized as a well-defined single layer of graphitic surfaces for MALDI target plates . The obtained graphene surfaces were examined by Raman microscopy; the characteristic Raman peaks of pristine graphene, G and 2D peaks, were clearly observed from all measured positions (Figure 1).
MALDI Using DHB on the Graphene Films
As a case for experimental samples other than standard materials, we applied this method for MALDI MS using DHB of N-glycans that we released directly from human serum using the PNGase F enzyme. DHB is a versatile matrix but requires certain amount of practice to achieve reproducibility due to its propensity to forming large crystals. In this study, we employed an optimized protocol, which we found to be reliable and routinely used for MALDI MS monitoring of N-glycans .
We successfully demonstrated a simple strategy for inducing spontaneous uniform crystallization of MALDI sample spots using large-area graphene films as attractive target surfaces. It is probably driven by interaction of graphene surfaces with excessive matrix molecules during the sample drying process, where the infinitely periodic graphitic surfaces may serve as the seeds for observed homogeneous crystallization of the organic matrixes. Indeed, the comparison of calibration curves between MALDI on SUS target plates and on graphene films revealed that the use of induced uniform spots offered a significant improvement in linearity (R2 > 0.99) and standard deviation (σ < 10%), which is beneficial for MALDI applications for quantification.
Detailed understanding of crystallization process induced by the graphene surfaces and its possible effects on the actual MALDI process still needs further study. Nevertheless, this study using graphene films clearly demonstrated that controlling the properties of target surfaces can be an important strategy in the development of sample preparation to achieve highly reproducible, and thus quantitative, MS using MALDI ionization.
S.Y.H. is grateful to Prof. Sunmin Ryu (POSTECH) for his kind assistance for characterization of graphene films on Cu foils using Raman spectroscopy.
This work was supported by MOTIE via KEIT (Grant No. 10063335). This work was also supported by MSIT via NRF (Grant No. NRF-2016R1D1A1B03931987).
- 1.Hillenkamp, F., Peter-Katalinic, J.: MALDI MS: a Practical Guide to Instrumentation, Methods and Applications, 2nd edn. Weinheim, Germany (2014)Google Scholar
- 20.Pabst, M., Fagerer, S.R., Köhling, R., Küster, S.K., Steinhoff, R., Badertscher, M., Wahl, F., Dittrich, P.S., Jefimovs, K., Zenobi, R.: Self-aliquoting microarray plates for accurate quantitative matrix-assisted laser desorption/ionization mass spectrometry. Anal. Chem. 85, 9771–9776 (2013)CrossRefPubMedGoogle Scholar
- 27.Rodríguez, C.E., Palacios, J., Fajardo, I., Urdiales, J.L., Le Guével, X., Lozano, J., Sánchez-Jiménez, F.: Conventional matrices loaded onto a graphene layer enhances MALDI-TOF/TOF signal: its application to improve detection of phosphorylated peptides. J. Am. Soc. Mass Spectrom. 27, 366–369 (2016)CrossRefPubMedGoogle Scholar
- 32.Ryu, J., Kim, Y., Won, D., Kim, N., Park, J.S., Lee, E.-K., Cho, D., Cho, S.-P., Kim, S.J., Ryu, G.H., Shin, H.-A.-S., Lee, Z., Hong, B.H., Cho, S.: Fast synthesis of high-performance graphene films by hydrogen-free rapid thermal chemical vapor deposition. ACS Nano. 8, 950–956 (2014)CrossRefPubMedGoogle Scholar