Investigation for quantitative MALDI Imaging and profiling with reproducible MALDI

Abstract

Matrix-assisted laser desorption ionization (MALDI) combined with time-of-flight (TOF) mass spectrometry is an useful technique to analyze biomolecules. When laser irradiate on the matrix containing analytes, gas-phase ions are generated. It is well known that generating reproducible analyte ion signals is difficult and therefore it has not become the method for quantification. To overcome this problem, adding an internal standard, e.g. an isotopically labeled internal standard such as 13C and 15N has been implemented. Recently, MALDI spectra for some peptides under several experimental conditions were collected and were tagged with the temperature in the early MALDI plume, Tearly. We found that the patterns of the spectra became similar, or reproducible, when those tagged with the same Tearly were collected. To get easily reproducible MALDI mass spectra, we invented some methods to fix Tearly. The similar MALDI mass spectra at the same Tearly stated above meant that not only the fragmentation patterns of analyte ion, [A + H]+, and matrix ion, [M + H]+, but also the analyte-to-matrix ion ratio, [A + H]+/[M + H]+, was thermally determined. To check this, we implemented experiments for samples with various matrix-to-analyte ratios, collected spectra tagged with the same Tearly, and measured the reaction quotient Q = ([M]/[A])([A + H]+/[M + H]+). For the matrix-to-analyte neutral ratio in the plume, [M]/[A], we used the ratio in the solid sample. Then, Q turned out to be independent of the neutral ratio, or was essentially the equilibrium constant K. The equilibrium relation can be written in the following form. ([A + H]+/[M + H]+) = K ([A]/[M]) This equation suggests that the concentration, or the amount, of an analyte in a sample can be measured from the abundances of analyte and matrix ions. We took a plot of the ion ratio versus the [A]/[M] as a calibration curve to quantify an analyte. We also defined the matrix suppression (S) as 1 – I([M + H]+)/I0([M + H]+). I([M + H]+) and I0([M + H]+) mean the matrix ion abundances in the spectrum of a matrix-analyte mixture and of a pure matrix, respectively. Once S exceeded a critical value, the linearity of calibration curve was broken. Mapping the spatial distributions of interesting analytes in biological samples has attracted a lot of interest. Imaging and profiling based on mass spectrometry are particularly attractive because of its capability to determine the distribution of lots of unknown chemicals in a single measurement. Two popular ionization techniques widely used in imaging and profiling are Secondary ion mass spectrometry (SIMS) and MALDI, especially in combination with TOF analyzer. MALDI is more useful to analyze large molecules, compared to SIMS. There are big two problems to be solved in MALDI imaging and profiling. The first is that mass spectra obtained by MALDI were irreproducible, from sample to sample, from spot to spot in a sample, and from shot to shot at a spot. Hence, there is no way to draw quantitatively meaningful image maps. The other is as in the following. The analytes are located in the sample from at first and a matrix solution is put on the sample surface and the solution dry. Finally crystals made of matrixes and analytes are formed. In this process, there is uncertainty that all analytes in the original sample are relocated in the matrix crystal, namely analyte transfer efficiency. To verify that the first issue can be solved by our MALDI quantification method, we investigated samples with known concentrations of an analyte in the matrix crystal. For accurate results, we prepared a mixture solution consisting of both matrixes and analytes, sprayed it on a cleaned tissue, and dried the tissue. We call this premixed sample. We quantified the analytes in the premixed samples using the method invented by us and showed the quantification results were agreed with the prepared concentrations. That is, taking the analyte-to-matrix ion ratio as the measure of the analyte concentration at a spot can be a solution to the first problem. However, the preparation method of premixed sample is far from how the MALDI imaging and/or profiling sample is prepared. In real situation, a matrix solution is put onto the sample surface and the analyte existing in the sample is extracted by the solvents. Finally the crystals containing matrixes and anayltes are formed after the solvents are dried. To check whether or not all of analytes are quantitatively extracted, we prepared a pseudo MALDI profiling sample in which the analyte and matrix solutions were loaded and dried one after another. Through the close and many studies, we found that an evaporation time of the solvents used is responsible for the analyte transfer efficiency from the sample to the matrix. We recently introduced that fluidic liquid matrixes can be made by the nonstoichiometric mixing of the organic acids and organic bases. Such a liquid is fluidic and homogeneous. More importantly, this characteristic property of the liquid matrix can be useful for the efficient extraction of analytes. The remedy for the second problem is that choice of proper solvent which does not quickly evaporate or using the liquid matrix. In this work, the effort to unravel the problems hindering the quantitative MALDI imaging and profiling is introduced.