Matrix-assisted laser desorption/ionization mass spectrometry imaging is a technique for direct analysis of tissue sections without the use of molecular tags or contrast agents. The combination of spatial and mass resolution results in large and complex data sets that require powerful and efficient analysis and interpretation tools. Conventional images, derived from a specific analyte mass, do not identify the spatially localized correlations between analytes that are latent in the data. A new approach to find and visualize these correlations is presented. Clustering methods are used to classify pixels by spectral similarity, facilitating definition of distinct spatial regions. Principal component and discriminant analyses are combined to comprehensively identify changes in the mass spectra between regions. Images are generated by projecting the spectra of each pixel on the discriminant spectra; contrast is then a function of multiple correlated peaks.

The matrix suppression effect (MSE) can lead to high-quality MALDI mass spectra: strong analyte signals and weak or negligible matrix background peaks. Experiment and theory suggest that MSE should be widespread and, therefore, generally applicable to measurement of low molecular weight (LMW) substances. These are otherwise impractical with MALDI due to interference from matrix. Appropriate conditions for MSE were investigated and tested on a variety of LMW substances. Straightforward and semiautomated interpretation was possible for 87.7% of these. Another 3.5% gave poor MSE due to sodium cationization rather than protonation of the analyte, but interpretation was possible. MALDI imaging shows that MSE varies significantly across a typical sample. Selective data accumulation could further increase the utility of the method. Samples containing more than one analyte were also studied. Analyte-analyte suppression was not found to be excessive, and moderately abundant minority species can be adequately detected.

Noncovalent complexes between chicken muscle adenylate kinase and two inhibitors, P 1 ,P 4 -di(adenosine-5')tetraphosphate (Ap4A) and P 1 ,P 5 -di(adenosine-5') pentaphosphate (Ap5A), were investigated with electrospray ionization mass spectrometry under non-denaturing conditions. The nonconvalent nature and the specificity of the complexes are demonstrated with a number of control experiments. Titration experiments allowed the association constants for inhibitor binding to be determined. Problems with concentration dependent ion yields are circumvented by a data evaluation method that is insensitive to the overall ionization efficiency. The K a values found were 9.0 ϫ 10 4 M Ϫ1 (Ap4A) and 4.0 ϫ S oft ionization mass spectrometry (MS), in particular matrix-assisted laser desorption/ionization (MALDI) and electrospray ionization (ESI), is well known for its ability to bring high molecular weight biomolecules into the gase phase and to even preserve noncovalent complexes [1][2][3][4][5][6][7][8][9][10]. Besides its established applications, a recurring question concerns the applicability of mass spectrometry to measuring noncovalent interaction strengths [11]. There are different approaches to measure the interaction strengths in the gas phase as well as in the liquid phase. In the former case, the mass spectra should ideally directly represent the solution phase chemistry. Gas-phase methods include cone voltage driven dissociation (called the "VC 50 " method) [12,13], collision-induced dissociation [14,15], guided ion beam tandem mass spectrometry [16,17], blackbody infrared radiative dissociation [18,19] and heated capillary dissocation [20,21].Some mass spectrometric studies have been reported in the literature for the determination of noncovalent interaction strengths. There are three basic methods to determine association constants that are reflective of those in solution: first, it is possible to record "melting curves" by raising the temperature of the analyte solution and to use MS to determine the percentage of intact complexes. Cheng et al. analyzed complementary DNA strands [22] and Fändrich et al. studied the chaperone GimC/prefolding homologue complex [13] with this method. Second, competition experiments can be used to evaluate the stability of noncovalent complexes. In Jørgensen's method [23,24], the relative intensities of the free host and the different complexes were used, whereas Kempen et al. employed the known concentration of a reference complex to determine the concentration of an unknown complex [25]. Third, a titration of a host with a guest (or vice versa) can be used. In this case, the mass spectrometric signals should accurately represent the concentrations of the species in solution. The titration method is documented in a number of publications, all of them convincingly showing the successful use of mass spectrometry for the quantitative determination of noncovalent binding constants of proteins and small molecules [26 -33]. This aspect was the main focus in most previous research; th...

No abstract

Electrosprayed spots of varying thickness were evaluated for use as reproducible, homogenous, high efficiency MALDI samples. Thin samples on stainless steel plates were found to give exceptionally strong signals, as did the last layers of thick samples, when ablated down to the steel substrate. A small enhancement was also observed for thin samples on a gold substrate, and with a few-nanometer gold coating on top of a thick sample. Ion yields and intensity ratios can be understood in the context of the previously described quantitative MALDI model including the matrix-metal interfacial ionization potential reduction effect (Knochenmuss, R.; Anal. Chem. 2004, 76, 3179 -3184). The absolute and relative stabilities of ion signals were found to be at least a factor of two better for the thin electrosprayed spots, , both the underlying ionization mechanisms and sample preparation methods remain objects of active study. These are coupled topics since mechanistic understanding can allow rational, rather than undirected, method development. Here, we make use of a recently developed quantitative model for UV-MALDI based on a two-step ionization process [2,3]. Primary laser-created matrix ions and analyte neutrals undergo ion-molecule reactions in the expanding ablation plume to yield secondary analyte ions. The observed mass spectrum is determined by the thermodynamics of these reactions. Because matrix and analyte are explicitly coupled, the model proved particularly suitable for understanding of practically relevant phenomena such as ion suppression effects [4 -6], which have been shown to be very common phenomena [7].A variant of this model has recently been developed for thin samples on metal substrates [8]. This was motivated by measurements of electron emission yields from MALDI samples on metallic versus insulating substrates [9], and suggestions that such electrons might neutralize (positive) analyte ions during the desorption process [10]. The model gave excellent agreement with the electron yield data, but this was not consistent with claims of higher analyte ion yields for samples on nonconductive versus conductive substrates [9]. The major difference between thick and thin samples is the interaction of matrix molecules with the substrate, which can lower the ionization potential of the combined system if the substrate is metal. It was, therefore, theoretically expected that thin samples on metal would yield higher, not lower, signals. The present study was undertaken to investigate this discrepancy and restore consistency to sample preparation strategy.The method selected here to prepare homogenous samples of variable thickness is electrospray (ES) deposition of analyte/matrix solution [11,12]. Electrospray produces more uniform, fine-grained sample spots than the more convenient and widespread dried droplet technique, which tends to form large matrix crystals in random locations. By varying the electrospray time, it is also possible to make samples as thin as a few hundred nanometers, but a few mm in diamete...