SLAS Discovery - Special Issue on Advances in MALDI Mass Spectrometry for Drug Discovery - Volume 22 Issue 10 - December 2017 - 1207

Beeman et al.

1207

Figure 2. SRCtide and p-SRCtide detection via MALDI-MS. The 1536-well automated MALDI measurement and peak analysis of the
substrate (SRCtide) and product (p-SRCtide) of the c-MET assay. (A) Schematic representation of the c-MET substrate and product.
(B) Example MALDI-TOF MS profiles for the substrate peak at m/z = 1679 and phosphorylated product peak at m/z = 1759, as well as
(C) for the c-MET kinase assay. (D,E) SRC and p-SRC titration shows the linear range of detection for the substrate and product of
the c-MET assay, either raw values of intensity (D) or normalized to the highest concentration (E). The blue and gray arrows mark the
substrate and product concentration representative of the c-MET assay.

probably require integration of another pipettor, like the
CiBio Well Vario positive displacement 1536-tip head system or TTP Mosquito. Increased use of Echo transfers during a screening run creates an additional limitation. Using
the Echo for MALDI target preparation, as well as compound transfers to the assay plate in a standard HTS workflow, creates a scheduling bottleneck and will limit
throughput of the screening process. Therefore, the pros and
cons for the use of the Echo liquid handler have to be
weighed carefully and considered for the individual screening platform setup.
Suppression of ionization is one of the biggest challenges in MALDI-MS sample preparation.11,16 Since we
expected such ionization problems with our established
c-MET assay buffer, we turned our attention to the adjustment of buffer ingredients and tested the c-MET kinase
assay performance with those different buffer conditions.
To be functional in an in vitro setting, biological systems
require a certain pH range and are dependent on salts and/or
certain metal cations for their activity. On the other hand,
sample ingredients, such as buffer substances (e.g., HEPES
and Tris), salts (e.g., NaCl, KCl, and MgCl2), and detergents
(e.g., Triton X-100 and Tween-20), are known to influence
ionization and negatively modulate the signals of MALDI
spectra. For the c-MET kinase assay, we started with the
known assay conditions of 20 mM HEPES (pH 7.5), 5 mM
MgCl2, 2 mM DTT, and 0.01% Triton X-100, which were

used to screen for c-MET kinase inhibitors with a Caliper
CE system. Since we wanted to keep the established assay
working within similar parameters, we started to adjust buffer ingredients step-by-step and monitored assay performance by looking at the percent of conversion by CE. It was
shown before that Triton-X100 has a major impact on
MALDI-MS signal intensity;11 therefore, we first used different concentrations of Brij-35 to replace Triton-X100
(Fig. 2A). As shown in Figure 2A, 0.003%-0.1% Brij-35
keeps assay performance in an acceptable range. To further
adjust the assay conditions for MALDI-MS analysis, we
chose 0.01% Brij-35 and worked on further adjusting the
buffer conditions.
As shown in Figure 2B, 0.3-10 mM HEPES was able to
support the c-MET kinase activity at a similar level (Fig.
2B, blue). On the other hand, variation in MgCl2 concentration had a bigger influence on assay performance. Even if 1
mM MgCl2 was in an acceptable range of 15%-30% conversion, a clear decrease between the starting conditions
and 1 mM was visible (Fig. 2B, gray). After evaluating the
influence of the single buffer components, we looked at a
combination of changes and compared them with the original assay conditions (Fig. 2C, dark gray). Those experiments led to the conclusion that lowering the HEPES and
MgCl2 concentrations and changing the feasible detergent
would achieve a more MALDI-MS-compatible buffer composition. We decided to move ahead and use 1 mM HEPES

Table of Contents for the Digital Edition of SLAS Discovery - Special Issue on Advances in MALDI Mass Spectrometry for Drug Discovery - Volume 22 Issue 10 - December 2017