Natural products play an important role in phenotypic approaches to drug discovery.
Recent years have seen the resurgence of these approaches, in particular due to their
pre-selection for cell-active compounds and their ability to identify novel druggable
nodes. Therefore, the identification of protein targets of natural products and the
elucidation of their mechanism of action remain areas of high interest. Quantitative
chemoproteomics has emerged for us as a principal strategy for generation of testable
target hypotheses. It typically combines competition-based compound affinity chromatography
from cell lysates with mass spectrometry-based protein identification and quantitation
using e.g. isobaric labeling tags. These studies enable the identification of the
full spectrum of cellular interactors of compounds under conditions approximating
the disease-relevant in vivo situation, including the efficacy target(s) and also
potential off-targets. However, for unbiased target deconvolution, the immobilization
of compound on solid support is required at some stage during the workflow. While
the introduction of a functionalized linker at a permissive site, as typically done
for synthetic small molecules, is in principle also possible for natural products,
it requires extensive knowledge of structure activity relationships which might not
always be easily obtainable. Therefore, an extended toolbox of approaches has proven
to be valuable, including site-nonselective compound immobilization on photocrosslinker
beads. Several variants of chemoproteomics as applied to natural products target deconvolution
and validation will be presented.
