Novel photocleavable intracellular heterodimerizer to manipulate protein dynamics with high spatiotemporal precision

Cells generate a large repertoire of signaling pathways out of a limited set of proteins through their ability to activate specific signaling proteins in a subcellular region for a short amount of time. A common mechanism to activate signaling pathways represents the translocation of proteins from the cytosol to internal membranes or the inner leaflet of the plasma membrane, where they meet their substrates and initiate downstream signaling pathways.
Chemical inducers of dimerization (CIDs) are powerful tools to manipulate proteins in a spatially and temporally confined region within the cell and to reproduce protein-protein and protein-membrane interactions, which trigger the activation of signaling pathways. These so called “chemical dimerizers” are small organic molecules, which bind simultaneously two specific dimerizing domains. In the presence of the dimerizer, two proteins fused to these dimerizing domains are brought into close proximity, which initiates a cellular effect linked to
the interaction of the protein partners.
We developed a new class of CIDs, based on an intracellular, covalent Halo- and SNAP-tag reactive dimerizer, called HaXS. Its modular synthetic strategy enables the relatively simple introduction of novel functional groups into the core module linking the Halo- and the SNAPtag substrates in order to generate HaXS derivatives with novel features. Through the introduction of a photocleavable methylnitroveratryl (MeNV) group, we developed the first,
cell-permeable photocleavable heterodimerizer, called MeNV-HaXS. Excitation at 360 nm cleaves MeNV-HaXS and reverses the MeNV-HaXS-induced SNAP- and HaloTag dimer complex.
HaXS unifies several important features, which are essential to mimic physiologic signaling pathways, such as the fast and selective induction of protein-protein interactions, the absent interferences with endogenous signaling pathways as well as the possibility to reverse an
induced dimerization event. HaXS was successfully used to target tagged proteins to selected intracellular organelles such as endosomes, lysosomes, the plasma membrane, mitochondria, the nucleus and the actin skeleton, which can be exploited to study the function
of a particular protein in different subcellular contexts. Furthermore, the manipulation of protein localizations can be used as a strategy to initiate signaling pathways at defined starting points and cellular locations. Through the HaXS-induced control of signaling protein
localizations, which function upstream of the targeted signaling pathway, HaXS is able to selectively control the activation resp. inactivation of an isolated signal transduction branches
in a complex signaling network. Additionally, the combination of chemical-induction and light induced reversion of MeNV-HaXS-induced dimers enables to inactivate any protein of interest through sequestering it away from its functional compartment, followed by an optically guided cleavage of the dimer complex, which releases the sequestered proteins and restores its function. Sine the release of anchored proteins occurs with high spatiotemporal precision (t < 1 sec, from single vesicles), this simple experimental setup enables to study translocation kinetics of trapped proteins back to their normal localization in live cells.
To get further insights into the dimerization behavior of the HaXS CID or CIDs in general, we analyzed the chemical induced dimerization reactions with a modeling software called CellDesigner. We demonstrated how various parameters of a CID, such as the ratio of the
rate constants of the dimerizing tags or the choice of the dimerizer concentration affects the speed as well as the efficiency of the dimer formation. This allows one to perform efficient dimerization experiments and to understand how parameters of a CID can be optimized to
improve its dimerization performance.
The most widely used CID system is based on rapamycin, which induces a tight binding between FKBP12 and the FKBP rapamycin domain (FRB) of mammalian target of rapamycin (mTORC1). The rapamycin CID profits from excellent kinetics, but the cross reactions with mTORC1 diminishes the utility to study cellular events involved in cell growth and metabolism. Analogs of rapamycin (so-called rapalogs), which only react with an engineered version of the FRB domain but not with the endogenous FRB domain of mTORC1, are developed to overcome the immunosuppressive properties of rapamycin while retaining their dimerization ability. The synthesis of rapalogs is challenging, since already minute rapamycin or rapamycin by-products impurities are sufficient to inhibit mTORC1. We successfully
established a new protocol for the synthesis and purification of a C16 phenyl carbamate (pcRap) rapalog, which induces dimerization of FKBP and FRB fusion proteins without interfering with the mTORC1 pathway.
Summing up, we provide three valid dimerizer molecules (HaXS8, MeNV-HaXS, pcRap) to the current toolbox of CIDs. Many important features necessary to reproduce physiological signaling pathways are unified in our CIDs and will enable to dissect many cellular signaling
events. Furthermore, the combination of a HaXS CID with a rapamycin or rapalog based CID, enables the simultaneous and orthogonal control of two different proteins within a single cell and thus greatly improves the possibilities for cellular interrogations.