Recent evidence has established a role for the small GTPase RAB25, as well as related effector proteins, in enacting both pro-oncogenic and anti-oncogenic phenotypes in specific cellular contexts. Here we report the development of all-hydrocarbon stabilized peptides derived from the RAB-binding FIP-family of proteins to target RAB25. Relative to unmodified peptides, optimized stapled peptides exhibit increased structural stability, binding affinity, cell permeability, and inhibition of RAB25:FIP complex formation. Treatment of cancer cell lines in which RAB25 is pro-oncogenic with an optimized stapled peptide, RFP14, inhibits migration, and proliferation in a RAB25-dependent manner. In contrast, RFP14 treatment augments these phenotypes in breast cancer cells in which RAB25 is tumor suppressive. Transcriptional profiling identified significantly altered transcripts in response to RAB25 expression, and treatment with RFP14 opposes this expression profile. These data validate the first cell-active chemical probes targeting RAB-family proteins and support the role of RAB25 in regulating context-specific oncogenic phenotypes.

Integration of chemical probes into proteomic workflows enables the interrogation of protein activity, rather than abundance. Current methods limit the biological contexts that can be addressed due to sample homogenization, signal-averaging, and bias toward abundant proteins. Here we report a platform that integrates family-wide chemical probes with proximity-dependent oligonucleotide amplification and imaging to quantify enzyme activity in native contexts with high spatial resolution. Application of this method, activity-dependent proximity ligation (ADPL), to serine hydrolase and cysteine protease enzymes enables quantification of differential enzyme activity resulting from endogenous changes in localization and expression. In a competitive format, small-molecule target engagement with endogenous proteins in live cells can be quantified. Finally, retention of sample architecture enables interrogation of complex environments such as cellular co-culture and patient samples. ADPL should be amenable to diverse probe and protein families to detect active enzymes at scale and resolution out of reach with current methods.

Supplemental Figure 1: Chemical structures of photoproximity probes used in this study. Supplemental Figure 3: Sub-cellular localization and photocleavage with the PhotoPPI system. A) Fluorescence microscopy images of HeLa cells transiently transfected with SNAP-FLAG-NLS, which exhibit strong nuclear localization of the SNAP protein. B) Fluorescence and brightfield microscopy images of HeLa cells transientlytransfected with SNAP-FLAG-NLS show nuclear-localization of FITC signal in cells treated with the non-cleavableFITC-BnG, or PF-BnG in the absence of UV irradiation (left three columns). Progressive exposure to UVirradiation at 365 nm results in complete loss of FITC signal in less than 10 min. Note that DAPI co-staining of the nucleus could not be performed due to overlap in the wavelength for DAPI excitation and nitroveratryl cleavage. C) Magnified view of cells in (B) above and reproduced from Fig. 1C here for comparison. Scale bars in A, B and C equal 8.3, 50 and 8.3 µm, respectively. Supplemental Figure 4: In vitro photoproximity labeling in the presence of whole cell proteome. Anti-biotin (streptavidin-800) and anti-mouse Western blot analyses of PP1 labeled SNAP-FLAG/ a-FLAG antibody complex with and without UV irradiation prior to analysis. Photolabeling was performed in the presence of whole cell lysate. Labels for individual proteins are included at appropriate molecular weights: LC, light chain; HC, heavy chain; "SNAP" label represents SNAP-Tag protein without the FLAG epitope. Supplemental Figure 5: Schematic depicting C-terminal (KEAP1-SNAP) and N-terminal (SNAP-KEAP1) genetic fusions used in photoproximity profiling of KEAP1 in cells. GxS represents a glycine-serine spacer, with X indicating the number of glycines. General synthetic methodsReagents purchased from commercial suppliers were analytical grade and used without further purification. All reactions were carried out in oven dried flasks using anhydrous solvents (Acros) unless otherwise specified. Reaction progress was monitored by thin-layer chromatography on Macherey-Nagel SIL G-25 UV254 TLC plates, visualized with UV light, ceric ammonium molybdate (CAM), p-anisidine, bromophenol blue, 2,4-dinitrophenyl hydrazine (DNP), or KMnO4 TLC stains. Nuclear magnetic resonance spectra were acquired using either a Bruker AVANCE II+ 500; 11.7 Tesla NMR or Bruker DRX 400; 9.3 Tesla NMR instrument. Accurate mass measurements were obtained using an Agilent 6224 Tof-MS instrument. When necessary, compounds were purified via flash column chromatography using Siliaflash F60 60 Å, 230-400 mesh silica gel (Silicycle) Chemical synthesis of photoproximity profiling (PhotoPPI)-probes 4-[[(2,2,2-trifluoroacetyl)amino]methyl]benzoic acid (2)Solid 4-(aminomethyl)benzoic acid 1 (15.1 g, 100 mmol) was dissolved in TFAA (42 mL) cooled to 0 °C. Once dissolved, the ice bath was removed and the reaction was allowed to stir at rt until starting material was consumed, ~2 hr. Upon completion, the reaction was quenched with H2O (100 mL) and precipitate collected via vacuum...

Chemoproteomic methods can report directly on endogenous, active enzyme populations, which can differ greatly from measures of transcripts or protein abundance alone. Detection and quantification of family-wide probe engagement generally requires LC-MS/MS or gel-based detection methods, which suffer from low resolution, significant input proteome requirements, laborious sample preparation, and expensive equipment. Therefore, methods that can capitalize on the broad target profiling capacity of family-wide chemical probes but that enable specific, rapid, and ultrasensitive quantitation of protein activity in native samples would be useful for basic, translational, and clinical proteomic applications. Here we develop and apply a method that we call soluble activity-dependent proximity ligation (sADPL), which harnesses family-wide chemical probes to convert active enzyme levels into amplifiable barcoded oligonucleotide signals. We demonstrate that sADPL coupled to quantitative PCR signal detection enables multiplexed “writing” and “reading” of active enzyme levels across multiple protein families directly at picogram levels of whole, unfractionated proteome. sADPL profiling in a competitive format allows for highly sensitive detection of drug–protein interaction profiling, which allows for direct quantitative measurements of in vitro and in vivo on- and off-target drug engagement. Finally, we demonstrate that comparative sADPL profiling can be applied for high-throughput molecular phenotyping of primary human tumor samples, leading to the discovery of new connections between metabolic and proteolytic enzyme activity in specific tumor compartments and patient outcomes. We expect that this modular and multiplexed chemoproteomic platform will be a general approach for drug target engagement, as well as comparative enzyme activity profiling for basic and clinical applications.

Macrocyclization can improve bioactive peptide ligands through preorganization of molecular topology, leading to improvement of pharmacologic properties like binding affinity, cell permeability, and metabolic stability. Here we demonstrate that Diels−Alder [4 + 2] cycloadditions can be harnessed for peptide macrocyclization and stabilization within a range of peptide scaffolds and chemical environments. Diels− Alder cyclization of diverse diene−dienophile reactive pairs proceeds rapidly, in high yield and with tunable stereochemical preferences on solid-phase or in aqueous solution. This reaction can be applied alone or in concert with other stabilization chemistries, such as ring-closing olefin metathesis, to stabilize loop, turn, and α-helical secondary structural motifs. NMR and molecular dynamics studies of model loop peptides confirmed preferential formation of endo cycloadduct stereochemistry, imparting significant structural rigidity to the peptide backbone that resulted in augmented protease resistance and increased biological activity of a Diels−Alder cyclized (DAC) RGD peptide. Separately, we demonstrated the stabilization of DAC αhelical peptides derived from the ERα-binding protein SRC2. We solved a 2.25 Å cocrystal structure of one DAC helical peptide bound to ERα, which unequivocally corroborated endo stereochemistry of the resulting Diels−Alder adduct, and confirmed that the unique architecture of stabilizing motifs formed with this chemistry can directly contribute to target binding. These data establish Diels−Alder cyclization as a versatile approach to stabilize diverse protein structural motifs under a range of chemical environments.