Tryptoline Stereoprobe Elaboration Identifies Inhibitors of the GRPEL1-HSPA9 Chaperone Complex - PubMed (original) (raw)

[Preprint]. 2025 Oct 20:2025.10.20.683548.

doi: 10.1101/2025.10.20.683548.

Raymond F Berkeley 1, Zijian Gao 2, Maximilian Garhammer 3, Marc A Morizono 4, Evert Njomen 1, Haoxin Li 1, Kristen E DeMeester 1, Victor Cociorva 1, Mark A Herzik Jr 4, F Ulrich Hartl 3, Bruno Melillo 1, Benjamin F Cravatt 1

Affiliations

Tryptoline Stereoprobe Elaboration Identifies Inhibitors of the GRPEL1-HSPA9 Chaperone Complex

Rachel E Hayward et al. bioRxiv. 2025.

Abstract

Activity-based protein profiling has identified hundreds of proteins from diverse classes that react at specific cysteine residues with stereochemically defined electrophilic compounds (stereoprobes) in human cells. The structure-activity relationships underlying these stereoprobe-protein interactions, however, remain poorly understood. Here we show that the protein interaction landscape of tryptoline acrylamide stereoprobes can be profoundly altered by structural modifications distal to the acrylamide reactive group. The majority of stereoprobe liganding events occurred at non-orthosteric sites and mostly evaded assignment by the machine learning-based co-folding model Boltz-2, which instead tended to misplace the stereoprobes in orthosteric pockets (an outcome we term "orthostery burnout"). We found that stereoprobes reacting with C124 in the nucleotide exchange factor GRPEL1 disrupt interactions with the mitochondrial HSP70 chaperone HSPA9/mortalin, leading to impairments in mitochondrial protein import and induction of mitophagy. Our results highlight tryptoline acrylamides as a versatile source of covalent ligands targeting non-orthosteric sites on proteins, including tool compounds that perturb the mitochondrial HSP70 chaperone system.

PubMed Disclaimer

Conflict of interest statement

Competing Interests The authors declare no competing interests.

Figures

Extended Data Fig. 1 |

Extended Data Fig. 1 |. Chemical proteomic analysis of elaborated tryptoline acrylamide stereoprobes.

a, Structures of parent and alkynylated tryptoline acrylamide stereoprobes used previously. b, Heatmap showing DFT calculations for elaborated tryptoline acrylamide stereoprobes compared to unelaborated counterparts (* parent stereoprobes, ** other analogs included for analysis), suggesting minimal impact of C6- and C7-substitution on tryptoline acrylamide core geometry. c, Workflow for cysteine-directed ABPP experiments where stereoprobe reactivity with cysteines is determined by blockade of iodoacetamide-desthiobiotin (IA-DTB) labeling, streptavidin enrichment, and identification and quantification by multiplexed (tandem mass tagging, TMT10plex) MS-based proteomics, as described previously. d, Bar graph showing the number of experiments in which stereoprobe-liganded cysteines were quantified (shown for all cysteines liganded by elaborated, parent, and/or alkyne stereoprobes). e, Bar graphs showing representative liganding profiles for cysteines preferentially engaged by elaborated (left) or parent stereoprobes (right) or showing no preference (middle). Data represent average values ± SD of four independent experiments. f, Pie chart showing number of cysteines preferentially engaged (> 1.5-fold) by elaborated stereoprobes or alkyne stereoprobes. Cysteines not quantified in either elaborated or alkyne stereoprobe datasets were excluded from the analysis (12 total cysteines). g, Bar graph comparing the number of liganded cysteines for each elaborated tryptoline acrylamide stereoprobe where blue and grey designate cysteines that were engaged solely by the indicated stereoprobe vs multiple stereoprobes, respectively.

Extended Data Fig. 2 |

Extended Data Fig. 2 |. Boltz-2 analysis of tryptoline acrylamide-liganded proteins with virtual competition by orthosteric ligands.

a, A virtual competition assay comparing Boltz-2 predictions for various allosteric stereoprobe ligands for CRBN with and without the orthosteric ligand lenalidomide. Allosteric ligands targeting CRBN_C219 (WX-01–12; WX-63d) or CRBN_C287 (ZL-9B) are shown (far left) alongside CRBN:DDB1ΔBPB predicted structures with WX-01–12 (left), WX-63d (center), and ZL-9B (right); generated with allosteric ligands alone (top), or with allosteric ligands and lenalidomide (bottom). Boltz-2 rank 0 models are light blue, rank 1–9 models are transparent. Liganded cysteines are colored yellow. Ligands colored per their assignment at far left, lenalidomide poses are coral. b, A virtual competition assay comparing Boltz-2 predictions for IKBKB and EV-97 with (right) and without (left) K252a, a staurosporine analog known to bind the ATP-binding pocket as shown in PDB 4KIK. Boltz-2 rank 0 models are light blue, rank 1–9 models are transparent. EV-97 is pink, K252a is plum.

Extended Data Fig. 3 |

Extended Data Fig. 3 |. Additional Boltz-2 analysis of tryptoline acrylamide-liganded proteins.

a, Rank 0 structure of HIRA co-folded with WX-62b aligned with the crystal structure of the WD40 domain protein Cdc20 in complex with the small molecule apcin (PDB 4N14). The WD40 domain of HIRA is light blue, while the remainder of HIRA is white. Candidate liganded cysteines are yellow spheres (note that the tryptic peptide for HIRA found to be engaged by WX-62b in cysteine-directed ABPP experiments has four cysteines (C295, C298-C300) representing candidates for the liganding event). WX-62b is pink. The 4N14 structure, including the ligand apcin, is green. b, Pie charts representing the distribution of stereochemistries for predicted stereoprobe ligands with respect to the stereochemistry of the experimentally determined stereoprobe ligands (determined by cysteine-directed ABPP) and specified in the input for all structural predictions for Boltz-2.2.0 (no template-aware stereochemical guidance) and Boltz-2.2.1 (adds template-aware stereochemical guidance). c, Waterfall plot showing cysteine-ligand distances as predicted by Boltz-2.2.1 for each protein identified as having a tryptoline acrylamide stereoprobe-liganded cysteine (Supplementary Dataset 1). The ten diffusion samples for the protein-stereoprobe pair with the closest stereoprobe-cysteine distance are shown. Proteins are sorted by closest stereoprobe distance and points are colored according to the mean pLDDT of all residues within 5 Å of the liganded cysteine.

Extended Data Fig. 4 |

Extended Data Fig. 4 |. Characterization of elaborated stereoprobe-cysteine interactions.

a, Gel-ABPP data showing stereoselective engagement of recombinant WT-FBL, but not a C268S-FBL mutant by WX-72d-yne (5 μM, 1 h) in lysates of HEK293T cells. UT, untransfected cells. b, Gel-ABPP data showing concentration-dependent and enantioselective blockade of WX-72d-yne (5 μM, 1 h) engagement of WT-FBL by WX-72d (top gel) but not parent probe EV-97 (bottom gel) (2 h pretreatment). Left, representative gel-ABPP data; right, quantification of gel-ABPP data. Data represent average values ± SD for 2 independent experiments. c, Gel-ABPP data showing increased enantioselective liganding of recombinant WT-FBL by WX-72d-yne in the presence of SAM (right) but not SAH (left). Experiments were performed in lysates of HEK293T cells stably expressing WT-FBL. d, Protein-directed ABPP data showing relative enrichment of endogenous FBL in Ramos cell lysates treated with WX-72d-yne or inactive enantiomer WX-72b-yne (5 μM, 1 h) in the presence and absence of SAM (50 or 100 μM). Data represent average values ± SD for two independent experiments. e, Boltz-2 prediction for WX-72d binding to FBL showing WX-72d placement in the presence of SAM, which occupies the SAM/SAH binding pocket, causing displacement of WX-72d to a site distal to liganded C268. f, Heatmap showing cysteine-directed ABPP data for TEX264, identifying decreases in the IA-DTB reactivity for C165 and C68 in cysteine-directed ABPP experiments of Ramos cells treated with the indicated stereoprobes. Blue and red text represent C6- and C7-elaborated stereoprobes. g, Gel-ABPP data showing concentration-dependent and enantioselective blockade of WX-62c-yne (5 μM, 1 h) engagement of recombinant TEX264 by WX-62c (left), but not parent stereoprobe EV-99 (right) (2 h pretreatment) in HEK293T cells. h, Cysteine-directed ABPP data for all quantified cysteines in NCKAP1L and all stereoprobes (elaborated, parent, and alkyne stereoprobes), identifying WX-73a (red) as the sole stereoprobe that liganded C338. The dashed line marks 66.7% IA-DTB blockade used to define a liganded cysteine event. Cysteines separated by a semi-colon (;) are located on the same tryptic peptide. i, Gel-ABPP data showing concentration-dependent and enantioselective blockade of WX-73a-yne (5 μM, 1 h) engagement of recombinant NCKAP1L by WX-73a (left) but not parent probe EV-98 (right) (2 h pretreatment) in HEK293T cells. For f and h, cysteine-directed ABPP data represent average values, and only cysteines quantified from at least two independent experiments are shown (from a total of 4–6 experiments performed per stereoprobe). For a, g, and i, data are from a single experiment representative of at least two independent experiments

Extended Data Fig. 5 |

Extended Data Fig. 5 |. Tryptoline acrylamide stereoprobes disrupt GRPEL1:HSPA9 interactions.

a, Gel-ABPP data showing enantioselective blockade of WX-01–07 (5 μM, 1 h) engagement of WT-GRPEL1 by WX-71b and WX-71c (20 μM, 1 h) in HCT-116 cells. UT, untransfected cells. b, IP-western blotting data showing concentration-dependent and enantioselective disruption of recombinant WT-GRPEL1 interactions with endogenous HSPA9 following treatments with WX-71b and WX-71c or enantiomeric controls WX-71d and WX-71a (indicated concentrations, 4 h) in HCT-116 cells. c, Heatmap of IP-MS data showing stereoselective disruption of protein interactions for recombinant WT-GRPEL1, but not C124A-GRPEL1, following treatment with WX-71b vs WX-71d (20 μM, 4 h) in HCT-116 cells. Displayed proteins exhibited at least 3-fold enrichment in WT-GRPEL1-expressing cells compared to untransfected HCT-116 cells. Protein signals were normalized to GRPEL1 signals within each treatment group and then to DMSO signals across treatment groups and are listed from left-to-right following GRPEL1 by average spectral count values. Data represent average values ± SD for four independent experiments. d, Gel-ABPP data showing time-dependent and enantioselective blockade of WX-01–07 (5 μM, 1 h) engagement of WT-GRPEL1 by WX-71b or WX-71d (20 μM, indicated times). e, Volcano plots of IP-MS data showing stereoselective disruption of protein interactions for recombinant HSPA9 following treatment with WX-71b vs WX-71d (5 or 20 μM, 4 h) in HCT-116 cells. Protein signals were normalized to HSPA9 within each treatment group and then to DMSO signals across treatment groups. Data represent average values ± SD of four independent experiments. Statistical significance was assessed using Welch two sample t-tests. f, Intact protein MS data for purified WT-GRPEL1 (3.5 μM) incubated with the indicated times with WX-71b or WX-71d (10 μM). Proteins were analyzed by time-of-flight (TOF)-LC/MS. g, Protein-directed ABPP data showing concentration-dependent and enantioselective blockade of WX-01–06 (5 μM, 1 h) engagement of GRPEL1, but not SF3B1 or GRPEL2, by WX-71b. Data represent average values ± SD of four independent experiments. For a, b, and d, data are from a single experiment representative of at least two independent experiments

Extended Data Fig. 6 |

Extended Data Fig. 6 |. GRPEL1 stereoprobes modulate mitochondrial protein import and function.

a, Colocalization of MTS-EGFP and Mitotracker Deep Red FM for 10 images from a single independent experiment performed in MTS-EGFP-inducible parental HCT-116 cells treated with doxycycline (0.5 μg/mL) and DMSO or WX-71b or WX-71d (20 μM, 8 h) as described in Fig. 5a. Data represent average values ± SD of ten technical replicates. b, Generation of sgGRPEL1 cells. HCT-116 cells stably expressed Flag epitope-tagged WT or C124A-GRPEL1 were subject to CRISPR/Cas9 disruption of endogenous GRPEL1 and analyzed at the population level. c, Workflow for pulsed-SILAC labeling with tandem-mass tag (TMT16plex)-based multiplexing. sgGRPEL1 HCT-116 cells expressing Flag-epitope tagged WT- or C124A-GRPEL1 were pretreated in light media with DMSO or WX-71b or WX-71d (20 μM, 4 h) followed by shifting the cells to heavy amino acid media in the continued presence of stereoprobes (5 μM, 8 h). Mitochondria were biochemically enriched and analyzed by quantitative proteomics. d, Violin plot showing heavy-labeled protein abundance for the indicated treatment groups in pulse-SILAC experiments. Protein signals were corrected to light-labeled protein signals and normalized to heavy-labeled DMSO-treated WT-GRPEL1 or C124A-GRPEL1 signals. Statistical significance evaluated with parametric, two-tailed, paired t-test. e, Violin plot showing relative heavy-labeled protein abundance for WT-GRPEL1 cells across the indicated submitochondrial localizations as annotated from MitoCarta3.0 in combination with information retrieved from Uniprot and Human Protein Atlas. OMM, outer mitochondrial membrane; IMS, mitochondrial intermembrane space; IMM, inner mitochondrial membrane. f, Histograms (cell count (y axis)) versus mt-mKeima excitation (x axis)) showing mitophagy induction in (left) parental HCT-116 cells or sgGRPEL1 cells expressing WT-GRPEL1 (middle) or C124A-GRPEL1 (right) and also expressing mt-mKeima and Parkin treated with DMSO or WX-71b or WX-71d (20 μM, 8 h). Data show a single experiment representative of three independent experiments (Fig. 5h, i present quantification).

Fig. 1 |

Fig. 1 |. Design and chemical proteomic analysis of elaborated tryptoline acrylamide stereoprobes.

a, Structures of original parent and alkynylated tryptoline acrylamide stereoprobes,. b, Heatmap showing representative cysteines displaying preferential stereoselective reactivity with C6-alkyne modified tryptoline acrylamide stereoprobes WX-01–10, −11, and −12. Cysteine-directed ABPP data were from Ramos cells, as reported previously, and represent average values from 4 independent experiments per stereoprobe. c, Structures of tryptoline acrylamides with elaborations at the C6 (blue) and C7 (red) positions. d, Total numbers of proteins/cysteines, and their stereoselectively liganded subsets, quantified by cysteine-directed ABPP for parent and/or elaborated stereoprobes in Ramos cells. e, Heatmap showing the 95 cysteines liganded by the elaborated stereoprobes compared to alkyne and parent stereoprobes. Cysteines separated by a semi-colon (;) are located on the same tryptic peptide. Cysteines are sorted alphabetically then by number of experiments quantified (9–10 experiments and 1–8 experiments; also see Extended Data Fig. 1d). Cysteine-directed ABPP data represent average values from 4–6 independent experiments per stereoprobe. f, Functional class distribution of elaborated stereoprobe-liganded proteins assigned as described previously using GO (Panther) and Uniprot annotations,. g, Pie chart showing number of stereoselectively liganded cysteines preferentially engaged (> 1.5-fold IA-DTB blockade) by elaborated versus parent stereoprobes of the same stereoconfiguration. Cysteines not quantified in either elaborated or parent stereoprobe datasets were excluded from the analysis (10 total cysteines). h, Pie chart showing number of stereoselectively liganded cysteines preferentially engaged (> 1.5-fold IA-DTB blockade) by C6-elaborated versus C7-elaborated stereoprobes with the same R group and stereoconfiguration. Cysteines not quantified in either C6- and C7-elaborated stereoprobe datasets of the same R group and stereoconfiguration were excluded from the analysis (13 total cysteines). i, Pie chart showing number of stereoselectively liganded cysteines preferentially engaged by a specific elaborated chemotype (R group) as determined by comparing data for stereoprobes having different R groups at the same position (C6 or C7) and of the same stereoconfiguration (> 1.5-fold IA-DTB blockade by the reference stereoprobe compared to the median IA-DTB blockade by the three R-group analogs). Cysteines not quantified in all datasets required for comparison were excluded from this analysis (20 total cysteines).

Fig. 2 |

Fig. 2 |. Boltz-2 analysis of tryptoline acrylamide-liganded proteins.

a, Waterfall plot showing cysteine-ligand distances for each protein identified as having a tryptoline acrylamide stereoprobe-liganded cysteine (Supplementary Dataset 1). The ten diffusion samples for the protein-stereoprobe pair with the closest stereoprobe-cysteine distance are shown. Proteins are sorted by closest stereoprobe distance and points are colored according to the mean pLDDT of all residues within 5 Å of the liganded cysteine. b, Left: Stacked bar graphs showing the fraction of proteins with and without orthosteric site annotations in UniProt colored by predicted stereoprobe proximity to the liganded cysteine (blue: < 5 Å (correct predictions); gray: > 5 Å (incorrect predictions)). Right: Stacked bar graphs, where each bar shows the fraction of liganded cysteines occupying (right) or not occupying (left) orthosteric sites for the subset of proteins with orthosteric site annotations in UniProt, and where each bar is then further color-coded for correct (blue) and incorrect (gray) predictions as defined in left panel. c, Scatter plot representing the relationship between the shortest distance between the stereoprobe and the liganded cysteine (x-axis) versus the stereoprobe and the orthosteric site (y-axis) for each protein with an unambiguous orthosteric assignment. d, Structures for ten diffusion samples of predictions for WX-74b binding to BTK. The rank 0 BTK structure is blue, ranks 1–9 are transparent. Orthosteric annotations on the rank 0 structure are denoted in green. The liganded cysteine BTK_C481 is yellow. All WX-74b ligands are pink. e, Structures for ten diffusion samples of predictions for EV-97 binding to IKBKB. The rank 0 IKBKB structure is blue, ranks 1–9 are transparent. Orthosteric annotations on the rank 0 structure are denoted in green. The liganded cysteine IKBKB_C464 is yellow. All EV-97 ligands are pink. Insets show the orthosteric site for IKBKB with incorrect predictions of EV-97 binding to this site (top inset) or with an orthosteric ligand K252a binding to this site in an experimental structure (PDB: 4KIK). f, Structures for ten diffusion samples of predictions of WX-63d binding to a CRBN:DDB1ΔBPB complex in the presence (left) or absence (right) of the orthosteric ligand lenalidomide. The rank 0 CRBN:DDB1ΔBPB structure is blue, ranks 1–9 are transparent. Orthosteric annotations on the rank 0 structure are denoted in green. The liganded cysteine CRBN_C219 is yellow. All WX-63d ligands are pink. g, Structures representing accurate Boltz-2 protein-ligand co-folding predictions for representative non-orthosteric stereoprobe liganding events with HIRA, ANKMY2, and INTS13 proteins (top), and the corruption of these predictions by fusing the proteins to a protein with a well-defined orthosteric pocket (EGFR kinase domain (green); bottom). The liganded cysteines in each image are yellow. Red dashed boxes highlight the location of stereoprobes in each co-folding prediction.

Fig. 3 |

Fig. 3 |. Characterization of elaborated stereoprobe-cysteine interactions.

a, Cysteine-directed ABPP results showing stereoselective engagement of FBL_C268 by WX-72d. Data represent average values ± SD for four independent experiments. b, Boltz-2 prediction for WX-72d binding to FBL showing ligand placement proximal to C268 (yellow) in the SAM/SAH binding pocket. c, Structures of alkynylated elaborated stereoprobes used to confirm liganding of FBL. d, Gel-ABPP data showing stereoselective engagement of recombinant FLAG epitope-tagged WT-FBL, but not a C268S-FBL mutant by WX-72d-yne (5 μM, 1 h) in HEK293T cells. UT, untransfected cells. e, Gel-ABPP data showing stereoselective blockade of WX-72d-yne (5 μM, 1 h) engagement of recombinant WT-FBL by WX-72d (20 μM, 1 h pretreatment) in HEK293T cells. f, Gel-ABPP data showing increased engagement of WT-FBL by WX-72d-yne in the presence of SAM but not SAH. Experiments were performed in lysates of HEK293T cells stably expressing WT-FBL. g, Quantification of gel-ABPP signals from f normalized to DMSO signals. Data represent average values ± SD for four independent experiments. h, Cysteine-directed ABPP results showing enantioselective engagement of TEX264_C165 by WX-62b and WX-62c. Data represent average values ± SD for four independent experiments. i, Structures of alkynylated elaborated stereoprobes used to confirm liganding of TEX264. j, Gel-ABPP data showing stereoselective engagement of recombinant FLAG epitope-tagged WT-TEX264 and a C68A-TEX264 mutant, but not a C165A-TEX264 mutant, by WX-62c-yne (5 μM, 1 h) in HEK293T cells. k, Gel-ABPP data showing enantioselective blockade of WX-62c-yne (5 μM, 1 h) engagement of recombinant WT-TEX264 by C6 elaborated probes WX-62b and WX-62c and C7 elaborated probe WX-72a (20 μM, 2 h) in HEK293T cells. l, Cysteine-directed ABPP results showing stereoselective engagement of NCKAP1L_C338 by WX-73a. Data represent average values ± SD for four independent experiments. m, Structures of alkynylated elaborated stereoprobes used to confirm liganding of NCKAP1L. n, Gel-ABPP data showing stereoselective engagement of recombinant FLAG epitope-tagged WT-NCKAP1L, but not a C338A-NCKAP1L mutant by WX-73a-yne (5 μM, 1 h) in THP1 cells. o, Gel-ABPP data showing stereoselective blockade of WX-73a-yne (5 μM, 1 h) engagement of WT-NCKAP1L by WX-73a (20 μM, 1 h pretreatment) in THP1 cells. p, Overlay of NCKAP1L AlphaFold-predicted structure (AF-P55160-F1-model_v4) with the WAVE regulatory complex crystal structure containing the NCKAP1L paralog NCKAP1 (PDB: 3P8C) showing location of C338 (yellow) relative to NCKAP1L hotspot mutations (red: M371V, R258L, P359L, V519L, R129W, V141F) that lead to human immunological disorder. q, Graph showing distance of NCKAP1L residues mutated in immunological disorders (red) to NCKAP1L_C338 (yellow) in AF-P55160-F1-model_v4. For d-f, j, k, n, o, gel-ABPP data represent stereoprobe treatments of cells transiently (TEX264) or stably (FBL and NCKAP1L) expressing epitope-tagged WT or C-to-A/S mutant. ABPP signals were measured by CuAAC conjugation to rhodamine-azide tag followed by SDS-PAGE and in-gel fluorescence scanning; data are from single experiments representative of at least two independent experiments.

Fig. 4 |

Fig. 4 |. Tryptoline acrylamide stereoprobes disrupt GRPEL1:HSPA9 interactions.

a, Cysteine-directed ABPP data showing enantioselective engagement of GRPEL1_C124 by WX-71b and WX-71c. Data represent average values ± SD of four independent experiments. b, Gel-ABPP data showing enantioselective engagement of recombinant FLAG epitope-tagged WT-GRPEL1, but not a C124A-GRPEL1 mutant by WX-01–06 and −07 (5 μM, 1 h) in HCT-116 cells. UT, untransfected cells. c, Gel-ABPP data showing concentration-dependent and enantioselective blockade of WX-01–07 (5 μM, 1 h) engagement of WT-GRPEL1 by WX-71b (top gel) and WX-71c (bottom gel) (2 h pretreatment). Top, representative gel-ABPP data; bottom, quantification of gel-ABPP data. Data represent average values ± SD for 2 independent experiments. d, Cryo-EM structure of GRPEL1 dimer (blue) in complex with HSPA9 (red/orange) (PDB: 9BLS) showing location of stereoprobe-liganded cysteine GRPEL1_C124 (yellow) at the GRPEL1:HSPA9 interface. SBD: substrate binding domain; NBD: nucleotide binding domain; GRPEL1-A, protomer 1; GRPEL1-B, protomer 2. e, IP-MS data showing stereoselective disruption of recombinant GRPEL1 interactions with endogenous HSPA9 following treatment with WX-71b or enantiomeric control WX-71d (20 μM, 4 h) in HCT-116 cells. HSPA9 signals for treatment groups were normalized to DMSO signals within each WT- and C124A-GRPEL1 group. Data represent average values ± SD for four independent experiments. f, IP-western blotting data showing time-dependent and enantioselective disruption of recombinant GRPEL1 interactions with endogenous HSPA9 and TIMM44 following treatment with WX-71b or enantiomeric control WX-71d (20 μM, 4–24 h) in HCT-116 cells. g, IP-MS data showing stereoselective disruption of recombinant FLAG epitope-tagged HSPA9 interactions with endogenous GRPEL1 following treatment with WX-71b or WX-71d (5 or 20 μM, 4 or 8 h) in HCT-116 cells. GRPEL1 signals for each treatment group were normalized to DMSO signals. Data represent average values ± SD for four independent experiments. h, Gel-ABPP data showing enantioselective and site-specific engagement of purified GRPEL1 (top) but not purified GRPEL1 in complex with HSPA9 (bottom) by WX-01–06 and −07 (10 μM, 2 h). i, Intact protein MS data for purified WT-GRPEL1 and C124A-GRPEL1 (3.5 μM) incubated with the indicated concentrations of WX-71b or WX-71d (4 h). Proteins were analyzed by time-of-flight (TOF)-LC/MS. Data show deconvoluted mass spectra from a single experiment representative of two independent experiments. j, Volcano plot for protein-directed ABPP data in parental HCT-116 cells showing blockade of proteins enriched by WX-01–06 by pre-treatment with WX-71b or WX-71d. Dashed vertical lines mark proteins showing greater than 3-fold stereoselective blockade by WX-71b or WX-71c. Proteins stereoselectively engaged by WX-71b that also show less than 2-fold engagement by WX-71c are marked in red. SF3B1 is in blue. Data represent average values ± SD for four independent experiments. Statistical significance was assessed using Welch two sample t-tests. For b, f, and h data are from a single experiment representative of at least two independent experiments.

Fig. 5 |

Fig. 5 |. GRPEL1 stereoprobes modulate mitochondrial protein import and function.

a, Microscope images showing localization of EGFP and Mitotracker Deep Red FM in MTS-EGFP-inducible parental HCT-116 cells treated with doxycycline (0.5 μg/mL) and WX-71b or WX-71d (20 μM, 8 h). b, Quantification of a where mitochondrial localization of EGFP was determined by colocalization with Mitotracker Deep Red FM using the Imaris 10.2 colocalization module. For each treatment group, 10 images per independent experiment were analyzed (Extended Data Fig. 5a shows one independent experiment with 10 images). Data represent average values ± SD of four independent experiments. Statistical significance was assessed using Welch two sample t-test. c, Microscope images showing localization of EGFP and Mitotracker Deep Red FM in MTS-EGFP-inducible sgGRPEL1 HCT-116 cells expressing recombinant FLAG epitope-tagged WT-GRPEL1 (top) or C124A-GRPEL1 (bottom) treated with doxycycline (0.5 μg/mL) and DMSO or WX-71b or WX-71d (20 μM, 8 h). d, Quantification of c, as described in panel b. Data represent average values ± SD of three independent experiments. Statistical significance was assessed using two-way ANOVA with Šídák’s multiple comparisons test (reported p values are adjusted p values). e, f, Mitochondrial respiration measured by a Seahorse mitochondrial stress test in (e) parental HCT-116 cells or (f) sgGRPEL1 HCT-116 cells expressing recombinant FLAG epitope-tagged WT-GRPEL1 (left) or C124A-GRPEL1 (right) treated with DMSO or WX-71b or WX-71d (5 μM, 8 h). Data represent average values ± SD of ten technical replicates from a single experiment representative of three independent experiments. g, Left, immunoblotting of integrated stress response (ISR) markers ATF4 and CHOP in sgGRPEL1 HCT-116 expressing recombinant Flag epitope-tagged WT- or C124A-GRPEL1 treated with DMSO or WX-71b or WX-71d (5 μM, 8 h). Right, quantification of blotting signals. Protein signals were normalized to FCCP-treated samples. Data represent average values ± SD of three independent experiments. Statistical significance was assessed using unpaired-multiple t-tests (Holm-Šídák approach to multiple comparisons; reported p values are adjusted p values). h, Parental HCT-116 cells stably expressing mt-mKeima and Parkin were treated with DMSO, FCCP (10 μM), or WX-71b or WX-71d (20 μM, 8 h), and mitophagy flux was measured by flow cytometry. Cells showing increased 561 nm/405 nm mt-mKeima ratios were considered mitophagy-positive. Data represent average values ± SD of three independent experiments. Statistical significance was assessed using ordinary one-way ANOVA with Dunnett’s multiple comparison (reported p values are adjusted p values). i, sgGREPL1 HCT-116 cells expressing mt-mKeima, Parkin, and Flag epitope-tagged WT- or C124A-GRPEL1 were treated with DMSO, FCCP (10 μM), or WX-71b or WX-71d (20 μM, 8 h), and mitophagy flux was measured by flow cytometry. Cells showing increased 561 nm/405 nm mt-mKeima ratios were considered mitophagy-positive. Data represent average values ± SD of three independent experiments. Statistical significance was assessed using two-way ANOVA with Šídák’s multiple comparisons test (reported p values are adjusted p values).

References

    1. Edfeldt F. N. B., Folmer R. H. A. & Breeze A. L. Fragment screening to predict druggability (ligandability) and lead discovery success. Drug Discov Today 16, 284–287 (2011). -PubMed
    1. Hopkins A. L. & Groom C. R. The druggable genome. Nat Rev Drug Discov 1, 727–730 (2002). -PubMed
    1. Spradlin J. N., Zhang E. & Nomura D. K. Reimagining Druggability Using Chemoproteomic Platforms. Acc Chem Res 54, 1801–1813 (2021). -PubMed
    1. Santos R. et al. A comprehensive map of molecular drug targets. Nat Rev Drug Discov 16, 19–34 (2017). -PMC -PubMed
    1. Dang C. V., Reddy E. P., Shokat K. M. & Soucek L. Drugging the ‘undruggable’ cancer targets. Nat Rev Cancer 17, 502–508 (2017). -PMC -PubMed

Publication types

Grants and funding

LinkOut - more resources