CRISPR screens and quantitative proteomics reveal remodeling of the aryl hydrocarbon receptor-driven proteome through PARP7 activity

Abstract

PARP7 is an enzyme that uses donor substrate NAD⁺ to attach a single ADP-ribose moiety onto proteins related to immunity, transcription, and cell growth and motility. Despite the importance of PARP7 in these processes, PARP7 signaling networks remain underresearched. Here, we used genome-wide CRISPR screens and multiplex quantitative proteomics in distinct lung cancer cell lines treated with a PARP7 inhibitor to better understand PARP7 molecular functions. We find that manipulating the aryl hydrocarbon receptor (AHR) transcriptional activity mediates PARP7 inhibitor sensitivity and triggers robust changes to the AHR-controlled proteome (AHR-ome). One of the striking features of such AHR-ome remodeling was the downregulation of filamins A and B concurrent with the induction of the corresponding E3 ubiquitin ligase ASB2. We also show that suppressor of cytokine signaling 3 (SOCS3) crosstalks to AHR. Inhibition of PARP7 in SOCS3 knockout cells leads to reduced viability compared to wild-type cells treated with a PARP7 inhibitor. Our results reveal signaling interplay between PARP7, AHR, and SOCS3 and establish an invaluable resource to study the role of PARP7 in the regulation of AHR signaling and innate immunity through its ADP-ribosyl transferase activity.

Keywords: ADP-ribosylation; CRISPR; PARP; aryl hydrocarbon receptor; proteomics.

Fig. 1.

A genome-wide CRISPR synthetic lethality screen with a panel of lung cancer cell lines identifies PARP7 inhibitor resistance hits. (A) Cell viability of selected cell lines treated with a range of concentrations of RBN2397 for 6 d measured in a CellTiter-Glo® assay. Data are shown as mean ± SEM of n = 3 biological replicates. (B) Schematic representation of the CRISPR screen workflow. (C) Overlap of resistance hits identified from the screen and compared to published studies (Dataset S4). The union of significant hits across the two timepoints (day 11 and 18) was used for each cell line of the screen. (D) Resistance hits identified from the genome-wide CRISPR screen in HCC44, SKMES1, and H838 cells with FDR < 0.1, shown as log₂ fold-change of RBN2397 treatment relative to DMSO. PARP7 (TIPARP) is shown in green.

Fig. 2.

Whole-proteome changes induced by PARP7 inhibition and AHR agonist and antagonist. (A) Cell viability of HCC44 cells treated with a range of concentrations of RBN2397 with or without 1 μM CH223191 or 1 μM tapinarof for 6 d, measured in a CellTiter-Glo^® assay. Data are shown as mean ± SEM of n = 3 biological replicates (2 technical replicates each). (B) TMT 18-plex quantitative proteomics workflow. (C–E) Quantitative proteomics (TMT 18-plex) volcano plots showing significantly upregulated (red) and downregulated (blue) proteins in HCC44 cells treated with 1 μM tapinarof, 1 μM RBN2397, and a combination of RBN2397 and tapinarof (both at 1 μM), respectively. (F) Western blot validation of up- and downregulated hits from proteomics experiments in panels C–E in HCC44 cells.

Fig. 3.

A genome-wide CRISPR synthetic lethality screen identifies SOCS3 as a conserved PARP7 inhibitor sensitivity hit. (A) Overlap of sensitivity hits identified from the screen and compared to published studies (Dataset S4). The union of significant hits across the two timepoints (day 11 and 18) was used for each cell line of the screen. (B) Sensitivity hits identified from the genome-wide CRISPR screen in HCC44, SKMES1, and H838 cells with FDR < 0.1, shown as log₂ fold-change of RBN2397 treatment relative to DMSO. (C) Western blotting analysis of SOCS3 knockout HCC44 cell lines compared to wild-type, treated with or without 1 µM RBN2397. STAT3 phosphorylation was detected with a phospho-STAT3-specific antibody. Western blot signal was quantified using Li-COR Odyssey ImageStudio software and used to calculate the pSTAT3/STAT3 ratio. Data are shown as mean ± SEM of n = 3 biological replicates, **P < 0.01. α-tubulin was used as a loading control. (D) CellTiter-Glo® assay on WT and SOCS3 knockout HCC44 cells. Cells were treated with a range of concentrations of RBN2397 for 6 d. Data are shown as mean ± SEM of n = 4 biological replicates (with 2 technical replicates each).

Fig. 4.

SOCS3 knockout boosts IFN- and AHR-regulated proteins upon PARP7 inhibition. (A) Volcano plot showing significantly upregulated (red) and downregulated (blue) proteins in SOCS3 KO compared to wild-type HCC44 cells. (B) Heatmap of proteomic changes to IFN-regulated proteins in SOCS3 knockouts compared to wild-type HCC44 cells with and without 1 µM RBN2397 treatment for 24 h. (C) Western blotting analysis of AHR and SOCS3 levels in HCC44 cells treated with 1 µM RBN2397 with or without 1 µM CH223191 and 1 µM tapinarof. Western blot signal was quantified using Li-COR Odyssey ImageStudio software. Data are shown as mean ± SEM of n = 3 biological replicates, *P < 0.05. α-tubulin was used as a loading control. (D) Western blotting analysis of AHR levels in wild-type and SOCS3 KO HCC44 cells treated with RBN2397 or DMSO control. Western blot signal was quantified using Li-COR Odyssey ImageStudio software. Data are shown as mean ± SEM of n = 3 biological replicates, **P < 0.01. α-tubulin was used as a loading control. (E) Heatmap of proteomic changes to AHR-regulated proteins (induced by the tapinarof/RBN2397 combination treatment in Fig. 2F) in SOCS3 knockouts compared to wild-type HCC44 cells with and without 1 µM RBN2397 treatment for 24 h.