Overexpression of the WWE domain of RNF146 modulates poly-(ADP)-ribose dynamics at sites of DNA damage

Abstract

Protein poly-ADP-ribosylation (PARylation) is a post-translational modification formed by transferring successive units of ADP-ribose to target proteins to form poly-ADP-ribose (PAR) chains. PAR plays a critical role in the DNA damage response (DDR) by acting as a signaling platform to promote the recruitment of DNA repair factors to the sites of DNA damage that bind via their PAR-binding domains (PBDs). Several classes of PBD families have been identified, which recognize distinct parts of the PAR chain. Proteins encoding PBDs play an essential role in conveying the PAR-mediated signal through their interaction with PAR chains, which mediates many cellular functions, including the DDR. The WWE domain, encoded in 12 human proteins, identifies the iso-ADP-ribose moiety of the PAR chain. PARylation is a heterogeneous structure that is highly dynamic in cells. Capturing the dynamics of PARylation is essential to understanding its role in the DDR, which can be achieved by expanding the tool kit for PAR detection and tracking mediated by the unique binding capability of various sensors. We recently described the WWE domain of RNF146 as a robust genetically encoded probe, when fused to EGFP, for the detection of PAR in live cells. Expanding on this, we used structural prediction tools to evaluate all of the WWE domains encoded in human proteins, evaluating each as molecular PAR probes in live cells. We demonstrate unique PAR dynamics when tracked by WWE-encoded PAR binding domains, in addition to an engineered macrodomain, that can be exploited for modulation of the PAR-dependent DNA damage response.

Keywords: DTX4; Macrodomain; PAR binding domains; Poly-ADP-ribose (PAR); RNF146; WWE domain.

Figure 1.. Evaluation of WWE domains as probes for the detection of PARylation

(A) Alignment of the amino acid sequences of the WWE domains in the twelve human proteins; (B) Graphic depicting a PAR binding domain (PBD) fused to EGFP that is used to detect levels of, and temporal dynamics of, PAR chains in live cells after induction of DNA damage by MNNG or laser micro-irradiation, leading to multiple fluorescent foci (MNNG) or a fluorescent focus (laser micro-irradiation).

Figure 2.. RNF146 encoded WWE domain-mediated detection of PARylation levels and temporal dynamics in live cells

(A) Confocal micrograph images of cells expressing different RNF146 encoded WWE-EGFP fusions in ES-2 cells following laser micro-irradiation at multiple time points, white scale bar denotes 20μm; (B) Plot depicting the recruitment dynamics of WWE domains RNF146(92–168) and RNF146(100–182) in ES-2 cells to sites of laser micro-irradiation (405nm laser), N ≥ 28 cells, recruitment intensity normalized to nucleus fluorescent intensity background at time t=0; (C) Relative recruitment peak intensity of the RNF146(100–182) and RNF146(92–168) WWE domains expressed in ES-2 cells. Each point represents a single cell recruitment focus, graph shows mean ± SEM; (D) Peak recruitment time of the RNF146(100–182) and RNF146(92–168) WWE domains expressed in ES-2 cells. Each point represents a single cell recruitment focus, graph shows mean ± SEM; (E) Plot depicting the dissociation dynamics of the RNF146(100–182) and RNF146(92–168) WWE domains foci for 20 minutes following laser-induced DNA damage in ES-2 cells, N ≥ 36 cells. Exclusion percentages were 0% for RNF146(100–182) and 33.3% (12 foci) for RNF146(92–168). NS: no significance, *p<0.05, **p<0.01, ***p<0.001, ****p<0.0001; a Mann-Whitney test for panels C and D and a Kaplan-Meier test for panel E.

Figure 3.. Analysis of RNF146, DTX4, PARP11, and DTX2 disorder prediction and structural organization

(A) Analysis of RNF146 disorder prediction and structural organization; PONDR plot illustrating predicted disorder based on residue number. Residues with a PONDR score above 0.5 (horizontal line) are predicted to be disordered, while those below the line are predicted to be ordered. A thick black line indicates a large region of disorder (top); AlphaFold-predicted structure of full-length RNF146, with the UniProt-defined WWE domain (residues Glu92–Arg168) highlighted in cyan. All other regions of the protein are shown in tan (middle); Full-length RNF146 structure with residues Val45–Arg182 highlighted in blue, representing the WWE domain and additional residues based on the AlphaFold prediction and PONDR plot data. All other regions of the protein are shown in tan (bottom); (B) Analysis of DTX4 disorder prediction and structural organization; PONDR plot illustrating predicted disorder based on residue number. Residues with a PONDR score above 0.5 (horizontal line) are predicted to be disordered, while those below the line are predicted to be ordered. A thick black line indicates a large region of disorder (top); AlphaFold-predicted structure of full-length DTX4, with the UniProt-defined tandem WWE domains (residues Met1-Arg155) highlighted in cyan. All other regions of the protein are shown in tan (middle); Full-length DTX4 structure with residues Met1–Gly166 highlighted in blue, representing the WWE domain and additional residues based on the AlphaFold prediction and PONDR plot data. All other regions of the protein are shown in tan (bottom); (C) Analysis of PARP11 disorder prediction and structural organization; PONDR plot illustrating predicted disorder based on residue number. Residues with a PONDR score above 0.5 (horizontal line) are predicted to be disordered, while those below the line are predicted to be ordered. A thick black line indicates a large region of disorder (top); AlphaFold-predicted structure of full-length PARP11, with the UniProt-defined WWE domain (residues Asn22-Arg106) highlighted in cyan. All other regions of the protein are shown in tan (middle); Full-length PARP11 structure with residues Asn22-Ser112 highlighted in blue, representing the WWE domain and additional residues based on the AlphaFold prediction and PONDR plot data. All other regions of the protein are shown in tan (bottom); (D) Analysis of DTX2 disorder prediction and structural organization; PONDR plot illustrating predicted disorder based on residue number. Residues with a PONDR score above 0.5 (horizontal line) are predicted to be disordered, while those below the line are predicted to be ordered. A thick black line indicates a large region of disorder (top); AlphaFold-predicted structure of full-length DTX2, with the UniProt-defined tandem WWE domains (residues Ser8-Arg174) highlighted in cyan. All other regions of the protein are shown in tan (middle); Full-length DTX2 structure with residues Met1–Pro178 highlighted in blue, representing the WWE domain and additional residues based on the AlphaFold prediction and PONDR plot data. All other regions of the protein are shown in tan (bottom).

Figure 4.. Structurally predicted extended WWE domains as probes for detection of MNNG-induced PARylation

(A) Confocal micrograph images of U2OS cells expressing RNF146(100–182)-EGFP fusion after treatment with MNNG (10μM), PARGi (PDD00017273, 10μM) and NRH (100μM) or MNNG (10μM), PARGi (PDD00017273, 10μM), NRH (100μM) and PARPi (Talazoparib, 10μM) compared to media only (left); Plot depicting PAR foci per nucleus after treatment with MNNG (10μM), PARGi (PDD00017273, 10μM), and NRH (100μM) or MNNG (10μM), PARGi (PDD00017273, 10μM), NRH (100μM), and PARPi (Talazoparib, 10μM), each point represents foci count per nucleus, n=3 (≥ 150 cells), graph shows mean ± SEM (right); (B) Confocal micrograph images of U2OS cells expressing RNF146(45–182)-EGFP fusion after treatment with MNNG (10μM), PARGi (PDD00017273, 10μM), and NRH (100μM) or MNNG (10μM), PARGi (PDD00017273, 10μM), NRH (100μM) and PARPi (Talazoparib, 10μM) compared to media only (left); Plot depicting PAR foci per nucleus after treatment with MNNG (10μM), PARGi (PDD00017273, 10μM), and NRH (100μM) or MNNG (10μM), PARGi (PDD00017273, 10μM), NRH (100μM), and PARPi (Talazoparib, 10μM), each point represents foci count per nucleus, n=3 (≥ 150 cells), graph shows mean ± SEM (right); (C) Confocal micrograph images of U2OS cells expressing DTX4(1–166)-EGFP fusion after treatment with MNNG (10μM), PARGi (PDD00017273, 10μM), and NRH (100μM) or MNNG (10μM), PARGi (PDD00017273, 10μM), NRH (100μM), and PARPi (Talazoparib, 10μM) compared to media only (left); Plot depicting PAR foci per nucleus after treatment with MNNG (10μM), PARGi (PDD00017273, 10μM), and NRH (100μM) or MNNG (10μM), PARGi (PDD00017273, 10μM), NRH (100μM), and PARPi (Talazoparib, 10μM), each point represents foci count per nucleus, n=3 (≥ 150 cells), graph shows mean ± SEM (right); (D) Confocal micrograph images of U2OS cells expressing PARP11(22–112)-EGFP fusion after treatment with MNNG (10μM), PARGi (PDD00017273, 10μM), and NRH (100μM) or MNNG (10μM), PARGi (PDD00017273, 10μM), NRH (100μM), and PARPi (Talazoparib, 10μM) compared to media only (left); Plot depicting PAR foci per nucleus after treatment with MNNG (10μM), PARGi (PDD00017273, 10μM), and NRH (100μM) or MNNG (10μM), PARGi (PDD00017273, 10μM), NRH (100μM), and PARPi (Talazoparib, 10μM), each point represents foci count per nucleus, n=3 (≥ 150 cells), graph shows mean ± SEM (right). NS: no significance, *p<0.05, **p<0.01, ***p<0.001, ****p<0.0001; a Kruskal-Wallis test was used for panels A-D.

Figure 5.. WWE domain of DTX4(1–166), PARP11(22–112), and DTX2(1–178) as molecular probes to track PARylation dynamics in live cells

(A) Confocal micrograph images of U2OS cells expressing EGFP-fused RNF146(100–182) and EGFP-fused DTX4(1–166) following laser micro-irradiation; (B) Recruitment of RNF146(100–182) and DTX4(1–166), to sites of laser micro-irradiation (405nm) following BrdU sensitization (10μM, 24 hours), N≥40 cells, recruitment intensity normalized to first frame (F/F₀: Maximum Fluorescence intensity / Fluorescence intensity at t₀); (C) Relative peak intensity of recruitment for RNF146(100–182) and DTX4(1–166) in U2OS cells. Each point represents a single cell recruitment focus, graph shows mean ± SEM. (F/F₀: Maximum Fluorescence intensity / Fluorescence intensity at t₀); (D) Peak recruitment time for RNF146(100–182) and DTX4(1–166) expressed in U2OS cells. Each point represents a single cell recruitment focus, graph shows mean ± SEM; (E) Plot depicting the dissociation dynamics of RNF146(100–182) and DTX4(1–166) foci in U2OS cells during 20 minutes following laser-induced DNA damage, N ≥ 40 cells; (F) Confocal micrograph images of U2OS cells expressing EGFP-fused RNF146(100–182) and EGFP-fused PARP11(22–112) following laser micro-irradiation; (G) Recruitment of RNF146(100–182) and PARP11(22–112), to sites of laser micro-irradiation (405nm) following BrdU sensitization (10μM, 24 hours), N≥40 cells, recruitment intensity normalized to the first frame (F/F₀: Maximum Fluorescence intensity / Fluorescence intensity at t₀); (H) Relative peak intensity of recruitment for RNF146(100–182) and PARP11(22–112) in U2OS cells. Each point represents a single cell recruitment focus, graph shows mean ± SEM. (F/F₀: Maximum Fluorescence intensity / Fluorescence intensity at t₀); (I) Peak recruitment time for RNF146(100–182) and PARP11(22–112) expressed in U2OS cells. Each point represents a single cell recruitment focus, graph shows mean ± SEM; (J) Plot depicting the dissociation dynamics of RNF146(100–182) and PARP11(22–112) foci in U2OS cells during 20 minutes following laser-induced DNA damage, N ≥ 40 cells; (K) Confocal micrograph images of U2OS cells expressing EGFP-fused RNF146(100–182) and EGFP-fused DTX2(1–178) following laser micro-irradiation; (L) Recruitment of RNF146(100–182) and DTX2(1–178), to sites of laser micro-irradiation (405nm) following BrdU sensitization (10μM, 24 hours), N≥40 cells, recruitment intensity normalized to the first frame (F/F₀: Maximum Fluorescence intensity / Fluorescence intensity at t₀); (M) Relative peak intensity of recruitment for RNF146(100–182) and DTX2(1–178) in U2OS cells. Each point represents a single cell recruitment focus, graph shows mean ± SEM. (F/F₀: Maximum Fluorescence intensity / Fluorescence intensity at t₀); (N) Peak recruitment time for RNF146(100–182) and DTX2(1–178) expressed in U2OS cells. Each point represents a single cell recruitment focus, graph shows mean ± SEM; (O) Plot depicting the dissociation dynamics of RNF146(100–182) and DTX2(1–178) foci in U2OS cells during 20 minutes following laser-induced DNA damage, N ≥ 40 cells; Recruitment foci having a relative peak intensity below 1.15 of the first frame were excluded from the experiment and from statistical analysis in graphs (B-E, G-J, L-O). Exclusion percentages were 2.22% (1 focus) for RNF146(100–182), 8.33% (4 foci) for DTX4(1–166), 37.5% (15 foci) for PARP11(22–112) and 50% (24 foci) for DTX2(1–178). NS: no significance, *p<0.05, **p<0.01, ***p<0.001, ****p<0.0001; a Mann-Whitney test was used for panels C, D, H, I, M, and N, and a Kaplan-Meier test for panels E, J and O.

Figure 6.. Effect of disorder prediction and structural organization of RNF146 and DTX4 on the tracking of PARylation dynamics

(A) Confocal micrograph images of cells expressing EGFP fusions with RNF146(100–182), RNF146(45–182) and RNF146(92–168) in U2OS cells, following laser micro-irradiation; (B) Recruitment of RNF146(100–182), RNF146(45–182) and RNF146(92–168) to sites of laser micro-irradiation (405nm) following BrdU sensitization (10μM, 24 hours), N≥32 cells, recruitment intensity normalized to the first frame (F/F₀: Maximum Fluorescence intensity / Fluorescence intensity at t₀); (C) Relative peak intensity of recruitment for RNF146(100–182), RNF146(45–182) and RNF146(92–168) in U2OS cells. Each point represents a single cell recruitment focus, graph shows mean ± SEM. (F/F₀: Maximum Fluorescence intensity / Fluorescence intensity at t₀); (D) Peak recruitment time for RNF146(100–182), RNF146(45–182) and RNF146(92–168) expressed in U2OS cells. Each point represents a single cell recruitment focus, graph shows mean ± SEM; (E) Plot depicting the dissociation dynamics of RNF146(100–182) and RNF146(92–168) foci in U2OS cells during 20 minutes following laser-induced DNA damage, N ≥ 40 cells; (F) Plot depicting the dissociation dynamics of RNF146(100–182) and RNF146(45–182) foci in U2OS cells during 20 minutes following laser-induced DNA damage, N ≥ 32 cells; (G) Confocal micrograph images of U2OS cells expressing EGFP fusions with DTX4(1–166), DTX4(1–155), and DTX4(1–78), following laser micro-irradiation; (H) Recruitment of DTX4(1–166), DTX4(1–155), and DTX4(1–78) to sites of laser micro-irradiation (405nm) following BrdU sensitization (10μM, 24 hours), N≥32 cells, recruitment intensity normalized to first frame (F/F₀: Maximum Fluorescence intensity / Fluorescence intensity at t₀); (I) Relative peak intensity of recruitment for DTX4(1–166), DTX4(1–155), and DTX4(1–78) in U2OS cells. Each point represents a single cell recruitment focus, graph shows mean ± SEM. (F/F₀: Maximum Fluorescence intensity / `Fluorescence intensity at t₀); (J) Peak recruitment time for DTX4(1–166), DTX4(1–155), and DTX4(1–78) expressed in U2OS cells. Each point represents a single cell recruitment focus, graph shows mean ± SEM; (K) Plot depicting the dissociation dynamics of DTX4(1–166) and DTX4(1–155) foci in U2OS cells during 20 minutes following laser-induced DNA damage, N ≥ 40 cells; (L) Plot depicting the dissociation dynamics of DTX4(1–166) and DTX4(1–78) foci in U2OS cells during 20 minutes following laser-induced DNA damage, N ≥ 32 cells. Recruitment foci having a relative peak intensity below 1.15 of the first frame were excluded from the experiment and from statistical analysis in graphs (B-F, H-L). Exclusion percentages were 9.38% (3 foci) for RNF146(45–182), 4.88% (2 foci) for RNF146(92–168), 7.5% (3 foci) for DTX4(1–166), 15% (6 foci) for DTX4(1–155) and 43.75% (14 foci) for DTX4(1–78). No exclusion was made in RNF146(100–182. NS: no significance, *p<0.05, **p<0.01, ***p<0.001, ****p<0.0001; a Kruskal-Wallis test was used for panels C, D, I, and J, and a Kaplan-Meier test for panels E, F, K and L.

Figure 7.. WWE domain of RNF146(100–182) and the engineered Af1521(K35E/Y145R) macrodomain as molecular probes to track PARylation dynamics in live cells

(A) Schematic representation of the RNF146(100–182) WWE domain (fused to EGFP) or of the Af1521 macrodomain (fused to EGFP) bound to PAR chains at sites of DNA damage; (B) Confocal micrograph images of EGFP-fused RNF146(100–182), Af1521(WT) or Af1521(K35E/Y145R) expressed in U2OS cells, white scale bar denotes 20μm; (C) Confocal micrograph images of cells expressing EGFP fusions with RNF146(100–182), Af1521(K35E/Y145R), and Af1521(WT) in U2OS cells, following laser micro-irradiation; (D) Recruitment of RNF146(100–182), Af1521(WT), and Af1521(K35E/Y145R), to sites of laser micro-irradiation (405nm) following BrdU sensitization (10μM, 24 hours), N≥25 cells, recruitment intensity normalized to first frame (F/F₀: Maximum Fluorescence intensity / Fluorescence intensity at t₀); (E) Relative peak intensity of recruitment for RNF146(100–182), Af1521(WT), and Af1521(K35E/Y145R), in U2OS cells. Each point represents a single cell recruitment focus, graph shows mean ± SEM. (F/F₀: Maximum Fluorescence intensity / Fluorescence intensity at t₀); (F) Peak recruitment time for RNF146(100–182), Af1521(WT), and Af1521(K35E/Y145R) expressed in U2OS cells. Each point represents a single cell recruitment focus, graph shows mean ± SEM; (G) Plot depicting the dissociation dynamics of RNF146(100–182), Af1521(K35E/Y145R), and Af1521(WT) foci in U2OS cells during 20 minutes following laser-induced DNA damage, N ≥ 40 cells. Recruitment foci having a relative peak intensity below 1.15 of the first frame were excluded from the experiment and from statistical analysis in graphs (D-G). Exclusion percentages were 5% (2 foci) for Af1521(K35E/Y145R) and 37.5% (15 foci) for Af1521(WT). No exclusion was made in RNF146(100–182). NS: no significance, *p<0.05, **p<0.01, ***p<0.001, ****p<0.0001; a Kruskal-Wallis test was used for panels E and F and a Kaplan-Meier test for panel E.

Figure 8.. Overexpression of the RNF146(100–182) WWE domain modulates PAR levels and dynamics at sites of laser micro-irradiation

(A) Schematic representation of the RNF146(100–182) WWE domain (fused to a myc-tag) and the engineered macrodomain Af1521(K35E/Y145R) (fused to EGFP). Af1521(K35E/Y145R)-EGFP is used for tracking PAR levels, and the modulation of PAR dynamics, impacted by the expression of RNF146(100–182)-myc; (B) Recruitment of Af1521(K35E/Y145R)-EGFP in U2OS cells (left) and in ES-2 cells (right), after overexpression of RNF146(100–182)-myc, to sites of laser micro-irradiation (405nm) following BrdU sensitization (10μM, 24 hours), N≥49 cells, recruitment intensity normalized to the first frame, (F/F₀: Maximum Fluorescence intensity / Fluorescence intensity at t₀); (C) Relative peak intensity of recruitment for Af1521(K35E/Y145R)-EGFP, after overexpression of RNF146(100–182)-myc, in U2OS cells (left) and in ES-2 cells (right). Each point represents a single-cell recruitment focus. Graph shows mean ± SEM (F/F₀: Maximum Fluorescence intensity/ Fluorescence intensity at t₀), graph shows mean ± SEM. Af1521(K35E/Y145R)-EGFP foci having a relative peak intensity below 1.15 of the first frame were excluded from the experiment and from statistical analysis in graphs (C-E). NS: no significance, *p<0.05, **p<0.01, ***p<0.001, ****p<0.0001; (D) Peak recruitment time for Af1521(K35E/Y145R)-EGFP, after overexpression of RNF146(100–182)-myc, in U2OS cells (left) and in ES-2 cells (right). Each point represents a single cell recruitment focus, graph shows mean ± SEM. Af1521(K35E/Y145R)-EGFP foci having a relative peak intensity below 1.15 of the first frame were excluded from the experiment and from statistical analysis in graphs (C-E). NS: no significance, *p<0.05, **p<0.01, ***p<0.001, ****p<0.0001; (E) Plot depicting the dissociation dynamics of Af1521(K35E/Y145R) foci after overexpression of RNF146(100–182) in U2OS cells (left) and in ES-2 cells (right) during 20 minutes following laser-induced DNA damage. N ≥ 48 cells. Af1521(K35E/Y145R) EGFP foci having a relative peak intensity below 1.15 of the first frame were excluded from the experiment and from statistical analysis in graphs (C-E). Exclusion percentages were 4% (2 foci) for Af1521(K35E/Y145R) and no exclusion for Af1521(K35E/Y145R) foci after overexpression of RNF146(100–182); (F) Immunoblot probing PAR levels in cells expressing Af1521(K35E/Y145R)-EGFP and/or the RNF146(100–182)-myc and after treatment with the PARG_i PDD00017273 (10μM, 8 hours) and MNNG (20μM, 1 hour); a Kruskal-Wallis test was used for panel C, a Mann-Whitney test for panel D and a Kaplan-Meier test for panel E.