Discovery of a paralog-selective p300 protein degrader with potent anti-cancer activity in hematological malignancies Authors: Marwa S. Asem, Yan Zhai, Xiaohong Song, Milad Rouhimoghadam, Sreenivas Punna, Fritz G. Buchanan, Daniel T. Cohen, Ryan McClure, Stephanie Sandoval, Anlu Chen, Shaun McLoughlin, Colin Woodford, Peter Kovar, Vlasios Manaves, Alla V. Korepanova, Justin M. Reitsma, Andrea Shergalis, Judith A. Ronau, Yifei Kong, Yu Shen, Jurgen Dinges Published: 2026 ABSTRACT The E1A-associated protein p300 (EP300) is a key regulator of oncogenic transcription factors, making it a promising target for cancer therapy. However, its high sequence similarity to its paralog, CREB-binding protein (CBP), has hindered the development of selective inhibitors, leading to dose-limiting toxicities. Here, we describe the discovery of a highly potent and selective p300 degrader. Unlike dual p300/CBP degraders, this compound forms a more stable ternary complex with p300, driving enhanced proteasomal recruitment and ubiquitination. Notably, our data uncover a previously unrecognized mechanism of paralog selectivity mediated by regioselective ubiquitination of a unique lysine residue on p300. Hematological malignancies, including multiple myeloma, non-Hodgkin lymphoma, and acute myeloid leukemia, exhibit marked sensitivity to selective p300 degradation, resulting in cell lethality and robust antitumor activity in xenograft models. These findings establish selective p300 degradation as a mechanistically distinct and promising therapeutic strategy in hematological malignancies. Introduction The E1A-associated protein p300 (p300) is a pivotal epigenetic regulator that modulates chromatin structure dynamics and the activity of several DNA-binding transcription factors1. This ~300 kDa protein can promote gene transcription directly through its histone acetyltransferase (HAT) domain, which can transfer acetyl groups to histone proteins at gene promoters, enhancers, and super-enhancers, leading to active transcription at these regions2. In a more indirect approach, p300 can acetylate transcription factors like p53 and MYC, transcriptional co-activators, and nuclear receptors such as the androgen receptor (AR), regulating their stability and activity in cells3. Furthermore, p300 contains a bromodomain (BRD), which can recognize acetylated lysine on histone proteins to facilitate the recruitment of multi-protein transcription complexes and serves as a scaffold between the transcription machinery proteins and the transcription factors2,4. As a result, p300 is involved in the regulation of active transcription and a multitude of cellular signaling pathways that regulate cell metabolism, proliferation, survival, and DNA replication2,3,5. Given the vast repertoire of p300’s physiological functions in the nucleus, it is often implicated in regulating the activity of oncogenic transcription factors, leading to tumor progression6–8. For example, in prostate cancer, p300 is overexpressed and acetylates the phosphorylated AR at the K609 position, resulting in AR stabilization and activation9,10. Moreover, p300 forms a complex with and acetylates S-phase kinase-associated protein-2 (Skp2) at K68 and K71, leading to enhanced Skp2 stability and activity, and promoting its downstream oncogenic signaling in prostate cancer11. In acute myeloid leukemia (AML), the clinical TARGET dataset shows a significant correlation between EP300 expression and poor overall survival time (OS) in MLL-r AML patients12. Additional studies show that p300 is a key co-activator of Myeloblastosis (Myb) proto-oncogene, and this interaction is essential for the ability of AML1-ETO and MLL fusions to induce self-renewal properties in myeloid progenitor cells and activate their downstream transcriptional programs13–15. On the other hand, p300 acetylates the fusion oncoprotein AML1-ETO in leukemia cells isolated from t(8;21) AML patients, and this acetylation process is indispensable for leukemogenicity and self-renewal in human cord blood CD34+ cells15,16. In multiple myeloma (MM), p300 inhibition led to the downregulation of the master transcriptional factor Interferon Regulatory Factor 4 (IRF4) at the mRNA and protein levels, and reduced its acetylation, leading to the abrogation of its transcriptional activity and inhibiting MM cell proliferation17. p300 was also found to regulate proliferation, cell cycle, and metabolism in several cancers, including gastric and breast cancers, melanoma, and lymphoma, making it an attractive target for cancer therapy1. p300 and its paralog protein, CREB-binding protein (CBP), are believed to originate from a common ancestral gene via gene duplication and have remarkably high sequence homology overall (70%), and in their domains (HAT: 88% and BRD: 96%)3. Based on their origin, they mostly have overlapping functions, and, due to their commonalities, they are often referred to as one entity (CBP/p300)18. Although a wide variety of small molecule inhibitors for many of the domains in CBP and p300 have been published, none of them have yet demonstrated clinical success with only two dual inhibitors, CCS1477 and FT-7051, are currently undergoing clinical evaluation, highlighting continued efforts to therapeutically target this axis6,19. Owing to the high sequence homology between the two proteins, no inhibitor has been able to achieve sufficient paralog selectivity, resulting in dose-limiting toxicities from dual CBP/p300 inhibition19,20. Bifunctional protein degrader technology has evolved as a promising modality in drug development, with 20 degraders having proceeded to the clinical stage21. Commonly, degraders consist of a target binding ligand (TBL), which is based on small molecule inhibitors of the target protein, a linker, and a ligase binding ligand (LBL) that recruits the ubiquitin–proteasome machinery to induce the degradation of the protein of interest (POI)22. A few degraders targeting p300/CBP have been published, with only three showing preferential degradation of p300 over CBP23–27. Here, we report the discovery of a highly potent and paralog-selective p300 degrader, Degrader 27. Biochemical analyses confirm that paralog selectivity arises from differences in degradation kinetics between p300 and CBP proteins. Degrader 27 preferentially ubiquitinates a unique lysine on the surface of p300 that is not present on CBP, leading to selective degradation of p300. Degrader 27 demonstrates potent antiproliferative activity in hematological cancer cells, with a strong inhibitory effect on the expression of master hematopoietic transcription factors, resulting in the induction of cell cycle arrest and subsequent apoptosis in hematological cancer cells. Furthermore, Degrader 27 achieves antitumor efficacy and better tolerability compared to the dual p300/CBP degrader in MM and non-Hodgkin lymphoma (NHL) xenograft models. Collectively, these results suggest that the selective degradation of p300 is sufficient for inducing potent and tolerable antitumor activity and may represent a promising strategy for hematological cancer therapy. Results Design and synthesis of p300 degrader molecules To maximize our chances of identifying paralog-selective degraders, we synthesized a large and diverse library of 1775 degrader molecules. The overall design strategy and experimental workflow are: (i) selection of diverse, literature-reported p300/CBP inhibitors that target the four key domains: HAT, BRD, the cysteine/histidine-rich 1 (CH1), and Nuclear Receptor Interaction domain (NRID)28; (ii) identification of suitable “exit vectors” (EVs) on each inhibitor for potential linker attachment, based on structural and modeling data; (iii) systematic installation of chemical linkers at the selected positions; (iv) combinatorial coupling with a panel of E3 ligase ligands; and (v) generation of the final library of 1775 potential degraders (Fig. 1a). Figure 1b provides essential context for the subsequent chemical structures, clarifying how the selected inhibitors serve as the foundation for library construction. Numerous CRBN- and VHL-based ligase binding ligands (LBLs, 16–22) were utilized in the synthesis of degrader compounds (Supplementary Fig. 1)29–38.Fig. 1Rational design and synthesis of paralog-selective p300 degraders.a Schematic overview of the structure-guided workflow to generate a diverse p300/CBP degrader library. Created in BioRender. Asem, M. (2026) https://BioRender.com/fn2v7aj. b The eleven most potent and selective CBP/p300 inhibitors, which bind to four different domains of p300 and CBP: Histone Acetyltransferase (HAT), Bromodomain (BRD), cysteine/histidine-rich 1 (CH1) domain and the Receptor Interaction domain (NRID). The possible EVs are marked with arrows. Rational design and synthesis of paralog-selective p300 degraders. a Schematic overview of the structure-guided workflow to generate a diverse p300/CBP degrader library. Created in BioRender. Asem, M. (2026) https://BioRender.com/fn2v7aj. b The eleven most potent and selective CBP/p300 inhibitors, which bind to four different domains of p300 and CBP: Histone Acetyltransferase (HAT), Bromodomain (BRD), cysteine/histidine-rich 1 (CH1) domain and the Receptor Interaction domain (NRID). The possible EVs are marked with arrows. For the TBLs, we selected the eleven most potent and structurally diverse p300/CBP inhibitors, providing a direct comparison across different domains (Fig. 1b). To determine suitable positions for linker attachment, we analyzed solvated and protein-bound crystal structures (sources: RCSB PDB and published models). Specifically, we identified all solvent-exposed positions referred to as exit vectors (arrows in Fig. 1b) where chemical modification is likely to preserve target engagement after conjugation. Structural information was accessible for eight out of eleven inhibitors (1: 7LJE, 2: 6V8N, 3: 5W0E, 4: 7XH6, 5: 4NR7, 6: 5I8G, 7: 7KPY, 8: 5J0D); published docking, homology models, or SAR informed the others. For example, FT-6876 (9) was docked into the available CBP-BRD domain due to lack of public structure (Supplementary Data 1), while for compound 26i (10), we used the previously described model by Wang et al.39. For the NRID-targeting scaffold (11), where no structural data are reported, the exit vector position was inferred from a biotinylated analog40. Exit vector selection is justified as follows: only those positions that are solvent-exposed in the protein-inhibitor complex or predicted to tolerate derivatization without affecting binding were chosen. This structural basis ensures that linker installation is both feasible and functionally rational. The variable number of exit vectors per scaffold directly reflects differences in core chemotypes, their binding modes, and the presence (or absence) of modifiable groups in solvent-accessible orientations. Overall, 1775 compounds were generated. Functional analysis of the compounds in cells To evaluate the cellular degradation potency and p300-selectivity of the produced compounds, we used CRISPR-Cas9 to knock in a luminescent HiBiT-tag at the C-terminus of the endogenous p300 or CBP proteins in U2OS cells and measured protein degradation after 24 h of treatment with the compounds. Only 21 compounds (1.2%) induced p300 degradation with DC50 values < 1 μM. Active degraders were identified with TBLs based on A-485 (1), CPI-637 (6) and FT-6876 (9), but the only degrader with paralog selectivity contained the GNE-781 (3) ligand (Fig. 1b). Degrader 23 utilized exit vector 2 and had a p300 DC50 = 524 nM with a Dmax = 80%, while no degradation of CBP was detected (Table 1, and Supplementary Fig. 2a). This result was confirmed by an additional immunoassay using capillary-based electrophoresis (Supplementary Fig. 2b). As expected, there was no selectivity in binding to the bromodomains of the two proteins (AlphaLISA™ p300 IC50 = 8.9 nM and CBP IC50 = 12.8 nM) (Table 1). For CRBN binding, 23 registered an IC50 of 45.5 nM (Table 1). The knock-in of a HiBiT-tag into BRD4 in HEK293T cells allowed to determine that 23 was fully selective over its closest BRD domain family member (DC50 > 10 μM) (Supplementary Fig. 2a).Table 1Optimization of LBL and linkerCompoundStructurep300CBPCRBNDC50a,b (nM)Dmaxa,c (%)IC50d (nM)DC50a,b (nM)Dmaxa,c (%)IC50d (nM)IC50d (nM)23524808.9>10,000012.845.524>10,0000N/T>10,0000N/T7810251518236.1>10,000032.448.3261458623.7>10,000021.259.02722.58620.0>10,000026.1190.02817.49012.610.28920.614.8aThe degradation of p300 and CBP was determined in 24 h HiBiT assays in U2OS cells.bDegrader concentrations needed to reduce the POI by 50%.cMaximum percent reduction of the POI.dThe binding of the degrader to the respective protein was assessed in an AlphaLISA™ assay. Table 1Optimization of LBL and linkerCompoundStructurep300CBPCRBNDC50a,b (nM)Dmaxa,c (%)IC50d (nM)DC50a,b (nM)Dmaxa,c (%)IC50d (nM)IC50d (nM)23524808.9>10,000012.845.524>10,0000N/T>10,0000N/T7810251518236.1>10,000032.448.3261458623.7>10,000021.259.02722.58620.0>10,000026.1190.02817.49012.610.28920.614.8aThe degradation of p300 and CBP was determined in 24 h HiBiT assays in U2OS cells.bDegrader concentrations needed to reduce the POI by 50%.cMaximum percent reduction of the POI.dThe binding of the degrader to the respective protein was assessed in an AlphaLISA™ assay. Optimization of LBL and linker aThe degradation of p300 and CBP was determined in 24 h HiBiT assays in U2OS cells. bDegrader concentrations needed to reduce the POI by 50%. cMaximum percent reduction of the POI. dThe binding of the degrader to the respective protein was assessed in an AlphaLISA™ assay. With respect to the degradation mechanism, ternary complex formation (TCF) was demonstrated in an AlphaLISA™ proximity assay41. The N-methylated version of 23, compound 24 (Table 1), sustained binding to the POIs, but exhibited significantly weakened affinity to CRBN and, consequently, no longer degraded p300. Together, these results substantiate that 23 triggers the cereblon-dependent ubiquitination of p300 and its subsequent proteasomal degradation. The potency and complete paralog-selectivity of compound 23 made it an excellent starting point for hit expansion work. Attempts to either modify the pomalidomide (13) or to completely replace it led to Degrader 25 with a naphthyl dihydrouracil (Table 1). The compound was about equipotent to 23 and maintained the paralog selectivity (Table 1, and Fig. 2a–c). Modifications of the linker showed that the “sweet spot” for the linker length was 8-9 heavy atoms, depending on the complexity of the linker (Supplementary Fig. 3). Otherwise, the structure-activity relationship (SAR) was very narrow, particularly for paralog selectivity, and only the replacement of the 3-aminocyclobutane carboxamide with its six-membered ring analog (26) preserved the activity and selectivity profile of the parent compound (Table 1, and Fig. 2a, b). Both rings had to have a cis configuration (Table 1).Fig. 2Degraders induce potent and selective p300 degradation in cells.a Dose-dependent degradation of p300-HiBiT, or CBP-HiBiT in U2OS cells, and BRD4-HiBiT in HEK293T cells following treatment with the indicated degraders for 24 h. HiBiT protein levels were then quantified using the lytic detection system. Data are represented as mean %RLU relative to DMSO-treated sample (n = 3; representative experiment shown, data are shown as mean ± SEM). b Immunoassay analysis of p300 and CBP protein levels in U2OS cells treated with the indicated degraders for 24 h in dose-response experiments (n = 3; representative experiment shown). c Volcano plots showing changes in protein abundance in MM.1S cells treated with degraders for 6 h, and proteins were analyzed by Orbitrap Mass Spectrometer. The X-axis displays the Log2 fold change in all identified proteins calculated relative to DMSO-treated sample. The Y-axis displays the descriptive P value (log10) from duplicate biological replicates. No statistical inference or hypothesis testing performed. d U2OS cells were treated with the indicated degraders for 24 h and then processed for H3K27Ac and H3 fluorescent imaging via high content microscopy. Data are represented as mean %fluorescent intensity relative to DMSO-treated sample (n = 3; representative experiment shown, data are shown as mean ± SEM). e Dose-dependent degradation of p300-HiBiT in U2OS cells following 1 h of pre-treatment with indicated inhibitor (GNE-781 (3), 10 µM; Pomalidomide, 1 µM; MG132, 1 µM) and 24 h of treatment with indicated degraders. Data are represented as mean %RLU relative to DMSO-treated sample (n = 3; representative experiment shown, data are shown as mean ± SEM). Degraders induce potent and selective p300 degradation in cells. a Dose-dependent degradation of p300-HiBiT, or CBP-HiBiT in U2OS cells, and BRD4-HiBiT in HEK293T cells following treatment with the indicated degraders for 24 h. HiBiT protein levels were then quantified using the lytic detection system. Data are represented as mean %RLU relative to DMSO-treated sample (n = 3; representative experiment shown, data are shown as mean ± SEM). b Immunoassay analysis of p300 and CBP protein levels in U2OS cells treated with the indicated degraders for 24 h in dose-response experiments (n = 3; representative experiment shown). c Volcano plots showing changes in protein abundance in MM.1S cells treated with degraders for 6 h, and proteins were analyzed by Orbitrap Mass Spectrometer. The X-axis displays the Log2 fold change in all identified proteins calculated relative to DMSO-treated sample. The Y-axis displays the descriptive P value (log10) from duplicate biological replicates. No statistical inference or hypothesis testing performed. d U2OS cells were treated with the indicated degraders for 24 h and then processed for H3K27Ac and H3 fluorescent imaging via high content microscopy. Data are represented as mean %fluorescent intensity relative to DMSO-treated sample (n = 3; representative experiment shown, data are shown as mean ± SEM). e Dose-dependent degradation of p300-HiBiT in U2OS cells following 1 h of pre-treatment with indicated inhibitor (GNE-781 (3), 10 µM; Pomalidomide, 1 µM; MG132, 1 µM) and 24 h of treatment with indicated degraders. Data are represented as mean %RLU relative to DMSO-treated sample (n = 3; representative experiment shown, data are shown as mean ± SEM). In pharmacokinetic (PK) studies, the exposure levels for 25 and 26 after a 1 mg/kg i.p. dose in CD1 mice were not sufficient to advance either degrader into in vivo proof-of-concept studies (25: AUC = 584 ng·h/ml, 26: AUC = 624 ng·h/ml). Based on the elevated intrinsic clearance values for the two degraders (CD1 mouse hepatocyte CL,int = 2.59 L/h/kg and 1.29 L/h/kg for 25 and 26, respectively), we speculated that insufficient metabolic stability could be responsible for the low exposure levels. Metabolite identification after a 30 min incubation of each compound with mouse and human liver S9 microsome fractions exposed the two amide bonds in the linkers as the primary metabolism sites. While trying to block these sites, the central amide turned out to be essential, and its modification led to the loss of degradation potency. However, removal of the amide that connected the linker to the LBL resulted in the more metabolically stable Degrader 27 (CL,int = 0.74 L/h/kg). Interestingly, the preference for the configuration of the cyclohexyl moiety switched to trans. The modification also increased the degradation activity by about sixfold (p300 DC50 = 22.5 nM, Dmax = 86%), while retaining the selectivity over CBP and BRD4 (DC50 > 10 μM) (Table 1, Fig. 2a, b). To further confirm the selective degradation of p300 over its paralog, CBP, we performed a proteome-wide analysis of protein levels in MM1.S cells treated with Degraders 25, 26 at 12 nM and 27 at 3 nM for 6 h. The results revealed a profound degradation of the p300 protein with no major impact on CBP protein level or on canonical CRBN substrates, including IKZF1, IKZF3, CK1α, and GSPT1 (Fig. 2c). When benchmarked against the previously reported p300-selective compounds JQAD1 and MC-1, our compounds demonstrated markedly greater p300 degradation potency, whereas JQAD1 and MC-1 showed no detectable activity against p300 or CBP at concentrations up to 10 µM (Supplementary Fig. 2c, d)24,27. Previous genetic studies showed that p300 acetylates H3 protein at lysine 27 (H3K27)42. To examine the downstream effect of p300 degradation in U2OS cells, we monitored H3K27 acetylation with a high-content microscopy-based assay. As predicted, after a 24-h treatment with 25, 26, and 27, the levels of H3K27Ac decreased in a dose-dependent manner (Fig. 2d). Furthermore, p300 degradation was abolished by co-incubation with the cereblon ligand pomalidomide, the proteasome inhibitor MG-132, and the bromodomain binder GNE-781 (3) in U2OS cells treated with the p300 degraders for 24 h, demonstrating that p300 degraders mechanistically function through the proteasome pathway and engage both CRBN and the p300 bromodomain (Fig. 2e). Taken together, our data show the successful synthesis of highly potent and paralog-selective p300 degraders. p300 degraders’ mechanism for selectivity To elucidate the molecular basis of p300 selectivity, we first assessed whether our degraders exhibited differential binding to the bromodomains of p300 and CBP using a biochemical AlphaLISA assay. As expected, Degraders 25, 26, and 27 showed comparable binding affinities to both bromodomains, indicating that selective p300 degradation arises from downstream mechanistic events rather than differences in target engagement (Table 1). Protein degradation is accomplished through sequential mechanistic steps, including the ternary complex formation (TCF), target protein ubiquitination, and the recruitment of the 26S proteasome non-ATPase regulatory subunit 3 (PSMD3) to the target protein43. To determine which step underlies selective p300 degradation, we developed cellular NanoBRET assays to quantify TCF, ubiquitination, and PSMD3 recruitment to p300 and CBP proteins in HEK293 cells (Fig. 3a–c)41,44. The selective Degraders 25, 26, and 27 formed a stronger and more stable TCF with p300, induced higher p300 ubiquitination, and achieved a higher level of PSMD3 recruitment to p300 compared to CBP protein. In contrast, the dual p300/CBP degrader 28 induced comparable levels of TCF formation, ubiquitination, and PSMD3 recruitment potency for both p300 and CBP (Fig. 3a–c, and Supplementary Fig. 2c, d). Consistent with their limited degradation efficacy in HiBiT assays, JQAD1 and MC-1 displayed minimal or no activity in these mechanistic assays, with MC-1 showing weak effects only at the highest concentration tested (Fig. 3a–c). Together, these findings indicate that the selective degraders achieve preferential p300 degradation through enhanced ubiquitination and proteasome engagement rather than differential binding affinity, and that their overall potency exceeds that of previously reported p300 selective degraders.Fig. 3Degraders selectively deplete p300 in cells through degradation.HEK293 cells were transfected with either p300- or CBP-NanoLuc, along with one of the following HaloTag fusion constructs: a HaloTag-CRBN, b HaloTag-Ubiquitin, or c HaloTag-PSMD3. After treatment with the indicated compounds and concentrations for 6 h, mean BRET ratios were measured to assess a ternary complex formation (TCF), b ubiquitination, and c proteasome recruitment. Data are presented as mean ± SEM from n = 3 independent biological replicates. Statistical significance was determined using a two-sided two-way ANOVA followed by Dunnett’s multiple comparisons test against DMSO control. p300-Ubiquitin: Degrader 25, 26, 27, and 28 (0.01, 0.5, and 1 µM), P = 1.5 × 10-13; MC-1 (1 µM), P = 1.5 × 10-13. p300-PSMD3: Degrader 25, 26, 27, and 28 (0.01, 0.5, 1 µM), P = 1.5 × 10-13; MC-1 (1 µM), P = 1.5 x 10-13. a, b, and c were created in BioRender.com. Rouhimoghadam, M. (2026) https://BioRender.com/1124da8, https://BioRender.com/3u81kjw, and https://BioRender.com/snoe3wz. d TMD8 cells were treated with 3 nM of the indicated compounds for 24 h, followed by quantification of p300 and CBP ubiquitination using the PTMScan® Ubiquitin Remnant Motif (K-ε-GG) kit, as described in “Methods”. Effect sizes were calculated from n = 2 independent experiments. e Cellular ubiquitination of p300 or CBP was examined in HEK293 cells co-expressing HaloTag-ubiquitin and either wild-type or lysine mutant p300- or CBP-NanoLuc. Cells were treated with 0.01 µM of Degrader 27 or Degrader 28. Data are represented as mean of normalized BRET ratios relative to DMSO-treated sample (n = 3; representative experiment shown, data are shown as mean ± SEM). Statistical significance was determined by two-way ANOVA with Dunnett’s multiple comparisons test. p300-Ubiquitin: Degrader 27 (K1637R), P = 5.4 × 10−13; (K1050R), P = 1.3 × 10−13, Degrader 28 (K1050R), P = 1.3 × 10−13. CBP-Ubiquitin: Degrader 28 (K1086R), P = 8.3 × 10−13. Degraders selectively deplete p300 in cells through degradation. HEK293 cells were transfected with either p300- or CBP-NanoLuc, along with one of the following HaloTag fusion constructs: a HaloTag-CRBN, b HaloTag-Ubiquitin, or c HaloTag-PSMD3. After treatment with the indicated compounds and concentrations for 6 h, mean BRET ratios were measured to assess a ternary complex formation (TCF), b ubiquitination, and c proteasome recruitment. Data are presented as mean ± SEM from n = 3 independent biological replicates. Statistical significance was determined using a two-sided two-way ANOVA followed by Dunnett’s multiple comparisons test against DMSO control. p300-Ubiquitin: Degrader 25, 26, 27, and 28 (0.01, 0.5, and 1 µM), P = 1.5 × 10-13; MC-1 (1 µM), P = 1.5 × 10-13. p300-PSMD3: Degrader 25, 26, 27, and 28 (0.01, 0.5, 1 µM), P = 1.5 × 10-13; MC-1 (1 µM), P = 1.5 x 10-13. a, b, and c were created in BioRender.com. Rouhimoghadam, M. (2026) https://BioRender.com/1124da8, https://BioRender.com/3u81kjw, and https://BioRender.com/snoe3wz. d TMD8 cells were treated with 3 nM of the indicated compounds for 24 h, followed by quantification of p300 and CBP ubiquitination using the PTMScan® Ubiquitin Remnant Motif (K-ε-GG) kit, as described in “Methods”. Effect sizes were calculated from n = 2 independent experiments. e Cellular ubiquitination of p300 or CBP was examined in HEK293 cells co-expressing HaloTag-ubiquitin and either wild-type or lysine mutant p300- or CBP-NanoLuc. Cells were treated with 0.01 µM of Degrader 27 or Degrader 28. Data are represented as mean of normalized BRET ratios relative to DMSO-treated sample (n = 3; representative experiment shown, data are shown as mean ± SEM). Statistical significance was determined by two-way ANOVA with Dunnett’s multiple comparisons test. p300-Ubiquitin: Degrader 27 (K1637R), P = 5.4 × 10−13; (K1050R), P = 1.3 × 10−13, Degrader 28 (K1050R), P = 1.3 × 10−13. CBP-Ubiquitin: Degrader 28 (K1086R), P = 8.3 × 10−13. To further investigate whether distinct ubiquitination sites influence degradation selectivity, we performed proteome-wide profiling of lysine ubiquitination in TMD8 cells treated with 3 nM of Degrader 27 or 28 for 24 h. The dual degrader 28 induced strong ubiquitination of p300 K1050 and the corresponding CBP residue K1086, whereas the paralog-selective Degrader 27 selectively ubiquitinated p300 K1167 without modifying the analogous CBP K1203 (Fig. 3d). These results suggested that regioselective ubiquitination of p300 K1167 may also contribute to paralog selectivity. To directly test this hypothesis, we generated lysine-to-arginine mutants of p300 (K1637R, K1050R, K1167R) and CBP (K1086R, K1203R) and assessed their ubiquitination following treatment with Degraders 27 or 28 using BRET-based assays (Fig. 3e). Mutation of p300 K1050 abolished ubiquitination induced by both degraders, confirming its essential role in degradation. Notably, the p300 K1167R mutation selectively reduced ubiquitination by Degrader 27 but had no effect on Degrader 28 activity, demonstrating that K1167 is a degrader-specific ubiquitination site critical for p300 selectivity. Consistent with weak TCF formation with CBP, Degrader 27 did not induce CBP ubiquitination. In contrast, Degrader 28-mediated CBP ubiquitination was blocked by the K1086R mutation but not K1203R, validating K1086 as the functional ubiquitin acceptor site for dual activity. Together, these findings demonstrate that paralog selectivity arises from both degradation kinetics and spatial determinants of ubiquitin transfer. Degrader 27 relies on regioselective ubiquitination of p300 at K1167, a unique lysine site not equivalently engaged in CBP, enabling efficient proteasome recruitment and selective p300 degradation. Degraders’ potency in p300-dependent cancer models p300 acts as a key coactivator for several oncogenic transcription factors and is indispensable for the survival and progression of multiple human malignancies6. To confirm the sensitivity of cancer cell lines with high p300-genetic dependency scores to Degrader 27, we conducted a drug-gene dependency association analysis. Using a high-throughput PRISM screen, 906 cancer cell lines were treated with Degrader 27 for 5 days. We then correlated the AUC of the dose-response curve with top gene dependencies identified through genome-wide CRISPR knockout screening in the Dependency Map (DepMap) database. Indeed, a strong correlation between the AUC of Degrader 27 and the EP300 dependency score was observed, confirming the degrader’s on-target activity (Fig. 4a).Fig. 4p300 degraders induce potent antiproliferative activity in hematological cancer cell lines.a Dot plot showing the correlation between Degrader 27 sensitivity (from PRISM screening) and the top 250 genetic dependencies (from DepMap CRISPR KO project) based on q value, with q value less than 0.1 are visualized. The Pearson correlation between top gene dependencies (x-axis) and associated -log10(q)-values (y-axis) are computed. b The median CERES dependency scores across all cancer cell lines by lineage type, derived for p300 based on CRISPR-based gene knockout across 1320 cancer cell lines. Cell lines are binned by tumor type as annotated by the DepMap database. All tumor types assessed are represented by at least 10 cell lines. c Mean IC50 values of all MM, NHL, AML cell lines treated with 8 doses of Degrader 27 (3-fold dilution) for 5 days in a PRISM assay. The screening was run in triplicate and each plate contained positive (Bortezomib, 20 µM) and negative (DMSO) controls. d Dose-response curves for treating NCI-H929, TMD8, and NOMO-1 cells with the indicated compounds for 7 days. Cell viability was measured by a CTG assay. Data are represented as mean %RLU relative to DMSO-treated sample. (n = 3; representative experiment shown, data are shown as mean ± SEM). p300 degraders induce potent antiproliferative activity in hematological cancer cell lines. a Dot plot showing the correlation between Degrader 27 sensitivity (from PRISM screening) and the top 250 genetic dependencies (from DepMap CRISPR KO project) based on q value, with q value less than 0.1 are visualized. The Pearson correlation between top gene dependencies (x-axis) and associated -log10(q)-values (y-axis) are computed. b The median CERES dependency scores across all cancer cell lines by lineage type, derived for p300 based on CRISPR-based gene knockout across 1320 cancer cell lines. Cell lines are binned by tumor type as annotated by the DepMap database. All tumor types assessed are represented by at least 10 cell lines. c Mean IC50 values of all MM, NHL, AML cell lines treated with 8 doses of Degrader 27 (3-fold dilution) for 5 days in a PRISM assay. The screening was run in triplicate and each plate contained positive (Bortezomib, 20 µM) and negative (DMSO) controls. d Dose-response curves for treating NCI-H929, TMD8, and NOMO-1 cells with the indicated compounds for 7 days. Cell viability was measured by a CTG assay. Data are represented as mean %RLU relative to DMSO-treated sample. (n = 3; representative experiment shown, data are shown as mean ± SEM). A deeper analysis of the DepMap dependency scores revealed that cancer cell lines from lymphoid and myeloid origins were the most sensitive to EP300 knockout (Fig. 4b). This trend was also found in the PRISM screen, where Degrader 27 exhibited potent antiproliferative activity in MM, AML, and NHL cell lines (Fig. 4c). Moreover, Degraders 25, 26, and 27 were >10-fold more potent than the small molecule p300/CBP HAT inhibitor A-485 (1) and the bromodomain inhibitor GNE-781 (3) in the MM NCI-H929, NHL TMD8, and AML NOMO-1 cell lines (Fig. 4d). In contrast, JQAD1 and MC-1 displayed only weak antiproliferative activity in these models (Fig. 4d). Selective targeting of p300 in dependent cell models As paralogs, p300 and CBP can partially compensate for each other if the function of one of the proteins is lost45. We therefore assessed whether the selective degradation of p300 alone would be sufficient to inhibit the proliferation of cancer cells, using a genetic approach. Knockout of EP300, but not CREBBP, effectively reduced the viability of H929, TMD8, and NOMO-1 cells (Fig. 5a, b). These data were consistent with Dependency scores retrieved from the DepMap database showing a higher dependency on EP300 compared to CREBBP in cancer cell lines from Myeloid and Lymphoid origins (Fig. 5c).Fig. 5Selective p300 targeting induces a lethal phenotype in hematological cancer cells and exhibits antitumor efficacy in MM and NHL xenograft models.a NCI-H929, TMD8, and NOMO-1 cells were transfected with CRISPR/Cas9-gRNA RNP targeting EP300 or CREBBP and cell viability was measured at 5 days post transfection with a CTG assay. Data are represented as mean %RLU relative to DMSO-treated sample (n = 3; representative experiment shown, data are shown as mean ± SEM). b Immunoassay of the indicated proteins run at 24 h post-transfection (n = 3; representative experiment shown). DepMap gene dependency scores with volcano plots represent CERES scores for c EP300 and CREBBP in Myeloid and Lymphoid Lineages, and d top essential genes in MM, NHL, and AML cell lines based on CRISPR-based gene knockout as annotated in DepMap database. All tumor types are represented by at least 20 cell lines. e Immunoassay of the indicated proteins in NCI-H929, TMD8, and NOMO-1 cells treated with Degrader 27 for 24 h in a dose-response experiment (n = 3; representative experiment shown). f Tumor growth curves in mice bearing NCI-H929 or TMD8 tumors treated twice daily with Degrader 27 (1, or 2.5 mg/kg), intraperitoneal injection for 14 days. Data are shown as mean ± SEM, n = 8 mice per group. Two-tailed Student’s t-test between groups on day 41 for NCI-H929 model and day 31 for TMD8 model. NCI-H929: Degrader 27 (1 mg/kg), P = 0.0003; (2.5 mg/kg), P = 1.52 × 10−6, TMD8: Degrader 27 (1 mg/kg), P = 4.9 × 10−7, (2.5 mg/kg), P = 2.46 × 10−7. g The measurement of body weights of the mice from the TMD8 in vivo study. h Immunoassay of the p300 and CBP proteins in tumors collected from the NCI-H929 xenograft model after intraperitoneal injection with Degrader 27 (2.5 mg/kg) once: tumors were collected at 6 and 24 h post-injection, or twice a day for 2.5 days: tumors were collected 6 h post-injection. n = 4 mice per group. Selective p300 targeting induces a lethal phenotype in hematological cancer cells and exhibits antitumor efficacy in MM and NHL xenograft models. a NCI-H929, TMD8, and NOMO-1 cells were transfected with CRISPR/Cas9-gRNA RNP targeting EP300 or CREBBP and cell viability was measured at 5 days post transfection with a CTG assay. Data are represented as mean %RLU relative to DMSO-treated sample (n = 3; representative experiment shown, data are shown as mean ± SEM). b Immunoassay of the indicated proteins run at 24 h post-transfection (n = 3; representative experiment shown). DepMap gene dependency scores with volcano plots represent CERES scores for c EP300 and CREBBP in Myeloid and Lymphoid Lineages, and d top essential genes in MM, NHL, and AML cell lines based on CRISPR-based gene knockout as annotated in DepMap database. All tumor types are represented by at least 20 cell lines. e Immunoassay of the indicated proteins in NCI-H929, TMD8, and NOMO-1 cells treated with Degrader 27 for 24 h in a dose-response experiment (n = 3; representative experiment shown). f Tumor growth curves in mice bearing NCI-H929 or TMD8 tumors treated twice daily with Degrader 27 (1, or 2.5 mg/kg), intraperitoneal injection for 14 days. Data are shown as mean ± SEM, n = 8 mice per group. Two-tailed Student’s t-test between groups on day 41 for NCI-H929 model and day 31 for TMD8 model. NCI-H929: Degrader 27 (1 mg/kg), P = 0.0003; (2.5 mg/kg), P = 1.52 × 10−6, TMD8: Degrader 27 (1 mg/kg), P = 4.9 × 10−7, (2.5 mg/kg), P = 2.46 × 10−7. g The measurement of body weights of the mice from the TMD8 in vivo study. h Immunoassay of the p300 and CBP proteins in tumors collected from the NCI-H929 xenograft model after intraperitoneal injection with Degrader 27 (2.5 mg/kg) once: tumors were collected at 6 and 24 h post-injection, or twice a day for 2.5 days: tumors were collected 6 h post-injection. n = 4 mice per group. To further elucidate the p300-induced downstream mechanism in these cell lines, we investigated the top dependency genes in each lineage identified in DepMap and assessed whether their expression was impacted by EP300 knockout (Fig. 5d). Indeed, EP300 knockout was accompanied by a loss of the expression of key transcription factors required for cancer cell survival and leukemogenicity in each indication, such as IRF4, myocyte enhancer binding factor 2B (MEF2B), cellular myelocytomatosis viral oncogene homolog (c-Myc), c-Myb, acute myeloid leukemia 1 protein (AML1) and specificity protein 1 proto-oncogene (SPI1) (Fig. 5b, d)8,46,47. Interestingly, the knockout of CREBBP had no impact on the expression of these oncogenes, except for MEF2B in TMD8 cells (Fig. 5b). Correspondingly, the treatment of the three cell lines with Degrader 27 for 24 h inhibited the expression of the same transcription factors in a dose-dependent manner (Fig. 5e), caused cell cycle arrest at G1 phase at 48 h of treatment, and induced apoptosis at day 5 (Supplementary Fig. 4a–c). To evaluate p300-dependent transcriptional programs, we performed RNA sequencing in NCI-H929 cells treated with Degrader 27 (3 nM, 6 h) and compared the results with those obtained from stable EP300 knockout cells generated by RNP editing. The two perturbations share a core set of downregulated p300-dependent pathways, including leukocyte and myeloid differentiation, cytokine/immune-regulatory modules, and growth factor-responsive programs. However, they diverged where acute p300 degradation induced downregulation of oxidative phosphorylation, electron transport, and hypoxia-response pathways, consistent with rapid metabolic perturbation and immediate disruption of p300-dependent transcriptional coactivation, and EP300-KO cells showed suppression of developmental and adhesion pathways, reflecting secondary compensatory adaptations that emerge following prolonged loss of p300 function (Supplementary Fig. 5a, b). These findings demonstrate that acute pharmacologic degradation of p300 elicits distinct, yet mechanistically coherent, transcriptional consequences compared with long-term genetic ablation, a phenomenon also observed with other chromatin regulators48. Collectively, these results establish that the selective degradation of p300 is sufficient to eradicate hematological cancer cells. In vivo activity of p300 degraders Since p300 degraders induced a potent antiproliferative effect in hematological cancer cell lines in vitro, we next evaluated the degraders for their in vivo antitumor activity using NCI-H929 and TMD8 xenograft models. Intraperitoneal injections of Degrader 27 (1 and 2.5 mg/kg b.i.d.) in SCID mice for 14 days achieved tumor growth inhibition (TGI) compared to vehicle controls in both models without inducing any substantial weight loss (<5%) (Fig. 5f, g). The intraperitoneal injections of the dual p300/CBP Degrader 28 (0.05 and 0.1 mg/kg b.i.d.) were not tolerated, as some mice treated with Degrader 28 exhibited signs of severe toxicity, including one death and additional moribund animals by day 6 of treatment. The immunoassays using capillary-based electrophoresis for p300 and CBP showed a potent and selective degradation of p300 protein in mouse tumor tissue treated with a single or five (b.i.d.) i.p. doses of Degrader 27 at 2.5 mg/kg with the earliest p300 degradation observed at the 6-h timepoint (Fig. 5h, Supplementary Fig. 6). These results confirm that selective p300 degradation is efficacious and more tolerable than dual p300/CBP degradation in selected in vivo settings and can lead to effective therapy in hematological malignancies. Discussion p300 and CBP are paralog proteins with approximately 70% sequence similarity, particularly within their functional domains, HAT and BRD3. These proteins are ubiquitously expressed across normal tissues and perform overlapping roles, leading to their frequent collective reference as CBP/p300. Both p300 and CBP have been implicated in the progression of various cancers4. However, their high homology posed challenges for the development of pharmacological compounds that selectively target each paralog, which made it difficult to investigate their functions independently. With the exception of CCS1477 and FT-7051, currently under clinical evaluation, all previous pharmacological inhibitors failed to progress to the clinic due to the systemic toxicities arising from simultaneous inhibition of both proteins6,20. Emerging studies leveraging genetic models and database resources such as DepMap revealed a preferential dependency on p300 over CBP in several cancers, supporting the notion that these proteins have distinct, context-dependent functions24. Thus, selective targeting of p300 could potentially eliminate cancer cells while sparing CBP in normal tissues, where it may compensate for the loss of p300 and improve the therapeutic index. In this study, we developed a paralog-selective p300 degrader, Degrader 27, that demonstrated potent and selective p300 degradation across multiple cellular and xenograft animal models. While several CBP/p300 degraders have been reported recently, only two, JQAD1 and MC-1, have shown preferential activity toward p30023–26. JQAD1 binds to the HAT domain of both proteins, but selectively degrades p300 in neuroblastoma cellular models, though its mechanism of selectivity remains unknown24. Similarly, MC-1 binds to the CH1 domain of p300 and recruits E3 ubiquitin-protein ligase Mdm2 (MDM2) as an E3 enzyme, showing activity in prostate cancer cells26. However, the selectivity mechanism for this degrader also remains undefined. Notably, both JQAD1 and MC-1 exhibited weak cellular p300 degradation, underscoring the need for more potent and mechanistically understood p300-selective degraders. Protein degradation involves several key mechanistic steps, which include (i) binding of the degrader to the target protein, (ii) formation of a ternary complex with a ubiquitin ligase, (iii) ubiquitination of the target protein, and (iv) the recruitment of PSMD3 to the target43,49. Although bromodomain binding affinities were comparable between p300 and CBP, only Degrader 27 promoted efficient ternary complex formation, robust ubiquitination, and PSMD3 recruitment to p300. Our data reveal a previously unrecognized determinant of paralog selectivity among CBP/p300 degraders, regioselective ubiquitination of a unique lysine residue (K1167) on p300. Quantitative ubiquitinomics identified K1167 as a major ubiquitination site selectively modified by Degrader 27. Functional validation using lysine-to-arginine mutants confirmed that K1167 is essential for Degrader 27-induced ubiquitination, establishing a direct causal role for this residue in driving paralog selectivity. These results suggest that p300 selectivity emerges from both kinetic and spatial control of ubiquitin transfer, where the degrader positions the E3 ligase in a conformation that favors transfer to p300 K1167 but not to analogous CBP sites. This spatial bias likely facilitates efficient proteasome recruitment and turnover of p300 while sparing CBP, offering a mechanistic framework for achieving selective degradation of closely related paralogs. Beyond clarifying the basis of p300 selectivity, this work highlights the potential of exploiting site-specific ubiquitination geometry to rationally design selective degraders targeting highly homologous proteins. The identification of p300 K1167 as a degrader-specific ubiquitination hotspot provides both a mechanistic explanation for Degrader 27’s selectivity and a blueprint for next-generation CBP/p300 degraders with improved safety and efficacy profiles. We further demonstrated that hematological cancers such as MM, NHL, and AML exhibited selective dependency on p300 rather than CBP, as evident by dependency scores from DepMap and genetic knockout approaches. In these models, p300 knockout led to loss of expression of key oncogenic transcription factors such as IRF4, MEF2B, c-Myc, c-Myb, AML1 and SPI1, which are known to regulate leukemogenicity, cell cycle progression, and survival. Interestingly, CBP knockout affected only the expression of MEF2B, underscoring the distinct roles of p300 and CBP. Fundamentally, Degrader 27 induced potent antiproliferative activity at a nanomolar level in a panel of hematological cell lines and outperformed the dual p300/CBP inhibitors A-485 and GNE-781 and the p300 selective degraders JQAD1 and MC-1. Furthermore, Degrader 27 inhibited the expression of the key transcription factors observed with p300 knockout, caused cell cycle arrest at G1 phase at 48 h, and triggered apoptosis by 72 h. In vivo, administration of Degrader 27 at 2.5 mg/kg (single or five b.i.d. doses) led to rapid and selective p300 degradation, detectable as early as 6 h after the final dose. Selective p300 degradation was efficacious and well tolerated, whereas the dual p300/CBP degrader 28 was poorly tolerated even at 0.05 mg/kg, leading to early termination of treatment. These findings highlight the broader therapeutic window achievable through selective p300 targeting. Degrader 27 induced significant tumor growth inhibition in NCI-H929 multiple myeloma and regression in TMD8 lymphoma xenografts. Tumor regrowth was observed only after treatment withdrawal, consistent with reversible pharmacologic effects rather than intrinsic resistance. The delayed regrowth likely reflects residual tumor survival in suboptimally exposed regions or compensatory transcriptional adaptations. Although subcutaneous, the NCI-H929 model provides a robust and reproducible system for pharmacologic evaluation. Future work employing orthotopic or patient-derived xenograft models and rational combination strategies, such as BCL2 antagonists, may further enhance and prolong the antitumor efficacy of selective p300 degradation. In summary, this study has led to the discovery of a highly potent and paralog-selective p300 degrader. Our data from hematological cells and xenograft models suggest that selective p300 degradation is effective in inhibiting cancer progression and represents an attractive strategy for the treatment of hematological cancers. Methods Ethics statement AbbVie is committed to ensuring the humane care and use of laboratory animals in the company’s research and development programs. Our programs exceed regulatory agency standards, and we are committed to the internationally accepted principles of the 3Rs (refinement, reduction, replacement). All animal studies were reviewed and approved by AbbVie’s Institutional Animal Care and Use Committee or Oversight Body (in accordance with national regulations). Animal studies were conducted under an AAALAC-accredited program, where veterinary care and oversight were provided to ensure optimal animal care. Synthesis of compounds Detailed description of the synthesis of the compounds in this study can be found in Supplementary file 1. Modeling studies, including docking calculations, were performed as described in Supplementary files 1 and 2. JQAD1 (HY-145765) was purchased from MedChemExpress. MC-1 was synthesized as described previously27. Cell lines and cell culture U2OS (CRL-3455), MM.1S (CRL-2974), NCI-H929 (CRL-3580) cells were obtained from the American Type Culture Collection (ATCC). NOMO-1 (ACC 542) cell line was obtained from the German Collection of Microorganisms and Cell Culture (DSMZ). TMD8 was obtained from Tokyo Medical and Dental University (Bunkyo, Japan). All cells were cultured in RPMI 1640 medium (22400-089, Thermo Fisher Scientific) supplemented with 10% (v/v) heat-inactivated FBS (A5256801, Gibco), 100 U ml-1 of penicillin and 100 µg ml-1 of streptomycin (penicillin-streptomycin, 15140122, Thermo Fisher Scientific). Animal studies All experiments were conducted in compliance with AbbVie’s Institutional Animal Care and Use Committee and the National Institutes of Health Guide for Care and Use of Laboratory Animals guidelines in a facility accredited by the Association for the Assessment and Accreditation of Laboratory Animal Care. Inoculation material was derived from NCI-H929 or TMD8 cells grown in tissue culture (2–3 passages), prepared in 1:1 Matrigel (BD) to deliver 5 × 106 cells in an injection volume of 0.1 mL per mouse. For the NCI-H929 study, female SCID-beige mice were obtained from Charles River Laboratories, and for the TMD8 studies, female NSG mice were obtained from Jackson Laboratories. Mice were injected subcutaneously into the shaved right hind flank. During the initial staging phase, tumor volume was monitored by use of calipers until the calculated volume reached approximately 150–280 mm3 for the NCI-H929 study and 150–225 mm3 for the TMD8 efficacy studies. Measurements of the length (L) and width (W) of the tumor were taken via electronic caliper, and the volume was calculated according to the equation: V = L × W2/2. At that time, mice were assigned to study groups based on a matched distribution of tumor volumes so that the mean tumor volume within all study groups was similar. Dosing was initiated within 24 h of reallocation of mice into treatment groups. All compounds were formulated in 2% (vol) DMSO + 5% (vol) Tween 80 + 20% (vol) PEG 400 + 73% (vol) D5W (DMSO, Tween 80 and PEG 400) (Sigma Aldrich) and were dosed by intraperitoneal injection 2 times per day (b.i.d) for 14 consecutive days. Throughout the course of the efficacy studies, tumors were measured, and animal body weights were determined twice per week. Tumor volume was calculated for Individual animals. Animals were removed from the study when their tumor volume reached 2000 mm3 or 3000 mm3, depending on the model, or were removed upon animal health issues dictated by IACUC protocols. For the pharmacodynamic (PD) study, mice were similarly dosed as indicated over a 1–3 day period, with tissue samples taken at 6 or 24 h after the final dose. Mice were anesthetized by inhalation with Sevoflurane (McKesson, 937614) and then humanely euthanized by cervical dislocation. The tumor tissues were dissected, placed in cryovials and snap frozen in liquid nitrogen. These tissues were stored at −80 °C until they were used in the assay for PD biomarker movement. Tumors were dissected and digested using an enzyme cocktail of Cell Lysis Buffer (Cell Signaling, 9803), PMSF (Cell Signaling, 8553), and Pierce Protease and Phosphatase Inhibitor (ThermoFisher, A32959) mix using Precellys Tissue Homogenizing (Bertin Corp.). The homogenized tissue was centrifuged at 14,000 × g at 4 °C for 15 min and the supernatant was transferred into a clean tube, and concentration was measured using Pierce™ Bovine Serum Albumin Standard Pre-Diluted Set (ThermoFisher, 23208). Samples were processed in WES immunoassay as described below. Mice were euthanized: (1) when the maximal permitted tumor burden was reached (2000 mm³ for NCI-H929 or 3000 mm³ for TMD8); (2) when general efficacy parameters were confidently determined; or (3) when other animal health issues were observed. Mice were housed in a specific pathogen–free facility under a 12 h light/12 h dark cycle at 22 ± 2 °C and 40–60% relative humidity, with ad libitum access to standard rodent chow and water. Animals used for experiments were 2–3 months old. Animals were group housed in ventilated cages with environmental enrichment in accordance with institutional guidelines. WES immunoassay Cells were plated into 12-well plates, followed by overnight incubation before treatment. Cells were harvested and lysed in CelLytic M (Sigma, C2978) supplemented with Benzonase (Millipore, 71206), 1 mM DTT (Sigma, 4381-10 ml, 1 M), 0.4 mM AEBSF (Roche, 11429868001), Protease Inhibitor cocktail (SIGMA, p8340) and Phosphatase Inhibitor cocktail 2 (Sigma, P5726). Capillary-based immunoassays were performed using a standard WES (Simple Western) protocol (ProteinSimple). Lysates were loaded onto WES plates at 0.6 mg/mL total protein. WES staining was conducted using the following antibodies: p300 (CellSignaling, 70088), CBP (Cell Signaling, 7389), Lamin-B (Cell Signaling, 12586), HSP90 (Cell Signaling, 4877), c-Myc (Cell Signaling, 13987S), c-Myb (Cell signaling, 59995S), PU.1/Spi1 (Abcam, ab302623), MEF2B (Invitrogen, PA5-75240), IRF4 (Active Motif, 61511), AML1 (Cell Signaling, 4334). All antibodies were diluted to 1:50 in WES antibody diluent. Antibodies were validated using manufacturer information and/or prior publication, and by detection of bands at the expected molecular weight in WES. Cell viability assay Cells were plated into 384-well tissue-culture plates with 2000 cells per well in 50 µL of the appropriate media. Plates were centrifuged at 350G to remove bubbles and placed in an incubator overnight. One day after seeding, cells were treated with the referenced compounds and resuspended in DMSO with D300e Digital Dispenser (Tecan Group Ltd, CH). Cells were incubated at 37 °C and 5% CO2. After 7 days of incubation, 25 µL of CellTiter-Glo® 2.0 (Promega, G924C) was added per well, and plates were put on an orbital shaker for 10 min to induce cell lysis. After lysis, the individual well luminescent signal was read on Agilent BioTek Synergy Neo2 plate reader (Agilent BioTek N2SN). Dose-response curves were generated by normalizing signal to DMSO-treated wells. AUC values were calculated using GraphPad PRISM software. Cell HiBiT assay U2OS cells modified to stably express the human HiBiT-tagged EP300 or HiBiT-tagged CREBBP, and HEK293T cells modified to stably express the human HiBiT-tagged BRD4 were obtained from Promega. Cells were plated into 384-well tissue-culture plates with 2000 cells per well in 50 µL of the appropriate media. The first and last column of each plate was seeded with media without cells. Plates were centrifuged at 350G to remove bubbles and placed in an incubator overnight. One day after seeding, cells were treated with the referenced compounds and resuspended in DMSO with D300e Digital Dispenser (Tecan Group Ltd, CH). Cells were incubated at 37 °C and 5% CO2. After a 24 h incubation, 25 µL of NanoGlo® HiBiT Lytic Reagent, containing a substrate and LgBiT protein (Promega, N3030), was added to each well, and cells were incubated in the dark for 10 min before measuring the endpoint luminescence on Agilent BioTek Synergy Neo2 plate reader (Agilent BioTek N2SN). Dose-response curves were generated by normalizing signal to DMSO-treated wells. AUC values were calculated using GraphPad PRISM software. NanoBRET ubiquitination, ternary complex formation (TCF), and proteasomal recruitment assays To directly assess the formation of p300-CRBN and CBP-CRBN complexes, as well as the recruitment of p300 or CBP to the proteasome, HEK-293 cells (1 × 106) were cultured in complete medium (Dulbecco’s Modified Eagle Medium supplemented with 10% FBS) and transfected in 6-well plates using FuGENE HD (Promega) with 20 ng of either p300-NanoLuc or CBP-NanoLuc and 2 μg of HaloTag-CRBN or HaloTag-PSMD3 vectors. For the study of live-cell p300 or CBP ubiquitination and identification of ubiquitination sites, HEK-293 cells (1 × 106) were similarly maintained and transfected with 20 ng of either wild-type or lysine mutant P300- or CBP-NanoLuc constructs, along with 2 μg of HaloTag-Ubiquitin vectors. The following day, transfected cells (2 × 104) were replated in triplicate into white 96-well tissue culture plates in 90 μl of assay medium (Opti-MEM Reduced Serum Medium without phenol red, supplemented with 4% FBS), with or without HaloTag NanoBRET 618 Ligand (Promega), and incubated overnight at 37 °C, 5% CO2. For endpoint analysis of ternary complex formation, ubiquitination, and proteasomal recruitment, 10 μl of 10X concentrated drugs were added to the respective wells and incubated for 6 h at 37 °C, 5% CO2. Afterwards, 25 μl of 5X NanoBRET Nano-Glo substrate (Promega) was added and BRET signals were measured using a VICTOR® Nivo™ Multimode Plate Reader (PerkinElmer, USA). Luminescence was collected through a dual-filter setup: a 460/480-nm bandpass filter for the donor (P300- or CBP-NanoLuc fusion proteins) and a 610-nm long pass filter for the acceptor (HaloTag NanoBRET 618 ligand), with an integration time of 0.5 s. NanoBRET ratios were background-subtracted and expressed in milliBRET units. AlphaLISA peptide binding assay p300 and CBP peptides previously generated were used in the binding assay using AlphaLISA (Revvity)42. The compounds were pre-dispensed into Revvity 384-well white proxiplates via Echo® Acoustic Liquid Handler (Beckman). For p300, a solution of 125 nM Enzyme, 12.5 nM Anaspec Lys(Ac)5/8/12/16]-Histone H4 (1-21)-GGK (Biotin, AS-64989-1), 2.5 μg/mL AlphaScreen® streptavidin donor beads (Revvity, 6760002S), and 2.5 μg/mL AlphaScreen® nickel chelate acceptor beads (Revvity, AL108C) were dispensed to each well and incubated for 1 h while protected from light. The CBP binding assay was performed the same except with 2.5 μg/mL AlphaScreen® anti-GST acceptor beads (Revvity, 6760603C). After incubation, the plates were read on an EnVision® (PerkinElmer) with excitation at 680 nm and emission at 620 nm. Cereblon binding assay Cereblon binding was measured in vitro using a binding assay developed internally employing AlphaLISA® beads (Amplified Luminescent Proximity Homogeneous Assay, PerkinElmer) in the detection step. The AlphaLISA® detection of Cereblon/DDB1 complex binding uses AlphaLisa® Nickel chelate acceptor beads (Perkin Elmer, AL108M) to capture the His-tagged CRBN-DDB1 protein complex and streptavidin-coated AlphaScreen® donor beads (Perkin Elmer, 6760002) to capture the biotinylated competitive ligand. Donor beads and acceptor beads come into proximity through ligand binding to the CRBN-DDB1 protein complex. Donor beads contain a photosensitizer, phthalocyanine, which upon illumination at 680 nm, converts ambient oxygen to an excited and reactive form of O2, singlet oxygen, and diffuses approximately 200 nm in solution within its 4 μs half-life. When an Acceptor bead is within that distance, energy is transferred from the singlet oxygen to rubrene as the final fluorophore, emitting light between 520 and 620 nm. Each well in the competition assay includes the following: sample or control; His-tagged CRBN-DDB1 protein complex; biotinylated ligand; AlphaLisa acceptor beads; and Streptavidin AlphaScreen® donor beads. For the competition assay, compounds dissolved in 100% dimethyl sulfoxide were plated onto white, low-volume 384-well plates (PerkinElmer, 6008289) using a Labcyte Echo acoustic dispenser (Beckman), in an 11-point dose curve and threefold dilution scheme. This resulted in a 10 µM top dose (starting concentration) and 0.1% DMSO final concentration after the addition of 10 µL of an assay mixture containing His-tagged Cereblon/DDB1 complex and 50 nM biotin-labeled ligand in a buffer consisting of 10 mM HEPES pH 7.4, 150 mM NaCl, 0.05 % Tween-20 and 10 mM DTT. Plates were equilibrated at room temperature for 30 min. For detection, a 10 µL detection reagents was added in low light conditions containing nickel chelate acceptor beads (Perkin Elmer, AL108M) and streptavidin donor beads (Perkin Elmer, 6760002), resulting in a final concentration of 20 µg/ml. Plates were incubated for 2 h at room temperature then read on a PerkinElmer™ Envision™ plate reader (680 nm excitation and 570 nm emission settings) (PerkinElmer, 2102-0010). The IC50 of each compound was determined by calculating percent inhibition from no drug and no protein controls, and the Ki of each compound was calculated using the Cheng-Prusoff transformation. High content microscopy U2OS cells were plated in Collagen I-coated 96-well view plates (Perkin Elmer, 6005810) and incubated overnight at 37 °C and 5% CO2. Cells were then treated with the referenced compounds resuspended in DMSO with D300e Digital Dispenser (Tecan Group Ltd, CH) for 24 h. Cells were fixed in 10% formaldehyde (Polysciences, Inc., 04018) at room temperature for 10 min, washed in PBS, and then permeabilized in 0.1% Triton X-100 for 10 min. Cells were then blocked in 1% BSA for 1 h and incubated with H3K27Ac antibody (Cell Signaling, 8173) and H3 antibody (Active Motif, 39763) in antibody dilution buffer (0.3% BSA in PBS) at a concentration of 1:100 overnight at 4 °C. Cells were washed three times in PBS and then incubated with a mixture of Alexa Fluor 488-conjugated goat anti-rabbit IgG antibodies (Life Technologies, A-11029), Alexa Fluor 555-conjugated goat anti-mouse IgG (Life Technologies, A-21424) antibodies, and Hoechst 33342 (Life Technologies, H3570) for 1 h at room temperature. After washing four times in PBS, the plates were scanned within 24 h of processing on a ThermoFisher CellInsight using the target activation algorithm, acquiring 15 fields per well. Fluorescence intensities were quantified using the average mean intensity function. Dose-response curves were generated by normalizing the signal to DMSO-treated wells. AUC values were calculated using GraphPad PRISM software. Cell cycle and apoptosis flow cytometry (FACS) analysis Cells were treated with the referenced compounds for the designated timepoints. Cells were harvested and processed using the PI/RNase Staining buffer diluted at 1:1000 (Invitrogen, P3566, 12091-021), or Annexin V diluted at 1:100 (Biotium, 29003). All FACS analyzes were performed on NovoCyte Quanteon Flow Cytometer and analyzed using FlowJo (v10.10.0). Knockout cell line generation Genomic deletions of EP300 and CREBBP in NCI-H929, TMD8 and NOMO-1 cells were performed using Alt-R CRISPR-Cas9 RNP system with SpCas9 Nuclease (Integrated DNA Technologies IDT, 1081058) and predesigned Alt-R™ CRISPR-Cas9 guide RNA targeting EP300 or CREBBP were obtained from Integrated DNA Technologies (IDT). EP300-sgRNA sequence: GTTCAATTGGAGCAGGCCGA, and CREBBP-sgRNA sequence: CAACTGTCGGAGCTTCTACG. The Cas9-gRNA mix was prepared according to the manufacturer's protocol and delivered into cells using the Neon Transfection System (Invitrogen, 10431915) according to the manufacturer's protocol. RNA-seq and data analysis Four independent replicates of the experiment were performed for each group. NCI-H929 cells were treated with Degrader 27 at 3 nM for 6 h or were transfected with CRISPR/Cas9-gRNA RNP targeting EP300 for 24 h and cells were then collected and flash-frozen in liquid nitrogen for subsequent storage and delivery. RNA extraction, sample QC, Library preparation and sequencing analysis were performed at GENEWIZ, LLC./Azenta US. Differential expression analysis was performed using DESeq2 (version 1.32.0) in R (version 4.1.1). Genes with an adjusted p-value < 0.05 and |log2 fold change| > 1 were considered significantly differentially expressed. Normalization of counts and dispersion estimation were done as described in the DESeq2 manual. Default settings for the DESeq function were employed unless otherwise specified. Significantly differentially expressed genes were plotted with EnhancedVolcano (version 1.10.0). Pathway analysis and gene ontology interpretations were carried out using the clusterProfiler package (version 4.0.5). The analysis was performed on differentially expressed genes against GO:BP pathways. Gene sets with a false discovery rate (FDR) < 0.05 were considered significantly enriched. Key pathways and genes of interest were further investigated using manual curation and literature review. Top enriched pathways were plotted as a dotplot using the enrichplot package (version 1.12.3). Whole proteome profiling Three independent replicates of the experiment were performed for each group. MM.1S cells treated with test compounds for 6 h were lysed using the iST-NHS kit (PreOmics, P.O. 00030). 150 µl of LYSE-NHS reagent was added into each sample in their original tubes to lyse, reduce, and alkylate. Tubes were vortexed and placed on a heating block for 10 min at 95 ॰C. After allowing tubes to cool to room temperature, lysates were transferred into 0.5 mL Bioruptor microtubes (Diagenode, C30010013) and sonicated for 10 min using the Bioruptor Plus Sonication device (Diagenode, B01020001), cycling on/off every 30 s. Total protein concentration per sample was measured using a BCA protein assay kit (Thermo Fisher Scientific, 23225) and normalized to 50 µg per 50 µl of LYSE-NHS reagent. Once protein concentrations were normalized, samples were digested using 50 µl of the Trypsin Lys-C buffer included in the PreOmics® iST-NHS kit. Samples were placed on a heating block for 2 h at 37 °C to digest. Following digestion, TMTpro label reagents (ThermoFisher Scientific, 44522) were prepared by resuspending with 70 µl of Acetonitrile. 34 µl of each label was then aliquoted into the corresponding samples. Labeled samples were vortexed and incubated for 1 h at room temperature. After incubation, TMT-labeled samples were quenched with 15 µl of 5% Hydroxylamine, vortexed, and placed on a mixer for 15 min at room temperature. Labeled samples were acidified with 100 µl of the STOP reagent buffer included in the PreOmics® iST-NHS kit and vortexed afterwards. Samples were then transferred to an equilibrated and conditioned PreOmics® iST cartridge and centrifuged at 2200 G for 5 min between washes, and for 1 min between elutions. All buffers used for desalting were included in the PreOmics® iST-NHS kit. The desalted elutions were combined in a 2 mL microcentrifuge tube (Eppendorf, 022431102), which was placed in the −80 ॰C freezer for an hour before being dried down using a vacuum concentrator (Labconco, 7310022). Once dried down, 300 µl of 0.1% TFA (Trifluoroacetic acid) was added to the tube before it was vortexed. The 300 µl of combined sample was transferred into a UPLC vial (Waters, 186005663CV) and fractionated into twelve fractions in a deepwell 96-well plate (Eppendorf, 186005837) using a Waters Acquity UPLC System. Corresponding fractions were combined in 1.5 mL microcentrifuge tubes (Eppendorf, 022431081). Leftover fractions in the 96-well plate were stored as backup in the −80 °C freezer. The fractions in the 1.5 mL tubes were placed in the −80 °C freezer for an hour before being dried down using a vacuum concentrator. Fractions were desalted once more using a positive pressure manifold (Waters, 186006961) and a SepPak tC18 µElution plate (Waters, 186002318). Plate conditioning consisted of flowing 1.2 mL of 90% MeOH at high pressure, followed by 1.2 mL of 0.1% TFA at high pressure. Dried down fractions were resuspended with 500 µl of 0.1% TFA and transferred into individual conditioned wells on the deepwell plate, and the positive pressure manifold was set on low pressure. Fractions were then washed with 500 µl of 5% MeOH on high pressure. The waste plate was switched out for a new deepwell 96-well plate (Eppendorf, 951033502) and fractions were eluted at low pressure with 300 µl of 50% Acetonitrile. The eluted samples were then transferred into individual 2 mL microcentrifuge tubes and placed in the −80 °C freezer for an hour before being dried down using a vacuum concentrator. Dried fractions were resuspended with 25 µL of 5% Acetonitrile/0.1% FA solution, loaded on a hard-shell PCR plate (Bio Rad, HSP9901) and placed in an EASY-nLC UHPLC System (ThermoFisher Scientific, LC140) coupled to a Thermo Orbitrap Mass Spectrometer for analysis. Data was searched through MaxQuant and after exporting the results, proteins with a significant number of missing values between the various conditions were filtered out. Lastly, we imputed remaining missing values using the KNN method and the imputeLCMD R package, and data were quantile normalized. Volcano plots were generated with a log FC cutoff at 3.5σ and p value cutoff at 0.01 using the limma R-package. Ubiquitin profiling with peptide enrichment Two independent replicates of the experiment were performed for each group. TMD8 cells were treated with the indicated compounds for 24 h and lysed for 10 min in 1 mL urea lysis buffer containing 8 M urea, 20 mM HEPES (pH 8.0), and 1 mM chloroacetamide. Cells were then sonicated using a tip sonicator and centrifuged at 20,000 G for 15 min at 10 °C. Soluble portion was transferred to a new tube. Total protein concentration per sample was measured using a BCA protein assay kit (ThermoFisher Scientific, 23225) and normalized to 10 mg. After protein normalization, protein disulfide bonds were reduced with 5 mM dithiothreitol (ThermoFisher Scientific, A39255) at 25 °C for 60 min and proteins were alkylated in the dark using 10 mM iodoacetamide (ThermoFisher Scientific, 122271000) at 25 °C for 45 min. Lysates were then diluted 1:4 using 20 mM HEPES, pH 8.0, to lower the urea concentration to 2 M. Samples were digested with Trypsin/LysC (Promega, V5072) at a 1:50 enzyme-to-substrate ratio at 25 °C overnight with mixing (1350 rpm). The digestion was quenched with 20% TFA to reach a volumetric concentration of 1% TFA and pH less than 3. pH was checked with pH paper. Samples were placed on ice for 15 min to facilitate urea precipitation. Samples were spun at 5000 G for 15 min to remove precipitated urea. Samples were de-salted using Sep-Pak 50 mg C18 columns (Waters, WAT054955). Sep-Pak columns were conditioned with 1 mL 90% ACN + 0.1% TFA and equilibrated with 2 mL 0.1% TFA. Each sample was then loaded onto a column and washed twice with 1 mL 0.1% TFA and 0.5 mL of 5% MeOH 0.1% TFA. Peptides were then eluted off the column with 0.6 mL 50% ACN/0.1% TFA, 0.2 mL at a time, and dried down using a vacuum concentrator (Labconco, 7310022). Ubiquitination profiling was conducted using the PTMScan® Ubiquitin Remnant Motif kGG Kit (CST, 5562) and included reagents. 1 mL of 1X HS IAP Bind Buffer #1 (from CST kit) was added to the dried peptides and pellets were resuspend mechanically by pipetting repeatedly. pH of peptide solution was adjusted to 7 before continuing. The peptide solution was then cleared by centrifugation for 5 min at 10,000 G at 4 °C and cooled on ice afterwards. The vial of antibody-bead conjugate was spun at 2000 G for 2–5 s to bring down any buffer and beads clinging to the sides and the cap of the vial. 20 µL of antibody-bead conjugate was pipetted into a new 1.5 mL microcentrifuge tube (Eppendorf, 022431102) for each sample, every time mixing the antibody-bead conjugate stock before each transfer. 1 mL of ice-cold 1X PBS was aliquoted into the 1.5 mL microcentrifuge tubes, followed by inverting each tube five times to allow the mixing of buffer with beads. The tubes were placed on a magnetic separation rack for 10 s, just until all beads attracted to the magnet. PBS buffer was then carefully removed with a pipette and bead washing was repeated with 1 mL of 1X PBS three times, for a total of four PBS washes per tube. The peptide solution was then transferred into the tubes containing antibody beads and incubated on an end-over-end rotator for 2 h at 4 °C. Any remaining insoluble pellet was discarded. After incubation, tubes were briefly spun at 2000 G for 2–5 s to bring down the beads and the solution clinging to the sides and the cap. Tubes were then placed in the magnetic separation rack for 10 s, and the unbound peptide solution was transferred to another microcentrifuge tube. To wash, 1 mL of chilled HS IAP Wash Buffer (1X) was added to the beads and mixed by inverting five times. The tube was briefly centrifuged and placed in the magnetic stand for another 10 s. All wash buffer was removed. The washing step was repeated three more times for a total of four washes with HS IAP Wash Buffer. 1 mL of chilled LCMS water was added to the beads and mixed by inverting five times. The tubes were briefly centrifuged and placed in the magnetic stand for 10 s. All water was then removed. This washing step was repeated one more time for a total of two water washes. 50 µl of IAP Elution Buffer (0.15% TFA) was added to the beads. This solution was placed on a mixer (1500 rpm) for 10 min at room temperature to keep beads in suspension. Tubes were then briefly centrifuged and placed in the magnetic rack. The eluted samples were transferred to a new microcentrifuge tube. Another 50 µL of IAP Elution Buffer (0.15% TFA) was added to the beads to repeat the elution step. Both eluents with kGG containing peptides were combined in the same microcentrifuge tube. For the second de-salting step, an adaptor was placed on the top of a lidless 1.5 mL tube (Eppendorf, 022431081) and a C18 tip (Thermo Scientific, 84850) was placed inside the adaptor. The C18 tips were equilibrated by pipetting 50 µL of 50% ACN + 0.1% TFA onto the top and centrifuging at 1000 G for 30 s. The C18 tips were transferred to clean collection tubes. 50 µL 0.1% TFA was added into the C18 tips which were then centrifuged at 1000 G for 30 s. This step was repeated once for two total equilibration and washing steps. Samples were loaded by passing IAP eluent through the C18 tip. The IAP eluent was added in two aliquots of 50 µL, for a total of 100 µL, and then centrifuged at 1000 G for 30 s. C18 tips were washed by passing 50 µL 0.1% TFA through two times and centrifuged at 1000 G for 30 s. Peptides were eluted off the C18 tips by passing 10 µL 50% ACN, 0.1% TFA through two times. The resulting eluents were pooled and dried down using a vacuum concentrator (Labconco, 7310022). Peptides were then re-dissolved in 20 µL of 5% ACN + 0.1% Formic Acid and loaded on a hard-shell PCR plate (Bio Rad, HSP9901). iRT peptide mix (Bruker Daltonics, 1816351) was also added to the final samples for retention time calibration. The plate was placed in an EASY-nLC UHPLC System (ThermoFisher Scientific, LC140) coupled to a Bruker®, TimsTOF Pro 2 Mass Spectrometer for analysis. We searched data through Spectronaut and after exporting the results, we filtered out proteins with a significant number of missing values between the various conditions. Lastly, we imputed remaining missing values using KNN method and the imputeLCMD R package, and the data was quantile normalized. DepMap dependency analysis Analysis of dependency data was retrieved from the DepMap portal (www.depmap.org) using the 24Q2 dataset. Dependency data were extracted as probability of dependency for all cell lines. Details of individual cell lines are available at www.depmap.org. PRISM assay Degrader 27 was screened in PRISM DNA-barcoded cell lines established by the Broad Institute. In brief, 20-25 cell lines per pool were plated in 384-well plates and treated with Degrader 27 at 8 doses in threefold dilutions starting at 10 µM for 5 days with 906 cancer cell lines passing QC. Three benchmark compounds – included in the MTS Validation set—were also tested to ensure high data quality. All compounds were run in triplicate, and each plate contained positive (Bortezomib, 20 μM) and negative (DMSO) controls. Cells were then lysed in TCL mRNA lysis buffer, and then PCR with reverse transcription was performed. Detection of the barcodes and univariate and multivariate analysis was then performed as previously described50. Statistical analysis All data are expressed as the mean ± SEM, and experimental group data were normalized to the DMSO-treated group. For analysis of CRISPR dependency datasets, cell viability experiments, and Cell HiBiT assay all statistical tests were performed using GraphPad PRISM software (v10). Details of the statistical tests used for each analysis are shown in the figure legends. Reporting summary Further information on research design is available in the Nature Portfolio Reporting Summary linked to this article. Supplementary information Supplementary Information Description of Additional Supplementary Files Supplementary Data 1 Reporting Summary Transparent Peer Review file Supplementary Information Description of Additional Supplementary Files Supplementary Data 1 Reporting Summary Transparent Peer Review file Source data Source Data Source Data