Chemoproteomics-Enabled Discovery of a Covalent Molecular Glue Degrader

Targeting NF-kB

Elizabeth A. King

1,2,3

, Yoojin Cho

1,2,3

, Dustin Dovala

2,4

, Jeffrey M. McKenna

2,5

John A. Tallarico

2,5

, Markus Schirle

2,5

, and Daniel K. Nomura

1,2,3,6

Department of Chemistry, University of California, Berkeley, Berkeley, CA 94720 USA

Novartis-Berkeley Center for Proteomics and Chemistry Technologies

Innovative Genomics Institute, Berkeley, CA 94704 USA.

Novartis Institutes for BioMedical Research, Emeryville, CA 94608 USA

Novartis Institutes for BioMedical Research, Cambridge, MA 02139 USA

Department of Molecular and Cell Biology, University of California, Berkeley, Berkeley, CA 94720 USA

*correspondence to dnomura@berkeley.edu

Keywords: activity-based protein profiling, targeted protein degradation, molecular glue, E2 ligase, UBE2D,

NFKB1, transcription factor

.CC-BY-NC-ND 4.0 International licenseavailable under a

was not certified by peer review) is the author/funder, who has granted bioRxiv a license to display the preprint in perpetuity. It is made

The copyright holder for this preprint (whichthis version posted May 18, 2022. ; https://doi.org/10.1101/2022.05.18.492542doi: bioRxiv preprint

Abstract

Targeted protein degradation using heterobifunctional Proteolysis-Targeting Chimeras (PROTACs) or molecular

glues has arisen as a powerful therapeutic modality for degrading disease targets. While PROTAC design is

becoming more modular and straightforward, the discovery of novel molecular glue degraders has been more

challenging. While several recent studies have showcased phenotypic screening and counter-screening

approaches to discover new molecular glue degraders, mechanistically elucidating the ternary complex induced

by the small-molecule that led to the initial phenotype—i.e. identifying the degraded target and relevant

components of the ubiquitin-proteasome system—has remained cumbersome and laborious. To overcome these

obstacles, we have coupled the screening of a covalent ligand library for anti-proliferative effects in leukemia

cells with quantitative proteomic and chemoproteomic approaches to rapidly discover both novel covalent

molecular glue degraders and their associated ternary complex components and anti-proliferative mechanisms.

We have identified a cysteine-reactive covalent ligand EN450 that impairs leukemia cell viability in a NEDDylation

and proteasome-dependent manner. Chemoproteomic profiling revealed covalent interaction of EN450 with an

allosteric C111 in the E2 ubiquitin ligase UBE2D. Follow-up quantitative proteomic profiling revealed the

proteasome-mediated degradation of the oncogenic transcription factor NFKB1 as a putative degradation target.

Subsequent validation studies demonstrated that EN450 induced the ternary complex formation between UBE2D

and NFKB1 and that both UBE2D and NFKB1 were important for the anti-proliferative mechanisms of EN450.

Our study thus puts forth the discovery of a novel molecular glue degrader that uniquely induced the proximity

of an E2 ligase with a transcription factor to induce its degradation and anti-proliferative effects in cancer cells.

Taken more broadly, our study showcases a rapid and modular approach for discovering novel covalent

molecular glue degraders and their respective ternary complex components in an unbiased fashion.

.CC-BY-NC-ND 4.0 International licenseavailable under a

was not certified by peer review) is the author/funder, who has granted bioRxiv a license to display the preprint in perpetuity. It is made

The copyright holder for this preprint (whichthis version posted May 18, 2022. ; https://doi.org/10.1101/2022.05.18.492542doi: bioRxiv preprint

Introduction

Most small-molecule drugs in the clinic operate through classical occupancy-driven pharmacology that

consists of small-molecules binding to deep active site binding pockets and resulting in functional modulation of

the target. However, many therapeutic target proteins have been deemed “undruggable” since they do not

possess well-defined, functionally relevant binding pockets, thus rendering these proteins inaccessible to

classical drug discovery approaches (Dixon and Stockwell, 2009; Spradlin et al., 2021). Targeted protein

degradation using heterobifunctional Proteolysis-Targeting Chimeras (PROTACs) or molecular glues has arisen

as a powerful alternative therapeutic modality aiming at degradation instead of inhibition of the disease target

(Bond and Crews, 2021; Burslem and Crews, 2020). While heterobifunctional PROTACs still require protein-

targeting ligands that are capable of binding to the target protein with decent potency, monovalent molecular

glue degraders can exploit shallower protein interfaces to induce ternary complex formation and subsequent

ubiquitination and degradation of specific proteins (Schreiber, 2021). While molecular glue degraders are

potentially more attractive and drug-like, most molecular glue degraders have been discovered fortuitously and

rational discovery of novel molecular glue degraders has remained challenging.

Recent studies by Mayor-Ruiz and Winter et al. have showcased innovative phenotypic screening

paradigms for rapidly discovering small-molecules that exert anti-cancer activity through molecular glue degrader

mechanisms (Mayor-Ruiz et al., 2020). These screens for anti-cancer small-molecules consisted of counter

screens for an attenuated phenotype in hyponeddylation cell lines to identify molecules that exerted their

phenotypes through a Cullin E3 ligase-dependent mechanism. However, the backend mechanistic elucidation

of the ternary complex components—identifying the degraded target and ubiquitin-proteasome component that

were brought together by the small-molecule—required whole genome-wide CRISPR screens which can

sometimes be laborious and may yield indirect targets in addition to direct targets of the small-molecule (Mayor-

Ruiz et al., 2020).

Covalent chemoproteomic approaches have arisen as powerful platforms for coupling phenotypic

screening of covalent electrophile libraries with rapid mechanistic deconvolution (Backus et al., 2016; Chung et

al., 2019; Meissner et al., 2022; Spradlin et al., 2019a, 2021; Vinogradova et al., 2020). As such, we conjectured

that coupling the screening of a covalent ligand library for novel molecular glue degraders with backend

chemoproteomic and quantitative proteomic approaches would provide rapid discovery of novel molecular glue

.CC-BY-NC-ND 4.0 International licenseavailable under a

was not certified by peer review) is the author/funder, who has granted bioRxiv a license to display the preprint in perpetuity. It is made

The copyright holder for this preprint (whichthis version posted May 18, 2022. ; https://doi.org/10.1101/2022.05.18.492542doi: bioRxiv preprint

degraders and their ternary complex components and downstream mechanisms. In this study, we phenotypically

screened a library of covalent ligands for antiproliferative compounds and used chemoproteomic platforms to

discovery a novel molecular glue degrader and associated ternary complex.

Results

To discover novel covalent molecular glue degraders, we screened a library of 750 cysteine-reactive

covalent ligands for anti-proliferative effects in HAP1 leukemia cancer cells (Figure 1a; Table S1). We identified

11 hits from our primary screen, of which 3 of these hits—EN222, EN450, and EN266—showed reproducible

impairments in HAP1 cell proliferation by greater than 90% (Figure 1b). Among these three hits, we next sought

to identify if these compounds may be exerting their anti-proliferative effects through a Cullin E3 ubiquitin ligase-

dependent mechanism. To this end, we counter-screened our hit compounds in UBE2M knockdown

hyponeddylation lines alongside dCeMM1, a positive control molecular glue degrader previously identified by

Mayor-Ruiz and Winter et al. which induces degradation of RBM39 in a CRL4

DCAF15

-dependent manner (Figure

1c-1d) (Mayor-Ruiz et al., 2020). NEDDylation is a critical post-translational modification necessary for the

activity of all Cullin E3 ligases (Baek et al., 2021). As such, compounds that show attenuated phenotypes in

these hyponeddylation cell lines would indicate a Cullin E3 ligase-dependent mechanism. EN266, EN450, and

the positive control glue degrader dCeMM1, but not EN222, showed significant attenuation of anti-proliferative

effects in HAP1 cells with EN450 showing the most robust effects (Figure 1d-1e). EN450 is a fragment ligand

containing a cysteine-reactive acrylamide warhead (Figure 1e). EN450 as well as dCeMM1 also showed

significant attenuation of anti-proliferative effects upon pre-treatment of HAP1 cells with the NEDDylation inhibitor

MLN4924 and the proteasome inhibitor bortezomib, further cementing that EN450 was exerting its anti-

proliferative effects through a Cullin E3 ligase and proteasome-dependent mechanism (Figure 1f-1g).

Based on these data, we conjectured that EN450 was a molecular glue inducing the proximity of a target

protein with a component of the ubiquitin-proteasome system (UPS) to form a ternary complex leading to the

ubiquitination and proteasome-mediated degradation of the target protein and subsequent anti-proliferative

effects in HAP1 cells. To identify proteome-wide covalent targets of EN450, we performed cysteine

chemoproteomic profiling using isotopic tandem orthogonal proteolysis-activity-based protein profiling (isoTOP-

ABPP) (Backus et al., 2016; Weerapana et al., 2010). Through competitive profiling of EN450 covalent protein

.CC-BY-NC-ND 4.0 International licenseavailable under a

was not certified by peer review) is the author/funder, who has granted bioRxiv a license to display the preprint in perpetuity. It is made

The copyright holder for this preprint (whichthis version posted May 18, 2022. ; https://doi.org/10.1101/2022.05.18.492542doi: bioRxiv preprint

targeting in HAP1 cells against cysteine-reactive alkyne-functionalized iodoacetamide probe labeling, we

identified 23 targets that showed significant engagement—control/EN450-treated ratio of >4 with adjusted p-

values <0.05—among 4501 cysteines quantified (Figure 2a; Table S2). Given that we and others have

previously shown that the UPS is highly ligandable with covalent ligands, that even partial covalent occupancy

of a UPS effector protein can lead to significant degradation of the target protein due to the catalytic nature of

degraders, and that EN450 is a small-molecule fragment that is not likely to engage in high-affinity interactions,

we conjectured that the covalent interaction of EN450 is likely occurring with a component of the ubiquitin-

proteasome system, rather than the target protein (Belcher et al., 2021; Henning et al., 2022b, 2022a; Spradlin

et al., 2019a; Zhang et al., 2018, 2021). Among these targets, only one protein was involved in the UPS and the

Cullin E3 ligase machinery—C111 of ubiquitin-conjugating enzyme E2D (UBE2D). C111 is an allosteric cysteine

in UBE2D that is distal from the cysteine (C85) that undergoes ubiquitin conjugation during ubiquitin transfer.

Because of the high degree of sequence identity between UBE2D isoforms, we could not distinguish whether

EN450 targeted C111 on UBE2D1, UBE2D2, UBE2D3, or UBE2D4. Nonetheless, UBE2D is one of the key E2

ligases that are coupled with the Cullin E3 ligase complex, which is consistent with attenuation of EN450 anti-

proliferative effects with a NEDDylation inhibitor of Cullin E3 ligases (Baek et al., 2021).

To further characterize interactions of EN450 with UBE2D, we synthesized an alkyne-functionalized

probe derivative of EN450—EK-1-8 (Figure 2b). EK-1-8 still impaired HAP1 cell viability and this effect was also

attenuated with the NEDDylation inhibitor, as was observed with EN450 (Figure 2c). EK-1-8 directly and

covalently labeled recombinant UBE2D1 C85S mutant protein in a dose-responsive manner (Figure 2d). We

further demonstrated that EK-1-8 engaged UBE2D1, but not an unrelated protein such as GAPDH, in cells and

that this engagement was outcompeted in-part by EN450, as assessed by enrichment of UBE2D1 in HAP1 cells

by EK-1-8 treatment, appendage of biotin through azide-alkyne cycloaddition (CuAAC), avidin-enrichment,

elution, and blotting (Figure 2e). To understand the contribution of the individual UBE2D isoforms to EN450

effects, we assessed dose-responsive effects of EN450 on HAP1 cell proliferation upon knockdown of UBE2D1,

UBE2D2, UBE2D3, and UBE2D4 and knockdown of all four enzymes (Figure 2f-2g, Figure S1). Knockdown of

UBE2D1 conferred the greatest degree of resistance to EN450 antiproliferative effects, followed by knockdown

of UBE2D4 and knockdown of all four enzymes. Intriguingly, knockdown of UBE2D2 and UBE2D3 led to

hypersensitization to EN450, suggesting divergent roles of each individual E2 ligase isoform in relation to

.CC-BY-NC-ND 4.0 International licenseavailable under a

was not certified by peer review) is the author/funder, who has granted bioRxiv a license to display the preprint in perpetuity. It is made

The copyright holder for this preprint (whichthis version posted May 18, 2022. ; https://doi.org/10.1101/2022.05.18.492542doi: bioRxiv preprint

interactions with EN450 despite their high degree of sequence identity (Figure 2f-2g, Figure S1). Nonetheless,

our data indicated that UBE2D1 and UBE2D4 are likely the UPS components covalently engaged by EN450 to,

presumably along with the recruitment of its associated Cullin E3 ligase complex, ubiquitinate and degrade a yet

unknown target protein to exert its antiproliferative effects.

To identify the degraded target protein, we next performed tandem mass tagging (TMT)-based

quantitative proteomic profiling to map protein level changes from EN450 treatment in HAP1 cells. This profiling

effort revealed only one protein—the NFKB1 p105 subunit—that was significantly reduced in levels by >4-fold

upon EN450 treatment compared to vehicle-treated controls (Figure 3a; Table S3). This loss of NFKB1 p105

was confirmed by Western blotting (Figure 3b-3c). We further confirmed that this reduction in NFKB1 p105

levels by EN450 treatment was attenuated by NEDDylation inhibitor pre-treatment (Figure 3d-3e). Further

corroborating EN450-mediated molecular glue interactions, we demonstrated that EN450 promoted ternary

complex formation between UBE2D1 and NFKB1, through pulldown studies with recombinant GST-NFKB1 and

UBE2D1 (Figure 3f-3g).

We next sought to link NFKB1 degradation as the mechanism underlying EN450 anti-proliferative effects.

Given that NFKB1 is a known oncogenic transcription factor that drives cancer cell proliferation, we conjectured

that the EN450-mediated degradation of NFKB1 led to impaired cell proliferation (Chaturvedi et al., 2011). Stable

overexpression of NFKB1 in HAP1 cells led to significant rescue of EN450-mediated anti-proliferative effects in

a dose-dependent manner (Figure 3h-3i). This rescue was relatively modest potentially due to a low degree of

NFKB1 overexpression or stable expression of NFKB1 in cells leading to oncogene addiction or other

compensatory changes. We thus also transiently transfected NFKB1 in HEK293T cells, which are more

amenable to transient transfection compared to HAP1 cells, where we also demonstrated NEDDylation-

dependent cell viability impairments and reduction in NFKB1 levels with EN450 treatment (Figure 3j-3k). This

transient transfection of NFKB1 in HEK293T cells led to a significant and more robust rescue of EN450-mediated

anti-proliferative effects (Figure 3l-3m). Overall, these data support the mechanism that EN450 brings together

UBE2D1 with NFKB1 to ubiquitinate and degrade NFKB1 leading to antiproliferative effects.

.CC-BY-NC-ND 4.0 International licenseavailable under a

was not certified by peer review) is the author/funder, who has granted bioRxiv a license to display the preprint in perpetuity. It is made

The copyright holder for this preprint (whichthis version posted May 18, 2022. ; https://doi.org/10.1101/2022.05.18.492542doi: bioRxiv preprint

Discussion

In this study, we have used covalent chemoproteomic approaches to discover a novel covalent molecular

glue degrader that induces the proximity of an E2 ubiquitin ligase UBE2D1 to the oncogenic nuclear transcription

factor NFKB1 to ubiquitinate and degrade NFKB1 in a proteasome-dependent manner, leading to anti-

proliferative effects in cancer cells. Our study also reveals that allosteric sites within E2 ubiquitin ligases, rather

than E3 ligases or E3 ligase substrate adaptor proteins, can be directly targeted by small-molecules to engage

in molecular glue interactions with neo-substrate proteins. While the neddylation-dependency indicates that

UBE2D still needs to be coupled to the Cullin E3 ligase complex to mark NFKB1 for degradation and confer anti-

proliferative effects based on the attenuation of these effects by a NEDDylation inhibitor, we do not yet know

whether recruitment of the E2 ligase bypasses the necessity for a substrate adapter protein to induce the

ubiquitination and degradation of NFKB1. The mechanism for NFKB1 degradation may also operate through

initial monoubiquitination of NFKB1 through the EN450-mediated UBE2D/NFKB1 molecular glue interaction that

subsequently sensitizes NFKB1 to polyubiquitination and degradation through a combination of or specific E3

ligases. Of future interest is the investigation of whether NFKB1 already interacts weakly with this allosteric site

on UBE2D, and whether EN450 thus just strengthens an already existing interaction or whether NFKB1 truly

represents a neo-substrate that engages UBE2D only upon EN450 binding. While we demonstrate here that

EN450 directly engages with UBE2D1 alone without NFKB1, it is unclear at this point whether EN450 has any

inherent reversible affinity for NFKB1 or whether UBE2D1 binding to EN450 is necessary to recruit NFKB1. Our

study is reminiscent of previously reported studies by Slabicki and Ebert et al. demonstrating that the CDK8

inhibitor CR8 acts as a molecular glue degrader of cyclin K through engaging a core adapter protein of the CUL4

E3 ligase machinery DDB1, thereby bypassing the requirement for a substrate receptor for ubiquitination and

degradation (Słabicki et al., 2020). Both our and the Ebert study demonstrate that core proteins within the

ubiquitin proteasome machinery, such as E2 ligases and DDB1, can be exploited for targeted protein degradation

application. This indicates that small-molecule recruiters can be developed against these proteins to be

deployed in heterobifunctional degraders to either act independently of substrate receptors or to potentially

swarm targets with more than one substrate receptor E3 ligase to degrade neo-substrate target proteins. Taken

more broadly, our study also demonstrates the utility of coupling covalent ligand screening with chemoproteomic

.CC-BY-NC-ND 4.0 International licenseavailable under a

was not certified by peer review) is the author/funder, who has granted bioRxiv a license to display the preprint in perpetuity. It is made

The copyright holder for this preprint (whichthis version posted May 18, 2022. ; https://doi.org/10.1101/2022.05.18.492542doi: bioRxiv preprint

and quantitative proteomic platforms to rapidly discover novel molecular glue degraders and their ternary

complex components.

Acknowledgement

We thank the members of the Nomura Research Group and Novartis Institutes for BioMedical Research for

critical reading of the manuscript. This work was supported by Novartis Institutes for BioMedical Research and

the Novartis-Berkeley Center for Proteomics and Chemistry Technologies (NB-CPACT) for all listed authors.

This work was also supported by the Nomura Research Group and the Mark Foundation for Cancer Research

ASPIRE Award for DKN, EAK. This work was also supported by grants by the National Science Foundation

Molecular Foundations for Biotechnology Award (2127788) (for DKN) and the National Science Foundation

Graduate Research Fellowship (for EAK). We also thank Drs. Hasan Celik, Alicia Lund, and UC Berkeley's NMR

facility in the College of Chemistry (CoC-NMR) for spectroscopic assistance. Instruments in the CoC-NMR are

supported in part by NIH S10OD024998.

Author Contributions

EAK, DKN conceived of the project idea, designed experiments, performed experiments, analyzed and

interpreted the data, and wrote the paper. EAK, YC, DD, DKN performed experiments, analyzed and

interpreted data, and provided intellectual contributions. EAK, DD, JAT, JMK, MS provided intellectual

contributions to the project and overall design of the project.

Competing Financial Interests Statement

JAT, JMK, DD are employees of Novartis Institutes for BioMedical Research. This study was funded by the

Novartis Institutes for BioMedical Research and the Novartis-Berkeley Center for Proteomics and Chemistry

Technologies. DKN is a co-founder, shareholder, and adviser for Frontier Medicines and Vicinitas

Therapeutics. DKN is also on the scientific advisory board of The Mark Foundation for Cancer Research,

Photys Therapeutics, and Apertor Pharmaceuticals, and is a consultant for MPM Capital and Droia Ventures.

.CC-BY-NC-ND 4.0 International licenseavailable under a

was not certified by peer review) is the author/funder, who has granted bioRxiv a license to display the preprint in perpetuity. It is made

The copyright holder for this preprint (whichthis version posted May 18, 2022. ; https://doi.org/10.1101/2022.05.18.492542doi: bioRxiv preprint

References

Backus, K.M., Correia, B.E., Lum, K.M., Forli, S., Horning, B.D., González-Páez, G.E., Chatterjee, S., Lanning,

B.R., Teijaro, J.R., Olson, A.J., et al. (2016). Proteome-wide covalent ligand discovery in native biological

systems. Nature 534, 570–574. https://doi.org/10.1038/nature18002.

Baek, K., Scott, D.C., and Schulman, B.A. (2021). NEDD8 and ubiquitin ligation by cullin-RING E3 ligases.

Curr. Opin. Struct. Biol. 67, 101–109. https://doi.org/10.1016/j.sbi.2020.10.007.

Belcher, B.P., Ward, C.C., and Nomura, D.K. (2021). Ligandability of E3 Ligases for Targeted Protein

Degradation Applications. Biochemistry https://doi.org/10.1021/acs.biochem.1c00464.

Bond, M.J., and Crews, C.M. (2021). Proteolysis targeting chimeras (PROTACs) come of age: entering the

third decade of targeted protein degradation. RSC Chem. Biol. 2, 725–742. https://doi.org/10.1039/d1cb00011j.

Burslem, G.M., and Crews, C.M. (2020). Proteolysis-Targeting Chimeras as Therapeutics and Tools for

Biological Discovery. Cell 181, 102–114. https://doi.org/10.1016/j.cell.2019.11.031.

Chaturvedi, M.M., Sung, B., Yadav, V.R., Kannappan, R., and Aggarwal, B.B. (2011). NF-κB addiction and its

role in cancer: ‘one size does not fit all.’ Oncogene 30, 1615–1630. https://doi.org/10.1038/onc.2010.566.

Chung, C.Y.-S., Shin, H.R., Berdan, C.A., Ford, B., Ward, C.C., Olzmann, J.A., Zoncu, R., and Nomura, D.K.

(2019). Covalent targeting of the vacuolar H+-ATPase activates autophagy via mTORC1 inhibition. Nat. Chem.

Biol. 15, 776–785. https://doi.org/10.1038/s41589-019-0308-4.

Dixon, S.J., and Stockwell, B.R. (2009). Identifying druggable disease-modifying gene products. Curr. Opin.

Chem. Biol. 13, 549–555. https://doi.org/10.1016/j.cbpa.2009.08.003.

Grossman, E.A., Ward, C.C., Spradlin, J.N., Bateman, L.A., Huffman, T.R., Miyamoto, D.K., Kleinman, J.I., and

Nomura, D.K. (2017). Covalent Ligand Discovery against Druggable Hotspots Targeted by Anti-cancer Natural

Products. Cell Chem. Biol. 24, 1368-1376.e4. https://doi.org/10.1016/j.chembiol.2017.08.013.

Henning, N.J., Manford, A.G., Spradlin, J.N., Brittain, S.M., Zhang, E., McKenna, J.M., Tallarico, J.A., Schirle,

M., Rape, M., and Nomura, D.K. (2022a). Discovery of a Covalent FEM1B Recruiter for Targeted Protein

Degradation Applications. J. Am. Chem. Soc. 144, 701–708. https://doi.org/10.1021/jacs.1c03980.

Henning, N.J., Boike, L., Spradlin, J.N., Ward, C.C., Liu, G., Zhang, E., Belcher, B.P., Brittain, S.M., Hesse,

M.J., Dovala, D., et al. (2022b). Deubiquitinase-targeting chimeras for targeted protein stabilization. Nat.

Chem. Biol. https://doi.org/10.1038/s41589-022-00971-2.

Mayor-Ruiz, C., Bauer, S., Brand, M., Kozicka, Z., Siklos, M., Imrichova, H., Kaltheuner, I.H., Hahn, E., Seiler,

K., Koren, A., et al. (2020). Rational discovery of molecular glue degraders via scalable chemical profiling. Nat.

Chem. Biol. 16, 1199–1207. https://doi.org/10.1038/s41589-020-0594-x.

Meissner, F., Geddes-McAlister, J., Mann, M., and Bantscheff, M. (2022). The emerging role of mass

spectrometry-based proteomics in drug discovery. Nat. Rev. Drug Discov. 1–18.

https://doi.org/10.1038/s41573-022-00409-3.

Schreiber, S.L. (2021). The Rise of Molecular Glues. Cell 184, 3–9. https://doi.org/10.1016/j.cell.2020.12.020.

Słabicki, M., Kozicka, Z., Petzold, G., Li, Y.-D., Manojkumar, M., Bunker, R.D., Donovan, K.A., Sievers, Q.L.,

Koeppel, J., Suchyta, D., et al. (2020). The CDK inhibitor CR8 acts as a molecular glue degrader that depletes

cyclin K. Nature 585, 293–297. https://doi.org/10.1038/s41586-020-2374-x.

Spradlin, J.N., Hu, X., Ward, C.C., Brittain, S.M., Jones, M.D., Ou, L., To, M., Proudfoot, A., Ornelas, E.,

Woldegiorgis, M., et al. (2019a). Harnessing the anti-cancer natural product nimbolide for targeted protein

degradation. Nat. Chem. Biol. 15, 747–755. https://doi.org/10.1038/s41589-019-0304-8.

.CC-BY-NC-ND 4.0 International licenseavailable under a

was not certified by peer review) is the author/funder, who has granted bioRxiv a license to display the preprint in perpetuity. It is made

The copyright holder for this preprint (whichthis version posted May 18, 2022. ; https://doi.org/10.1101/2022.05.18.492542doi: bioRxiv preprint

Spradlin, J.N., Hu, X., Ward, C.C., Brittain, S.M., Jones, M.D., Ou, L., To, M., Proudfoot, A., Ornelas, E.,

Woldegiorgis, M., et al. (2019b). Harnessing the anti-cancer natural product nimbolide for targeted protein

degradation. Nat. Chem. Biol. 15, 747–755. https://doi.org/10.1038/s41589-019-0304-8.

Spradlin, J.N., Zhang, E., and Nomura, D.K. (2021). Reimagining Druggability Using Chemoproteomic

Platforms. Acc. Chem. Res. 54, 1801–1813. https://doi.org/10.1021/acs.accounts.1c00065.

Vinogradova, E.V., Zhang, X., Remillard, D., Lazar, D.C., Suciu, R.M., Wang, Y., Bianco, G., Yamashita, Y.,

Crowley, V.M., Schafroth, M.A., et al. (2020). An Activity-Guided Map of Electrophile-Cysteine Interactions in

Primary Human T Cells. Cell 182, 1009-1026.e29. https://doi.org/10.1016/j.cell.2020.07.001.

Weerapana, E., Wang, C., Simon, G.M., Richter, F., Khare, S., Dillon, M.B.D., Bachovchin, D.A., Mowen, K.,

Baker, D., and Cravatt, B.F. (2010). Quantitative reactivity profiling predicts functional cysteines in proteomes.

Nature 468, 790–795. https://doi.org/10.1038/nature09472.

Xu, T., Park, S.K., Venable, J.D., Wohlschlegel, J.A., Diedrich, J.K., Cociorva, D., Lu, B., Liao, L., Hewel, J.,

Han, X., et al. (2015). ProLuCID: An improved SEQUEST-like algorithm with enhanced sensitivity and

specificity. J. Proteomics 129, 16–24. https://doi.org/10.1016/j.jprot.2015.07.001.

Zhang, X., Crowley, V.M., Wucherpfennig, T.G., Dix, M.M., and Cravatt, B.F. (2018). Electrophilic PROTACs

that degrade nuclear proteins by engaging DCAF16. BioRxiv 443804. https://doi.org/10.1101/443804.

Zhang, X., Luukkonen, L.M., Eissler, C.L., Crowley, V.M., Yamashita, Y., Schafroth, M.A., Kikuchi, S.,

Weinstein, D.S., Symons, K.T., Nordin, B.E., et al. (2021). DCAF11 Supports Targeted Protein Degradation by

Electrophilic Proteolysis-Targeting Chimeras. J. Am. Chem. Soc. 143, 5141–5149.

https://doi.org/10.1021/jacs.1c00990.

.CC-BY-NC-ND 4.0 International licenseavailable under a

was not certified by peer review) is the author/funder, who has granted bioRxiv a license to display the preprint in perpetuity. It is made

The copyright holder for this preprint (whichthis version posted May 18, 2022. ; https://doi.org/10.1101/2022.05.18.492542doi: bioRxiv preprint

Figure Legends

Figure 1. Discovery of a novel covalent molecular glue degrader with anti-proliferative activities in

HAP1 leukemia cancer cells. a) HAP1 cell viability screen of cysteine-reactive covalent ligands. DMSO

vehicle or cysteine-reactive covalent ligands (50 µM) were treated in HAP1 cells for 24 h and cell viability was

assessed by Hoescht staining. Compounds highlighted in red and labeled are those that impaired HAP1 cell

viability by greater than 90 % compared to vehicle-treated controls. This screen was performed with n=1

biological replicate/group. (b) The 11 hit compounds from (a) were rescreened with n=3 biologically

independent replicates/group in HAP1 cells under the same conditions. Individual replicate values shown.

Among these hits EN222, EN226, and EN450 showed reproducible impaired HAP1 cell viability of greater than

90 % compared to DMSO vehicle-treated controls. (c) Knockdown of UBE2M. HAP1 cells were transiently

transfected with siControl or siUBE2M oligonucleotides and knockdown of UBE2M was assessed by Western

blotting compared to GAPDH loading control. (d) HAP1 cell viability in siControl and siUBE2M cells treated with

DMSO vehicle, the positive control molecular glue degrader dCEMM1, EN222, EN266, or EN450 (50 µM) for

24 h, assessed by Hoescht staining. (e) structure of EN450 with the reactive acrylamide handle highlighted in

red. (f) Attenuation of HAP1 cell viability impairments by NEDDylation inhibitor MLN4924. HAP1 cells were pre-

treated with DMSO vehicle or MLN4924 (1 µM) for 1 h prior to treatment of cells with DMSO vehicle, dCEMM1,

or EN450 (50 µM) for 24 h, and cell viability was assessed by Hoechst staining. (g) Attenuation of HAP1 cell

viability impairments by proteasome inhibitor bortezomib. HAP1 cells were pre-treated with DMSO vehicle or

bortezomib (1 µM) for 1 h prior to treatment of cells with DMSO vehicle or EN450 (50 µM) for 24 h, and cell

viability was assessed by Hoechst staining. Data in (d, f, g) are average ± sem of n=3-6 biologically

independent replicates/group with individual replicate values also shown. Statistical significance is expressed

as *p<0.05 compared to siControl or DMSO control groups in (d, f, g), #p<0.05 compared to corresponding

siControl or DMSO pre-treated treatment groups, and calculated with Student’s unpaired two-tailed t-tests.

Figure 2. Chemoproteomic profiling and validation of EN450 targets in HAP1 cells. (a) Cysteine

chemoproteomic profiling of EN450 targets in HAP1 cells using isoTOP-ABPP. HAP1 cells were treated with

DMSO vehicle or EN450 (50 µM) for 3 h, after which resulting cell lysates were labeled with an alkyne-

functionalized iodoacetamide cysteine-reactive probe (200 µM) for 1 h, and an isotopically light (for DMSO) or

.CC-BY-NC-ND 4.0 International licenseavailable under a

was not certified by peer review) is the author/funder, who has granted bioRxiv a license to display the preprint in perpetuity. It is made

The copyright holder for this preprint (whichthis version posted May 18, 2022. ; https://doi.org/10.1101/2022.05.18.492542doi: bioRxiv preprint

heavy (for EN450) biotin-azide handle bearing a TEV protease recognition peptide was appended by CuAAC.

Control and treated proteomes were combined in a 1:1 ratio, taken through the isoTOP-ABPP procedure and

light/heavy probe-modified peptides were analyzed by LC-MS/MS and quantified. Shown in blue are probe-

modified peptides with light/heavy ratios >4 with adjusted p<0.05. Shown in red is C111 of UBE2D. Data

shown are average ratios from n=4 biologically independent replicates/group. (b) structure of alkyne-

functionalized probe derivative of EN450—EK-1-8 with cysteine-reactive acrylamide warhead highlighted in

red. (c) Attenuation of HAP1 cell viability impairments by NEDDylation inhibitor MLN4924. HAP1 cells were

pre-treated with DMSO vehicle or MLN4924 (1 µM) for 1 h prior to treatment of cells with DMSO vehicle, EK-1-

8 (50 µM) for 24 h, and cell viability was assessed by Hoechst staining. (d) EK-1-8 labeling of recombinant

pure human UBE2D1 C85S mutant protein. UBE2D1 C85S mutant protein was labeled with DMSO vehicle or

EK-1-8 for 30 min, after which an azide functionalized rhodamine handle was appended onto probe-labeled

protein by CuAAC, after which proteins were resolved by SDS/PAGE and visualized by in-gel fluorescence.

Protein loading was assessed by silver staining. Gels are representative of an n=3 biologically independent

replicates/group. (e) UBE2D1 pulldown from HAP1 cells with EK-1-8 probe. HAP1 cells were pre-treated with

DMSO vehicle or EN450 (50 µM) for 1 h prior to treatment of cells with DMSO vehicle or EK-1-8 (10 µM) for 3

h. Probe-labeled proteins from resulting cell lysates were subsequently appended to an azide-functionalized

biotin handle by CuAAC, avidin-enriched, eluted and separated by SDS/PAGE and blotted for UBE2D1 and

unrelated protein GAPDH. Shown on the blot are also input UBE2D1 and GAPDH protein levels between the

three groups shown. Blots are representative of an n=3 biologically independent replicates/group. (f)

Confirmation of UBE2D1 knockdown. HAP1 cells were transiently transfected with siControl or siUBE2D1

oligonucleotides and UBE2D1 and loading control GAPDH expression were assessed after 72 h by Western

blotting. Blot is representative of n=3 biologically independent replicates/group. (g) Percent HAP1 cell

proliferation in siControl and siUBE2D1 cells treated with EN450 for 24 h compared to DMSO vehicle-treated

controls. Data shown in (c) and (g) are average ± sem of n=6-30 biologically independent replicates/group with

individual replicate values also shown. Statistical significance in (c) is expressed as *p<0.05 compared to

DMSO-pre-treated control groups and #p<0.05 compared to DMSO-pre-treated EK-1-8-treated groups and in

(g) is expressed as *p<0.05 compared to the corresponding treatment group in siControl cells.

.CC-BY-NC-ND 4.0 International licenseavailable under a

was not certified by peer review) is the author/funder, who has granted bioRxiv a license to display the preprint in perpetuity. It is made

The copyright holder for this preprint (whichthis version posted May 18, 2022. ; https://doi.org/10.1101/2022.05.18.492542doi: bioRxiv preprint

Figure 3. Identification of the protein degraded by EN450. (a) TMT-based quantitative proteomic profiling of

EN450 in HAP1 cells. HAP1 cells were treated with DMSO vehicle or EN450 (50 µM) for 24 h. Shown in red is

the only protein reduced in levels by >4-fold with p-value <0.05—NFKB1. Data shown are from n=3 biologically

independent replicates/group. (b) Western blotting analysis of NFKB1 and loading control vinculin levels in

HAP1 cells treated with DMSO vehicle or EN450 (50 µM) for 24 h. (c) Quantification of experiment described in

(b). (d) NEDDylation inhibitor attenuates NFKB1 loss by EN450 treatment in HAP1 cells. HAP1 cells were pre-

treated with DMSO vehicle or MLN4924 (1 µM) for 1 h prior to treatment of cells with DMSO vehicle or EN450

(50 µM) for 24 h. NFKB1 and loading control vinculin levels were assessed by Western blotting. (e)

Quantification of experiment described in (d). (f) Ternary complex formation between UBE2D1, EN450, and

NFKB1. Recombinant pure human GST-NFKB1 and UBE2D1 proteins were incubated with DMSO vehicle or

EN450 (50 µM) for 1 h, after which GST-NFKB1 was enriched and the pulldown eluate was blotted for NFKB1

and UBE2D1. (g) Quantification of pulldown experiment described in (f). (h) Stable lentiviral overexpression of

FLAG-NFKB1 in HAP1 cells assessed by Western blotting for NFKB1, FLAG, and loading control GAPDH. (i)

HAP1 eGFP control or FLAG-NFKB1 overexpressing cell viability from cells treated with DMSO vehicle or

EN450 for 24 h. (j) Attenuation of HEK293T cell viability impairments by NEDDylation inhibitor MLN4924.

HEK293T cells were pre-treated with DMSO vehicle or MLN4924 (1 µM) for 1 h prior to treatment of cells with

DMSO vehicle or EN450 (50 µM) for 24 h, and cell viability was assessed by Hoechst staining. (k) NFKB1 and

loading control vinculin levels in HEK293T cells pre-treated with DMSO vehicle or MLN4924 (1 µM) 1 h prior to

treatment of cells with DMSO vehicle or EN450 (50 µM) for 24 h, assessed by Western blotting. (l) Transient

transfection and overexpression of FLAG-NFKB1 in HEK293T cells assessed by Western blotting of NFKB1,

FLAG, and loading control GAPDH. (m) HEK293T eGFP control or FLAG-NFKB1 overexpressing cell viability

from cells treated with DMSO vehicle or EN450 (50 µM) for 24 h. Blots shown in (b, d, f, h, k, l) are

representative of n=3 biologically independent replicates/group. Data shown in (c, e, g, j, m) are average ±

sem of n=3-8 biologically independent replicates/group with individual replicate values also shown. Statistical

significance in (c, e, g, j) are expressed as *p<0.05 compared to DMSO vehicle-treated controls or eGFP-

expressing control cells and in (e, l, j) as #p<0.05 compared to DMSO pre-treated EN450-treated groups in (e)

or EN450-treated eGFP-expressing cells in (m). Significance is expressed as *p<0.05 in (i) compared to the

corresponding eGFP-expressing treatment groups.

.CC-BY-NC-ND 4.0 International licenseavailable under a

was not certified by peer review) is the author/funder, who has granted bioRxiv a license to display the preprint in perpetuity. It is made

The copyright holder for this preprint (whichthis version posted May 18, 2022. ; https://doi.org/10.1101/2022.05.18.492542doi: bioRxiv preprint

Materials and Methods

Materials

Cysteine-reactive covalent ligand libraries were either previously synthesized and described or for the

compounds starting with “EN” were purchased from Enamine, including EN450.

Cell Culture

HAP1 cells were purchased from Horizon Discovery and were cultured in IMDM containing 10% (v/v) fetal

bovine serum (FBS) and maintained at 37 °C with 5% CO

. HEK293T cells were obtained from the American

Type Culture Continued. HEK293T cells were cultured in DMEM containing 10% (v/v) fetal bovine serum (FBS)

and maintained at 37° C with 5% CO2. Unless otherwise specified, all cell culture materials were purchased

from Gibco. It is not known whether HEK293T cells are from male or female origin.

Preparation of Cell Lysates

Cells were washed twice with cold PBS, scraped, and pelleted by centrifugation (1,400 g, 4 min, 4° C). Pellets

were resuspended in PBS, sonicated, clarified by centrifugation (21,000 g, 10 min, 4° C), and lysate was

transferred to new low-adhesion microcentrifuge tubes. Proteome concentrations were determined using BCA

assay and lysate was diluted to appropriate working concentrations.

Cell Viability

Cell viability assays were performed using Hoechst 33342 dye (Invitrogen, H3570) according to the

manufacturer’s protocol. For survival assays, cells were seeded into 96-well plates (20,000 per well) in a

volume of 100 µL and allowed to adhere overnight, cells were treated with an additional 50 µL of media

containing DMSO vehicle or compound (150 µM) in 0.1% DMSO for 24 h. For rescue experiments, cell media

was changed to media (100 µL per well) containing Bortezimib (1 µM) or MLN4924 (1 µM) for 1 h prior to

compound treatment. For dose response experiments, dilutions of compound were prepared in DMSO prior to

dilution in media. After incubation, the media was aspirated from each well, and 100 µL of staining solution

containing 10% formalin and Hoechst 33342 dye was added to each well and incubated for 25 min in the dark

.CC-BY-NC-ND 4.0 International licenseavailable under a

was not certified by peer review) is the author/funder, who has granted bioRxiv a license to display the preprint in perpetuity. It is made

The copyright holder for this preprint (whichthis version posted May 18, 2022. ; https://doi.org/10.1101/2022.05.18.492542doi: bioRxiv preprint

at room temperature. Staining solution was then removed, and wells were washed with 3x with PBS before

fluorescent imaging.

Production of Recombinant UBE2D1

A UBE2D1(1-147) expression plasmid was synthesized by Twist Biosciences with an N-terminal 8xHistidine

tag and HRV 3C cleavage site. The C85S mutation was produced using a Q5 Site Directed Mutagenesis Kit

(New England Biolabs) and standard cloning techniques. Hi Control BL21(DE3) E. coli cells (Lucigen) were

transformed with expression plasmid and a single colony was used to start an overnight culture in LB media,

followed by inoculation of a 1 L culture in Terrific Broth supplemented with 50 mM sodium phosphate

monobasic pH 7.0 and 50 µg/mL kanamycin. This culture grew at 37 °C until the OD600 reached

approximately 1.2, at which point the temperature was reduced to 19 °C along with immediate addition of 0.5

mM (final) IPTG. The cells were allowed to grow overnight prior to harvesting via centrifugation. The cell pellet

was resuspended in IMAC_A Buffer (50 mM Tris pH 8.0, 400 mM NaCl, 1 mM TCEP, 20 mM imidazole) and

lysed with three passages through a cell homogenizer at 18,000 psi. Cell lysate was then clarified with

centrifugation at 45,000 x g for 30 minutes. The clarified lysate was flowed through a 5 mL HisTrap Excel

column (Cytiva) pre-equilibrated with IMAC_A Buffer. After loading, the resin was washed with IMAC_A Buffer

until the UV absorbance reached baseline, after which the protein was eluted with a linear gradient of IMAC_B

Buffer (50 mM Tris pH 8.0, 400 mM NaCl, 1 mM TCEP, 500 mM imidazole). The eluate was treated with HRV

3C protease until cleavage of the histidine tag was complete as determined by ESI-LC/MS. Cleavage

proceeded while the protein dialyzed into IMAC_A Buffer, enabling reverse-IMAC purification which was

conducted in batch using 3 mL of Ni-NTA resin pre-equilibrated with IMAC_A Buffer. The flow-through was

collected, concentrated, and subjected to size exclusion chromatography using a Superdex 75 16/60 column

pre-equilibrated with SEC Buffer (25 mM HEPES pH 7.5, 150 mM NaCl, 1 mM TCEP). Fractions within the

included volume were analyzed by SDS-PAGE and those containing pure protein were pooled and

concentrated. This protocol was used to produce both wild-type and C85S variants of UBE2D1 with an

approximate yield of ~90 mg/L.

Labeling of Recombinant UBE2D1 with EK-1-8 Probe

.CC-BY-NC-ND 4.0 International licenseavailable under a

was not certified by peer review) is the author/funder, who has granted bioRxiv a license to display the preprint in perpetuity. It is made

The copyright holder for this preprint (whichthis version posted May 18, 2022. ; https://doi.org/10.1101/2022.05.18.492542doi: bioRxiv preprint

For in vitro labelling of UBE2D1 C85S, recombinant pure human protein (0.1 µg per sample) was treated with

either DMSO vehicle or EK-1-8 at 37° C for 30 min in 25 µL PBS. Each sample was incubated with 1 µL of 30

µM rhodamine-azide (in DMSO), 1 µL of 50 mM TCEP (in water), 3 µL of TBTA ligand (0.9 mg/mL in 1:4

DMSO/t-BuOH), and 1 µL of 50 mM Copper (II) Sulfate for 1 h at room temperature. Samples were then

diluted with 10 µL of 4x reducing Laemmli SDS sample loading buffer (Alfa Aesar), boiled at 95° C for 5 min,

and separated by SDS/PAGE. Probe-labeled proteins were analyzed by in-gel rhodamine fluorescence using a

ChemiDoc MP (Bio-Rad). Protein loading was assessed by silver staining.

Pulldown of UBE2D1 from HAP1 Cells with EK-1-8 Probe

HAP1 WT cells were pretreated at 70% confluency with DMSO or 50 µM EN450 in situ for 1 h followed by

treatment with DMSO or 10 µM EK-1-8 in situ for 3 h . Cells were harvested, lysed via sonication, and the

resulting lysate normalized to 6 mg/mL per sample. Following normalization, 30 µL of each lysate sample was

removed for Western blot analysis of input, and 500 µL of each lysate sample was incubated for 2 h at room

temperature with 10 µL of 5 mM biotin picolyl azide (in DMSO) (Sigma Aldrich 900912), 10 µL of 50 mM TCEP

(in water), 30 µL of TBTA ligand (0.9mg/mL in 1:4 DMSO/t-BuOH), and 10 µL of 50 mM Copper (II) Sulfate.

Proteins were precipitated, washed 3 x with cold MeOH, resolubilized in 200 µL of 1.2% SDS/PBS (w/v),

heated for 5 min at 98 C, and centrifuged to remove any insoluble components. Each resolubilized sample was

then transferred to 1.5 mL eppendorf low-adhesion tubes containing 1 mL PBS with 30 µL high-capacity

streptavidin resin (Thermo Scientific 20357) to give a final SDS concentration of 0.2%. Samples were

incubated with the streptavidin beads at 4° C overnight on a rotator. The following day the samples were

warmed to room temperature and washed with 0.2% SDS and further washed 3 x with 500 µL PBS and 3 x

with 500 µL water to remove non-probe-labeled proteins. For western blot analysis, the washed beads were

resuspended in 30 µL PBS, transferred to 1.5 mL eppendorf low-adhesion tubes, combined with 10 µL

Laemmli Sample Buffer (4 x), heated to 95° C for 5 min, and analyzed by Western blotting to look for enriched

UBE2D1 (Abcam ab32072) versus non-enriched control Vinculin (Bio-Rad MCA465GA).

.CC-BY-NC-ND 4.0 International licenseavailable under a

was not certified by peer review) is the author/funder, who has granted bioRxiv a license to display the preprint in perpetuity. It is made

The copyright holder for this preprint (whichthis version posted May 18, 2022. ; https://doi.org/10.1101/2022.05.18.492542doi: bioRxiv preprint